Bio-Inspired Surfaces and Appli - Eddie Y K NG

May 2, 2013 14:6 BC: 8831 - Probability and Statistical Theory PST˙ws
This page intentionally left blank

Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
Library of Congress Cataloging-in-Publication Data

Names: Luo, Yuehao, editor. | Ng, Y. K. Eddie, editor.
Title: Bio-inspired surfaces and applications / edited by Yuehao Luo
(The George Washington University, USA), Eddie Y.K. Ng (NTU, Singapore).
Description: Hackensack, New Jersey : World Scientific, [2016]
Identifiers: LCCN 2016020391 | ISBN 9789814704489 (hardcover : alk. paper)
Subjects: LCSH: Biomimetic polymers. | Biomedical materials. | Hydrophobic surfaces.
Classification: LCC TP248.65.P62 B519 2016 | DDC 610.28/4--dc23
LC record available at https://lccn.loc.gov/2016020391
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
Copyright © 2016 by World Scientific Publishing Co. Pte. Ltd.

All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.
Typeset by Stallion Press

Email: enquiries@stallionpress.com
Printed in Singapore
Devi - Bio-Inspired Surfaces and Applications.indd 1 19/5/2016 11:10:21 AM

May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page v
Preface
The book addresses the latest technology developments in the area of

biomimetic drag-reducing surfaces and their application to fluid and soil
engineering. For examples, the book looks into the naturally occurring
drag-reducing shark skin, where accurate 3-D digital models of shark
skin are built through a highly accurate scanning method based on
real biological shark skin. It also discusses possible application of the
bio-inspired drag-reducing technology in natural gas pipelines, where
adopting the internal coating technology for use in “hydraulic smooth
pipes” reduces friction and increases transmission capacity. In brief,
the book highlights many other important topics such as “shape
imitation” and “spirit imitation” in bio-inspired systems and insect–
machine hybrid systems and even bio-inspired increasing of friction as
another extreme aspect, etc.
The aims of this book are mainly focused on the following aspects:
(1) understanding different bio-inspired technologies; (2) compre-
hending bio-inspired mechanisms; (3) expanding more bio-inspired
technologies into different fields. The book covers wide-ranging topics
as follows:
Bionics/biomimetics/bio-inspiration studies the working regula-
tions of biological systems and the functions developed in the process
of evolution and natural selection; it builds the models in order to apply
them to design and fabricate new and novel equipment. The ultimate
purpose of bionics is not to copy every detail of a biological system, but
to analyze the structures and functions of biology, better understand
the working principles, and finally realize the applications in the future
designs.
v
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page vi
vi Bio-Inspired Surfaces and Applications
Chapter 1 examines different aspects of the self-cleaning function

and super-hydrophobic surfaces, including advanced wetting theory,
methods of fabricating different super-hydrophobic surfaces and the
potential applications of super-hydrophobic surfaces. This helps readers
better understand recent progress in research involving the self-cleaning
effect and super-hydrophobic surfaces.
Sharks have been living on the Earth for more than 400 million
years, and they appeared 100 million earlier years than the dinosaurs,
which demonstrate their perfect properties for surviving the environ-
ment as compared to dinosaurs. The treatments of biological shark
skin, which ensure the mechanical intensity and integrity of structure,
are systematically introduced in Chapter 2. The digital model of shark
skin is built through a highly precise scanning method that provides
the intuitive impression where the digital model of a whole shark is
constructed, and the data is processed.
Many researchers have investigated and explored different fabri-
cating methods and applied biomimetic drag-reducing technology to
fluid engineering. Different approaches to manufacture low viscous
resistance drag with biomimetic textures are proposed and discussed
in Chapter 3, including direct bio-replicated forming, micro-rolling,
large-scale solvent-swelling, additive low-releasing design, synthetic
drag-reduction bio-replication, trans-scale enlarged 3-D imprinting,
flexible 3-D imprinting, etc. Experiments on drag-reduction efficiency
are conducted in water tunnel to validate the forming quality and effect,
and satisfactory results are demonstrated.
Chapter 4 presents the method for fabricating artificial composite
drag-reduction surface based on biological shark skin morphology,
where mucus nano-long chains are investigated. The chemical, mechan-
ical and hydrodynamic properties are explored in depth from various
angles; they are important for understanding the status of recent
research and for expanding the applications of shark skin in fluid
engineering.
For purposes of increasing the transportation capacity of natural
gas pipelines and protecting them from corrosion, the internal coating
technology has been vastly applied, leading to remarkable benefits.
Chapter 5 proposes such effective application of biomimetic shark skin
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page vii
Preface vii
surface in the natural gas pipelining industry with the actual working
circumstances of the West-to-East Project in China. To verify the drag
reduction and forming effect, experiments with air are first carried out,
following which the experiment with actual natural gas in the field
transmitting pipeline is performed. The results of the experiments are
consistent with those in the numerical simulation.
Lately, there has been an exciting revival in engineering bio-inspired
robust dropwise condensation surfaces. This is partially due to the
progress in the fundamental understanding of previously unidentified
physical phenomena during the phase change process, as well as
advances made in the preparation of various structured surfaces.
Chapter 6 develops an appreciation for recent bio-inspired strategies for
condensation enhancement as well as to provide a basis for controlling
or predicting the condensation performance of the biomimetic surfaces.
With the improvement of micro/nano fabrication capability, sur-
face structures have become more sophisticated than ever to meet
serious demands on surface functions. Chapter 7 presents a discussion
on the large-scale fabrication of biomimetic drag reduction functions
surface via bio-replication of shark skin. Bio-replicated shark skin
performs maximum drag reduction of about 12% only when the relative
fluid flow is around 5m/s, which agrees well with the normal swimming
speed of a shark. The maximum drag reduction however declines with
improvement of fluid flow speed beyond 5m/s. A proper adjustment
of the surface morphology of the shark skin to accommodate different
fluid flow environments is thus necessary for spreading industrial
applications of the shark skin effect.
Chapter 8 provides an analysis on how drag reduction of flow over a
circular cylinder can be done through having a dimpled surface. Com-
putational fluid dynamics (CFD) simulations are done to investigate the
degree of drag reduction by dimpled cylinder as compared to smooth
cylinder, which corresponded to 8.39% and 9.16% respectively, and the
results are validated by wind tunnel experimentations. With a reliable
and accurate CFD model, the work can be expanded to optimization of
the sizing, shape and arrangement of dimples on the cylinder. Studies
on the effects of dimpled cylinder on vortex shedding and vortex-
induced-vibration are also of great interest. Low-Re analysis on dimples
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page viii
viii Bio-Inspired Surfaces and Applications
on micro tubes may be of interest to the biomedical industry. When

those challenges are addressed, the usage of dimpled cylinders can be
far ranging in many different industries and applications.
Riblets are groove lines or indentations that are protruded into a
structure such as an aerofoil or duct walls such that flow over it can
be made as laminar (Blasius) as possible. The reduction of turbulence
is enhanced by channeling air through the grooves and allowing the
air region above the flat face to stick more closely to the surface
instead of diverting from it. Chapter 9 investigates the optimum ratio
for fin-height versus riblet spacing, physical dimensions of the riblet,
along with the optimum shape using the (L, U, or V) riblet design
configuration.
Chapter 10 describes the usefulness of the 3-D modelling of
biological systems for biomimetics. 3-D modelling plays a key role
in the geometrical characterisation and representation, especially in
computer graphical visualization. The problems associated with the
existing techniques and systems when they are applied to 3-D modelling
of biological systems are discussed. Three case studies are presented to
illustrate the applications of 3-D modelling to biomimetics research and
some areas of future research interests in 3-D modelling of biological
systems and for biomimetics are proposed.
Many methods that are used to create manmade superhydrophobic
surfaces are inspired by the “lotus effect.” Recent research has indi-
cated that there are even more superhydrophobic surfaces in nature,
and this helps to promote the application of biomimetic ideas into
practical fields. Chapter 11 describes the effect of superhydrophobic
surfaces made from SiO2-PDMS films. Water-repelling properties are
rendered by the composite films, and superhydrophobic surfaces are
formed. The SEM observation reveals that the hierarchical structure
in micro and nano scale can be created among the composite films on
superhydrophobic surfaces.
Chapter 12 deals with the application of bio-inspired macro-
morphologic surface modifications to reduce soil–tool adhesion. Keep-
ing in view the enormous associated losses, scientists and engineers have
been trying to solve the problem to reduce the burden of such excessive
energy consumption in sticky soils with unfavorable liability on the
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page ix
Preface ix
power source to develop additional effort for overcoming adhesive

forces. Principles involved under theories of soil adhesion as well as
remedial measures are elaborated. Microscopic forces at the soil–tool
interface are theoretically analyzed, and this opens a window to further
evaluate these concepts for various soil types with different tool shapes.
Application of bio-inspired surfaces can reduce adhesion to the
surfaces of soil-engaging components of agricultural and earth-moving
machinery. Chapter 13 provides a brief history of some methods
employed to improve soil-scouring properties of soil-engaging com-
ponents of different machines and tools. Each of these techniques has
its advantages and disadvantages and their development, especially in
recent years. It discusses the soil adhesion preventing mechanism of
soil animals’ cuticles, and in some situations a combination of these
techniques may be used to achieve the desired outcome.
Adhesion and resistance are mainly problems for soil-engaging
tillage components. However, soil-burrowing or soil-digging animals
give inspiration to resolve those problems. Chapter 14 introduces
some applications of bionic technologies for soil-engaging tillage
components in Northeast China, such as moldboard, subsoiler com-
ponents, furrow opener, roller, and bionic rototilling-stubble-breaking
blade. It reviews biological structures of some soil-burrowing or soil-
digging animals, like the beetle, mole cricket, earthworm, mole, vole,
pangolin, and snake. It also presents their mechanism of anti-adhesion
or resistance reduction. Future development trends are also discussed.
Many esteemed authors have contributed generously and made this
book possible through their diligent hard work and valuable time.
We thank them wholeheartedly for their significant contributions.
The book represents the latest bio-inspired technologies and helps
readers better understand bio-inspired mechanisms and expand their
applications.
E Y K Ng
Yuehao Luo

June 2, 2016 11:0 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xi
Contents
Preface v
List of Contributors xv
1. Self-Cleaning and Superhydrophobic Surfaces 1

G.G. Li, Y.T. Zhao, L. Zhang, B.D. Liu, Y. Luo,
B.Y. Li, E.Y.K. Ng
2. Treatments and Constructing Digital Model

of Biological Shark Skin/Shark 45
G.G. Li, Y.T. Zhao, L. Zhang, Y. Luo, E.Y.K. Ng
3. Different Approaches to Manufacture Low Viscous

Resistance Drag with Biomimetic Textures 69
J. Wang, Y.T. Zhao, L. Zhang, Y. Luo, E.Y.K. Ng
4. Different Characteristic Analysis of Drag-Reducing

Surface with Biological Morphology 113
J. Wang, Y.T Zhao, L. Zhang, Y Luo, E.Y.K. Ng
xi
June 1, 2016 15:27 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xii
xii Bio-Inspired Surfaces and Applications
5. Application of Biomimetic Shark Skin Surface

in Natural Gas Pipelining 143
J. Wang, Y.T Zhao, L. Zhang, Y. Luo, E.Y.K. Ng
6. Biomimetic Surfaces for Enhanced Dropwise

Condensation Heat Transfer: Mimic Nature
and Transcend Nature 185
Youmin Hou, Zuankai Wang, Shuhuai Yao
7. Large-Scale Fabrication of Biomimetic

Drag-Reduction Surface via Bio-Replication
of Shark Skin 229
Huawei Chen, Deyuan Zhang, Xin Zhang,
Da Che
8. Study of Flow over Dimpled Cylinder

for Drag Reduction 271
Tan S.P., Koh J.H. and Ng Y.K. Eddie
9. Fluid Flow in Biomimetics Simulated Vessel Having

a Grooved Surface: An Investigation of the Effect
of Riblets in Drag Reduction 295
Guangming Hu
10. 3-D Modelling of Biological Systems for Biomimetics 325

Shujun Zhang, Donghui Chen, Kevin Hapeshi
and Xu Zhang
11. Superhydrophobic Surfaces with Hierarchical

Structures Inspired by Nature Leaves 393
Yuying Yan and Nan Gao
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xiii
Contents xiii
12. Bio-Inspired Macro-Morphologic Surface

Modifications to Reduce Soil–Tool Adhesion 421
Peeyush Soni and Vilas M. Salokhe
13. Application of Bio-Inspired Surfaces in Reducing

Adhesion to the Surfaces of Soil-Engaging
Components of Agricultural and Earth-Moving
Machinery 485
Rashid Qaisrani and Li Jianqiao
14. Application of Bionic Technologies

for Soil-Engaging Tillage Components
in Northeast China 555
Ji-yu Sun, Zhi-jun Zhang, Jin Tong,
and Hong-lei Jia
Index 579

May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xv
List of Contributors
Da Che
School of Mechanical Engineering, Beihang University, Beijing, China
Donghui Chen
Key Laboratory of Bionics Engineering, Ministry of Education, Jilin
University, China
Huawei Chen
Nan Gao
Faculty of Engineering, University of Nottingham, UK
Guangming Hu
Department of Mechanical and Electrical Engineering Guangdong
Engineering Polytechnic, No. 18, Yuxing Road, Tian he District,
Guangzhou, 510520, China
Kevin Hapeshi
School of Computing and Technology, University of Gloucestershire,
UK
Youmin Hou
Department of Mechanical and Aerospace Engineering, The Hong
Kong University of Science and Technology, Hong Kong, China
Hong-lei Jia
Key Laboratory of Bionic Engineering (Ministry of Education), Jilin
University, Changchun, 130022, China
xv
June 2, 2016 11:0 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xvi
xvi Bio-Inspired Surfaces and Applications
Li Jianqiao
Jilin University, Changchun, China
Koh J.H.
School of Mechanical & Aerospace Engineering, Nanyang Technolog-
ical University, 50 Nanyang Avenue, Singapore 639798
B.D. Liu
School of Mechanical Engineering, Northwestern Polytechnical Uni-
versity, Xi’an, 710072, China
B.Y. Li
G.G. Li
School of Mechanical Engineering, Xi’an Jiaotong University, Xi’an,
710069, China
Y. Luo
School of Engineering, The George Washington University, Washing-
ton D.C., 20052, USA
Eddie Yin-Kwee Ng
School of Mechanical and Aerospace Engineering, Nanyang Techno-
logical University, Singapore 639798
Rashid Qaisrani
Department of Agriculture and Water Resources Canberra, Australia
Ji-yu Sun
Peeyush Soni
Asian Institute of Technology, Thailand
Vilas M. Salokhe
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xvii
List of Contributors xvii
Jin Tong
Tan S.P.
School of Mechanical & Aerospace Engineering, Nanyang Technolog-
ical University, 50 Nanyang Avenue, Singapore 639798
J. Wang
Zuankai Wang
Department of Mechanical and Biomedical Engineering, City Univer-
sity of Hong Kong, Hong Kong, China
Shuhuai Yao
Department of Mechanical and Aerospace Engineering, The Hong
Kong University of Science and Technology, Hong Kong, China
Yuying Yan
Faculty of Engineering, University of Nottingham, UK
Deyuan Zhang
L. Zhang
710069, China
Shujun Zhang
School of Computing and Technology, University of Gloucestershire,
UK;
Key Laboratory of Bionics Engineering, Ministry of Education, Jilin
University, China
Xin Zhang
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-fm page xviii
xviii Bio-Inspired Surfaces and Applications
Xu Zhang
CED Group of Rolls-Royce UTC, Faculty of Engineering and the
Environment, University of Southampton, UK
Y.T. Zhao
710069, China
Zhi-jun Zhang
Institute of Mechanical Science and Engineering, Jilin University,
Changchun 130022, China
May 24, 2016 13:34 Bio-Inspired Surfaces and Applications 9in x 6in b2378-ch01 page 1
Chapter 1
Self-Cleaning
and Superhydrophobic Surfaces
G.G. Li,∗ Y.T. Zhao,∗ L. Zhang,∗ B.D. Liu,† Y. Luo,‡

B.Y. Li,† E.Y.K. Ng§
∗ School of Mechanical Engineering, Xi’an Jiaotong University,
Xi’an, 710069, China
† School of Mechanical Engineering, Northwestern Polytechnical
University, Xi’an, 710072, China

‡ School of Engineering, The George Washington University,
Washington D.C., 20052, USA

§ School of Mechanical and Aerospace Engineering, Nanyang
Technological University, 639798, Singapore
Recently superhydrophobic surfaces have attracted considerable inter-

est due to their potential applications in the design and preparation of
self-cleaning surfaces. Generally speaking, a superhydrophobic surface
is defined as possessing water contact angle greater than 150 degrees.
Such surfaces already commonly exist in the biological world, as
especially exhibited by the lotus leaf, that is, the so-called lotus effect.
To the best of our knowledge, many plant leaves and specific surfaces
of animals in nature all exhibit this so-called lotus effect, where the
interplay of surface microstructure and chemical composition causes
water droplets to remain spherical on the surface. In this chapter,
different aspects of the self-cleaning function and superhydrophobic
1
2 Bio-Inspired Surfaces and Applications
surfaces are systematically illustrated, including advanced wetting

theory, methods of fabricating different superhydrophobic surfaces
and the potential applications of superhydrophobic surfaces, which can
help readers better understand recent progress in investigating the self-
cleaning effect and superhydrophobic surfaces.
1.1. Introduction
A proverb originating in ancient China states that “The lotus grows

in the mud, but is never contaminated by it,” which implies that lotus
leaf has a self-cleaning and superhydrophobic effect, and that small
droplets can roll off from lotus leaf easily,1−3 as shown in Fig. 1.1.
Self-cleaning technology has been progressively and quickly developing
since the late 20th century, and some important achievements have
led to practical applications in different fields. The applications of self-
cleaning technology are very broad and vary from window glass to
solar cell panels. A self-cleaning surface keeps itself clean by the action
of water, resulting in a reduction in maintenance cost and effort and the
use of detergents. A large number of studies have been conducted to
further develop self-cleaning technologies including anti-smudge, anti-
fouling, low adhesion, low drag, anisotropic wetting, anti-reflection,
Fig. 1.1. Lotus leaf.

Self-Cleaning and Superhydrophobic Surfaces 3
Surface height map and 2D profile of lotus leaf

Droplet on a Lotus leaf
60 µm 25 µm
Section A-A
0 µm −15
Height (µm)
A A
0
50 µm 60 µm
−15
0 0 25 50 µm
0 50 µm
Fresh louts
Surface height maps of micro bump on lotus leaf

SEM images of lotus leaf
50 µm 7 µm 50 µm 7 µm
0 µm 0 µm
20 µm
0 0
0 Top scan 50 µm 0 50 µm
Bottom scan
50 µm scan
2 µm 1 µm
0 µm
2 µm
5 µm
2 µm
0 0 0
0 2 µm
2 µm scan
Fig. 1.2. Superhydrophobicity and morphology of lotus leaf.
directional adhesion, photocatalysis, self-sterilizing, and anti-fogging

surface technologies as an environmentally friendly green technology.
The superhydrophobicity and morphology of the lotus leaf is illustrated
in Fig. 1.2 — it can be seen that the contact angle can surpass 150
degrees, and there are nano/micro structures on the lotus surface.4−8
The diversity of plant surface structures, evolved over 460 mil-
lion years, has led to a large variety of highly adapted functional
structures.9,10 The plant cuticle provides structural and chemical
modifications for surface wetting, ranging from superhydrophilic to
superhydrophobic. Superhydrophobic surfaces are widely found in
nature,11−13 such as on plant leaves, and insect legs and wings. As shown
in Fig. 1.3, the contact angle can be greater than 150 degrees. Water on
these surfaces forms small beads with a high contact angle in excess of
150◦ and drops off rapidly when the surfaces are slightly inclined, while
Fig. 1.3. Different superhydrophobic surfaces of insects.
taking away powder-like contaminants. This interesting phenomenon

has stimulated extensive research to make artificial superhydrophobic
surfaces and use them for a variety of applications.14 Combined with
hydrophobic surface layers, a superhydrophobic state was observed.
Water droplets falling on the surface bead up and roll off the surface
instantaneously. Superhydrophobicity requires that both the surface
hydrophobic layer and rough surface structures are present.15−17 This
phenomenon is quite common in the biology area, as shown in Fig. 1.4.
Artificial superhydrophobic surfaces were prepared before the
phenomenon was widely acknowledged and realized. It was reported
that many water repellent films could be obtained by depositing
the materials of hexamethyldisiloxane on silicon chips in high-energy
plasma at low temperature.17−19 Drops of water had contact angles as
high as 180 degrees and rolled off the slightly inclined surfaces, which
suggested a near-zero roll off angle. Morra et al.20 had prepared a
superhydrophobic surface by O2 plasma etching on PTFE (Polytetraflu-
oroethylene) surfaces in 1989. After five minutes of treatment, water
drops rolled easily across the surface. Due to the absence of hysteresis,
there was no obstacle to their movement. The first super water- and
oil-repellent surface was developed on anodically oxidized aluminum
after hydrophobic treatment.21−23 The effect of roughness on surface
wetting is scale-dependent. Besides the surface hydrophobicity, the
Fig. 1.4. Water droplets on plant leaf.
surface roughness can be more important in achieving dewetting

surfaces even without the most hydrophobic materials. On flat surfaces
oil repellency is difficult to achieve. However, with appropriate surface
roughness with special geometry designs, a superoleophobic surface
may result. This signifies the importance of surface wetting/dewetting
control via flexible, fine-tuned surface structure design at micro and
nano scales.
1.2. Advances in Wetting Theory
It is usually believed that wettability is the tendency of a liquid to

spread on a solid substrate and is generally characterized by the contact
angle, which is defined geometrically as the angle formed by a liquid
at the three-phase boundary where a liquid, gas and solid intersect
each other,24 as shown in Fig. 1.5. If the contact angle is less than
90 degrees, the surface is called a hydrophilic surface, and it has good
adhesiveness and wettability. Meanwhile, if the contact angle is greater
than 90 degrees, the surface is called a hydrophobic surface, and it has
poor adhesiveness and wettability, as shown in Fig. 1.6. This is one
of the most important properties that can be employed to assess the
hydrophobic capability of a surface, and has a significant influence in
agriculture, industry and daily life.
Fig. 1.5. Contact angle.
Fig. 1.6. Hydrophilic and hydrophobic surfaces.
Furthermore, it has been confirmed by researchers all over the

world that the wetting behavior of a liquid on a solid surface depends
not only on the chemical composition but also on the morphological
structure. Thus attention in this research field is mainly paid to the two
following aspects. On the one hand, systematic detailed thermodynamic
analyses on the free energy and free energy barrier have been investi-
gated for the purpose of reducing the surface free energy. On the other
hand, researchers found that micro- and nanostructures play a key role
in the wetting of a surface, regardless of chemical composition, on the
basis of the established theoretical models and experimental results.
In this regard, a wealth of investigations with respect to geometry
models have been made in the past decades to make the mechanism
of superhydrophobic surfaces more clear and further guide the design
and fabrication of biomimetic superhydrophobic surfaces.25−27 These
studies resulted in a number of models including the Wenzel model,
the Cassie–Baxter model, the composite structure with double-scale
Fig. 1.7. Wenzel’s model and Cassie’s model.
cylindrical groove model, a model with paraboloids of revolution and

so forth,28−30 as shown in Fig. 1.7. Although the theoretical models
proposed have been used to elucidate the principles of superhydropho-
bic surfaces, they are suitable only for specific conditions and still
contain a lot of experiential parameters. Therefore, they don’t have
universal applicability for all superhydrophobic states, and these models
need to be improved before being used. It is desirable to establish the
typical theories and models with widespread applicability to explain
the relationships between structures and properties quantitatively. In
this section, we mainly concentrate on the recent advances in the
theory of superhydrophobic surfaces, and particularly focus on the
geometry models originating from Young’s, Wenzel’s and Cassie–
Baxter’s equations, so as to quantitatively obtain the contact angle as a
function of various surface morphologies.
The contact angle of liquid on a solid surface is an important
parameter for characterizing and defining the surface wettability and
predicting its wetting behavior in practical applications.31 Therefore,
the value of the water contact angle on a surface can be used to
determine its hydrophobicity, denoting a surface with a contact below
90 as hydrophilic. On the basis of this definition and the deep investi-
gation of surfaces, the wetting behavior of a surface can be subdivided
in to two classes: one in which the contact angle on the surface is
between 90 and 150 degrees can be denoted as a hydrophobic surface,
and the other in which the surface has a contact above 150 degrees,
Lower Surface Tension Substrate Lower Surface Tension Substrate Low Surface Tension Substrate
High Surface Tension Substrate
Non-wetting
Complete wetting Partial wetting Partial wetting
=0
Fig. 1.8. Different wetting surfaces.
is denoted superhydrophobic.32−34 In terms of a superhydrophobic

surface, the other important limiting condition is low contact hysteresis,
which is usually below five degrees, and different wetting surfaces are
presented in Fig. 1.8. The definitions stated above are based on an ideal
smooth surface without considering the effect of surface morphology
and other factors. Given the potential applications of wetting/non-
wetting surfaces, it is desirable to construct rational theoretical models
to explain the mechanism and establish the quantitative equations to
predict the wetting behavior of a surface.
The difference between advancing and receding contact angles is
termed hysteresis; this has been the subject of significant literature over
the past few decades and has reemerged as a popular topic because
of recent interest in superhydrophobic and self-cleaning surfaces.35−37
When a small water drop encounters a solid surface, for example, a
raindrop on the hood of a car, a droplet is formed that consists of
a sphere of water sectioned by the surface at a discrete, measurable
contact angle. The shape of the droplet is not reproducible, and on
most surfaces, the contact angle will vary by 20 degrees or more.
If a droplet on a surface is allowed to evaporate in a low humidity
environment or if water is carefully withdrawn from the droplet with a
syringe, the droplet decreases in volume and contact angle, maintaining
the same contact area with the surface until it begins to recede. It
recedes with a constant contact angle, θR , characteristic of the surface
chemistry and topography. If the surface is cooled to below the dew
point and water condenses on the droplet or if water is carefully
added to the droplet with a syringe, the droplet volume and contact
angle increase, and again, the same contact area is maintained until
the droplet begins to advance. It does so at a constant advancing
contact angle, θA , which is also characteristic of the surface chemistry
Fig. 1.9. Droplets on different kinds of surfaces.
and topography. A metastable droplet can be formed with any angle

between the advancing and the receding contact angles. This is one
reason it is important that both advancing and receding contact angles
be reported to characterize a surface; one static, metastable angle is less
meaningful and only designates an angle somewhere between θR and
θA . In order for a droplet to move on a tilted surface, the droplet must
both advance and recede; it must also distort from a section of a sphere
to a complex shape with different contact angles around the entire
perimeter of the droplet.38 The relationship between these angles, θR
and θA , is not simple, as shown in Fig. 1.9.
About 200 years ago, it was proposed for the first time that
an interface between two materials had specific energy, the so-called
interfacial energy, which is proportional to the interfacial area.39,40
This concept has laid the foundations for the field of wettability, which
has become an extremely hot topic in the last two decades, thanks to
biological and high-technology applications ranging from self-cleaning
surfaces, microelectronics and thin film coatings, to image formation
that involves the spreading of liquids on solid surfaces. It is well known
that the surface atoms or molecules of liquids or solids have higher
energy than similar atoms and molecules in the interior, which results
in surface tension or surface free energy so as to reach a stable state
with a relatively lower energy.
As a rule, the surface tension or free surface energy can be used to
evaluate the surface properties quantitatively. Its value is equal to the
surface work which is required to create one unit area of the surface
at constant volume and temperature. The units of g are J m−2 or N
m−1 , and it can be interpreted either as energy per unit surface area
or as tension force per unit length of a contact line at the surface.
When a solid is in contact with liquid, the molecular attraction will

reduce the energy of the system so as to reach an equilibrium value
which is less than the sum of two separated surfaces. The formation or
establishment of an interface during the wetting process is governed
by the thermodynamic principle of energy reduction. The elimination
of two surfaces to form an interface reduces the total energy of the
system.40,41 Hence, the thermodynamic condition for wetting to occur
can be written as:
γSA ≤ γSL ≤ γLA , (1.1)
in which, γSA and γLA are the surface energies of the solid and liquid
against air, and γSL is the interface energy between the solid and liquid.
Then the work of adhesion may be expressed by the Dupre equation:
WSL = γSA + γLA − γSL . (1.2)
In this equation, WSL is the work of adhesion per unit area
between two surfaces. If a droplet of liquid is placed on a solid surface,
the liquid and solid surfaces come together under equilibrium at a
characteristic angle called the static contact angle. The contact angle
can be determined by the condition of the total energy of system being
minimized, in which, the total energy is depicted as following:
Etotal = γLA (ALA + ASL ) − WSL ASL , (1.3)
in which, ALA and ASL are the contact areas of the liquid-air and
solid-air interfaces respectively. It is assumed that the droplet is small
enough so that gravitational potential energy can be neglected. At the
equilibrium dEtotal = 0, this yields
γLA (dALA + dASL ) − WSL dASL = 0. (1.4)
For a droplet of constant volume, it is easy to show from geometrical
considerations that
dALA
= cos θ0 . (1.5)
dASL
Therefore, according to the above-mentioned formulas, the so-
called Young’s equation for the contact angle can be obtained, as
following:
γSA − γSL WSL
cos θ0 = = − 1 or γSL = γSA − γLA cos θ0 . (1.6)
γLA γLA
It can be seen that Eqs. (1.5) and (1.6) provide the value of the
static contact angle for given surface tensions. In fact, for an ideal solid
surface, the influences of roughness, chemical heterogeneity, surface
reconstruction, swelling and dissolution are neglected. The complete
wetting state corresponds to θ0 = 0, and typically happens for liquids
with low surface tension γLA , and on solids with high surface energy
γSA . When γSA > γLA , 0◦ < θ0 < 90◦ , and at the same time, when
γSA < γLA , 90◦ < θ0 < 180◦ . According to the Young’s equation, it
is clear that the interfacial tension between the solid and liquid γLA is
lower than γSA only when θ0 < 90◦ ; this happens only in the case of
wetting.
However, γSL > γSA can happen only when θ0 > 90◦ ; the contact
area of the liquid–solid interface will be reduced. In this case the liquid
behaves in a non-wetting manner because γLA is always finite and
positive, then the non-wetting behavior minimizes the total surface/
interfacial energy of the liquid. Thus the contact angle depends on the
optimization of the contact area of the solid–liquid and the liquid–air
interfaces. More importantly, the wetting property of a liquid on a flat
solid substrate can also be understood from the thermodynamics of
surfaces by evaluating the work of adhesion, thus leading to Young’s
equation. Another simple derivation of Young’s equation can be
obtained by balancing the forces at the line of contact where all three
media meet. Thus, the equilibrium contact angle of the liquid drop
upon a smooth and flat substrate depends on the value of the difference
between γSA and γSL as expressed above. It is inferred that liquids
with high surface tension on surfaces with low surface energy tend to
form droplets with high contact angle. However, there are no absolute
smooth ideal surfaces, structured or rough surfaces exist commonly in
reality. So it’s necessary to know how the contact angle functions on a
rough surface.42−45
With deep investigations into the effect of wettability in industrial
and agricultural fields, Young’s equation begins to be invalid because its
application is limited to an ideal smooth solid surface. In the handling
of practical problems, perfect smoothness of a surface is only an
assumption due to the existence of roughness resulting from synthesis
and processing. In 1936, Wenzel46 first introduced the concept of
surface roughness into the theoretical study of wetting behavior.
LA
Liquid
LA
h
Liquid 0
SL SA SL 0
SA
(a) (b)
Fig. 1.10. Young’s equation using surface tension vectors for a liquid on a solid
surface.
He modified Young’s equation based on comprehensive considerations

of the influence of roughness on wettability. Since then, when one
considered the wetting behavior of a liquid droplet deposited on a
rough solid surface, the Wenzel mode could be used as the analytical
model to quantitatively evaluate the wetting/non-wetting. Recently,
more and more experimental results have indicated that the roughness
of a solid surface can greatly magnify the wetting properties of the solid
even without taking account of chemical composition.
As forces are more easily visualized than energy values, it is conven-
tional, in the analysis of wetting problems, to define force concepts as
numerically equal to the characteristic interfacial specific energy values,
and to deal with these forces as vector quantities constant in magnitude
and variable in direction. It is in this translation from energies to forces
that the importance of the physical conditions of the solid surfaces is
likely to be obscured. Then a distinction between the total or “actual
surface” and the superficial or “geometric surface” of an interface
should be recognized. The latter is the surface as measured in the plane
of the interface. Perfect smoothness is an acceptable assumption, where
the actual surface and geometric surface are identical, but at the surface
of any real solid the actual surface will be greater than the geometric
surface because of surface roughness. This surface ratio is termed the
“roughness factor” and designated as following47−49 :
Areaactual
roughness factor = r = . (1.7)
Areageometric
The roughness factor is a non-dimensional parameter which is

always greater than unity. Thus as r increases, the total surface/interface
energy increases gradually due to the increase in the actual area. This
has been incorporated directly into the Wenzel model and plays an
important role in wetting properties. But the Wenzel state prevails
when the surface is completely wetted. It is evident that a rough surface
provides an additional interfacial area for the spreading liquid, thus the
true contact angle would be different to the nominal contact angle.
The additional surface area provided by roughening the surface results
in the increase of surface energy. Wenzel46 studied the effect of surface
roughness on the equilibrium contact angle and proposed an equation
that gives a relation between the equilibrium contact angle and the
apparent angle formed on a rough surface:
cos θw = r cos θ0 , (1.8)
in which, θ0 is the equilibrium contact angle, θw is the apparent contact
angle on a rough surface and r is the average roughness ratio. Hence
r is the ratio of actual wetted surface area to geometric surface area
calculated from radius of the wetted base. Wenzel’s equation directly
suggests that if θ0 < π/2 then θw < θ0 , and hence in this case the
introduction of roughness will enhance the tendency of the liquid to
wet further. On the other hand, if θ0 > π/2 then θw > θ0 , the non-
wetting tendency will be enhanced as shown in Fig. 1.11. Thus in
Wenzel’s model, the roughness can amplify the wetting property of
the solid surface if the liquid is initially wetting on a smooth surface,
resulting in greater hydrophilicity. But if the liquid is non-wetting on
the smooth surface, then the introduction of the roughness can make
Fig. 1.11. Schematic drawing of Wenzel model.

the surface of the solid substrate even more non-wetting, leading to the
enhancement of hydrophobicity. Compared to Young’s equation, the
Wenzel equation extends the smooth contact surface into a rough one,
which induces a new realization of the mechanism of surface wetting
behavior.
Moreover, it introduces the concept of surface roughness, which
makes researchers realize the significance of surface texture to the
surface wettability. It has been demonstrated that surface roughness
plays an even more important role than chemical composition with low
surface energy with regard to superhydrophobic surfaces. However,
it should be noted that the Wenzel equation should be confined to
the thermodynamic stable equilibrium state. An energy barrier often
needs to be surmounted during the spreading of the liquid due to the
heterogeneous nature of rough surfaces. If the vibrational energy is
less than the energy barrier, the liquid will maintain a meta-stable state
instead of a stable state.
The Wenzel equation has revealed the relationship between the
equilibrium contact angle and the apparent contact angle on the rough
surface, but it is not valid if the solid surface is porous or in the case of
varying chemical composition. In 1944, Cassie50 further amended the
equations proposed previously by introducing the surface coefficient so
as to take the surface chemistry into account. He sufficiently considered
the effect of other factors on wettability, proposing that if the substrate
is rough and consists of randomly distributed “n” different types of
materials on the surface, it will produce a composite model where the
surface properties of each part of these materials is characterized by
their own surface energies/tensions, for example, γi,SL and γi,SA , with
the separated area fraction fi on the substrate surface satisfying the
equations below:
n n
f1 + f2 + · · · + fn = 1, γSA = fi (γi,SA ) and γSL = fi (γi,SL ).
i=1 i=1
(1.9)
Thus, a modified equation substituting for Young’s equation and
extending the Wenzel equation can be given as following:
n
fi (γi,SA − γi,SL )
n
cos θCB = = fi cos θi . (1.10)
i=1
γLA i=1
This is known as the Cassie–Baxter equation. From this equation it is

concluded that the wetting/non-wetting properties of solid surfaces
are the result of the synergetic effect of surface roughness and chemical
composition. For a simple situation in which a composite surface is
just composed of two different types of material, specific parameters
can be expressed with the fractional area f1 , θ1 , f2 and θ2 respectively,
the apparent contact angle of this heterogeneous interface can be
calculated from equation as following:
cos θCB = f1 cos θ1 + f2 cos θ2 . (1.11)
For a composite interface consisting of a solid–liquid fraction and
liquid–air fraction, air can be trapped between the asperities as in a
lotus leaf surface so that the liquid droplet rides on a surface with
a distribution of solid asperities and air-pockets, whose respective
fractions are fSL and fLA (fSL + fLA = 1). In Eq. (1.11), by means of
combining the roughness factor r with the Cassie–Baxter equation and
sequentially endowing f1 (the solid–liquid fraction), cos θ1 , f2 (the air-
pocket fraction) and cos θ2 with fSL , cos θ0 , fLA and −1 (due to non-
wetting), the CA is given by a weighted average of cosines of the contact
angles on the solid and air using the following equations:
cos θCB = rfSL cos θ0 + fSL −1 or cos θCB = r cos θ0 − fLA (r cos θ0 +1).
(1.12)
When fSL = 1, the CB equation turns into the Wenzel equation.
If a surface is covered by holes impregnated with water, it corresponds
to the water-on-water contact instead of the water-on-air contact. We
make a following transformation to the Cassie equation:
cos θCB = 1 + fSL (cos θ0 − 1). (1.13)
In the above mentioned equation, fSL represents the solid–liquid
fractions under the contact area. Equation (1.13) is used for the
homogeneous interface instead of Eq. (1.9), where the rough surface
is covered by holes filled with water. It is not identical to the Wenzel
regime as the roughness must be filled with liquid before the drop
reaches it. It is the so-called “impregnating” Cassie wetting regime.
According to the Cassia–Baxter equation, it is possible for us to tailor
the surface topography to let the solid fraction fSL approach zero by
introducing asperities on the surface and making θCB close to 180
Fig. 1.12. Heterogeneous Cassie–Baxter regime and impregnating Cassie wetting

regime.
degrees. Figure 1.12 presents the dependence of the contact angle

on the roughness and fractional liquid–air area for hydrophilic and
hydrophobic surfaces with a heterogeneous interface. The CB angle
θCB as a function of the roughness factor, and it can be observed from
the diagram that the value of fLA increases with an increasing value
of θCB . When θ is less than 90 degrees, i.e., when the smooth and
flat substrate is inherently hydrophilic as shown in Fig. 1.12a, θCB can
transit from hydrophilic to hydrophobic upon increasing the value of
r. By contrast, if θ is greater than 90 degrees, i.e., when the smooth
and flat substrate is inherently hydrophobic as shown in Fig. 1.12b, θCB
cannot convert from hydrophobic to hydrophilic with increasing the
value of r.
The Cassia–Baxter equation is based on the assumption that the
heterogeneous surface is composed of well-separated distinct patches
of different materials, so that the free surface energy can be averaged.
It is also argued that when the size of the chemical heterogeneities is
very small, the quantity that should be averaged is not the energy, but
the dipole moment of a molecule, and Eq. (1.11) should be replaced
as following:
(1 + cos θCB )2 = f1 (1 + cos θ1 )2 + f2 (1 + cos θ2 )2 . (1.14)
Experimental studies of polymers with different functional groups
showed a good agreement with the above-mentioned equations. As
the Wenzel equation has shown how a rough surface can influence
the apparent contact angle at the boundary between a liquid and the
surface, Cassie and Baxter extended the analysis of apparent contact

angles from rough surfaces to porous surfaces. Both the Wenzel and
Cassie–Baxter equations are good approximations of contact angles on
imperfect surfaces, but it should be recognized that they are valid
when the size ratio of the liquid drop to the wavelength of the
roughness or chemical heterogeneity is sufficiently large. Moreover,
both of the models are based on an idealized model, which simplified
the theoretical analysis. Recent experimental findings and theoretical
analyses have made it clear that the initial Wenzel and Cassie–Baxter
models do not explain the complexity of interactions during the wetting
of a rough surface, which can follow several different scenarios. As
a result, there are various types of wetting state for rough surfaces
corresponding to different kinds of surface energy, therefore wetting
cannot be characterized by a single parameter such as the contact angle.
Further improvements have been made by researchers subsequently
which will be addressed in detail in the next section.
The Wenzel and Cassie–Baxter models can be compared because
they have many similar features and they deal with the same type
of interactions; however, there are some fundamental differences as
discussed above. Frankly, there are still some challenges for these two
models to explain real conditions between droplets and solid surfaces.
1.3. Methods of Fabricating Superhydrophobic

Surfaces
Nature is an inexhaustible source of functional surfaces. For millennia,
plants and animals have developed surfaces with special wettability
ranging from very highly slippery to highly sticky surfaces. Surface
wettability, as regards to water, is measured by the contact angle
which a water droplet forms when it is put on a surface. Among
many intriguing phenomena happening in nature, self-cleaning has
excited the scientist’s curiosity. Self-cleaning surfaces exhibit special
non-wetting properties owing to a water contact angle greater than
150 degrees and an effortless rolling off of water droplets. Interest
in superhydrophobic surfaces has risen after the discovery of the self-
cleaning properties of lotus leaves depending on their hierarchical
roughness. I has been found that the micro/nanostructures combined

with the low surface energy bring superhydrophobic properties. On a
slippery surface, the water droplets do not penetrate into the asperities
of surface, and the interactions with the surfaces are lowered. On the
other hand, superhydrophobic surfaces showing highly sticky behavior
have also fascinated scientists. These surfaces exhibit high water contact
angle, which is greater than 150 degrees, and high roughness and
water droplets are less prone to roll off them. In this case, the water
droplet penetrates into cavities of the surface, the high roughness
increases the contact area, and as a consequence, the liquid–solid
interactions are increased. Mimicking nature, scientists have developed
many methods to fabricate superhydrophobic surfaces. The efforts were
focused on surface texturing, in order to create a suitable micro/nano-
roughness able to generate the superhydrophobic properties. The
literature reveals the great interest for both fundamental research and
practical applications. The main challenge is to fabricate robust and
durable superhydrophobic surfaces suitable for applications in many
fields ranging from barrier, anti-icing/fogging, to water/oil separation
and anti-bioadhesion materials.
The techniques used to prepare superhydrophobic surfaces can
be divided into top-down and bottom-up approaches. Top-down
approaches include lithography, template-based techniques and plasma
treatments of surface. Bottom-up approaches involve mainly self-
assembly and self-organization. Examples of bottom-up techniques
are chemical deposition, layer-by-layer (LbL) deposition and colloidal
assemblies,51−53 as shown in Fig. 1.13. There are also methods based
on the combination of both top-down and bottom-up techniques such
as casting of polymer solution, phase separation, and electro-spinning
method.
(1) Lithographic method
Lithographic methods have been mostly used to prepare superhy-
drophobic surfaces. This technique has the advantage of allowing the
control of the structure and the morphology of the surface. With this
method, surfaces patterned with circular pillars, squared pillars, star-
shaped pots, indented square pots, with different diameters, height,
and spacing have been prepared. The possibility to design and to
Lithography method
Template method
Electro spinning method
Sol-gel method
Layer-by-layer method
Method of fabricating
superhydrophobic surfaces
Etching method
Chemical vapor method
Electroless galvanic deposition method
Anodic oxidation method
Electrochemical deposition method
Fig. 1.13. Different methods of fabricating superhydrophobic surfaces.
compare features allows investigation of the relationship between

the structure and the wetting state of the surface. Bhushan et al.54
fabricated micro/nano and hierarchical structures by replication of
a micro-patterned silicon surface, using an epoxy resin. The nanos-
tructure was then created by self-assembling of alkanes resulting in
a hierarchical double scale structure. The morphologies on different
kinds of superhydrophobic surfaces are shown in Fig. 1.14; it can be
seen that there are nano/micro structures on them.
Superhydrophobic surfaces with dual-scale micro/nanofeatures
have been prepared by coupling lithographic methods with wet and
chemical etching. Shieh et al.55 fabricated a superhydrophobic Si surface
characterized by nanopillar arrays with nanograss features. The nano-
pillars were fabricated using e-beam lithography and dry etching,
followed by hydrogen plasma etching to form the nanograss on the
surface, and finally, the surface was hydrophobized with CHF3 plasma.
These surfaces showed a high superhydrophobic stable state, large
contact angle and low friction. Surfaces with ordered arrays of micro-
pillars showing nanofeatures on the top and on the sidewalls were
prepared by He and coworkers.56 The micro-pillars were prepared via
Fig. 1.14. Different kinds of superhydrophobic surfaces.
photolithographic method followed by a deep reactive ion etching,

which allows inducing an anisotropic etching. The nanoscale pattern
was formed by silver nanoparticles galvanic etching. This approach
allowed them to obtain a hierarchical structure bringing the super-
hydrophobicity to the surface after a hydrophobization treatment.
Colloidal lithography is a soft lithography technique growing
in interest for fabrication of two-dimensional and three-dimensional
ordered structures. This process uses soluble particles as a remov-
able template. Superhydrophobic surfaces were prepared based on a
polystyrene nanoparticles mixture. The mixture was deposited onto a
glass slide, and subsequently, the PS-NPs (polystyrene nanoparticles)
were removed. They observed that using 400 nm PS-NPs, the surface
showed a water contact angle as high as 170 degrees. In another
approach, silver was deposited onto a monolayer of PS spheres.
Subsequently, a PVA aqueous solution was cast on the surface and
dried. After peeling the glass slide, the PS spheres were removed by
dissolution, and the obtained film presented a double hierarchical
structure made of silver nano-particles on PVA micro-bowl array. After
modifying the surface with 1H, 1H, 2H, 2H- perfluorodecanethiol,
the surface showed a water contact angle of about 153 degrees and a
sliding angle of less than 3 degrees.
In order to produce superhydrophobic surfaces from hydrophilic
materials or superoleophobic surfaces from oleophilic materials, the
presence of reentrant structures on the surfaces was found to be the
Fig. 1.15. Double hierarchical structure on the superhydrophobic surface.
more important parameter. Superhydrophobic and superoleophobic

surfaces with a triple-scale structure were produced on PMMA using
PS microparticles. The PS microparticles were deposited onto PMMA
plates by spin coating. The PS microparticles deposition was followed
by an oxygen plasma deep ion etching treatment. Because the PMMA
is etched at a rate of 2 with respect to PS, pillars were formed
on the PMMA surface. Depending on the etching time, the height
and the cross-sectional diameter of pillars was controlled. Two-step
etching treatment led to a better reentrant profile of pillars. This
method created a double hierarchical structure, and the surface, after
hydrophobization, was superhydrophobic with low adhesion, as shown
in Fig. 1.15. In the same way, surfaces were obtained on silicon and
SiO2 substrates by colloidal lithography of PS particles assisted by a
photolithographic mask and plasma etching. The use of mesh mask
permitted a more uniform deposition of particles and avoided the
formation of multilayers of particles.
For optical applications, the colloidal lithography was found to
be a competitive technique to fabricate ordered three-dimensional
structures. Near-infrared anti-reflective and improved transmittance
surfaces were prepared by colloidal lithography of PS nanospheres onto
fused silica substrates. The two-dimensional array of PS particles was
etched by a reactive ion etching using a mixture of Ar and CF4 as process
gas. The resulting surfaces were characterized by the formation of silica

arrays whose shape, height, and diameter depend on etching time.
The surfaces showed high transmittance in the NIR region and anti-
reflectivity properties. Besides, the surfaces showed superhydropho-
bicity after modification with a fluorosilane. For optical applications,
optical interference lithography is another attractive technique to
obtain periodic arrays. Accardo et al.57 prepared superhydrophobic
PMMA surfaces with high optical and X-ray transparency by optical
lithography and deep reactive ion etching. The surface contained a
microscale pattern of pillars and a random nanofibrillar pattern. The
water contact angle of surfaces was as high as 170 degrees with a sliding
angle less than 5 degrees. Using this technique, different 3-D patterns
were obtained on glass substrates by varying the incident angle of a
laser beam. After development of the photoresist, the 3-D features
were etched by a reactive ion CF4 plasma treatment. The structure
and the wettability of surface was function of the incident angle of laser
beam and of etching time. Superhydrophobic high sticky surfaces were
obtained with an incident angle of 0.8 degrees and after 18 minutes
of RIE (reactive ion etching). On the other hand, untilted surfaces
reached lotus state after 9 minutes of RIE. Park et al.58 designed high
transmission antireflective silica surfaces. They prepared nano-textured
silica surfaces presenting conical feature by interference lithography.
The as-prepared surfaces were rendered superhydrophobic after
chemical vapor deposition of 1H, 1H, 2H, 2H-perfluorodecyltrichlo-
rosilane. To obtain better optical performances, the fused silica was
patterned on both sides. The excellent transmissivity, low reflection
over a wide range of incident angles, wavelengths, and polarization
states were associated with a robust superhydrophobicity and anti-
wetting properties. Lithographic methods allow scientists to fabricate
microfluidic channels with the control of wettability of microfluidic
network. Kang et al.59 developed a one-step method to modify super-
hydrophobic surfaces by soft lithographic technique, micromolding in
capillaries. The superhydrophobic surface was an anodic aluminum
oxide membrane post-modified with a fluorosilane. A PDMS mold
was used as mask to pattern the AAO (anodic aluminum oxide)
superhydrophobic surface. After immersion, the mask was peeled
off. The polydopamine modified surface remained superhydrophobic

exhibiting highly sticky behavior.
(2) Template method

Templates are an effective means to structure surfaces. A pattern or
shape is replicated using a templating method, wherein a material is
printed, pressed, or grown against the voids of a template. Often, the
template is removed, leaving the inverse of its pattern: This can be used
as template to achieve a replica of the original.
Due to its cylindrical nanopores closed-packed in a hexagonal
geometry, anodic aluminum oxide is used to produce vertically aligned
nanofibers with control over their spacing. Superhydrophobic surfaces
were prepared on a Ti/Al substrate via fabrication of platinum nanofiber
array. The platinum nanofiber array was obtained by electrodeposition
of platinum into the pores of an AAO template generated on the
substrate followed by removal of the template in alkaline solution. The
result is a network of nanofibers that form a highly rough surface, as
shown in Fig. 1.16. After modification with a fluoroalkylsilane, the
water contact angle achieved 158 degrees. Sheng et al.60 prepared
superhydrophobic nano-structured high density polyethylene by extru-
sion inside an AAO (anodic aluminum oxide) template. Through
tuning the diameter of AAO templates and extruded pressure, vari-
ous superhydrophobic HDPE (high density polyethylene) nanofiber
surfaces were prepared with water contact angle of about 150 degrees
and low sliding angle.
Fig. 1.16. Fabricating superhydrophobic surface with the special template.

The AAO can also be removed from the aluminum substrate and
used as a membrane. Cheng et al.61 prepared polystyrene nanopillars
with different tip geometry and studied the sticky behavior of the
surface. The nanopillars presented flat, concave tips, and nanotubes
form, respectively. They were prepared using an alumina membrane
as template. The PS film and the membrane were sandwiched, and
the different geometries of nanopillars were obtained by varying the
temperature and the contact time. The alumina membrane was finally
dissolved with a NaOH solution. All surfaces were superhydrophobic
with high adhesion to water, but the surface with concave tip had
the highest adhesion, whereas the nanopillars showed a relative low
adhesion. Cheng et al.61 supposed that the volume of air between water
and surface plays a crucial role in the adhesion properties.
Very recently, a method to synthesize superhydrophobic hollow
polyaniline bars from nanorod arrays of metal-monomer template was
developed. When aniline is added to 0.75 M Cd(Ac)2 solution, a
complex between aniline and Cd(Ac)2 precipitates as white bars. When
ammonium persulfate (APS) is added, the polymerization of aniline
occurs. The precipitate disappears and PANI (polyaniline) hollow bars
are formed. During the oxidative polymerization of aniline, it is released
from the complex, and the polymer growths along the surface of
precipitate resulting in hollow bars that replace the bars of precipitate.
When the hollow bars were cast on a glass slide, the water contact angle
measured 153 degrees.
(3) Electrospinning method
Electrospinning is a technique that allows the fabrication of continuous
fibers with diameters down to a few nanometers. This method can
be applied to synthetic and natural polymers, polymer alloys, and
polymers loaded with chromophores, nanoparticles, or active agents,
as well as to metals and ceramics. Whereas electrospinning was widely
used with polymers, the use of ceramics allows an enhancement of
the thermal stability of the surface. Electrospinning has been widely
used to produce superhydrophobic surfaces, and various strategies have
recently emerged. In order to improve the adherence of the electrospun
polymer with the substrate, a fluorinated diblock copolymer, made
of poly(heptadecafluorodecylacrylate-co-acrylic acid) (PFDA-co-AA)
random copolymer as the first block and polyacrylonitrile (PAN) as the

second one, was used. The fluorinated block promoted hydrophobicity
to the surface by reducing the surface tension, while its carboxylic acid
functions anchored the polymer film onto aluminum substrate after
annealing at 130 degrees Celsius. The PAN block of this copolymer
insured the stability of the structuration of the surface during annealing,
thanks to its infusible character. The films showed water contact
angle as high as 150 degrees, excellent corrosion resistance, and
good adhesion to the aluminum substrate. Other groups tried to
control the surface hydrophobicity by adding additives in polymers.
For example, polystyrene in presence of a series of fluoroalkyl end-
functionalized polystyrene additives were used. Additives migrated
during the electrospinning process, leading to an increase in hydropho-
bicity of the surface. The effect of the number of fluoroalkyl end-groups
and the molecular weight of fluoroalkyl end-functionalized additives
were studied. Results showed that low molecular weight additives or
increasing the number of fluoroalkyl end-group resulted in an increase
of hydrophobicity. However, for all additives, a maximum water contact
angle was observed at the concentration of about 4%. At concentrations
higher than 4%, water contact angle decreased. The additives molecules
probably form aggregates when concentration increases that are not
able to migrate to the surface. Because one of the key parameters of
superhydrophobic properties is the surface morphology, various strate-
gies were employed in order to control this parameter. Superhydropho-
bic surfaces were prepared using a statistical terpolymer containing
a commercial perfluoroethyl alkyl methacrylate, methyl methacrylate,
and butyl acrylate. The morphology of coating changed from uniform
fibers to beads-containing fibers as functions of terpolymer molecular
weight, its concentration, and electrospinning conditions. The largest
water contact angle was obtained for the morphology where bead-
containing fibers are formed, as shown in Fig. 1.17. Superhydrophobic
coatings were obtained by electrospinning of a dual layer of polyvinyli-
dene fluoride and fluorinated silane molecules onto a glass slide. It is
observed that the size of formed beads increased as the concentration
of PVDF (polyvinylidene fluoride) increased.
Therefore, in the first step, the glass surface was coated with a solu-
tion containing 5% of PVDF (polyvinylidene fluoride). In the second
Fig. 1.17. SEM images of superhydrophobic surface with electrospinning method.
step, a mixture containing 10% FSM (fluorinated silane molecules)

and 1% PVDF was deposited at lower flow rate than the first step.
The as-prepared coating showed a spike-like morphology with dual
hierarchical structure. For this surface, the water contact angle achieved
170 degrees and the roll off angle was less than 1 degree. Another group
used solutions containing PVDF and epoxy-siloxane modified SiO2
nanoparticles. During the electrospinning process, the epoxy-siloxane
modified silica nanoparticles were assembled on the micro-sized PVDF
beads, creating a dual-scale structure. The hydrophobicity of materials
varied as function of mass ratio of modified silica particles and PVDF.
Superhydrophobic polymethylsilsesquioxane (PMSQ) surfaces were
fabricated by electrospinning at the sol-gel transition. The PMSQ was
prepared from methyltrimethoxysilane and to make PMSQ undergo
hydrolysis and gelation zinc acetate was added.
The PMSQ in DMF (dimethylformamide) solution with different
gelation time was electrospun on aluminum foils. The surfaces were
dried and then cured at 150 degrees for two hours. The surface
morphology depended on the solution viscosity. The surface with a
water contact angle of 151 degrees and a sliding angle of 8 degrees
showed nanobeads-fibers associated with microbeads. After a thermal
treatment at 300 degrees Celsius, the as-prepared surface also showed
high stability in several organic solvents. Finally, for specific drug
delivery applications, superhydrophobic meshes were produced by
electrospinning of poly(e-caprolactone) in presence of poly(glycerol
monostearate-co-e-caprolactone) as hydrophobic polymer dopant. The
contact angle determined the rate of water penetration. For a dopant
concentration of 10%, the apparent water contact angle was 153
degrees, and the surface was in the metastable Cassie–Baxter state.

When the meshes were loaded with an anticancer agent, the trapped air
was slowly displaced by water, slowing the release of anticancer agent.
(4) Sol-gel method
Sol-gel methods have been used to fabricate superhydrophobic surfaces

by several groups and on various substrates including textiles and
wood. In a typical sol-gel process, the precursor is converted into a
glassy material through a series of hydrolysis and polycondensation
reactions. By varying the system conditions and reaction mixtures,
the surface roughness can be controlled. Superhydrophobic sol-gel
coatings usually present good resistance to temperature. Self-cleaning
surfaces were fabricated from an alumina sol using aluminum tri-sec-
butoxide (Al(O-sec-Bu) 3) as raw material. The sol was doped with
alumina nanoparticles with sizes of 80 nm. Then, it was spin-coated
onto a glass slide and treated at 400 degrees Celsius. The presence
of alumina nanoparticles led to a rough surface characterized by the
formation of papillae of about 1µm. To create a nanoscale roughness,
the surface was immersed in boiling water for 5 minutes. The resulting
surface showed a flower-like structure with micro/nano dual features.
After hydrophobization, non-sticky surface was achieved. Mahadik
et al.62 prepared superhydrophobic coatings by sol-gel process, using
methyltrimethoxysilane as source material. The sol was dip-coated
onto glass plates, sintered at 150 degrees Celsius, and rendered
superhydrophobic by modification with trimethylchlorosilane.
The surface showed a water contact angle of about 170 degrees.
The constant angle was stable up to 550 degrees Celsius and rapidly
decreased after 600 degrees Celsius. After a heat treatment of two
hours, the surface became superhydrophilic, but the superhydrophobic
behavior was recovered after another treatment with trimethylchlorosi-
lane. To produce superhydrophobic surfaces in one step, Lakshmi
et al.63 incorporated a perfluoroalkylmethacrylic polymer in a hybrid
sol-gel matrix containing fumed silica particles. The presence of
fluoropolymer resulted in more accentuated oleophobic properties of
coating that exhibited contact angles of 146 degrees with ethylene
glycol and 113 degrees with lubricant oil. Kim et al.64,65 fabricated
superhydrophobic silica coatings by electro spraying of sol-gel solu-

tion. In a recent paper, they prepared high rough silica coating by
electrospray deposition of a solution containing tetraethoxysilane and
methyltriethoxysilane as precursors. The electrosprayed SiO2 coating
was modified with 1H, 1H, 2H, 2H-perfluorooctylsilane to lower the
surface energy. The as-prepared surface showed superhydrophobic, self-
cleaning properties, and good repellency toward different liquids, such
as milk, coffee and juice.
To improve the adherence between the particles and the substrate,
the particles can be functionalized in order to react with the substrate by
forming covalent bonds. For example, superhydrophobic copper wafers
were modified by chemical etching in order to increase the roughness
of the surface and to create hydroxyl groups on the copper surface
(Cu-OH). Then, using vinyltrimethoxysilane as precursor, the coating
was grown on copper. During the sol-gel process, the hydroxyl groups
present on the copper surface are able to react with silanol groups
resulting in a covalent linkage between the coating and the copper.
The as-prepared surface exhibited water contact angle as high as 155
degrees showing good corrosion resistance.
Superhydrophobic anti-reflective or transparent surfaces were pre-
pared by sol-gel technique. Kavale et al.66 prepared superhydrophobic
optically transparent silica based coatings using methyltrimethoxysilane
and adding PMMA to enhance superhydrophobicity. The silica coating
prepared with 7% PMMA showed a water contact angle as high as
170 degrees with nearly 91% in the visible region. Superhydrophobic
and highly transparent coatings were also prepared from hollow
silica nanoparticles (HSN). The coatings were obtained via dip-
coating of 3-aminopropyltriethoxysilane (APTS)-modified HSNs on
glass substrate followed by thermal annealing and chemical vapor
deposition of 1H,1H,2H,2H-perfluorooctyltrimethoxysilane. Tuning
the APTS concentration and the withdrawing speed of dip-coating, a
superhydrophobic coating (water contact angle 150 degrees, sliding
angle 4 degrees with good transparency (about 90%)) was obtained.
(5) Layer-by-Layer method
Layer-by-Layer (LbL) technique has been widely used to fabricate
superhydrophobic surfaces. LbL is a method by which thin films,
particularly of oppositely charged layers, are deposited (polyelec-

trolytes, for example). In general, LbL assembly is described as sequen-
tial adsorption of positive or negative charged species by alternatively
dipping into the solutions. The excess or remaining solution after
each adsorption step is rinsed with solvent leading to a thin layer of
charged species on the surface ready for the next adsorption step.
LbL can also be utilized for nanoparticle assembly. Hierarchically-
organized composite surfaces were obtained by LbL assembly having
silica nanoparticles with different size and functionality. The wettability
of the surface was lowered by hydrophobization of the last layer. It is
pointed out that the hydrophobicity of the surface depends on the
number of layers of silica nanoparticles and surface topology.67−70
Sun et al.71 assembled silica nanoparticles on a honeycomb-patterned
PDMS substrate. Both high- and low-adhesion superhydrophobic
surfaces were obtained with different bilayers of assembly (with thermal
cross-linking and surface fluorination by chemical vapor deposition),
indicating that different levels of nano- to micro-structural hierarchy
can be realized using this method. Kim et al.72 used the spontaneous
wrinkling phenomenon of polyelectrolyte multilayer to create hierarchi-
cal structured surfaces. They realized a composite material composed
by a LbL-assembled polyelectrolyte multilayer (PEM) containing Ag
nanoparticles. The accumulation of Ag nanoparticles in the PEM matrix
causes a compressive stress that generates the spontaneous wrinkling of
the surface. The PEM layer was composed with linear polyethylenimine
(LPEI) and poly(acrylic acid) (PAA) that allows cation exchange within
the film when it is immersed in a solution of silver acetate. The
Ag+ ions are incorporated into the PEM matrix and successively
reduced to obtain Ag nanoparticles. A plasma-ashing process was
then applied to remove the polymeric phase. The surface formed
wrinkles with micrometer size and the Ag structures exhibited plate-
like or flake-like features with size of nanometers. After modification
with a fluorinated compound, the surface exhibited superhydrophobic
properties (water contact angle of 170 degrees, sliding angle <1
degree). Other strategies include the use of specific functions in order
to produce sensitive surfaces. A LbL method was developed to make
superhydrophobic water-sensitive substrates. This method includes the
covalent multilayer assembly of polyethyleneimine (PEI) and an amine

reactive polymer containing azolactone functionality (PVMDA) in
polar-aprotic solvents. The coating was finally modified by reaction
with a hydrophobic amine resulting in superhydrophobic coating that
resisted water penetration. To produce thin and compact films, Zhang
et al.73 investigated the co-deposition via LbL of polyelectrolytes
with UV-sensitive groups. They used positively charged diazoresin-
poly(acrylic acid) complexes (DAR-PAA) and free diazoresin (DAR).
The LbL deposition of PECs (polyelectrolyte complexes) led to a
formation of a hierarchical structure. After UV cross-linking via the
diazonium groups and vapor deposition of a layer of fluoroalkylsilane,
the coating was converted in superhydrophobic. Depending of the
number of deposition cycles, both sticky and anti-wetting coatings
were obtained with sliding angle as low as 1 degree. Li et al.74
fabricated self-healing superhydrophobic coatings. The strategy was to
create a porous flexible coating with micro- and nanofeatures. This
coating was prepared by LbL assembly of polyelectrolyte complexes of
poly(allylaminehydrochloride) and sulfonated poly(ether ether ketone)
(SPEEK) with poly(-acrylic acid) (PAA). The multilayered (PAH-
SPEEK/PAA)n[check n??] exhibited high porosity and roughness with
hierarchical micro-nanostructures. This coating, after thermal cross-
linking process, was made superhydrophobic by chemical vapor depo-
sition of 1H,1H,2H,2H perfluorooctylsilane (POTS). The resulting
water contact angle was 157 degrees, and the tilt angle was as low as 1
degree. In order to test the self-healing ability of the surface, the coating
was exposed to O2 plasma to remove the POTS from the surface. The
surface became superhydrophilic. After being exposed under a humid
environment (relative humidity of 40%), the superhydrophobicity of
coating was completely recovered. The self-healing property of coating
is due to the possibility of fluoroalkylsilane to migrate to the surface.
(6) Etching method
Etching treatments can increase the roughness of surface creating a
certain micropattern. Superhydrophobicity is reached either straight
away or by post-treatment with a hydrophobizing agent, depending
on the nature of the substrate. The removal of the outmost layers of
materials can be realized in different way.
One popular method to generate roughness on a polymeric

substrate is plasma ablation. Plasma etching was used to increase
roughness of polymers. Several papers reported on the increase in water
contact angle of PTFE (polytetrafluoroethylene) when the surface
was exposed to a RF (radio frequency) oxygen plasma or electron
irradiation. Vandecasteele et al.75 pointed out that treatment of PTFE
with low pressure O2 plasma induced an etching of the surface and
emphasized the importance of the high energy electrons and oxygen
in the surface modification. In a similar manner, the wettability of
polyimide and fluorinated ethylene propylene (FEP) surfaces was
modified. By increasing the treatment time, the PI (polyimide) became
more hydrophilic, whereas the hydrophobicity of FEP was enhanced.
A water contact angle higher than 150 degrees was measured after
60 minutes of treatment, and the water droplet rolled off on the
surface after 75 minutes of treatment. Superhydrophobic PET was
obtained by O2 plasma treatment. Wohlfart et al.76 observed that
surface morphology significantly changed as function of plasma param-
eters. They pointed out that the water contact angle decreased after
treatment, independently of the treatment time. Coating the sample
with a perfluoroalkylsilane resulted in an increase of water contact
angle that reached 160 degrees after 20 minutes of plasma treatment.
Water droplets also showed low adhesion rolling off the perfluorinated
surface after 20 minutes of treatment. The plasma treatment induced
the formation of fibrils on the surface. As the treatment time increased,
the fibrils collapsed, forming pyramid-like bundles containing several
fibrils inducing the transition from Wenzel to Cassie– Baxter behavior.
Superhydrophobic surfaces have also been obtained using plasma
treatment in presence of some fluorinated gases such as CF4 , C2 F6 , and
SF6 . Tsougeni et al.77 fabricated superhydrophobic PMMA and PEEK
by means of O2 plasma etching followed by C4 F8 plasma deposition.
The water contact angle was 153 degrees and contact angle hysteresis
lower than 5 degrees. The same approach was followed by Gnanappa
et al.78 who fabricated robust superamphiphobic PMMA surfaces. They
showed that the formed nanofibers coalesced in smaller and more
compacted nanostructures after immersion in water.
PDMS (polydimethylsiloxane) substrates with micro-protrusion
structures were obtained by microwave plasma etching and F4 /O2
gas flux. The micro-protrusion structures were created by etching of

PDMS with a Ni mask, and the water contact angle of the surface varied
from 104 to 148 degrees. Another widespread approach to create a
rough surface on a metal is chemical etching, which represents a quite
inexpensive and simple way to obtain superhydrophobic surfaces.
1.4. Potential Applications of Superhydrophobic

Surfaces
Superhydrophobicity is gradually becoming the hot term that has
been used to describe extreme water repellency of rough hydrophobic
surfaces. However, such techniques have been practiced for a long time.
Early products include breathable fabrics and certain membrane filters,
which use the same structure effects as superhydrophobic surfaces, but
these are rarely referred to as being superhydrophobic.79−81
The number and scope of techniques for generating superhy-
drophobic surfaces has greatly increased in recent years with possible
applications expanding with them. The main barrier that prevents
implementation of self-cleaning surfaces is that the surfaces are easily
damaged or contaminated during normal use, such as abrasive wear
and washing cycles. Despite this limitation, over 200 patents have been
granted that use or produce superhydrophobicity in some form and
products are becoming commercially available. Specific applications
that do not involve abrasion or contact with oils that may cause
contamination, such as self-cleaning building walls and windows, have
made use of these super-water-repellent surfaces. Commercially, one of
the greatest limitations to the use of such surfaces is the stability over
time under outdoor conditions. Investigations to improve resistance
to scratching are underway which could lead to useful products. As
research continues, methods of production and materials used to form
superhydrophobic surfaces will broaden and offer new applications.
Potential uses for self-cleaning surfaces are wide-ranging. Indeed,
we are surrounded by many surfaces that we want to keep clean, from
window panels to roof tiles to bathroom surfaces and house walls. For
example, cleaning of windows is expensive and cumbersome, especially
if the windows are on skyscrapers. Water-related corrosion processes can
also be prevented by such superhydrophobic surfaces. The hydrophilic

route to self-cleaning windows has been realized by several companies
and has just been released to market. To what extent these windows will
be a commercial success clearly remains to be seen; to some, a water film
running off the glass may appear to be a nuisance. The main advantage
of these surfaces is the combined hydrophilicity-photodegradation
effect, which significantly aids in the cleaning process by decomposing
organic moieties on the surface when illuminated under UV radiation.
Furthermore, the fact that a controlled roughness is not needed
in this approach is a clear technological advantage. Although the
superhydrophilic effect is reversible in principle, the ageing of these
surfaces under actual conditions is not known. Others argue that
coatings for windows and mirrors should be superhydrophilic, causing
water to form a transparent film which does not block the vision. In
contrast, superhydrophobic coatings on transparent windows may be
problematic due to the fact that raindrops rolling off the surface can
be distracting.
The alternative superhydrophobic approach has not yet been
realized in practice, but work in this direction is now underway. An
ambitious attempt has been made by combining the basic elements of
superhydrophobicity with biomimetic surface design. A system under
industrial development contains a reservoir with a hydrophobic poly-
mer that is intended to mimic the wax of the lotus leaf: The outcome
would be a self-cleaning surface that heals itself. The concept relies
on a replenishment layer embedded in a glass that serves as a surface-
repairing reservoir, refueling a hydrophobic cover layer when it has
been depleted at the surface, and hence restoring superhydrophobicity
to the roughened surface.
Another important application for superhydrophobic surfaces may
be in the field of oil/water separation. It is demonstrated that a mesh
coated with PTFE showed two extremes in wettability for water and oil.
For water, it is an extremely nonwetting surface (superhydrophobic),
while for oil, it changed to complete wetting (superoleophilic).82−84
The extreme difference in wettability for water and oil showed the
possibility to separate oil and water. When a mixture of oil and
water is put on the mesh film, the water remains on the upper part
of the film, while the oil penetrates through the mesh and can be
collected underneath. Corrosion-induced failure plays an important
role in microelectronic devices and electronic packaging failures.
In order for microelectronic components and devices to perform their
functions properly, highly reliable packaging is necessary. Failures of
microelectronic devices and packages not only cause malfunction of
the devices but also sometimes lead to catastrophic consequences for
entire systems. Among all microelectronic device failures, corrosion-
related failure is responsible for more than 20%. Corrosion in micro-
electronic packaging depends on the package type, electronic materials,
fabrication and assembly processes, and environmental conditions such
as moisture condensation, ionic or organic contaminants, temperature,
residual and thermal stress and electrical bias. With the ever-shrinking
feature sizes of microelectronic components and devices, susceptibility
to corrosion-induced failures increases. Better performance and relia-
bility requirements drive improved corrosion-resistance of packaging
systems.85,86
There are basically three types of packages for microelectronic
components and devices: ceramic, metal and plastic. Ceramic and metal
packages are hermetic packages mainly used in military, aerospace
and automobile applications where high reliability is required. Plastic
packaging is non-hermetic, but is widely used because of its low cost
and easy manufacturability. Compared to other package types, plastic
packaging systems have more corrosion-related problems because the
polymeric materials used in plastic package systems are more permeable
to moisture which can then reach die, wires, bond pads, lead frames
and solder joints. Therefore, corrosion problems are long-term issues
in microelectronic packages.87−89
Corrosion involves electrochemical processes unless oxidation can
take place at elevated temperature and dry environments. The basic
requirements for electrochemical corrosion include an electrically
conductive anode, cathode, interconnecting electrolyte (humidity envi-
ronment) and driving force. There are three major corrosion types:
galvanic corrosion, pitting corrosion and stress corrosion cracking.
Under most circumstances, water/moisture is a key in the corrosion
process. Corrosion prevention is a matter of preventing moisture
condensation on sensitive areas such as die interfaces, wires, bond pads,

and solder joints. In order to prevent corrosion for improved device
reliability, superhydrophobic coatings on microelectronic devices offer
great potential for water-repellent and moisture-resistant applications.
In microelectromechanical systems (MEMS), adhesion failure due
to bonding and capillary force between moving parts and substrates is
the major failure mechanism. Superhydrophobic surfaces are promising
to achieve antistiction in these applications due to the low surface
energy. Reduction of friction using superhydrophobic surfaces to
achieve slip flow in microfluidic channels is also of considerable interest.
Control or transfer of fluid droplets containing biologically relevant
molecules (DNA and proteins) while minimizing contamination is of
much significance in biotechnology and biomedical fields. Specifically,
the construction of complementary DNA (cDNA)-microarrays pre-
pared by spotting techniques requires specific wetting properties of
the substrates. Glass slides — the substrates most commonly used —
are usually only mildly hydrophobized, such that the drops of drying
cDNA-solution produce unwanted ring-like structures, a nuisance
known in the field as the “doughnut-effect,” as shown in Fig. 1.18.
This effect is related to the well-studied “coffee-stain effect”: an
evaporation-driven convection mechanism that drives dissolved or dis-
persed particles inside the drop to the surface-pinned contact line where
evaporation occurs rapidly. Superhydrophobicity is advantageous in this
Fig. 1.18. Potential applications of superhydrophobic surfaces.

application since the nearly spherical drops on a superhydrophobic

surface can shrink exactly like a drop in air. The positioning and
shape of spotted drops can be steered by combining hydrophilic and
hydrophobic patterns. The utility of this idea can be demonstrated
by consideration of an actual biosystem: Some desert beetles capture
their drinking water with a hydrophobically/hydrophilically structured
back. By prepatterning a substrate, hydrophilic prespotting (anchoring)
on an otherwise hydrophobic or superhydrophobic surface results in
new possibilities for improvements in spotting and analyzing DNA and
proteins by avoiding wall contact.90
Beyond the possibility of improving fluid shapes on substrates, the
possibility of guided motion of droplets on superhydrophobic surfaces
offers the opportunity to develop a droplet-based microfluidics system,
in contrast to the classical concept based on microfluidic channels.
Driving the liquid along the channels and causing merger at predefined
locations offers a novel way to mix reactants or direct biochemical
reactions; this approach defines the concept of a “liquid microchip”
or “surface-tension confined Microfluidics.” One advantage of the
open structures over capillaries, in addition to their ease of cleaning,
is that blocking of the capillary by unforeseen chemical reactions is
inhibited.91,92
To conclude this introduction into the vast number of approaches
and applications for superhydrophobicity, an indication of the greatest
challenge facing all self-cleaning or contaminant free surface applica-
tions must be offered: ageing and degradation of the artificial surfaces.
For biotechnology applications, this is not so relevant; the surfaces will
often be used for analytic purposes and hence designed as disposables.
If one thinks of contaminant-free surfaces for use in medicine, the
issues of reusability and hygiene are complex and at present unresolved.
However, a different issue is the use of self-cleaning surfaces in
outdoor applications. Examples are known already in which an initially
improved product lost its advertised self-cleaning property too rapidly,
and thereby did not justify the initial investment (Lotusan, private
communication). Clearly, failed products bear the risk of discrediting
an entire field of applications. Ageing will remain difficult to foresee —
pure empiricism reigns in its description — but the benefits to investors
and product users require that ageing be under control. In addition,

stability under high humidity conditions also needs improvement by
designing small features with cost-effective approaches. The dew point
may be reduced on superhydrophobic surfaces with small features.
Oil or surfactant repellent or superoleophobic surfaces may also be
desirable which add more constraints to superhydrophobicity design.
As coatings improve and prices fall, use of superhydrophobic surfaces
is likely to increase, particularly as understanding of many of the
limitations is developed and perhaps mitigated by careful selection of
coatings.93,94
1.5. Conclusions
In this chapter, different aspects of self-cleaning and superhydrophobic

surfaces are systematically illustrated, and the most recent advances
in wetting theory are introduced, which can help readers better
understand the progress made in this area. Different methods of fab-
ricating superhydrophobic surfaces are elaborated in detail, including
lithography, template, electronspinning, sol-gel, layer-by-layer, etching,
chemical vapor deposition and so on. In addition, the potential appli-
cations of superhydrophobic surfaces in different fields are expounded,
such as drag reduction, anti-icing, oil/water separation and so on.
References
[1] Koch K, Bhushan B, Jung YC and Barthlott W. (2009) Fabrication of Artificial
Lotus Leaves and Significance of Hierarchical Structure for Superhydrophobicity
and Low Adhesion. Soft Matter 5: 1386–1393.
[2] Jin MH, Feng XL, Feng L, Sun TL, Zhai J, Li TJ and Jiang L. (2005) Superhy-
drophobic Aligned Polyacrylonitrile Nanotubes Film with High Adhesive Force.
Adv. Mater. 17: 1977–1981.
[3] Feng L, Zhang Y, Xi J, Zhu Y, Wang N, Xia F, and Jiang L. (2008) Petal Effect: A
Superhydrophobic State with High Adhesive Force. Langmuir 24: 4114–4119.
[4] Jung YC and Bhushan B. (2006) Contact Angle, Adhesion, and Friction . . .
Polymers for Superhydrophobicity. Nanotechnology 17: 4970–4980.
[5] Shin YM, Hohman MM, Brenner MP and Rutledge GC. (2001) Electrospin-
ning: A Whipping Fluid Jet Generates Submicron Polymer Fibers. Appl.Phys.
Lett. 78: 1149–1151.
[6] Gentsch R, Boysen B, Lankenau A and Borner HG. (2010) Single-step

Electrospinning of Bimodal Fiber Meshes for Ease of Cellular Infiltration.
Macromol. Rapid Commun. 31: 59–64.
[7] Wang LF, Pai CL, Boyce MC and Rutledge GC. (2009) Wrinkled Surface
Topologies of Electrospun Polymer Fibers. Appl. Phys. Lett. 94: 151916.
[8] Kanehata M, Ding B and Shiratori S. (2007) Nanoporous Ultra-high Specific
Surface Inorganic Fibres. Nanotechnology 18: 315602.
[9] Nosonovsky M and Bhushan B. (2007) Biomimetic Superhydrophobic Surfaces:
Multiscale Approach. Nano Lett. 7: 2633–2637.
[10] Gao LC and McCarthy TJ. (2007) How Wenzel and Cassie Were Wrong.
Langmuir 23: 3762–3765.
[11] Mognetti BM and Yeomans JM. (2010) Modeling Receding Contact Lines on
Superhydrophobic Surfaces. Langmuir 26: 18162–18168.
[12] Extrand CW. (2003) Contact Angles and Hysteresis on Surfaces with Chemically
Heterogeneous Islands. Langmuir 19: 3793–3796.
[13] Drelich J. (1997) Static Contact Angles for Liquids at Heterogeneous Rigid
Solid Surfaces. Pol. J. Chem. 71: 525–549.
[14] Moulinet S, Guthmann C and Rolley E. (2002) Roughness and Dynamics of a
Contact Line of a Viscous Fluid on a Disordered Substrate. Eur. Phys. J. E 8:
437–443.
[15] Nosonovsky M. (2007) On the Range of Applicability of the Wenzel and Cassie
Equations. Langmuir 23: 9919–9920.
[16] Zhao Y, Lu QH, Li M and Li X. (2007) Anisotropic Wetting Characteristics on
Submicrometer-scale Periodic Grooved Surface. Langmuir 23: 6212–6217.
[17] Gao LC and McCarthy TJ. (2007) Reply to “comment on How Wenzel and
Cassie were Wrong by Gao and McCarthy. Langmuir 23: 13243.
[18] Sacher E, Sapieha JK, Schrieber HP, Wertheimer NR, et al. (1986) Surface and
Interface. Taylor & Franics: London.
[19] McHale G. (2007) Cassie and Wenzel: Were They Really So Wrong? Langmuir
23: 8200–8205.
[20] Morra M, Occhiello E, Garbassi F. (1989) Contact Angle Hysteresis in Oxygen
Plasma Treated Poly (tetrafluoroethylene). Langmuir 5: 872–876.
[21] Marmur A and Bittoun E. (2009) When Wenzel and Cassie are Right:
Reconciling Local and Global Considerations. Langmuir 25: 1277–1281.
[22] Erbil HY and Cansoy CE. (2009) Range of Applicability of the Wenzel and
Cassie-Baxter Equations for Superhydrophobic Surfaces. Langmuir 25: 14135–
14145.
[23] Leroy F and Muëller-Plathe F. (2011) Rationalization of the Behavior of Solid-
Liquid Surface Free Energy of Water in Cassie and Wenzel Wetting States on
Rugged Solid Surfaces at the Nanometer Scale. Langmuir 27: 637–645.
[24] Verplanck N, Galopin E, Camart J-C and Thomy V. (2007) Reversible Elec-
trowetting on Superhydrophobic Silicon Nanowires. Nano Lett. 7: 813–817.
[25] Krupenkin TN, Taylor JA, Wang EN, Kolodner P, Hodes M and Salamon
TR. (2007) Reversible Wetting-dewetting Transitions on Electrically Tunable
Superhydrophobic Nanostructured Surfaces. Langmuir 23: 9128–9133.
[26] Liu G, Fu L, Rode AV and Craig VSJ. (2011) Water Droplet Motion Control
on Superhydrophobic Surfaces: Exploiting the Wenzel-to-Cassie Transition.
Langmuir 27: 2595–2600.
[27] Boreyko JB and Chen CH. (2009) Restoring Superhydrophobicity of Lotus
Leaves with Vibration-induced Dewetting. Phys. Rev. Lett. 103: 174502.
[28] Nosonovsky M. (2007) Multiscale Roughness and Stability of Superhydrophobic
Biomimetic Interfaces. Langmuir 23: 3157–3161.
[29] Morita M, Koga T, Otsuka H and Takahara A. (2005) Macroscopic-wetting
Anisotropy on the Line-patterned Surface of Fluoroalkylsilane Monolayers.
Langmuir 21: 911–918.
[30] Chung JY, Youngblood JP and Stafford CM. (2007) Anisotropic Wetting on
Tunable Micro-wrinkled Surfaces. Soft Matter 3: 1163–1169.
[31] Panchagnula MV and Vedantam S. (2007) Comment on How Wenzel and Cassie
were Wrong by Gao and McCarthy. Langmuir 23: 13242.
[32] Gao XF, Yan X, Yao X, Xu L, Zhang K, Zhang JH, Yang B and Jiang L.
(2007) The Dry-Style Antifogging Properties of Mosquito Compound Eyes and
Artificial Analogues Prepared by Soft Lithography. Adv. Mater. 19: 2213–2217.
[33] Koch K, Bhushan B and Barthlott W. (2009) Multifunctional Surface Structures
of Plants: An Inspiration for Biomimetics. Prog. Mater. Sci. 54: 137–178.
[34] Tadanaga K, Morinaga J, Masuda A and Minami T. (2000) Superhydrophobic-
superhydrophilic Micropatterning on Flowerlike Alumina Coating Film by the
sol-gel Method. Chem. Mater. 12: 590–592.
[35] Zhang X, Jin M, Liu Z, Nishimoto S, Saito H, Murakami T and Fujishima A.
(2006) Preparation and Photocatalytic Wettability Conversion of TiO2 -based
[36] Zhang X, Jin M, Liu Z, Tryk DA, Nishimoto S, Murakami T and Fujishima
A. (2007) Superhydrophobic TiO2 Surfaces: Preparation, Photocatalytic Wetta-
bility Conversion, and Superhydrophobic-superhydrophilic Patterning. J. Phys.
Chem. C 111: 14521–14529.
[37] Nakata K, Nishimoto S, Yuda Y, Ochiai T, Murakami T and Fujishima A.
(2010) Rewritable Superhydrophilic-Superhydrophobic Patterns on a Sintered
Titanium Dioxide Substrate. Langmuir, 26: 11628–11630.
[38] Guenthner A, Khombhongse S, Liu W, Dayal P, Reneker D and Kyu T. (2006)
Dynamics of Hollow Nanofiber Formation During Solidification Subjected to
Solvent Evaporation. Macromol. Theory Simul. 15: 87–93.
[39] Bognitzki M, Czado W, Frese T, Schaper A, Hellwig M, Steinhart M, Greiner
A and Wendorff JH. (2001) Nanostructured Fibers via Electrospinning. Adv.
Mater. 13: 70–72.
[40] Megelski S, Stephens JS, Chase DB and Rabolt JF. (2002) Micro-and Nanos-
tructured Surface Morphology on Electrospun Polymer Fibers. Macromolecules
35: 8456–8466.
[41] Dayal P, Liu J, Kumar S and Kyu T. (2007) Experimental and Theoretical Inves-
tigations of Porous Structure Formation in Electrospun Fibers. Macromolecules
40: 7689–7694.
[42] Lin JY, Ding B, Yu JY, Wu GC, Yang JM and Sun G. (2010) Effect of Porous
Structure of Electrospun Fibers on their Specific Surface Area: Theoretical
Analysis and Experimental Verification. Int. J. Nonlinear Sci. Numer. Simul. 11:
523–527.
[43] Li Y, Liu F and Sun J. (2009) A Facile Layer-by-layer Deposition Process
for the Fabrication of Highly Transparent Superhydrophobic Coatings. Chem.
Commun. 2730–2732.
[44] Cao LL, Gao, D. (2010) Transparent Superhydrophobic and Highly Oleophobic
Coatings. Faraday Discuss 146: 57–64.
[45] Ogihara H, Xie J, Okagaki J, Saji T. (2012) Simple Method for Preparing
Superhydrophobic Paper: Spray-deposited Hydrophobic Silica Nanoparticle
Coatings Exhibit High Water-repellency and Transparency. Langmuir 28:
4605–4608.
[46] Wenzel RN. (1936) Resistance of Solid Surfaces to Wetting by Water. Industrial
& Engineering Chemistry 28: 988–994.
[47] Golovin K, Lee DH, Mabry JM, Tuteja A. (2013) Transparent, Flexible,
Superomniphobic Surfaces with Ultra-Low Contact Angle Hysteresis. Angew.
Chem. Int. Ed. 52: 13007–13012.
[48] He ZK, Ma M, Lan XR, Chen F, Wang K, Deng H, Zhang Q, Fu Q.
(2011) Fabrication of a Transparent Superamphiphobic Coating with Improved
Stability. Soft Matter 7: 6435–6441.
[49] Han Y, Alsayed AM, Nobili M, Zhang J, Lubensky TC, Yodh AG. (2006)
Brownian Motion of an Ellipsoid. Science 314: 626–630.
[50] Cassie ABD, Baxter S. (1944) Wettability of Porous Surfaces. Transactions of the
Faraday Society 40: 546–551.
[51] Celia E, Darmanin T, Givenchy ET, Amigoni S, Guittard F. (2013) Recent
Advances in Designing Superhydrophobic Surfaces. Journal of Colloid and
Interface Science 402: 1–18.
[52] Zhou H, Wang HX, Niu HT, Gestos A, Wang XG, Lin T. (2012) Fluoroalkyl
Silane Modified Silicone Rubber/Nanoparticle Composite: A Super Durable,
Robust Superhydrophobic Fabric Coating. Adv. Mater. 24: 2409–2417.
[53] Du FM, Fischer JE, Winey KI. (2005) Effect of Nanotube Alignment on
Percolation Conductivity in Carbon Nanotube/polymer Composites. Phys. Rev.
B 72: 4–12.
[54] Bhushan B, Koch K, Jung YC. (2009) Fabrication and Characterization of
the Hierarchical Structure for Superhydrophobicity and Self-cleaning. Ultra-
microscopy 109(8): 1029–1034.
[55] Shieh J, Hou F, Chen YC. (2010) Chen HM, Yang SP, Chen CC, Chen UL.
(2010) Robust Airlike Superhydrophobic Surfaces. Adv. Mater. 22: 597–601.
[56] He Y, Jiang C, Yin H, Chen J, Yuan W. (2011) Superhydrophobic Silicon
Surfaces with Micro–nano Hierarchical Structures via Deep Reactive Ion Etching
and Galvanic Etching. J. Colloid Interface Sci. 364: 219–229.
[57] Accardo A, Gentile F, Mecarini F, Angelis FD, Burghammer M, Fabrizio ED,
Riekel C, In situ X-ray Scattering Studies of Protein Solution Droplets Drying
on Micro- and Nanopatterned Superhydrophobic PMMA Surfaces. Langmuir
26: 15057–15064.
[58] Park KC, Choi HJ, Chang CH, Cohen RE, McKinley GH, Barbastathis G.
(2012). Nanotextured Silica Surfaces with Robust Superhydrophobicity and
Omnidirectional Broadband Supertransmissivity. ACS Nano 6: 3789–3799.
[59] Kang, SM, You I, Cho WK, Shon HK, Lee TG, Choi IS, Karp JM, Lee H. (2010)
One-Step Modification of Superhydrophobic Surfaces by a Mussel-Inspired
Polymer Coating. Angew. Chem. Int. Ed. 49: 9401–9404.
[60] Sheng X, Zhang J. (2009) Biomimetic Polyimide Nanotube Arrays with Slippery
or Sticky Superhydrophobicity. Langmuir 25: 6916–6922.
[61] Cheng Z, Gao J, Jiang L. (2010) Tip Geometry Controls Adhesive States of
[62] Mahadik SA, Kavale MS, Mukherjee SK, Rao AV, Transparent Superhydrophobic
Silica Coatings on Glass by Sol–gel Method. Appl. Surf. Sci. 257: 333–339.
[63] Lakshmi RV, Bharathidasan T, Bera P, Basu BJ, Fabrication of Superhydrophobic
and Oleophobic sol-gel Nanocomposite Coating. Surf. Coat. Technol. 206:
3888–3894.
[64] Kim EK, Hwang T, Kim SS. (2011) Change in Microstructure and Surface
Properties of Electrospray-synthesized Silica Layers. J. Colloid Interface Sci. 364:
561–565.
[65] Kim EK, Lee, CS, Kim SS. (2012) Superhydrophobicity of Electrospray-
synthesized Fluorinated Silica Layers. J. Colloid Interface Sci. 368: 599–602.
[66] Kavale MS, Mahadik DB, Parale VG, Wagh PB, Gupta SC, Rao AV, Barshilia
HC. (2011) Optically Transparent, Superhydrophobic Methyltrimethoxysilane
Based Silica Coatings Without Silylating Reagent. Appl. Surf. Sci. 258: 158.
[67] Guvendiren M, Heiney PA, Yang S. (2009) Precipitated Calcium Carbonate
Hybrid Hydrogels: Structural and Mechanical Properties. Macromolecules 42:
6606–6613.
[68] Mihut AM, Sanchez-Ferrer A, Crassous JJ, Hirschi LA, Mezzenga R, Dietsch
H. (2013) Enhanced Properties of Polyurea Elastomeric Nanocomposites with
Anisotropic Functionalised Nanofillers. Polymer 54: 4194–4151.
[69] Ebert D, Bhushan B. (2012) Transparent, Superhydrophobic, and Wear-
resistant Coatings on Glass and Polymer Substrates Using SiO2 , ZnO, and ITO
Nanoparticles. Langmuir 28: 11391–11399.
[70] Jung YC, Bhushan B. (2009) Mechanically Durable Carbon Nanotube-

Composite Hierarchical Structures with Superhydrophobicity, Self-Cleaning,
and Low-Drag. ACS Nano 3: 4155–4163.
[71] Sun W, Shen L, Wang J, Fu K, Ji J. (2010) Netlike Knitting of Polyelectrolyte
Multilayers on Honeycomb-Patterned Substrate. Langmuir 26: 14236–14240.
[72] Kim YH, Lee YM, Lee JY, Ko MJ, Yoo PJ, (2012) Hierarchical Nanoflake Surface
Driven by Spontaneous Wrinkling of Polyelectrolyte/metal Complexed Films.
ACS Nano 6: 1082–1093.
[73] Zhang L, Sun J. (2010) Layer-by-Layer Codeposition of Polyelectrolyte Com-
plexes and Free Polyelectrolytes for the Fabrication of Polymeric Coatings.
Macromolecules 43: 2413–2420.
[74] Li Y, Li L, Sun J. (2010) Bioinspired Self-healing Superhydrophobic Coatings.
Angew. Chem., Int. Ed. 49: 6129–6133.
[75] Vandencasteele N, Broze B, Collette S, De Vos C, Viville P, Lazzaroni R, Reniers
F. (2010) Evidence of the Synergetic Role of Charged Species and Atomic
Oxygen in the Molecular Etching of PTFE Surfaces for Hydrophobic Surface
Synthesis. Langmuir 26: 16503–16509.
[76] Wohlfart E, Fernandez-Blazquez JP, Arzt E, del Campo A. (2011) Nanofibrillar
Patterns on PET: The Influence of Plasma Parameters in Surface Morphology.
Plasma Process. Polym. 8: 876–884.
[77] Tsougeni K, Vourdas N, Tserepi A, Gogolides E. (2009) Mechanisms of
Oxygen Plasma Nanotexturing of Organic Polymer Surfaces: From Stable Super
Hydrophilic to Super Hydrophobic Surfaces. Langmuir 25: 11748–11759.
[78] Gnanappa AK, Papageorgiou DP, Gogolides E, Tserepi A, Papathanasiou AG,
Boudouvis AG. (2012) Hierarchical, Plasma Nanotextured, Robust Superam-
phiphobic Polymeric Surfaces Structurally Stabilized through a Wetting-drying
Cycle. Plasma Process. Polym. 9: 304–315.
[79] Lafuma A, David Quéré. (2003) Superhydrophobic States. Nat. Mater. 2: 457–
460.
[80] Bahadur V, Garimella SV. (2007) Electrowetting-based Control of Droplet
Transition and Morphology on Artificially Microstructured Surfaces. Langmuir
23: 4918–4924.
[81] Bormashenko E, Pogreb R, Whyman G, Bormashenko Y, Erlich M. (2007)
Vibration-induced Cassie-Wenzel Wetting Transition on Rough Surfaces. Appl.
Phys. Lett. 90: 201917.
[82] Krupenkin TN, Taylor JA, Wang EN, Kolodner P, Hodes M, Salamon TR.
(2007) Reversible Wetting-Dewetting Transitions on Electrically Tunable Super-
hydrophobic Nanostructured Surfaces. Langmuir 23: 9128–9133.
[83] Boreyko JB, Chen CH. (2009) Wetting and Dewetting on Superhydrophobic
Surfaces with Two-Tier Roughness. Phys. Rev. Lett. 103: 174502.
[84] Dorrer C, Rühe J. (2009) Some Thoughts on Superhydrophobic Wetting. J.
Soft Matter 5: 51–61.
[85] Nosonovsky M, Bhushan B. (2007) Biomimetic Superhydrophobic Surfaces:

Multiscale Approach. Nano Lett. 7: 2633–2637.
[86] Inyushkin AV, Taldenkov AN, Yakubovsky AY, Markov AV, Moreno-Garsia L,
Sharonov BN. (2003) Thermal Conductivity of Isotopically Enriched 71 GaAs
Crystal. Semicond. Sci. Technol. 18: 685–688.
[87] Srinivasan U, Houston MR, Howe RT, Maboudian R. (1998) Alkylsiloxane-
based Self-assembled Monolayers for Stiction Reduction in Silicon Microma-
chines. J. Microelectromech. Syst. 7: 252–260.
[88] Bormashenko E, Pogreb R, Whyman G, Erlich M. (2007) Cassie-Wenzel
Wetting Transition in Vibrated Drops Deposited on the Rough Surfaces: Is
Dynamic Cassie-Wenzel Transition 2D or 1D Affair? Langmuir 23: 6501–6503.
[89] Li W and Amirfazli A. (2007) Microtextured Superhydrophobic Surfaces: A
Thermodynamic Analysis. Adv. Colloid Interface Sci. 132: 51–68.
[90] Gao L and McCarthy TJ. (2007) A Commercially Available Perfectly Hydropho-
bic Material (theta(A)/theta(R)=180 degrees/180 degrees). Langmuir 23:
9125–9127.
[91] Zhou H, Wang HX, Niu HT, Gestos A, Lin T. (2013) Self-healing Superam-
phiphobic Fabrics Prepared by Two-step Coating of Fluoro-containing Polymer,
Fluoroalkyl Silane and Modified Silica Nanoparticles. Adv. Funct. Mater. 23:
1664–1669.
[92] Yunker PJ, Still T, Lohr MA, Yodh AG, (2011) Suppression of the Coffee-ring
Effect by Shape-dependent Capillary Interactions. Nature 476: 308–311.
[93] Yang S, Horibe Y, Chen CH, Mirau P, Tatry T, Evans P, Grazul J, Dufresne EM.
(2002) Ordered Hydrophobic Organosilicates Templated by Block Copolymers.
Chem. Mater. 14: 5173–5180.
[94] Park SG, Moon JH, Lee SK, Shim J, Yang SM. (2010) Bioinspired Holograph-
ically Featured Superhydrophobic and Supersticky Nanostructured Materials.
Langmuir 26: 1468–1472.
Chapter 2
Treatments and Constructing Digital

Model of Biological Shark Skin/Shark
G.G. Li,∗ Y.T. Zhao,† L. Zhang,∗ Y. Luo,‡ E.Y.K. Ng§

∗ School of Mechanical Engineering, Xi’an Jiaotong University,
Xi’an, 711069, China
† School of Mechanical Engineering, Northwestern Polytechnical

‡ School of Engineering, The George Washington University,

§ School of Mechanical and Aerospace Engineering, Nanyang
Sharks have been living on the Earth for more than 400 million
years, and they appeared 100 million years earlier than the dinosaurs.
Although dinosaurs have become extinct for more than 60 million
years, different kinds of sharks still live on Earth now, which demon-
strates their perfect properties to survive in the environment. In this
chapter, the characteristics and properties of shark skin are investigated
and explored, and the model of shark skin is built, which can give
us intuitive impressions. Additionally, treatments for biological shark
skin in order to keep its good mechanical properties are illustrated; the
research can help readers better understand the fundamental knowledge
of shark skin, which has great significance.
45
2.1. Introduction
Despite huge changes in their environment, sharks have survived on
Earth for an incredibly long time. This has a lot to do with their per-
fect body features, strong environmental adaptability, special/strong
immune system and other characteristics.1−3 Fish can usually be divided
into bony and chondrichthyes categories; sharks belong to the latter
group. Skin friction drag reduction has been re-emphasized in recent
decades due to its practical value in engineering applications, including
vehicles, aircraft, ships and fuel pipelines. Many control methods and
technologies have been suggested to reduce friction drag, but most
of them are active approaches that require additional power input and
complicate the devices.4−8
To explore the underlying drag reduction mechanism and to
further reveal the exceptional features of animals in nature would
provide us a novel approach for friction drag reduction. Along with
the rigorous selection process of evolution, surviving living organisms
have successfully developed extraordinary abilities to adapt to the
environment, and offer multiple examples of surfaces that are optimized
to control friction. Based on bionics theories, many drag-reduction
devices such as micro-structured bionic surfaces like riblets and other
non-smooth surfaces have been developed and are regarded as a most
portable way to be implemented in real applications.
The shark is the swimming king of the ocean, and its cruising
speed is about 5 km/h; it can surpass 70 km/h when hunting prey
and escaping predators, and its speed is inferior only to the tunny
(80 km/h). The fascinating shark skin morphology and mucus endow
it with an advantage, which can be called the “shark skin effect.” Shark
skin surface and morphology are shown in Fig. 2.1; the rough skin was
used as abrasive paper in ancient Roman times.
Placoid scales are unique to the chondrichthyes fish, and are made
of dermis and epidermis. Because the shaping process of placoid scales is
the same as that of teeth, their micro-hardness is about 300–350 HV —
therefore, they are also called “toothed scales.” Placoid scales are
composed of base plate and spine, as shown in Fig. 2.2. The scales are
arranged according to the diamond-shaped, and the individual scales do
Treatments and Constructing Digital Model of Biological Shark Skin/Shark 47
Fig. 2.1. Sharkskin and micro-grooved morphology on shark skin.
Fig. 2.2. Structure of placoid scale.
not grow in size with the shark. The shapes and parameters of shark skin
scales are different on different species of sharks; even on the same shark,
the shapes and sizes of scales are still different, as shown in Fig. 2.3.9,10
The living biological shark skin surface can secrete lots of mucus.
Via placoid scales and mucus, shark skin can achieve several different
functions: protection, drag reduction, anti-fouling, noise reduction,
anti-wearing and so on.11,12
Fig. 2.3. Different morphologies on shark skin surface.
2.2. Treatments of Biological Shark Skin
We can obtain whole shark from aquatic products companies. As shown

in Fig. 2.4 (a) and (b), its total length is about 1.45 m, and its weight is
about 22.5 kg. It is widely distributed in the tropical layer of the world,
and is a common high-speed cruising shark.
The first step is to collect the shark skin samples, taking the differ-
ences in structure of the scales on different parts into consideration; the
more parts on sharkskin should be selected for purpose of ensuring the
integrity and reliability of parameters on shark skin scales.10−15 After
the shark specimen is completely thawed, the shark skin is carefully
peeled off with a scalpel. Attention should be paid to ensure that the
subcutaneous tissue and attachments are cleaned up completely, and the
shark skin should never be scraped in order to ensure its flatness. Shark
skin in its natural state cannot be adopted as testing samples — there
is lots of mucus and impurities, and the biological shark skin will be
dehydrated or deformed due to external force, heat, chemical materials
and so on, and the mechanical properties will be impacted seriously.
Therefore, biological shark skin should be treated before adaptation.
The purpose of treating shark skin is to maintain the biological original
Fig. 2.4. (a) Shark specimen (b) Sample of biological shark skin.
Cleaning Chemical fixation Re-cleaning Dehydration Drying
Fig. 2.5. Pre-treatment steps for biological shark skin.
type and increase the mechanical strength, and the main steps are as
following: cleaning, chemical fixation, re-cleaning, dehydration, drying
and so on,16,17 as illustrated in Fig. 2.5.
(1) Cleaning
Clean the shark skin samples with the clear water three to five times,
then wash the samples with de-ionized water for two or three times —
the purpose is to get rid of the mucus, dirt and blood on the surface.
To protect the tissue of the shark skin, alkalescent solution cannot be
used.
(2) Chemical fixation
The purpose of chemical fixation is to make the various cell organelles

and macromolecular structure remain the live station, and they should
be firmly fixed at the original position. The methods of fixation
can be divided into two categories: physical fixation and chemical
fixation. For physical fixation, freezing and drying methods are put
into application; for chemical fixation, certain chemical materials are
adopted to fix the cell structure. The most commonly used fixatives
include osmium tetroxide and glutaraldehyde (C5 H8 O2 ). Osmium
tetroxide is a kind of strong oxidant which has strong affinity with
the nitrogen atom. It can also react with unsaturated fatty acid so
that the fat can be fixed. In addition, osmium tetroxide can also fix
lipoproteins to make the phospholipoprotein more steady and stable.
Glutaraldehyde (C5 H8 O2 ) is another important fixative; it has good
fixing functions on glycogen, glycoproteins, microtubules, endoplasmic
reticulum, extracellular matrix and so on.
In this chapter, glutaraldehyde fixation is put into application.
After cleaning, the biological shark skin is completely immersed in
2.5% glutaraldehyde solution then placed in a constant temperature
environment over 6 hours at 4◦ C to completely fix the shark skin. Take
note that shark skin will become hard after fixation, and wrinkles cannot
be smoothed easily; the only way is to keep it smooth and flat during
the fixation process. In this chapter, the biological shark skin sample is
nailed on the plate, as shown in Figs. 2.6 and 2.7, the purpose of which
is to ensure the flatness of the shark skin and to avoid the destruction of
Fig. 2.6. Sample of biological shark skin.

Fig. 2.7. Schematic drawing of fixing shark skin.
the structure integrity of biological shark skin adopting conventional

methods.
(3) Re-cleaning
After fixing the shark skin, the biological sample should be rinsed with
phosphate buffer solution and then rinsed with the dionized water 3–5
times, the purpose of which is to flush out the residual glutaraldehyde
attached to the biological shark skin.
(4) Dehydration
In order to avoid shrinkage and deformation of the shark skin during
the heating process, the sample should undergo a dehydration step.
Commonly used dehydrating agents include alcohol (C2 H6 O) and
acetone (C3 H6 O). Because rapid dehydration can lead to adverse
shrinkage and deformation, the biological shark skin should be gradu-
ally dehydrated before heating. In addition, excessive dehydration can
lead to the sample becoming crisp, so the level of dehydration should
be appropriate. In this chapter, alcohol with mass ratio of 50%, 75%,
90%, 95% was adopted to complete the dehydration.
(5) Drying
After the dehydration, in order to remove the moisture in the biological
shark skin to the greatest extent, the purpose of which is to keep the
biological shark skin without deterioration, deformation, etc. at room
temperature, it should be placed into the oven at about 60◦ C for 5–6
Fig. 2.8. SEM of sharkskin surface.
hours. Take note that the relevant steps should be conducted to prevent
the dried shark skin from damp deformation.
After the above-mentioned steps, the SEM and three-dimensional
images of shark skin surfaces can be obtained,17−20 as shown in Figs. 2.8
and 2.9.
It can be seen that the shark skin scales are approximately 0.1 mm ×
0.1mm − 0.2mm × 0.2mm in size, and the height and width of grooves
on shark skin scale are about 20um and 50um respectively. Additionally,
the size and shape of scales are different on different parts of the shark
skin, and there are no two uniform scales in shape and size on the whole
surface of the shark. The different scales on different parts of the shark
skin are shown in Fig. 2.10. Therefore, it is also necessary to explore
the drag reduction efficiency of shark skin with different scales in the
same area.
Scales on different parts of the shark skin all possess their own
unique characteristics, including shape, size and so on. In this chapter,
the typical scales are analyzed, as shown in Table 2.1. The velocities of
external flow on different parts of the shark skin are different. This
could be one of the important factors for producing the different
characteristics of scales, but further validation and exploration are
required.
Fig. 2.9. Three-dimensional image of shark skin scale.
Fig. 2.10. 3-D morphology of scales on different parts of shark skin.

Table 2.1. Analysis of Typical Scales on Different Parts of Shark Skin.
Location Head Back Abdomen
cross-section
curve
width and Width of groove: Width of groove: Width of groove:

depth of 115.0µm, Depth of 88.3µm, Width of 85.0µm, Width of
groove groove: 15.5µm groove: 6.43µm groove: 17.1µm
Primary char- Large in size and Smaller in size and less Sharp pointed on the
acteristics saponaceous on the depth of groove edge and larger depth
edge of groove
2.3. Constructing Digital Model of Shark

Skin Surface
Shape and attack angles can be regarded as the basic elements of
biological shark skin scales. The former factor has been explored com-
prehensively and deeply, however, the latter factor also has an important
influence on the drag-reduction effect. Therefore, for the purpose of
building an accurate digital model, both the above-mentioned factors
should be taken into consideration. Due to the limitations of science
and technology, the research at this point is limited; in order to build
an accurate digital model, the more scientific and reliable methods are
put into application in this chapter.
In this chapter, qualitative observations and precision measure-
ments of the structure of shark scales are made with several major
techniques21−24 : (1) with the biological microscope shark skin scale
morphology was observed during various parts of the analysis;
(2) preparation of metallographic borrow shark skin sample by means
of a sample of a longitudinal sectional and cross-sectional samples,
and by means of an optical microscope of its longitudinal scales, the
cross-section, in order to observe its shape characterized in detail;
(3) by means of a surface profiler shark skin surface topography is
precisely scanned as the sample, and after scanning the resulting image
is processed using supporting software to obtain the individual scales
of the digital topography image, on the basis of the extracted shark
Fig. 2.11. Microscope.
trench biological prototype cross-sectional profile curve. According

to the groove width, height scales ridge, groove shape and other
characteristics Round Valley quantifies the work will be the next step
for biological prototype shark skin trench to lay the foundation of
mathematical modeling and shaping biological replicates. Different
observing tools and equipment are shown in Figs. 2.11 to 2.13.
The optical photograph of the biological shark skin surface is shown
in Fig. 2.14, and the two-dimensional image of biological shark skin
surface made using the Taylor Hobson surface profile scanner is shown
in Fig. 2.15; it can be seen that the micro-grooves sit on the scales, the
size of the scales varies from 0.1mm × 0.1mm − 0.2mm × 0.2mm, and
the scales are always overlapping with each other. In order to construct
the digital model of the shark skin surface, the more precise approach
should be put into application further.
Figure 2.16 shows the process flow sheets of analyzing the micro
flow field on the real shark skin surface. The main steps are as
follows24−28 : (1) pretreat the sampled shark skin and gain the clean
Fig. 2.12. Taylor Hobson surface profile scanner.
Fig. 2.13. Phase Shift MicroXAM-3D.
surface of biological templates for scanning, in which, for purpose

of obtaining the clean and neat surface of biological shark skin, it
should include cleaning, chemical fixation, re-cleaning, dehydration
and desiccation; (2) sputter the surface of biological shark skin for
good optical reflection and electric conduction. The subtle particles
of Au should be uniformly distributed on the sharkskin surface, and
the depth of Au layer is less than 20um with the purpose of protecting
Fig. 2.14. Optical image of biological shark skin.
Fig. 2.15. Image of biological shark skin by TH surface profile scanner.
the original morphology; (3) scan the morphology of shark skin scale
with highly accurate equipment and obtain sufficient and precise data;
(4) build the three-dimensional and accurate digital model of shark skin
scale; (5) carry out the numerical simulation of micro flow field with
CFD software and explore the drag-reduction mechanism.
In the process of building the digital model, to obtain the good
effect of optical reflection and electric conduction, the shark skin should
Sampling of Pretreatment Templates Sputtering Digital Numerical

sharkskin of sharkskin of sharkskin and scanning modeling simulation
Sputtering
Fig. 2.16. Process flow sheet for analyzing micro flow field on real shark skin.
Fig. 2.17. Image of Phase Shift MicroXAM-3D.
go through the following main steps: cleaning, chemical fixation, re-

cleaning, dehydration, drying, sputtering nano-particles of Au. The
morphology of shark skin is complicated and sophisticated, and a super-
highly accurate scanning method should be adopted to obtain sufficient
and precise data. In this chapter, the applied high precision instrument
is the Phase Shift MicroXAM-3D, as shown in Fig. 2.17. Its RMS (root-
mean-square) repeatability is 1nm, the minimum of vertical scanning
resolution is 0.1nm and calibration accuracy is less than 0.1%; the
relevant measuring parameters are illustrated in Table 2.2. Via the
highly accurate scanning, the three-dimensional image and relevant
data of the shark skin can be obtained.
Building the 3-D digital model of the real shark skin surface is
one of the most important aims of this chapter, and the model should
be exactly consistent with the biological prototype to ensure accuracy
and reliability. The two-dimensional image of a single scale surface and
corresponding cross-section curve of signed line are shown in Fig. 2.18.
By moving the signed lines, all the digital information of morphology
on the shark skin scale can be received.
Table 2.2. Measuring Parameters of Phase Shift MicroXAM-3D.
Resolution ratio 480 × 752
X, Y, Z stroke 00mm × 100mm × 100mm

Vertical resolution 0.1nm
Horizontal resolution 8.8 − 0.11µm
Field of view 0.084 × 0.063mm2 ∼ 8 × 10mm2
Calibration precision <<0.1%
Reflectivity requirements 1%–100%
109
Line No 41
10.0
0.0
12 M1 10.918 8.392
M2 125.56 6.420
M2-N1 114.64 –1.972
10 dy/dx –0. 0172 ~ –0.9853˚
8 M3 38.213 1.495
0 M4 68.238 12.13
M4–N3 30.025 10.64
6 dy/dx 0.354 ~ 19.51˚
0
4 M5 68.238 12.13
M6 98.263 0.4843
M6–N5 30.025 –11.65
2 dy/dx –0.388 ~ –21.21˚
Physical Image Coord:

0 34.12 32.80 484.3
20 40 60 80 100 120
0 86.0
Fig. 2.18. Biological shark skin scale and corresponding cross-section curve.
Shark skin scale has many sharp edges on its surface and the results
received from SPIP (Scanning Probe Image Processor) software are
relatively close together, so the information of all points cannot be
obtained directly. In this chapter, the data of the cross-section curve is
fitted first, and then imported into the Solidworks software, and the 3-D
digital model of shark skin scale can be built using the lofting forming
method. Additionally, the longitudinal sections of the scales are not
parallel to the flowing direction, but at a particular attack angle. In order
to inspect and confirm the attack angles, the samples are fabricated, as
shown in Fig. 2.19. The attack angle (α) is illustrated in Fig. 2.20; it
can be seen that it ranges from 10◦ to 35◦ .
According to the preliminarily built model and attack angles of
shark skin scale, the digital model of a single shark skin scale can be
Fig. 2.19. Samples for inspecting the attack angles.
Fig. 2.20. Microscopic image of shark skin attack angle.
constructed ultimately,29−32 as shown in Fig. 2.21. By observing the

distribution of scales on shark skin, the model of shark skin surface in a
large area is also built, as shown in Fig. 2.22. Because the morphology
of shark skin scale is so complicated and sophisticated, it is very difficult
to suture different surfaces with no gaps or overlaps, and the continuity
of shark skin surface and computational domain cannot be ensured to
Fig. 2.21. Digital model of single shark skin scale.
Fig. 2.22. Real shark skin surface in a large area.
some extent. Therefore, the real shark skin surface should be simplified
in a reasonable way, holding the basic features, including attack angle,
size, distribution and so on; the model of shark skin without gaps or
overlaps is illustrated in Fig. 2.23.
2.4. Constructing Digital Model of Shark
In this section, the digital model of the whole shark is investigated and
studied, and the testing equipment is the binocular vision detection
system based on sinusoidal structured light, which is mainly comprised
of the vision module and the displacement module. The vision module
Fig. 2.23. Shark skin surface without gaps and overlaps.
Fig. 2.24. Binocular vision detection system based on sinusoidal structure light.
is composed of two charge-coupled devices (CCD) and a projector,

as shown in Fig. 2.24. The projector is used to project the sinusoidal
grating, the purpose of which is to capture the matching points of
projector. By calibrating the structure parameters and direction rela-
tionship between the two projectors, the three-dimensional coordinate
Data Data Data Point Building Curved Surface Curved Surface

Collection Pretreatment Treatment Cloud Curves Modification Generation
Zoning
Fig. 2.25. Flowing sheet of building digital model of shark.
can be obtained. The vision module is fixed on the stepping motor with
5 free degrees, which can realize the movements along the x axis, y axis,
z axis and revolving the x axis and z axis.
The process of measuring the digital model of the shark is illustrated
in Fig. 2.25. It is mainly composed of data collection, data pre-
treatment, data treatment, point cloud zoning, building curves, curved
surface modification, curved surface generation and so on.
The total length of shark the shark used was about 1.4m; the
weight is about 23 kg. Because the shark was directly purchased from
the aquatic product market, it is in the frozen state. In order to
obtain the digital model of the shark in its natural state, the thawing
process should be carried out first. The purpose of data collection is to
capture the digital point cloud, which can be regarded as the basis of
building model, manufacture, numerical simulation and characteristic
analysis. Therefore, the scanning precision has a direct relationship with
the analysis. Increasing the optical grating can increase the density
of the data point cloud and capture more information, which has
important significance for acquiring more information. In this chapter,
the optical grating is set as 768 × 576. Figure 2.26 shows the process
of scanning the shark, and Fig. 2.27 shows the point cloud of the
shark.
In the process of data collection, errors due to some uncertain
factors can be produced, which can lead to deviation of data point
cloud, even if the serious distortion is generated. All in all, the data
treatment step is very necessary, and is mainly composed of error points
processing, smooth processing and simplification processing. After the
above-mentioned data processes, the digital model of shark is shown in
Fig. 2.28. For the shark, the tail is very important, and enough attention
should be given32,33 ; the digital model of the shark tail is illustrated in
Fig. 2.29.
Fig. 2.26. Scanning the shark.
Fig. 2.27. Point cloud of shark.

Fig. 2.28. Digital model of shark from different directions.
Fig. 2.29. Digital model of shark tail.
2.5. Conclusion
In this chapter, the treatment of biological shark skin is systematically

introduced. Such treatment can ensure the mechanical intensity and
integrity of structure; the digital model of shark skin is built through
a highly precise scanning method, which can give us an intuitive
impression; the digital model of the whole shark is constructed, and
the data processes are also carried out.
References
[1] Luo YH. (2015) Recent Progress in Exploring Drag Reduction Mechanism of
Real Sharkskin Surface: A Review. J Mech Med Bio 15(3): 1530002.
[2] Luo YH. (2014) Recent Research Progress of Biological Sharkskin Surface: A
Review. J Surf Interf Mater 2(3): 167–181.
[3] Bushnell DM, Moore KJ. (1991) Drag Reduction in Nature. Annu Rev Fluid
Mech 23: 65–79.
[4] Luo YH, Zhang DY, Liu YF, et al. (2015) Chemical, Mechanical and Hydro-
dynamic Properties Research on Composite Drag Reduction Surface Based on
Biological Sharkskin Morphology and Mucus Nano-long Chain. J Mech Med Bio
15(5): 1530084.
[5] Bhushan B, June YC, Koch K. (2009) Micro-, nano- and hierarchical structures
for superhydrophobicity, self-cleaning and low adhesion. Philos Trans Roy Soc A
367: 1631–1672.
[6] Fish FE. (2006) The myth and reality of Gray’s paradox: implication of dolphin
drag reduction for technology. Bioinspiration & Biomimetics 1(2): 17–25.
[7] Fish FE, Hui CA. (1991) Dolphin swimming — a review. Mammal Rev 21(4):
181–195.
[8] Pavlov VV. (2006) Dolphin skin as a natural anisotropic compliant wall.
Bioinspiration & Biomimetics 1(2): 31–40.
[9] Shephard KL. (1994) Functions for fish mucus. Rev Fish Biol Fisher 4: 401–429.
[10] Flik G, Vanrijs JH, Bonga SEW. (1984) Evidence for the presence of calmodulin
in fish mucus. Eur J Biochem 138(3): 651–654.
[11] Bhushan B. (2009) Biomimetics: lessons from nature-an overview. Philos Trans
Roy Soc A 367: 1445–1486.
[12] Luo YH, Liu YF, Zhang DY. (2015) Advanced Progress in Nature Gas Pipelining
Applying different Drag Reduction/Energy Saving Technologies: A Review. Eur
J Environ Civil Eng 19(8): 931–949.
[13] Luo YH, Liu YF. (2014) Numerical Simulation of Micro Flow Field on
Biomimetic Sharkskin Micro-grooved Surface. Adv Mater Res 884–885:
378–381.
[14] Naresh MD, Arumugam V, Sanjeevi R. (1997) Mechanical behavior of shark
skin. J Bioscience 22: 431–437.
[15] Bechert DW, Bruse M, Hage W, Meyer W. (2000) Fluid mechanics of bio-
logical surfaces and their technological application. Naturwissenschaften 87(4):
157–171.
[16] Luo YH, Zhang DY. (2013) Investigation on fabricating continuous vivid
sharkskin surface by bio-replicated rolling method. Appl Surf Sci 282: 370–375.
[17] Luo YH, Zhang DY. (2011) Study on the Micro-replication Precision of Shark
Skin. Appl Mech Mater 44–47: 1151–1157.
[18] Bechert DW, Bruse M, Hage W, et al. (1997) Experiments on drag reducing
surfaces and their optimization with an adjustable geometry. J Fluid Mech 338:
59–87.
[19] Bechert DW, Bartenwerfer M. (1989) The viscous flow on surfaces with
longitudinal ribs. J Fluid Mech 206: 105–129.
[20] Choi H, Moin P, Kim J. (1993) Direct numerical simulation of turbulent flow
over riblets. J Fluid Mech 255: 503–539.
[21] Dean B, Bhushan B. (2010) Shark-skin surfaces for fluid-drag reduction in
turbulent flow: A review. Philos Trans Roy Soc A 368: 4775–4806.
[22] Lang A, Motta P, Habegger ML, et al. (2011) Shark Skin Separation Control
Mechanisms. Mar Technol Soc J 45: 208–215.
[23] Lang AW, Motta P, Hidalgo P, Westcott M. (2008) Bristled shark skin: a
microgeometry for boundary layer control? Bioinspiration & Biomimetics 3:
046005.
[24] Han X, Zhang DY. (2008) Study on the micro-replication of shark skin. Sci
China Ser E 51(7): 890–896.
[25] Han X, Zhang DY, Li X, Li YY. (2008) Bio-replicated forming of the biomimetic
drag-reducing surfaces in large area based on shark skin. Chinese Sci Bull 53(10):
1587–1592.
[26] Luo YH, Zhang DY, Liu YF, Ng EYK. (2015) Bio/Micro-Rolling Fabrication
of Biological Sharkskin Morphology on Semi-Cured Coating and Drag Force
Experimental Research. J Mech Med Bio 16(2): 1650016.
[27] Zhang DY, Luo YH, Li X, Chen HW. (2011) Numerical Simulation and Exper-
imental Study of Drag-Reducing Surface of A Real Shark Skin. J Hydrodynamics
23(2): 204–211.
[28] Luo YH, Liu YF, Zhang DY, Ng EYK. (2014) Influence of Morphology for
Drag Reduction Effect of Sharkskin Surface. J Mech Med Bio 14(2): 1450029.
[29] Luo YH, Liu YF, Anderson J, et al. (2015) Improvement of water repellent and
hydrodynamic drag reduction property on bio-inspired surface and exploring
sharkskin effect mechanism. Appl Phys A 120(1): 369–377.
[30] Koeltzsch K, Dinkelacker A, Grundmann R. (2002) Flow over convergent and
divergent wall riblets. Exp Fluids 33(2): 346–350.
[31] Wen L, Weaver JC, Lauder GV. (2014) Biomimetic shark skin: design, fabrication
and hydrodynamic function. J Exp Biol 217: 1656–1666.
[32] Motta P, Habegger ML, Lang A, et al. (2012) Scale Morphology and Flexibility
in the Shortfin Mako Isurus oxyrinchus and the Blacktip Shark Carcharhinus
limbatus. J Morphol 273: 1096–1110.
[33] Luo YH, Liu YF, Zhang DY. (2015) Prediction the variation of shark scale’s
attack angles in swimming. Indian J Anim Res 49(3): 295–302.
Chapter 3
Different Approaches to Manufacture

Low Viscous Resistance Drag with
Biomimetic Textures
J. Wang,∗ Y.T. Zhao,∗ L. Zhang,∗ Y. Luo,† E.Y.K. Ng‡

∗ School of Mechanical Engineering, Northwestern Polytechnical
† School of Engineering, The George Washington University,

‡ School of Mechanical and Aerospace Engineering, Nanyang
Through millions of years of natural selection, the surfaces of some

natural creatures have evolved into unique hierarchical micro-structures
with superior function to perfectly adapt to their survival environment.
Shark is the fast swimming animal in the ocean, and it is well-
known for the “shark skin effect.” Shark skin is covered with tiny
and rigid scales which can stick out of the viscous sublayer and have
the function of effectively inhibiting the occurrence of turbulence
and reducing wall friction. However, conventional methods cannot
realize the fabrication due to its complicated and subtle morphology,
and how to fabricate vivid shark skin has developed into a hot
topic nowadays. Many researchers have investigated and explored
different fabricating methods and put the biomimetic drag-reducing
technology into applications in fluid engineering; the results are very
69
satisfactory and striking. In this chapter, different methods to fabricate

real shark skin surfaces are presented in detail, including direct bio-
replicated forming, micro-rolling, large-scale solvent-swelling, additive
low-releasing design, synthetic drag-reduction bio-replication, trans-
scale enlarged 3-D imprinting, flexible 3-D imprinting and so on.
3.1. Why a Need to Fabricate Real Sharkskin?

It has become an undisputable fact that the biological and biomimetic
shark skin has the obvious effect of drag reduction in turbulent flow.
In the 1970s, Walsh at the NASA Langley Research Center found
that the biomimetic sharkskin ribbed surface can effectively reduce the
viscous friction for certain flowing conditions.1 It completely broke
with conventional thinking and opened up new approaches to drag-
reduction technology for mankind. Since then, many researchers have
explored and investigated the drag-reduction mechanism and exploited
its application, which has brought great advantages in a wide variety of
applications in daily life, industry and agriculture.
Bechert et al.2−4 manufactured the grooved surfaces with different
sizes and shapes by the optimization method, and then carried out the
experiments in water tunnel, and the results revealed that the maximum
drag-reducing efficiency was 9.9%. Reif and Dinsklacker5 discovered
that there were different morphologies on different shark skin scales,
and they had the drag-reduction effect on certain turbulent stations.
Luo and Zhang6−8 put the biomimetic shark skin micro-grooved drag
reduction technology into application on natural gas pipelining, and
the field experiments in Dongying revealed that the pressure loss could
be decreased more than 8% with the same transporting circumstance.
Bhushan9−11 explored different hierarchical nano/micro structured
biomimetic surfaces that had the obvious and striking effect of drag
reduction, self-cleaning, low adhesion and so on. Haecheon Choi12
studied the direct numerical simulation (DNS) of micro flow field over
the simplified and direct grooved surface for the first time, and the
drag-reduction mechanism was investigated and explored in depth. Liu
et al.13 carried out the experiments of 25.4 mm and 50.8 mm diameter
pipes lined with a film of micro-grooved equilateral triangles of base
Different Approaches to Manufacture Low Viscous Resistance Drag 71
0.11 mm, and the drag-reduction efficiency could reach about 7% in

the completely developed turbulent flow of water. German researchers
Koeltzsch et al.14 investigated the velocity field over convergent
and divergent riblet patterns by hot-wire measurements in turbulent
pipe flow. Significant changes in the near wall velocity field were
found, and drag-reduction efficiency up to about 10% was measured.
Xiang15 fabricated the drag-reducing biomimetic triangle simplified-
grooved surface efficiently using ultrasonic elliptical vibration cutting
(UEVC) technology, and the drag-reduction skin could be put into
application with different velocities. Viswanath16 reported that riblets
in symmetric V-grooves (height equal to spacing) with adhesive-backed
film manufactured by the 3M corporation in the United States had been
widely investigated. The results suggested that, with optimized riblets,
skin friction drag-reduction in the range of 5–8% had been measured
on 2-D airfoils at low incidence and in mild adverse pressure gradients,
and strong evidence existed at low speeds to indicate that riblets were
more effective in adverse pressure gradients.
Nature has given us so many lessons, and it has not been surpassed
so far. As far as the shark skin surface is concerned, the vivid shark skin
surface is superior to the simplified straight micro-grooves. Harvard
professors Lauder et al.17,18 investigated the self-propelled swimming
speed of foils on different surfaces, and the conclusion was that the
flexible shark skin foils showed a substantial improvement in swimming
performance of an average of 12.3% as compared with the same foils
with the surface denticles sanded off, and they proposed that the shark
skin denticles might thus enhance thrust as well as reduce drag. Luo
and Zhang19−21 investigated the distribution of micro flow field on real
shark skin surface by the direct numerical simulation (DNS) method for
the first time, and the results indicated that the drag-reducing efficiency
could surpass 12%, which was greater than that of the simplified micro-
textured surface. Additionally, they also explored different aspects of
the drag-reduction mechanism. Lang22,23 investigated the influence of
bristled shark skin on boundary layer control, concluding that shark
scales were pliable and might erect passively, which was helpful to keep
the high drag-reducing effect. All in all, the vivid shark skin surface
is more advanced than the simplified straight micro-grooved surface in
Fig. 3.1. Different methods of fabricating drag-reducing surfaces with biological

morphology.
drag-reduction effect, anti-fouling effect, wear resistance and so on, and

therefore, it is very necessary to manufacture the vivid shark skin. The
drag-reduction surfaces with biological shark skin morphology possess
many advantages compared with simple micro-grooved surfaces, and
many researchers have explored different accessible approaches, mainly
including direct bio-replicated, synthetic bio-replicated, continuous
micro-rolling, large-scale solvent-swelling, additive low-releasing, rapid
ultraviolet light painting, trans-scale 3-D imprinting, flexible 3-D
imprinting methods and so on, as illustrated in Fig. 3.1. For the purpose
of comprehending the shark skin effect, the drag-reduction mechanism
is also discussed systematically and comprehensively in this chapter.
3.2. Direct Bio-Replicated Method
Direct bio-replicated forming is a novel and practical method for

fabricating special micro-textured, original-scaled and special functional
surfaces, which is an effective tool to realize the transition from “shape-
imitation” to “sprit-imitation,” and it is one kind of shortcut when the
relevant mechanism has not been understood completely.
Guoming24 fabricated the artificial nano/micro-structured cicada
wing surface via the micro-embossing method with the biological origin
as the template. Dusan25 manufactured the micro-porous surface by
Fig. 3.2. (a) Conventional demolding; (b) demolding of shark skin scale.
Sharkskin Bio-replicated Replication Mold

Sampling Pretreatment template forming mold replicating
Fig. 3.3. Process flowsheet of the morphology replication of shark skin.
adopting the direct nano-imprinting the diatom biosilica. Nagaraja26

replicated the convex texture on the head of the dung beetle using the
micro-electro forming method, the result of which could transform the
hydrophilic surface into a hydrophobic surface. However, the above-
mentioned commonly adopted replication technologies are only fit for
objects with an obtuse bevel angle, while shark scales are slantwise
cuneiform, as shown in Fig. 3.2, and it raises the new requirements
on material liquidity and demolding.
The direct bio-replicated method can be exploited to fabricate
biomimetic shark skin surface with the original morphology. The
process flow sheets of the morphology replication of vivid shark skin
surface are illustrated in Fig. 3.3, and the step of bio-replicated forming
is the most important.
The main process includes the following steps27−28 : (3.1) Cut
out a large sheet of fresh shark skin and remove the subcutaneous
tissues. The biological shark skin should be continuous and the
apparent and visible defects should not exist on the skin. For purpose
of holding the biological shark skin prototype and enhancing the
mechanical strength, the pretreatments of cleaning, chemical fixation,
re-cleaning, dehydration and desiccation should be performed in turn,
as shown in Fig. 3.4. (2) Carry out the direct bio-replicated forming
method and get the micro-replication molds, the negative template
of the shark skin surface is shown in Fig. 3.5; (3) Pour or spray the
Fig. 3.4. Biological shark skin.
Fig. 3.5. Negative template of shark skin.
polymers in liquid or flowing condition, such as silicon rubber, PDMS

(polydimethylsiloxane) and so on, on the negative shark skin template;
(4) The artificial shark skin surface with the real morphology can be
obtained by turning the template over after the polymer is fully cured.
As illustrated in Fig. 3.6, the direct bio-replicated method based
on biological shark skin involves the following steps: substrate heating,
template tacking and isostatic pressing, flexibility demolding and mold
replicating. The flat PMMA (polymethyl methacrylate) plate was cho-
sen as the substrate. First, it was heated up to its glass temperature in the
hot and drying oven, which was about 105◦ C, and then the biological
shark skin template was stacked on the PMMA plate scale side down.
June 1, 2016 15:38 Bio-Inspired Surfaces and Applications 9in x 6in b2378-ch03 page 75
PMMA (Tg) PMMA (Tg)
Substrate heating
Shark skin
Stacking and pressing
Replication
Flexibillity demoulding mold
Biomimetic
skin
Model replicating
Transverse vtew Longitudinal view
Fig. 3.6. Illustration of bio-replicated method.
Second, the isostatic pressure was applied on the shark skin template
for 30 minutes. Third, the temperature was lowered to 70◦ C smoothly
and slowly, with the pressure maintained. Fourth, the demolding of
the PMMA plate was done at room temperature. It should be noted
that the scale root and the opposite direction of the scale should be
taken respectively as the starting point and the demolding direction.
Depending on the draft angle formed by the scale tip and the elasticity
of shark scales as well as the fastness of the shark skin, the shark scales
can bend both in transverse and longitudinal directions. Therefore,
the flexibility demolding is feasible. The demolding of the biomimetic
shark skin is alike. Finally, vulcanization silicon rubber RTV-II 5230
was selected as the replication material. The pre-polymer and curing
agent were mixed in a mass ratio of 100:1 and degassed in a desiccator,
and then the mixture was poured onto the mold surface and degassed
Fig. 3.7. (a) SEM of the mold; (b) Image of replicated shark skin.
again. After curing and demolding, the biomimetic shark skin was
obtained. The SEM image of negative sharkskin template is presented
in Fig. 3.7 (a), and the image of replicated shark skin with silicon rubber
is shown in Fig. 3.7 (b); it can be seen that a good forming effect was
obtained.27,28
It is worth taking care that the longitudinal section of the shark
skin scale is not parallel to the direction of water flow but at a particular
attack angle, as shown in Fig. 3.8.29 In the process of pouring silicon
rubber on the negative mold, there will be some air bubbles remaining
at the bottom of the hollows until the silicon rubber is cured completely,
as shown in Fig. 3.9. The consequence of this is that the shark skin
surface with real morphology cannot be perfectly replicated, perhaps
Fig. 3.8. Attack angle of shark skin scales.
Fig. 3.9. Air bubbles staying at the bottom of hollows.
affecting the drag-reducing efficiency. Of course, this should be further

verified.
In order to investigate the influence of not completely filling into
the hollows on drag reduction, the analysis of micro flow field over the
real shark skin surface is carried out comprehensively and in depth via
direct numerical simulation in software, and the velocity distribution
on a particular section can be obtained, as shown in Fig. 3.10. It can
be seen that the “back flowing” phenomenon exists on the slop surface
of scale, which can make the flow in the boundary layer more steady, in
turn lowering turbulence intensity and wall resistance. This is perhaps
Fig. 3.10. Numerical simulation of attack angles on back of scale.
Fig. 3.11. Schematic drawing of attack angle’s influence.
an important factor in producing higher drag-reducing efficiency

than simple grooved surfaces without attack angles. Therefore, if the
integrity of the attack angle on shark skin scales cannot be ensured, the
back flowing phenomenon will be weakened, as shown in Fig. 3.11,
and the drag-reducing efficiency will also be reduced to some extent.
In order to eliminate the air bubbles at the bottom of hollows,
the process of pouring silicon rubber on the negative mold should be
carried out in the vacuum oven; the schematic drawing of the vacuum
pouring mechanism is shown in Fig. 3.12. This is mainly composed of
1– basement of mechanism, 2 – supporting bracket, 3 – positioning shaft, 4 – rotating shaft, 5 – supporting shaft, 6 – slipping
shaft, 7 – sloping surface, 8 – connection line, 9 – timing vibrator, 10 – rotating box, 11 – beaker, 12 – supporting bracket,
13 – dring stick, 14 – negative shark skin template, – vacuum oven
Fig. 3.12. Schematic drawing of vacuum pouring mechanism.
the basement, supporting bracket, connection line, driving stick, timing

vibrator, rotating box, shafts with different functions and so on. The
two supporting brackets are fixed on the basement symmetrically; the
two ends of positioning shaft, rotating shaft and supporting shaft are
embedded into the supporting brackets. First, blend the silicon rubber
and curing agent completely using the stirring method with the weight
ratio of 100:1; Second, put the mixture into the beaker and place the
rotating box is horizontally on the supporting shaft; Third, vacuum
pump the oven until the pressure is less than 0.01 bar; Finally, the
timing vibrator will vibrate and slip along the slope surface, the box
will rotate under the drive of connection line, and the mixture will be
poured on the negative shark skin template without the air bubbles
staying at the bottom of the hollows.
According to the above-mentioned analysis, the bio-replicated
method in manufacturing biomimetic surface of shark skin has very
high replication-precision and good forming effect. In order to validate
the forming quality and drag-reduction effect achieved using the
vacuum casting method, an experiment is performed in water tunnel
in which the parameters are as following: (1) The length of the test
section in vacuole water tunnel is 3.2 m with diameter 0.8 m; (2) The
Water tunnel wall
Supporting part
Electronic scale
Testing model
(a)
(b)
Fig. 3.13. (a) Schematic drawing of experiment testing (b) Field image of experiment
testing.
flowing velocity in the water tunnel is adjustable from zero to 20 m/s;

(3) The central pressure of the test section can be changed between
8kPa∼400kPa; (4) The vacuole index is 0.15. The test instruments
also include resistance sensor with 300N measurement range and 0.1N
resolution, two SA55 DC amplifier, NI-PCI4472 digital sampling panel
with sampling frequency 200KHZ, and IBM-PIV computer.
The testing model with hollow elliptical sphere at front and hollow
cylinder at posterior segments are made of aluminum alloy LY12. The
size of test model with the length of the cylinder section 500 mm and
external diameter 90 mm are determined according to the size of strain
gauge balances of test system. And the external diameter of hollow ellip-
tical sphere is designed at 94 mm to guarantee smooth outer surface.
The wall resistance experiments are conducted according to the
rules for vacuole water tunnel test (Q/702J0301-2008). The water
temperature is set to 28◦ C and the test system degassed for more than
one hour before testing. The water velocity can change from 2 m/s to
10 m/s. The water velocity of each test continuously enhances until
the skins are washed off. The schematic drawing and field image of the
experiment are shown in Fig. 3.13, and the results are illustrated in
Table 3.1.
Table 3.1. Testing Results of Water Tunnel.
Flowing velocity (m/s) Smooth skin (N) Shark skin (N)
3.3 2.42 2.11

3.6 3.11 2.76
3.9 3.89 3.50
4.2 4.78 4.32
4.5 5.76 5.22
5.0 7.61 6.91
5.5 9.75 9.10
6.0 12.15 10.99
6.5 14.8 13.4
16
14
12
Wall resistance/ N
10
4
Result of smooth skin
2 Result of sharkskin with vivid morphology
0
3 4 5 6 7
Flowing velocity m/s
Fig. 3.14. Experimental results in water tunnel.
The fitted curves of wall resistance on smooth skin and shark skin
surface are shown in Fig. 3.14; the drag-reducing efficiency at different
velocities is plotted in Fig. 3.15. It can be seen that the drag-reducing
efficiency can surpass 12% under certain circumstances. Although the
vivid artificial shark skin surface has been manufactured and the high
drag-reduction effect is also validated in the water tunnel, there are still
some minor errors in the experiment as following: (1) The biological
scale is very hard, but the artificial shark skin is made of silicon rubber,
Experimental results
Drag-reducing efficiency/ %
Fitting curve
14
12
10
8
3 4 5 6 7
Flowing velocity/ m/s
Fig. 3.15. Testing result of drag-reducing efficiency.
which is very soft — the different hardness of the scales will have some
influence on the flow field; (2) Limited by the area of whole shark, the
samples are spliced together — the different seams will produce errors
in the test results.
3.3. Fabrication of Synthetic Drag-Reduction

Shark Skin
Natural shark skin has excellent drag-reduction performance because
of a hierarchical structure built up by microgrooves and nano-long
chain mucus drag-reduction interface around the shark body, and just
like most of the fish, shark skin can secrete mucus, which has the
function of keeping out of fungus and reducing viscous friction — it
was reported that the drag-reduction efficiency of mucus could surpass
60%. The morphology of shark skin could be replicated initially by direct
bio-replicated forming, however, how to realize the synthetic drag-
reduction effect by adopting the micro morphology and mucus has
become an important issue to be resolved. Synthetic drag-reduction has
been explored and investigated for the past several decades, however,
the drag-reduction morphology was built simplified and bio-drag
reduction interface was formed by injecting drag reduction agent
(DRA) or coating polymer, which has serious problems, such as the
morphology of micro-grooves being different from the biology, waste
of DRA and so on.
Fig. 3.16. Mucus and goblet secretory cell.7

Friction reduction compared with fresh water (%)
80
small mouth bass
70
White crappie
60
50 Kamioods
rainbow trout
40
German
30 brown
trout
20
bluegill
10
0
0 5 10 15 20 25 30 35 40 45 50 55
Percent time in solution in fresh water
Fig. 3.17. Drag-reduction efficiency of different fish mucus.
For majority of the fish, the goblet cells are embedded in the skin,
and their secretions are mainly composed of polysaccharide — protein
and fiber materials, which combined with water can become mucus after
the release of these substances from the mucous cells. The mucus and
goblet cells are illustrated in Fig. 3.16,7,30 and the drag-reduction effect
curves of different fish mucus in fresh water are shown in Fig. 3.17 —
it can be seen that the maximum can reach about 60%.
graft copolymer
PAM solution synthetic bio-replication
long-chain of DRA shark skin after pre-treatment
of shark skin
bubble
water-based epoxy
resin emulsion
pouring of
elastic demoulding
silicone rubber
graft copolymer
prepolymer
curing
curing agent
and plasticizer
degassing
filtering and pre-degassing
vacuum degassing
prepolymer
demoulding
pouring of prepolymer
preparation of prepolymer formation of synthetic drag reduction surface moulding of elastic female die
Fig. 3.18. Process of synthetic bio-replication of shark skin.
A novel synthetic bio-replication forming approach was pro-

posed for simultaneously grafting nano-long chains to bio-replicated
micro-grooves of shark skin, which maintained the simplicity of direct
bio-replicated forming. Synthetic bio-replication combined soft die
forming and grafting technology, which had advantages such as result-
ing in high replication precision and saving of drag-reduction agent.
It provided an effective approach for the manufacture of synthetic
drag-reduction shark skin surface with high replication precision and
excellent-drag reduction performance.
Figure 3.18 shows the process for synthetic bio-replication of the
shark skin surface, which mainly includes the following steps: (1) Pre-
treatment of the shark skin, involving rigidity fixing, cleaning, chemical
fixing, rinsing, dehydration and drying; (2) Molding of the flexible
female die; (3) Formation of the synthetic drag-reduction shark skin, in
which a graft copolymer of water-based epoxy resin and Polyacrylamide
(PAM) was used as the substrate of the synthetic drag-reduction shark
skin.30
The skin of the shortfin mako shark (Isurus oxyrinchus ), which
is a typical fast shark with three-tips and three-ridges scales, is taken
as the stamp of bio-replication. Condensation bi-component silicone
Fig. 3.19. Water-based bisphenol-A epoxy resin.
Fig. 3.20. Polyacrylamide (PAM).
rubber RTV-2 875 (Room Temperature Vulcanized), which has A,

B component, is an excellent silicone rubber for dies because of
advantages such as low heat shrinkage, less surface energy, and excellent
demolding performance. Waterborne epoxy resin is composed of water-
based epoxy resin emulsion (AB-EP-44) and waterborne epoxy curing
agent (AB-HGF). Water-based epoxy resin emulsion is water-based
bisphenol-A epoxy resin, whose chemical structural formula is shown
in Fig. 3.19. Polyacrylamide (PAM) is a water-soluble linear polymer
drag-reduction agent with molecular weight more than 3 million and
chain segment length 0.25nm. The chemical structural formula of PAM
is presented in Fig. 3.20.30,31 Di-n-butyl phthalate (DBP), C16 H22 O4 ,
is a stable colorless and transparent liquid which is used as a plasticizer.
The synthetic drag reduction shark skin surface can be obtained, as
shown in Fig. 3.21, and its size is about 18 cm × 15 cm. The replication
precision was also analyzed by the marking key point method, and the
conclusion could be obtained that the vertical section curve of synthetic
drag-reduction shark skin manufactured by synthetic bio-replication
was almost coincident with the shark skin sample, with a height offset
Fig. 3.21. (a) Biomimetic synthetic drag reduction shark skin; (b) SEM of synthetic
drag-reduction shark skin.
Supporting part
balance
model
the wall of water tuunel
Fig. 3.22. Schematic diagram of the test section.
randomly distributed and an average deformation rate of only 3.7%.

Therefore, the high replicating precision can also be obtained for the
synthetic drag-reduction shark skin surface.30,31
To validate the drag-reduction effect of synthetic drag-reduction
shark skin, various experiments were conducted in a vacuole water
tunnel at the China Ship Scientific Research Center (CSSRC, in Wuxi
city, Jiangsu Province, the People’s Republic of China); the schematic
diagram of the testing section and the testing results are illustrated in
Figs. 3.22 and 3.23, respectively. The drag-reduction rate of synthetic
drag-reduction shark skin surfaces increased steadily with the increase of
flow velocity, and they had a remarkable drag-reducing effect, reaching
up to 24.6% during the test process — much greater than the maximal
drag-reduction rate, 8.25%, of the single bio-shark skin micro-groove
drag-reduction skin.30,31
24
DR2
22
20
Drag reduction rate DR (%)
18 DR1
16
14
12
10
8
DR3
6
4 The sum of micro-groove and nano-long chain
2 Synthetic drag reduction shark skin
Bio-replication shark skin
0
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
Flow velocity v (ms–1)
Fig. 3.23. Experimental results in water tunnel.
3.4. Micro-Rolling Bio-Replicated Method

The vivid shark skin surfaces have been fabricated by the direct bio-
replicated micro-imprinting method, and their good drag-reducing
effect has been validated in water tunnel. However, for the surface
of fluid which is larger than the whole skin of a shark, splicing and
affixing steps are absolutely necessary. The complexity of the process
will be increased significantly, and the stress concentration produced
by external flow will be produced on joining seams, especially for those
which are perpendicular to the flow direction; the joining surface will be
destroyed at a high speed of flowing, and the drag-reducing efficiency
could be affected. Therefore, it is important to explore a new way to
manufacture the continuous vivid shark skin for a large area.32
In order to realize manufacturing of continuous shark skin in large
area, the rolling method is the best selection; many researchers have
explored the micro-rolling method to fabricate the biomimetic drag-
reducing surface. Bechert et al.2−4 adopted the micro-rolling method
to manufacture the suitable surface, and the micro-grooved texture
was printed onto the interior surface of a pot, which was covered with
a bee’s wax-paraffin mixture. Fritz Klocke33 machined the micro riblets
on the Ti-6Al-4V surface with the rolling process, and the height and
Sharkskin Sharkskin
Coating Scales
Cured coating
Fig. 3.24. Elastic demolding of shark skin.
Demolding direction
Rolling
Semi-cured Roller Scales

Coating Enlarged
Fig. 3.25. Model of rolling shark skin surface.
width of the riblets was 150 µm and 340 µm respectively. Hirt et al.34
explored the method of rolling the riblet structure on a large area, and
additionally, the feasibility of a new winding concept for the continuous
patterning of rolls with small negative riblet structures was pointed out.
In previous studies, due to the complexity of the shark skin surface,
the direct bio-replicated imprinting method was put into application,
the flat polymethyl methacrylate (PMMA) plate used as a negative
template is heated to glass state, the biological shark skin template is
stacked on it scale side down, and isostatic pressure is applied on the
template until the coating is cured completely. The elastic demolding
of shark skin can be carried out, as shown in Fig. 3.24. In order to
realize manufacturing of continuous shark skin in large area, the rolling
method is the best selection as illustrated in Fig. 3.25.32 As far as the
scale is concerned, its size is about 0.1 mm × 0.1 mm, and it is so tiny
with respect to the diameter of the shark skin roller and can be seemed
as vertical demolding, but the coating is semi-cured and not shaped
completely, so it will be destroyed by the wedge angle on the back of
scale, which can lead to the failure of the rolling process, therefore, the
new method should be explored.
Sharkskin
Sealing of wedge angle on back
Fig. 3.26. Demolding model of shark skin surface.
Fig. 3.27. Surface of biological shark skin.
As illustrated in Fig. 3.26, if the wedge angle on the back of the

scale is seamed completely, the rolling process can be successfully carried
out without the semi-cured coating being destroyed. How to avoid
the influence of the wedge angle on the scale’s back is progressively
becoming the crucial problem in the bio-replicated rolling process.
Fortunately, the photo lithography process can realize the function —
the positive photoresist can be cured by baking, and it can then
be decomposed into some substance by UV light with the given
wavelength, which can be rinsed by developing solution and deionized
water. But the shark skin scales are transparent, as shown in Fig. 3.27.
If the photo lithography process is carried out without any protection,
Pretreatment Fixation Sputtering Lithography Roller Rolling Turnover
Fig. 3.28. Process of fabricating continuous vivid shark skin by rolling.32
the cured photoresist under the scale will also be decomposed, and
therefore, the sputtering step should be adopted to ensure the scales
are lightproof.32
The process of fabricating continuous shark skin with the original
morphology by the bio-replicated rolling method is illustrated in
Fig. 3.28; the main steps are as following: (1) For purpose of holding
the biological prototype and enhancing the mechanical strength, the
pretreatment of cleaning, chemical fixation, re-cleaning, dehydration
and desiccation is first carried out; (2) To ensure the good planeness
of the shark skin surface, it should be fixed on a flat plate; (3) Sputter
the metal nano-particles on the shark skin, the purpose of which is to
ensure the scales are completely lightproof; (4) Carry out the photo
lithography step, which involves spin coating, baking, rinsing and so
on; (5) Manufacture of the biological shark skin roller; (6) When the
coating reaches the time-zone of best plastic, the rolling process is
performed, and the negative template can be received after it is cured
completely. (7) The continuous vivid shark skin can be obtained after
turnover forming by casting silicon rubber.
In the sputtering process, the vacuum magnetron sputtering instru-
ment of JF-4500 is applied, as shown in Fig. 3.29, when the vacuum
degree in the working room is descending to 5×10−3 Pa, the sputtering
step is performed immediately, and the particles of metal (Ni) will cover
the shark skin. The entire sputtering time is over 60 minutes, and the
image of the shark skin after sputtering is shown in Fig. 3.30 with the
relevant parameters given in Table 3.2.
The depth of the sputtering metal layer is less than 1µm, so the
morphology on the shark skin cannot be affected. The positive photo-
resist PR1-12000A (produced by Futurrex Corporation) is prepared on
the shark skin by spin-coating technology at the speed of 400r/min for
60s, and then they are baked at the temperature of 120◦ C for 180s and
exposed to the UV light of 436nm wavelength for about 6 min, and they
can be rinsed by developing solution and deinoized water; the depth
Fig. 3.29. Sputtering equipment (JF4500).
Fig. 3.30. Image of shark skin after sputtering.
of lithography is more than 20 µm. After removing the photoresist

completely, the shark skin with the wedge angle on back being sealed
can be obtained, and the SEM image and 3-D morphology are shown
in Fig. 3.31.
After the photo lithography process, the biological shark skin is
moist and soft, and can be pasted on the surface of the metal roller
Table 3.2. Parameters of Sputtering (spell out ‘Ar’).
Pressure Flowing rate of Ar Voltage Current Purity of Ar Time
0.4Pa 40L/min 350–360V 0.24–0.25A >99.9% >60 min
Fig. 3.31. SEM image of shark skin and 3-D morphology after lithography.
easily. However, it is prone to shrinkage and deformation in the course

of drying, therefore, the 704 silicon rubber and straps rich in elasticity
are used to ensure the good pasting effect and to join them closely,
and the smooth and soft membrane of plastic is covered on the shark
skin for protection at the same time. Put them in the drying box until
the shark skin is dried and the binder of 704 silicon rubber is cured
completely, and the roller of shark skin can be obtained by wiping out
the protection membrane and wraps, as shown in Figs. 3.32 and 3.33.32
For the given AW-01 epoxy resin, when the coating has been curing
at a temperature of 45◦ C for 145–150 minutes, it has the best plasticity,
so the rolling step should be carried out at this time-zone. The diameter
of the roller is 6 cm, and the rolling velocity is about 0.3 m/min.
After the rolling process, the coating can be cured completely at room
temperature, and the negative template of continuous vivid shark skin
can be obtained — the SEM image and 3-D morphology are shown in
Fig. 3.34. Then the continuous vivid shark skin can be fabricated by
casting silicon rubber.
The lithography is one of the crucial steps to realize the rolling
process, so it is necessary to analyze the forming quality and effect.
Fig. 3.32. Roller of shark skin with protection.
Fig. 3.33. Roller of shark skin.
Only is the little interference of sealing produced under the biological

scale, the photolithography step can be seemed as successful, that is the
angles between edge and lateral sides of scales must be more than 90◦ .
As shown in Fig. 5.35, the angles are all between 90◦ to 95◦ ; therefore,
the good lithography effect is guaranteed.32
Fig. 3.34. SEM image of negative template and 3-D morphology.
92.9˚
92.5˚
93.0˚
92.7˚
Fig. 3.35. Analysis of removing wedge angle on back of scale.
The rolling forming precision is one of the most important

evaluation factors. But it is very difficult to find the same scale on
the biological shark skin roller and negative template. Considering that
the structure of placoid scales lies only on their location and species, the
scales from the same part of a piece of shark skin have approximately
the same structure. Therefore, the randomly sampled shark skin scales
and model from the same part are technically comparable.
Fig. 3.36. Cross-section curves of biological scale and negative template.
The cross-section curves of biological scale and negative template

with the similar shape are shown in Fig. 3.36. The depth of the
biological scale and negative template are 14.5 µm and 12.9 µm
respectively, and the widths are congruously about 115 µm, so the
forming precision on the depth direction is about 89%, and there is
almost no deformation on the width direction. It is concluded that the
continuous vivid shark skin with good forming effect can be fabricated
by the bio-replicated rolling method.
3.5. Enlarged Bio-Replicated Method
The direct bio-replicated forming method and other methods which

put the original shark skin into application as replica template to
1:1 transfer surface morphology has been widely applied for drag
reduction. However, if the application environment differs from the
shark’s environment, the drag-reduction function could be eliminated,
or it could lead to adverse results. Pan Junfeng et al.35 explored the
large-scale equal-proportional amplification bio-replication of shark
skin based on solvent-swelling PDMS.
For the purpose of realizing the large-scale solvent-swelling forming
of the shark skin surface, controlling the swelling of polydimethyl-
siloxane (PDMS) had the significant importance of accurately amplify
the natural functional surface micro-morphology, and the bio-scaling
forming process mainly included the following steps: the molding of the
PDMS, the large-scale swelling of the PDMS in gaseous n-hexane and
the formation of amplified shark skin as shown in Fig. 3.37. The second
step, the swelling of the PDMS mold, was the most important for
accurate bio-scaling formation. The whole experiment was conducted
Fig. 3.37. Steps of large-scale solvent-swelling forming of shark skin.
Fig. 3.38. SEM of original and amplified shark skin surface.
at 28◦ C; the saturated vapor pressure of n-hexane at this temperature

was set as 24 kPa, indicating that the highest n-hexane concentration
was approximately 0.85 × 10−3 g ml−1 .
Figure 3.38(a) shows the original real shark skin surface, and
Fig. 3.38(b) shows the second-time amplified shark skin surface
adopting the large-scale solvent-swelling forming method.35,36 The

replication precision testing experimental results indicated that about
140% solvent-swelling ratio was achieved at one-time amplification.
To validate the amplified effect further, drag-reducing effect experi-
ments were performed in a water tunnel — the results showed that
the optimal drag-reduction velocity region could be adjusted by the
large-scale solvent-swelling bio-replication method, and the peak of
the drag-reduction ratio translational was varied by controlling the bio-
replication amplification.
3.6. Additive Low-Releasing Design

on Shark Skin
It has been well known that the addition of a tiny amount of high
molecular weight linear flexible polymer into a turbulent flow can
cause a considerable reduction in skin friction drag. With the high
molecular weight polymers giving better drag-reduction performance,
only parts per million levels of the polymers in the working fluid
suppress the formation of turbulent bursts in the buffer region and
in turn suppress the formation and propagation of turbulent eddies.
To achieve improved drag-reduction performance, it becomes more
necessary to efficiently combine drag-reduction riblets with polymer
additive.
Chen and Zhang37 explored the additive low-releasing with micro
pores design on vivid shark skin; the model is shown in Fig. 3.39.
Diffusion of polymer additive into the whole fluid via micro pores was
one integration approach of vivid shark skin with polymer additive.
The structural size and distribution of micro pores determined the
concentration of the polymer additive on the outer surface. The
polymer additive was diffused around the body surface through the
elaborated micro pores array under gentle external pressure, whose
schematics are shown in Fig. 3.40.
In the process of designing the additive low-releasing with micro
pores design on vivid shark skin, the diffusion of micro pores was opti-
mized first, including distribution, hole, and channel size. According to
Fick’s law of diffusion, the concentration C (x, y, z) of polymer additive
Fig. 3.39. Model of additive low-releasing with micro pores.
Fig. 3.40. Schematic drawing of polymer additive diffusion.
that is diffused through one micro pore can be given as37 :

m
· e −ux (y +z )/4xEx ,
2 2
c(x, y, z) = (3.1)
2πxEx
where Ex is the turbulent diffusion coefficient, m is the mass of polymer
additive diffused out per unit of time, and ux is the flow direction along
x of fluid.
Moreover, the aperture and channel size of the micro pores were
optimally determined to ensure that the minimal external pressure
diffuses sufficient polymer additives. The detailed structure of the drag-
reduction micro pores array was chosen as 13.5 mm × 5 mm; micro
pore diameter 0.8 mm; main channel 11 mm; and branch channel
Fig. 3.41. Micro wire mold fabrication of the micro pores array.
Fig. 3.42. Drag reduction based on additive low-releasing.
6 mm. Overlying the bio-replicated shark skin with micro pores array
upon a fluid channel layer which was fabricated via micro wire mold
was one effective way to obtain synthetic drag-reduction skin based on
polymer additive diffusion, as shown in Fig. 3.41. The drag-reduction
experiments were carried out in a water tunnel, and the maximum drag-
reducing efficiency could surpass 70%, as shown in Fig. 3.42, which was
very striking.37
3.7. Direct 3-D Imprinting Method

The shark possesses a low-resistance surface; its skin is covered with
many small scales arranged in a diamond pattern. The extended
direction of the scale is proximately parallel with its swimming direction,
and the groove tips of the scale can stick out of the viscous sublayer,
which will inhabit the occurrence of turbulence and reduce wall friction.
In manufacturing vivid trans-scale shark skin, the digital model should
be built first, and the key steps are as following38 : (1) Highly accurate
scanning of single biological scale; (2) Data processing and analyzing;
(3) Building three-dimensional digital model of single scale; (5) 3-D
printing vivid scales. Because all the relevant data for the model is
received from the real biological shark skin prototype, the method can
realize the vivid enlargement of shark skin scales. Shark skin scale digital
models can be amplified and stretching deformed without any error on
any direction. The different digital models of sharkskin scales are shown
in Fig. 3.43.
To meet the requirements of drag reduction in fluid medium with
high viscosity, the shark skin should be manufactured in a larger scale.
Although the “shark skin effect” is well known, not all people have
an intuitive understanding and awareness of shark skin. Additionally, a
shark skin scale is too tiny to be observed with the naked eye; therefore,
it is an urgent to be solved problem. Mold manufacturing is a high-cost
and complex step in the traditional machining process, especially for
parts with complex shape. As mentioned above, shark skin has complex
surface morphology, so solid freeform fabrication (SFF) technology
should be considered first, and it is applied in this chapter.
Fig. 3.43. Digital models of shark skin (a) original prototype of scale (b) stretching
on length direction (c) stretching on width direction (d) stretching on depth direction.
SFF technologies are manufacturing/prototyping technologies

that are characterized by layer-by-layer addition of material to fabricate
components. These techniques are also known as layered manufactur-
ing and rapid prototyping (RP). The layer-by-layer building approach
allows significantly more complex parts to be built in one fabrication
step than was previously possible thus simplifying process planning.
SFF technology can therefore automate the process planning and
fabrication of a part under computer control so that the only input
needed is a solid model of the part. Rapid prototyping is a computerized
fabrication technology that additively builds highly complex three-
dimensional physical objects layer by layer using data generated by
computer, for example CAD or digital graphic. Three-dimensional
printing (3DP) is one of such technologies that employ printing
technology for processing powder materials; it has the characteristics
of being low cost, short cycle, modified simple, dimensionally stable
and so on. This technology is divided into three types according to the
different spraying materials: bonding material forming, photo-sensitive
materials, molten material forming.
The applied machining equipment in the work is the 3-D printing
machine (Eden 250) produced in Israel; the image of Eden 250 and its
working diagram is shown in Fig. 3.44.38
The printing head of Eden 250 is similar to that in a linear printer —
it slides along the X axis in the printing process, and a super-thin layer
of photosensitive resin is paved in the molding room. After paving each
layer, the UV ray is emitted immediately; each layer of photosensitive
X Axis
Priating head
Y Axis
UV Ray
Forming material
Supporting material
Z Axis
Working table
Fig. 3.44. Eden 250 and its working diagram.

resin is cured and hardened rapidly. Therefore, this step can prevent
some unnecessary post-processing courses, but others. After each layer
of resin is printed, the working table will be dropped down very
precisely, and the printing head can work layer by layer sequentially,
until the completion of the prototype parts is done. Accurate tools can
ensure the harmonious operation of the printing heads and the same
amount of material can be sprayed simultaneously, which can create a
smooth and flat surface for the prototype part. In the forming process,
two different photosensitive resin materials are put into application:
one is mold forming material of parts, and the other is the supporting
material used to support the models. The skeleton of digital models is
arranged ahead of pre-program matched the complex parts. After the
mold is completely done, the supporting materials can be removed with
only a water nozzle, leaving a smooth and perfect surface.
The forming resolution of Eden 250 in x, y, z direction is
respectively 600dpi, 300dpi and 1600pdi. During fabrication, the
printer head is used to print thin layers of powder following the object’s
profile as generated by the system computer, by the relative movements
of print heads and carrier, the part is realized manufacture. For the
manufacture of vivid shark scales, the forming material is Full Cure
720, and the supporting material is Full Cure 705. Based on the models
built in the former steps, the manufactured physical scales are shown
in Fig. 3.45.38
Fig. 3.45. Physical vivid enlarged shark skin scales (a) enlarged 100 times on all
directions; (b) enlarged 100 times on the direction of length and 50 times on others.
3.8. Flexible Manufacture of Real Shark Skin

Prof. George V. Lauder and his team at Harvard University17,18 studied
the self-propelled swimming speed of different surfaces, and it was
concluded that flexible sharkskin foils actually showed a substantial
improvement in swimming performance of an average of 12.3%. Wen
and Lauder17,18 created a high-resolution view of the shark skin surface
by scanning the skin, then they zoomed in on a single identical to build
a detailed model of its structure, and hence reproduced it thousands of
times in a computer model.
Lauder et al. obtained the fresh dead specimen of a male shortfin
mako shark from fishermen near Boston, and a small area was extracted
using dissection instruments and carefully cleaned with water. They
picked a single representative skin identical from the scan on the sample
at a resolution of 1.583 um (proper unit) and constructed a 3-D model.
The reconstructed identical model was duplicated and linearly arrayed
in a controlled pattern on a membrane substrate, as shown in Fig. 3.46.
Lauder et al.17,18 adopted the Object Connex500 3-D printer
to manufacture the synthetic shark skin membrane, in which two
different materials were put into application to fabricate the denticles
and membrane substrate, and the supporting material was carefully
removed by water jet after the entire shark skin membrane was printed.
Fig. 3.46. 3-D reconstructed model of shark skin scale.

Fig. 3.47. Images of the fabricated synthetic shark skin.
Membranes in curved and flattened states are shown in Figs. 3.47 (a)
and (b) respectively. A single synthetic identical on finger is shown in
Fig. 3.47 (c). The experiments in water tunnel illustrated that at heave
frequency of 1.5 Hz and amplitude of ±1 cm, swimming speed was
increased by 6.6% and the energy cost-of-transport was reduced by
5.9%.17,18
3.9. Large-Proportional Shrunken

Bio-Replication of Shark Skin
As the classic product of natural selection, sharks have evolved unique
hierarchical surface morphology on their skins for survival in their
living environment. Since adaptation to the living environment is the
sole driving force of evolution, sharks show the best drag-reduction
function only in their living environment. In fact, the natural surface
function is impaired or even reversed when they stay far from their
living environment. Bio-replicated shark skin performs with a maximum
drag reduction of about 12% only when the relative fluid flow is
around 5 m/s, which agrees well with the normal swimming speed
of sharks. The maximum drag reduction commonly declines with
an improvement of the fluid flow speed beyond 5 m/s. Therefore,
adjusting the surface morphology of the shark skin to accommodate a
different fluid flow environment is necessary to enhance the industrial
applications of the shark skin effect. According to Walsh’s1 observation
that the maximum drag reduction of micro-riblets is dependent on
optimum spacing s + and h + , which is affected by the depth-to-width
ratio h/s, equal-proportional shrinkage of the surface morphology is
Fig. 3.48. Process of equal-proportion shrunken bio-replication.
one way to guarantee that the optimal drag-reduction function of shark

skin is preserved at much higher speeds of fluid flow.
The proposed large-proportional shrunken bio-replication process
is divided into three sub-steps as presented in Fig. 3.48: preparation
of the shark skin negative mold using UV-curable material, shrinking
of the UV-curable mold and the bio-replication formation of shark
skin from the shrunken negative mold.39 Accurately controlling the
shrinkage of the UV-curable material mold is most important in
achieving large-proportional shrunken bio-replication of the shark skin.
Degassing is conducted for 20 min in a vacuum chamber. The distance
between the sample of shark skin and the LED UV lamp (power 3
mow cm−2 ; Wavelength 395 nm, Shang Hai Machine Optoelectronic
Technology Co Ltd) is about 20 cm and the irradiation time is 1 min.
All the processes are conducted at 20◦ C.
Fig. 3.49. SEM images of the original shark skin and its corresponding shrunken
replication when the diluent evaporates.
The quality of surface morphology is another important factor for

bio-replicated shark skin. SEM photos of original shark skin template
and large-proportional shrunken bio-replicated sharkskin are compared
in Fig. 3.49. In order to obtain the same scale, a micro-hole was drilled
in the original shark skin template initially as the anchor point. We can
see that the bio-replicated shark skin scale from the UV-curable material
is remarkably shrunken as compared with the corresponding original
shark skin scale, whose shrinking ratio is also about 23%. Moreover, the
microstructures of the shark skin scale are well preserved even though
the large-proportional shrinkage of the bio-replicated sharkskin occurs
after curing, indicating the large-proportional shrunken bio-replication
based on UV-curable materials can provide a desirable microscale
surface quality. Moreover, the accurate microstructure of the shark skin
is hardly observed as the diluent content exceeds about 50 wt%.39
3.10. Bio-Replication Coating Device Based

on UV Curable Paint
Micro-embossing or -imprinting has become more prevalent for low-
cost and mass-production of micro- or nano-structures. However
micro-embossing or -imprinting was generally limited to regular 2-D
micro-patterns, not suitable for complex 3-D micro-patterns like shark
skin due to considerable difficulty in demolding. In order to make
transfer of natural shark skin onto a free-form surface possible, the
flexible bio-replication coating approach of shark skin is proposed
as shown in Fig. 3.50, whose mechanism is based upon micro-
embossing.40
Fig. 3.50. Schematics of natural function surface morphology coating technology.
Fig. 3.51. Natural function surface morphology bio-replication coating instrument.
The mechanism of large-area bio-replication coating instrument

is shown in Fig. 3.51. A long strip of the soft negative mold with
micro-morphology of shark skin tightly enwraps three plastic wheels,
and an ultraviolet lamp is set inside. The soft negative mold strip is
500 mm in length and 185 mm in width. The distance between the
strip and the UV lamp (power 300W and wavelength 395nm LED UV
lamp, Shanghai Machine Optoelectronic Technology Co. Ltd) is set
at 20 cm. An ultrasonic vibration PZT is mounted on the wheel closes
to the spraying side of UV curable paint. High-frequency ultrasonic
vibration (frequency 20kHz and amplitude 1um micro) is generated
throughout coating processes to assist UV curable paint filling in soft

negative mold. This is because high frequency vibration has significant
liquefaction effect. Curing rate of UV curable paint is greatly dependent
on the power of the UV lamp, the thickness of the soft negative mold
and the depth of UV curable paint. After minimum curing time Tin , is
determined by various coating experiments, the relation between the
distance L between two wheels and the speed of coating ν is decided
as v = Tmin
L 40
.
To replicate shark skin in large area, a piece of fresh shark skin
(200 mm × 550 mm) with good morphology was selected as the shark
skin template and pretreated as follows. Shark skin template is rinsed
three to five times in clear water, and then two to three times in
deionized water to remove the blood or other impurities. After being
flattened and nailed to a rigid plate, the shark skin template is put
into 2.5% glutaraldehyde (C5 H8 O2 ) over 3h to maintain its original
micro-morphology via chemical fixation. The chemical fixed shark skin
template is rinsed 1h by three to four times change of 0.1 mol/L
pH7.2 phosphate buffer solution. The residual solution upon template
is rinsed off three to four times in the clear water and deionized
water. Gradient dehydration process is conducted to prevent template
deformation from excessive water loss in drying. Chemical fixed shark
skin samples are immersed in a concentration of 30%, 50%, 75%, 80%,
95% and 100% ethanol solution to dehydrate step by step, each for
15–30 min. After keeping inside a drying oven over 12 hours at 60◦ C,
the pretreatment of the shark skin template is completed. The SEM of
pretreated shark skin is shown in Fig. 3.5240 ; the soft negative mold
is fabricated via soft lithography, in which the pretreated shark skin is
Fig. 3.52. SEM images of shark skin, soft negative mold and bio-replication coating.
taken as a template of the soft lithography. Pouring of PDMS upon

the template and de-molding of the soft negative mold from template
are the predominant procedures of soft lithography. In order to ensure
PDMS fully pouring into the micro-cavities of the shark skin template,
all the processes of soft negative mold preparation are conducted in a
vacuum chamber to squeeze air from the cavities of the shark skin.
3.11. Conclusion
In this chapter, several novel methods to fabricate vivid shark skin

surfaces were illustrated in detail, and the experiments on drag-
reduction efficiency were also conducted in water tunnel to validate
the forming quality and effect. The results were all very promising and
satisfactory, and the main conclusions can be obtained as follows:
(1) Biomimetic and bio-aided fabrication methods are very effective for
exploiting the application of shark skin, even if the drag-reduction
mechanism is not understood completely and absolutely.
(2) The mucus secreted by shark skin is one of the most important
factors leading to the high drag-reducing efficiency, but how to
maintain the low-releasing effect for a long time still needs to be
explored.
(3) The drag-reduction effect of vivid shark skin is better than that
of the simplified and straight micro-grooved surface, in which
the attack angle of scales can produce some certain influence.
Therefore, it is very necessary to explore the ways to manufacture
the vivid shark skin.
(4) How to protect the biomimetic drag-reduction surface in the
process of sustainable application is another issue to be investigated
and resolved.
References
[1] Walsh MJ (1983) Riblets as viscous drag reduction technique. AIAA J 21(4):
485–486.
157–171.
[4] Bechert DW, Bruse M, Hage W, et al. (1997) Experiments on drag-reducing
59–87.
[5] Reif WE, Dinkelacker A. (1982) Hydrodynamics of the squamation in fast
swimming sharks. Neues Jahrbuch fuer Geol Palaeontol 164: 184–187.
[6] Luo YH, Zhang DY. (2012) Experimental research on biomimetic drag-
reducing surface application in natural gas pipelines. Oil Gas-Eur Mag 38(4):
213–214.
[8] Zhang DY, Luo YH, Chen HW, Jiang XG. (2011) Exploring drag-reducing
grooved internal coating for gas pipelines. Pipeline Gas J 238(3): 58–60.
[9] Bhushan B. (2003) Adhesion and stiction: mechanisms, measurement tech-
niques and methods for reduction. J Vacuum Sci Technol B 21(6): 2262–2296.
[10] Bhushan B. (2009) Biomimetics: Lessons from nature-an overview, Philos Trans
R Soc A 367: 1445–1486.
[11] Bhushan B, June YC, Koch K. (2009) Micro-, nano- and hierarchical structures
for superhydrophobicity, self-cleaning and low adhesion. Philos Trans R Soc A
367: 1631–1672.
[12] Haecheon C, Parviz M, John K. (1993) Direct numerical simulation of turbulent
flow over riblets. J Fluid Mech 255: 503–539.
[13] Liu KN, Christodoulout C, Ricciust O, Joseph DD. (1999) Drag reduction in
pipes lined with riblets. AIAA J 28(10): 1697–1698.
[15] Li X, Jun Cai J, Zhang DY. (2010) Study on the Manufacturing Method of
the Biomimetic Drag Reducing Morphology Replication Mold. Adv Mater Res
97–101: 2533–2537.
[16] Viswanath PR. (2002) Aircraft viscous drag reduction using riblets. Prog Aerosp
Sci 38: 571–600.
[17] Wen L, Weaver JC, Lauder GV (2014) Biomimetic shark skin: design, fabrication
[18] Johannes O, George VL. (2012) The hydrodynamic function of shark skin and
two biomimetic applications. J Exp Biol 215: 785–789.
[19] Zhang DY, Luo YH, Li X, et al. (2011) Numerical simulation and experimental
study of drag-reducing surface of a real shark skin. J Hydrodyn 23(2): 204–211.
[20] Luo YH, Liu YF, Zhang D, Ng EYK (2014) Influence of morphology for drag
reduction effect of sharkskin surface. J Mech Med Biol 14: 1450029.
[21] Luo YH. (2015) Recent Progress in Exploring Drag Reduction Mechanism of
Real Sharkskin Surface: a Review. J Mech Med Biol 15(3): 1530002.
[22] Lang AW, Motta P, Hidalgo P, Westcott M. (2008) Bristled shark skin: A
microgeometry for boundary layer control? Bioinspir Biomim 3: 046005.
[23] Lang AW, Motta P, Habegger ML, et al. (2011) Shark skin separation control
mechanisms, Marine Technol Soc J 45: 208–215.
[24] Zhang GM, Zhang J, Xie GY, et al. (2006) Cicada Wings: A Stamp from Nature
for Nanoimprint Lithography. Small 12(2): 1440–1443.
[25] Losic D, Mitchell J, Lal R, Voelcker N. (2007) Rapid Fabrication of Micro- and
Nanoscale Patterns by Replica Molding from Diatom Biosilica. Adv Fun Mater
17: 2439–2446.
[26] Nagaraja P, Yao DG. (2007) Rapid pattern transfer of biomimetic surface
structures onto thermoplastic polymers. Mater Sci Eng C 27(4): 794–797.
[27] Han X, Zhang D. (2008b) Study on the micro-replication of shark skin. Sci
China Technol Sci 51(7): 890–896.
[28] Han X, Wang J. (2011) A Novel Method for Fabrication of the Biomimetic
Shark-Skin Coating. Adv Mater Res 239–242: 3014–3017.
[29] Luo YH, Liu YF, Zhang DY. (2015) Hydrodynamic Testing of a Biological
Sharkskin Replica Manufactured Using the Vacuum Casting Method. Surf Rev
Lett 22(2): 1550030.
[30] Zhang DY, Li YY, Han XX, et al. (2011). High-precision bio-replication of
synthetic drag reduction shark skin. Chinese Science Bulletin 56(9): 938–944.
[31] Luo YH, Zhang DY, Liu YF, et al. (2015). Chemical, Mechanical and Hydro-
dynamic Properties Research on Composite Drag Reduction Surface Based
on Biological Sharkskin Morphology and Mucus Nano-long Chain. Journal of
Mechanics in Medicine and Biology 15(5): 1530084.
[32] Luo YH, Zhang DY. (2013). Investigation on fabricating continuous vivid
sharkskin surface by bio-replicated rolling method. Appl Surf Sci 282: 370–375.
[33] Fritz K, Feldhaus B, Sebastian M. (2007) Development of an incremental rolling
process for the production of defined riblet surface structures, Engineering
Research and Development 1(3): 233–237.
[34] Hirt G, Thome M. (2007) Large area rolling of functional metallic micro
structures. Production Engineering Research and Development 1(4): 351–356.
[35] Pan JF, Chen HW, Zhang DY, et al. (2013) Large-scale solvent-swelling-
based amplification of microstructured sharkskin, Journal of Micromechanics and
Microengineering 23: 075018.
[36] Chen HW, Zhang X, Zhang DY, et al. (2013) Large-Scale Equal-Proportional
Amplification Bio-Replication of Shark Skin Based on Solvent-Swelling PDMS.
J Appl Polym Sci 130(4): 2383–2389.
[37] Chen HW, Zhang X, Da C, et al. (2014). Synthetic Effect of Vivid Shark Skin and
Polymer Additive on Drag Reduction Reinforcement. Advances in Mechanical
Engineering 2014: 42571.
[38] Luo YH, Zhang DY, Chen HW. (2012) Research on Manufacturing Vivid Trans-
scale Shark Skin Surface and Drag-Reducing Effect Simulation, Adv Sci Lett
5(1): 49–55.
[39] Chen HW, Che D, Zhang X, et al. (2015) Large-proportional shrunken bio-
replication of shark skin based on UV-curing shrinkage. J Micromech Microeng
25: 017002.
[40] Chen HW, Zhang X, Ma LX, et al. (2014) Investigation on large-area fabrication
of vivid shark skin with superior surface functions. Appl Surf Sci 316: 124–131.
Chapter 4
Different Characteristic Analysis

of Drag-Reducing Surface
with Biological Morphology
J. Wang,∗ Y.T Zhao,∗ L. Zhang,∗ Y Luo,† E.Y.K. Ng‡


Natural selection, survival of the fittest — through millions of years

of evolution, different animals have developed their own functional
surfaces, for example, the shark has become one of the fastest swimming
animals in the ocean, and it is very well-known for the shark skin effect,
especially for the “shark skin swimsuit.” Due to its superior properties
in drag reduction, anti-wear, self-cleaning and so on, investigations
on its essential mechanisms and fabricating methods have attracted
much attention from all over the world, and the achievements have
been widely put into application in industry, agriculture, transportation,
airspace and so on, and profits have been obtained so far. In this chapter,
the method of fabricating artificial composite drag-reduction surface
based on biological shark skin morphology and mucus nano-long
113
chain is investigated and studied, and the chemical, mechanical and

hydrodynamic properties are explored from different angles in depth.
Experimental results in water tunnel revealed that the drag-reducing
efficiency could surpass 20% with smooth skin as reference. The drag-
reduction mechanism is systematically explained and discussed from
different angles. This has significance for understanding the status of
recent research and for expanding the use of shark skin in the fluid
engineering.
4.1. Replicating Precision of Biological

Shark Skin
In the previous study, the process of researching the micro-replicated
precision of shark skin was done according to the theory that “the
structure and shape of shield scales depend only on their location
and type, and the characteristics of the scale structure in the same
place should be very similar.” The statistical method was used, and
the randomly sampled shark scales and scale mold from the same
part were technically comparable. But the shark skin scales are very
small, about 0.1 mm × 0.1 mm–0.2 mm × 0.2 mm, moreover, they are
always overlapping with each other. Therefore, even in a very small
area on the shark skin, it is difficult to find out two absolutely uniform
scales. Due to this, the statistical method will inevitably produce some
errors, and the evaluating method should be improved and further
modified.1−3
In this chapter, the marking key point method is adopted to search
for the corresponding biological shark skin scale and negative template.
An obvious and visual mark (such as drilling a micro hole in the
biological sharkskin) is made on the biological shark skin, and the
corresponding mark can be located on the negative template.1 Based on
the visual marks, the corresponding biological scales and negative scale
template can be matched, even though it is so small. In gaining precise
and sufficient data, super accurate scanning method is done on the
corresponding biological scale and negative template scale1 ; the images
are illustrated in Figs. 4.1 and 4.2. The highly precise instruments
include the Phase Shift MicroXAM-3D, with root-mean-square (RMS)
Different Characteristic Analysis of Drag-Reducing Surface 115
Fig. 4.1. Single scale of shark skin.
Fig. 4.2. Corresponding negative template.
repeatability of 1 nm, minimum vertical scanning resolution of 0.1 nm

and calibration accuracy of less than 0.1%.
For the replicating precision analysis of the shark skin scales, the
width and depth of groove are the most important factors. The 2-D
Fig. 4.3. Biological scale and particular cross section curve.
Fig. 4.4. Corresponding negative template and cross section curve.
image and parameters of the marked line can be obtained, as shown in

Figs. 4.3 and 4.4; the same location of corresponding scales should be
ensured. It can be seen that the width of groove on corresponding scales
is uniformly 115 µm and the depth of groove on the biological are 10.6
µm and 11.7 µm respectively, that on the negative template are 10.4 um
and 10.9 um respectively, so the replicating precision can be calculated
as 115/115 = 100%, 10.4/10.6 = 98.1% and 10.9/11.7 = 93.2%.
The corresponding cross section curves at the same location can
be exactly determined by moving the movable marked lines1,4 ; the
replicating precision on the whole scale can be obtained. As illustrated
in Table 4.1, the replicating precision on the vertical direction is greater
Table 4.1. Replicating Precision of Single Scale.
Biological scale unit: µm Negative template unit: µm Replicating precision

Depth Depth Depth Depth Depth Depth
Width (l) (r) Width (l) (r) Width (l) (r)
111.906 — — 111.9105 — — 100% — —

111.906 10.6 11.5 111.9105 9.99 10.9 100% 94.25% 94.78%
111.906 10.8 11.7 111.9105 10.3 11.3 100% 95.37% 96.58%
114.636 10.8 11.9 114.641 10.5 11.8 100% 97.22% 99.16%
114.636 10.9 12.0 114.635 10.8 11.8 100% 99.08% 98.33%
114.636 11.1 12.3 114.635 10.9 11.5 100% 98.20% 93.50%
114.636 11.4 12.5 114.642 11.2 11.6 100% 98.25% 92.80%
114.636 12.0 12.7 114.635 11.5 11.7 100% 94.26% 92.13%
114.636 12.0 12.8 114.635 11.6 12.2 100% 96.67% 95.31%
117.365 12.6 12.9 117.365 12.3 12.5 100% 97.62% 96.90%
117.365 13.6 13.6 117.365 12.9 12.6 100% 94.85% 92.65%
117.365 13.6 13.7 117.365 13.1 13.3 100% 96.32% 97.08%
114.636 14.3 13.9 114.645 13.9 13.4 100% 97.20% 96.40%
114.636 14.4 14.8 114.645 13.8 14.0 100% 95.83% 94.59%
111.906 15.1 14.9 111.912 14.6 13.9 100% 96.69% 93.29%
111.906 15.5 15.2 111.915 15.0 14.5 100% 96.77% 95.39%
106.447 12.1 13.0 106.454 11.9 12.6 100% 98.35% 96.92%
106.447 12.1 12.9 106.454 11.7 12.4 100% 96.69% 96.12%
95.533 11.9 12.9 95.533 11.3 12.4 100% 94.96% 96.12%
90.078 11.8 12.2 90.075 11.2 11.9 100% 94.92% 97.54%
than 95%, and at the meanwhile, there are no forming errors on the
horizontal direction.
The replication precision of single scale has been studied in detail,
and the precision is very high. However, for the surface of fluid
engineering, the forming quality of large area is of greater concern,
so it should be taken in more consideration. For research on the
replication precision of regional scales, the marking key-point method is
also applied; the corresponding regional biological scales and negative
template are shown in Figs. 4.5 and 4.6. In the two figures, the
biological scale (1) corresponds to the template scale (1’), and other
scales also correspond by analogy. In the course of bio-replicated
forming of shark skin, although the scales are very hard and it is difficult
to produce deformation, because of the flexibility of the biological
tissue attached to the scales, extension and deformation under the
Fig. 4.5. SEM image of regional scales.
Fig. 4.6. SEM image of corresponding negative template.
external force are produced, and the relative positions of the different
scales will also change. Therefore, the variation of relative position
between different scales is the main factor in assessing the replication
precision of large-area shark skin.
By analyzing the distances changes of corresponding scales, the
forming quality of regional scales can be evaluated. The measured
distances of different scales are shown in Figs. 4.7 and 4.8, and the
Fig. 4.7. Different distances of biological scales.
Fig. 4.8. Different distances of template scales.
corresponding distances of different scales are marked in the two

pictures; the data of distances and replication error are tabulated in
Table 4.2.
The shark skin scale is of enamel microstructure and is very hard,
so it is difficult to produce deformation. In the replicating process of
biological shark skin surface, there are two main factors causing: One
Table 4.2. Distances of Corresponding Scales and Replication Error.
No. Biological scales [µm] Template scales [µm] Replication error [%]
1 177.41 183.25 +3.29

2 270.03 275.69 +2.10
3 359.83 376.20 +4.55
4 399.48 406.15 +1.67
5 117.97 123.85 +4.98
6 176.11 183.08 +3.96
7 198.21 200.64 +1.23
8 266.73 273.47 +2.53
9 132.55 127.50 −3.81
10 236.37 236.50 +0.055
11 159.68 155.59 −2.56
12 161.60 166.45 +3.00
13 158.57 156.04 −1.60
14 348.41 357.13 +2.50
15 216.01 223.75 +3.58
16 112.60 116.36 +3.34
17 147.13 150.14 +2.05
18 121.07 125.35 +3.54
Fig. 4.9. Schematic drawing of epoxy resin deformation.
is the deformation of epoxy resin during the curing process. The epoxy
resin is constrained by the substrate in the horizontal direction, so it
almost cannot produce any deformation in the width direction, but
there is no constriction in the depth direction, and deformation will be
produced, as illustrated in Fig. 4.9. The other reason is that although
Fig. 4.10. Schematic drawing of relative-position changes.
hard scales cannot produce deformation, they are embedded in the

epidermal tissue, where it is easy to produce extension and deformation,
therefore, the relative-position changes between different scales are
unavoidable, as shown in Fig. 4.10.
4.2. Chemical Properties of Artificial Composite

Shark Skin Surface
From the drag-reduction viewpoint, the micro-groove of composite
bio-replicated hierarchical shark skin could reduce energy consumption
induced by longitudinal vortex, and the long-chain extending into
buffer layer could inhibit the radial pulsation of the fluid micelle to
decrease the energy consumption. As a result of such composite effect,
the composite bio-replicated hierarchical shark skin has excellent drag-
reduction effect superior to only micro-groove drag-reduction surface.
The composite bio-replicated hierarchical shark skin was formed by
the graft co-polymer of water-based epoxy resin and PAM. Graft
copolymerization, one of the main methods for chemical modification,
is one kind of reaction to bind appropriate branch chain or functional
lateral groups to the macro-molecule chain by chemical bond.5
In order to validate the grafting mechanism mentioned above, the
Fourier transform infrared spectrometer AVATAR 360 FT-IR is applied
to determine the infrared spectrum of water-based epoxy resin film
Fig. 4.11. Infrared spectrum of water-based epoxy resin film.
before and after grafting. Through spectrum, the functional groups

and structure of compounds can be identified from the functional group
region and fingerprint region, respectively.
Figure 4.11 shows the infrared spectrum of water-based epoxy
resin film before and after grafting. One sharp absorption peak near
1,666 cm−1 , i.e. the characteristic absorption peak of carbonyl groups
on amide groups, appears after grafting, which indicates the grafting
reaction generates a great number of amide groups. The magnitude
of obtuse round absorption peak near 3,397 cm−1 , which is the
characteristic absorption peak of hydroxyl groups, becomes larger after
grafting. That is to say, more hydroxyl groups were formed in the
grafting process. The characteristic absorption peak of epoxy groups
near 830 cm−1 declines after grafting, which implies that amount of
epoxy groups descend. Therefore, the grafting mechanism mentioned
above can be validated by appearance of carbonyl groups, increment of
hydroxyl groups and decrease of epoxy groups after grafting.
In order to further prove the correctness of the grafting mechanism,
experiments of infrared spectrum are also conducted for thin film
generated just by mixture of the graft co-polymer of water-based epoxy
Fig. 4.12. Infrared spectrums of thin film and PAM.
resin emulsion and PAM solution without curing agent. The infrared
spectrums of thin film and PAM are shown in Fig. 4.12 (a) and (b),
respectively. By comparison between them, we can see that double
peaks near 3,370 cm−1 , which are the stretching vibrational spectrum of
amino groups of amide groups, almost disappear and the characteristic
absorption peak of hydroxyl groups near 3,410 cm−1 arises. And the
carbonyl groups of amide groups whose absorption peak near 1,666
cm−1 in Fig. 4.12(a) is considerably lower than of Fig. 4.12(b). Thus,
amide groups and epoxy groups apparently conduct grafting reaction as
mentioned above. Therefore, the nano-long chain of DRA can be intro-
duced to micro-groove of bio-replicated shark skin by grafting copoly-
merization of water-based epoxy resin emulsion and PAM solution.
4.3. Mechanical Property of Composite

Drag-Reduction Shark Skin
If any accessory ingredient is not added or supplemented in the
artificial composite drag-reduction shark skin, the cured co-polymer
Fig. 4.13. Tensile testing experiment and testing samples.
completing the grafting reaction between waterborne epoxy resin and

polyacrylamide (PAM) is very brittle and easily broken. Therefore, it
is very necessary to add a certain proportion of the additives into the
above-mentioned co-polymer to improve the plasticity and flexibility.5
In this chapter, dibutylo-phthalate is selected to improve the mechanical
performance of composite drag-reduction shark skin.
To verify the correctness and feasibility of applying the dibutylo-
phthalate to improve the mechanical performance, the tension test of
artificial composite drag-reduction shark skin is carried out; the testing
equipment and testing samples are shown in Fig. 4.13. The 5565-5KN
electronic testing machine is used, and the maximum load is 500N, the
stretching speed can be set as 2mm/min, the load precision is 0.4%,
and when the sample is developing into tension fracture, the experiment
should end at once.5 When the amounts of additives are set as different
values, the relationship between tension stress and rate of deformation
can be obtained, as shown in Fig. 4.14. And the conclusion can be
reached that when the amount of dibutylo-phthalate is 0.75 g (which
means that the mass ratio is 0.75%), the composite shark skin has the
best mechanical performance.
4.4. Building of Water Tunnel

Large water tunnel is the common method for testing the drag of a
surface under flowing conditions, however, the testing fee is costly, so
it is necessary to construct the small tunnel.6−9
8
0
7
6
Tension stress/ MPa
3
1.25
2 1.5 0.25
0.5
1
1 0.75
0
0 5 10 15 20 25 30
Rate of deformation/ %
Fig. 4.14. Relationship between tension stress and rate of deformation.
To validate the drag-reducing efficiency of shark skin by micro-

rolling method, the experiment using water tunnel was tested in this
chapter. The schematic drawing and field image of the water tunnel
system are illustrated in Figs. 4.15 and 4.16, respectively. The system
includes water pump, water tank, pressure differential gauge, pipe
section, relief valve, throttle valve, butterfly valve, check valve and
so on. The drag resistance of different skins can be determined by
pressure difference between two measurement points, and the buffer
pipe section was set longer than 50 mm, which aimed to stabilize water
flow in front of the test pipe section. The inner wall of the test pipe
section is covered by prepared drag-reduction skins under testing.
The fluid experiment in water tunnel for validating forming quality
and the drag-reducing efficiency of artificial shark skin is discussed in
this chapter. The testing sample of shark skin is shown in Fig. 4.17.
In the process of testing drag-reduction effect, smoothness is the
principal requirement for the covering process to eliminate the adverse
effect of surface wrinkling and other defects on drag reduction. It can
be controlled by the electric cabinet, and the flowing velocity of the
water tunnel can be set by adjusting the flow meter. The whole testing
duration is about one hour.
Fig. 4.15. Schematic drawing of water tunnel.
Fig. 4.16. Field image of water tunnel.
When the velocity of the water tunnel is set to different values, the
drag reduction efficiency of artificial composite shark skin compared
with smooth skin can be obtained,10,11 as given in Table 4.3; the fitted
curve is shown in Fig. 4.18. It can be seen that the maximum drag-
reduction efficiency can surpass 20%.
Fig. 4.17. Testing samples of artificial shark skin.
Table 4.3. Drag-Reducing Efficiency of Artificial Composite Shark Skin (%).
Velocity (m/s) 1.2 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 3.9
Drag-reducing 3.2 5.1 6.2 8.1 9.9 11.5 13.4 15.1 16.3 18.2
rate
Velocity (m/s) 4.2 4.5 4.8 5.1 5.4 5.7 6.0 6.3 6.6 6.9
Drag-reducing 19.1 20.3 21.1 21.5 21.8 22.1 22.3 22.6 22.9 23.1
rate
Fig. 4.18. Experimental results of drag-reduction effect.

4.5. Building of Wind Tunnel

(1) Theoretical basis
The flowing conditions of fluids can be interpreted by the Navier–
Stokes equation. In practical engineering, the flowing stations generally
exhibit turbulence, which is the randomness in spatial. The average
Navier–Stokes equation can be called the Reynolds equation. For
incompressible turbulent pipe flow, the Reynolds equation in cylin-
drical coordinate system (as shown in Fig. 4.19) can be written as
following10−13 :
∂ū ∂ū ∂ū w̄ ∂ū

+ ū + v̄ +
∂t ∂x ∂r r ∂θ

1 ∂p̄ µ ∂u 2 1 ∂ru v 1 ∂u w
=− + ∇ 2 ū − + +
ρ ∂x ρ ∂x r ∂r r ∂θ
∂v̄ ∂v̄ ∂v̄ w̄ ∂v̄ w̄ 2

+ ū + v̄ + −
∂t ∂x ∂r r ∂θ r

1 ∂p̄ µ v̄ 2 ∂w̄
=− + ∇ 2 v̄ − 2 − 2
ρ ∂r ρ r r ∂θ

∂v u 1 ∂rv 2 1 ∂v w w 2
− + + −
∂x r ∂r r ∂θ r
∂w̄ ∂w̄ ∂w̄ w̄ ∂w̄ v̄ w̄
+ ū + v̄ + +
∂t ∂x ∂r r ∂θ r
v=0+v’
w=0+w’ θ
– –
u(y) u =u+u’
r y
Fig. 4.19. Pipe turbulence in the cylindrical coordinate system.

Fig. 4.20. Evolution of turbulent flowing near the entrance of the pipe.

1 ∂p̄ µ 2 ∂w̄ w̄
=− + ∇ w̄ + 2
2
− 2
ρr ∂θ ρ r ∂θ r

∂w u ∂w v 1 ∂w 2 2w v
− + + − ,
∂x ∂r r ∂θ r
in which
∂2 ∂2 1 ∂ 1 ∂2
∇2 =
+ + + .
∂x 2 ∂r 2 r ∂r r 2 ∂θ 2
After fluids flowing into the pipe, as shown in Fig. 4.20, the
following steps should be experienced14−19 :
(1) The velocity is uniformly distributed in the pipe near the entrance;
(2) The depth of boundary layer is becoming thicker;
(3) The boundary merged in the center of pipe, and the flow is
developed further;
(4) The developed turbulent flowing.
For the fully developed turbulent flowing in the pipe (the fourth
section, as shown in Fig. 4.20), the statistics of velocity and pressure
meet the demands of constant conditions and axial, circumferential
symmetry conditions, as following:
∂ ∂
= 0, =0
∂t ∂θ
∂velocity
=0
∂x
v̄ = w̄ = 0
uw = v w = 0.
On the above-mentioned basis, the Reynolds equation can be simplified
as following:
2
∂p̄ d ū 1 d ū 1 d
=µ + − ρru v (4.1)
∂x dr 2 r dr r dr
∂p̄ 1 dρrv 2 w 2
=− +ρ . (4.2)
∂r r dr r
To find the partial derivative of Eq. (4.2) based on x, the equation
∂2 p̄ ∂2 p̄
can be obtained: = 0 = , so the left-hand side of the
∂r∂x ∂x∂r
equation has no relation with r. Integrating Eq. (4.1) based on r, and
the boundary condition r = 0 is adopted, and then the boundary
condition r = R is introduced, the following results can be obtained;
r ∂p̄ d ū
= −ρu v + µ (4.3)
2 ∂x dr
R 2
v − w 2
p̄(x, r) + ρv − ρ
2 dr = p̄(x, R). (4.4)
r r
Integrating Eq. (4.3) based on x, and the boundary conditions are
introduced as following;

2 d ū 2
p̄(x, R) − p̄(0, R) = µ x = − τw x. (4.5)
R dr R
r=R
The pressure of fluid is descending with the distance increasing in
the axial direction, which means that the fluid friction stress can be
obtained by measuring the pressure along the pipes, as following:
p R
τw = . (4.6)
l 2
The static pressure in the pipe can be depicted as following:
R 2
2 v − w 2
p̄(x, r) − p̄(0, R) = − τw x − ρv + ρ
2 dr. (4.7)
R r r
The above-mentioned pressure drop law is aiming at the fully
developed turbulent flowing in the pipe, but it does not coincide
with the actual circumstance at the entrance of the pipe.20 And it
Fig. 4.21. Diagram of pipes testing and distribution of pressure tapping holes.
was discovered that the flowing is disturbed at the entrance; the fully
developed turbulence can be obtained at the points where x is larger
than 40 times of diameter.21
(2) Building of wind tunnel
Based on the theory that “the pressure potential on the inner wall of
pipelines is decreasing along with the axial direction; the sheer stress
(frictional resistance) is proportional to the pressure drop at the same
length of pipe,” so the frictional efficiency of pipes with different
Reynolds numbers can be obtained through measuring the pressure
drop and the fluid velocity of pipes.22−26 Considering the feasibility
and convenience of experiment in lab, the inner diameter of the steel
pipe is chosen as 105 mm, and the total length of testing pipe is 36 m,
with six pipes each 6 m long. The diagram of the pipes testing and
distribution of pressure tapping holes are indicated in Fig. 4.21. To
obtain the fully developed turbulent flow, the first testing point should
be far away from the inlet of gas where the distance is 40 times that of
the inner diameter. Moreover, the gas tightness of the system should
be ensured to allow the reliability of test results.
As shown in Figs. 4.22 and 4.23, the testing system for the
drag-reduction experiment is mainly composed of coated pipes, high-
speed impeller fan, frequency converter, multiple pressure gauge, pitot
tube and inclined micro-pressure gauge. The high-speed impeller fan
provides air flow supply, the frequency converter is for regulating
the velocity of air, the multiple pressure gauge is used to test the
pressure at various lengths, and the pitot tube and inclined micro-
pressure gauge are used to measure the air flow velocity in the pipes.
In the testing system, the testing error of pressure and wind speed
mainly derives from the naked eye’s reading error of the alcohol
column of the micro-manometer. In the process of experiments, the
Fig. 4.22. Frictional coefficient testing system of pipes.
(a) (b)
(c) (d)
Fig. 4.23. Details of testing system.

tilt angle (β) of the micro-manometer is set as 45◦ , and the working
medium is industrial alcohol, whose density is 0.8 × 103 kg/m3 , gravity
acceleration 9.8 m/s2 , and the reading error of micro-manometer
is 0.5 mm, so the absolute error is 2.8 Pa, which is calculated as
ρg l sin(α) = 0.806 × 9.8 × 0.5 × sin(π/4) ≈ 2.8 Pa.
In the drag-reduction experiments of coated pipes, the seam
between the different pipes should be calibrated exactly, as shown
in Fig. 4.23(a), and then the sealant is smeared on the joining seam
and the rigid hoop is used to encapsulate the pipes, as presented in
Fig. 4.23(b), for the purpose of ensuring air-tightness. The adjusted
stents are necessary to ensure a proper alignment, as illustrated in
Fig. 4.23(c). For the purpose of clean visual sense, the industrial alcohol
is colored in the micro-pressure gauge, as shown in Fig. 4.23(d).
4.6. Experimental Research in Water Tunnel
In order to validate the drag-reduction effect of the real sharkskin

surface, the drag-reducing experiment is conducted in vacuole water
tunnel (China Ship Scientific Research Center (CSSRC)), as shown in
Figs. 4.24 and 4.25. The parameters used in the test system are as
follows27−29 :
(1) The test section in vacuole water tunnel is 3.2 m in length and 0.8
m in diameter;
Fig. 4.24. Image of water tunnel.

Fig. 4.25. Schematic drawing of water tunnel. 1-Resistance scale, 2-Workding

section, 3-Direct current motor, 4-Pump, 5-Air compressor machine, 6-Pressurized
bucket, 7-vacuum pump, 8-vacuum box, 9-fliter apparatus, 10, 12-Degasssing tower,
11-water pump, 13-Pressure adjusting chamber.
(2) The flow velocity in test section is adjustable between 0 m/s–20

m/s;
(3) The central pressure of the test section can be varied between 8
kPa–400 kPa;
(4) The vacuole index is 0.15.
The test instruments also include a resistance sensor of 300 N in

measurement range and 0.1 N in resolution, two SA55 DC amplifiers,
NI-PCI4472 digital sampling panel with sampling frequency of 200
KHZ, and IBM-PIV computer.30,31
The testing model with hollow elliptical sphere at front and hollow
cylinder at posterior segments is made of aluminum alloy LY12. The
size of test model with the length of the cylinder section 500 mm and
external diameter 90 mm are determined according to the size of strain
gauge balances of test system. And the external diameter of hollow
elliptical sphere is designed as 94 mm to guarantee smooth external
surface after pasting the test skin sample. The vivid shark skin surface can
be fabricated by bio-replicated method, as shown in Fig. 4.26. In the
Fig. 4.26. Schematic drawing of experimental research and image of field testing.
process of pasting, the direction of micro-groove of skins should be

consistent with the direction of flow. The experiments are conducted
according to the rules for vacuole water tunnel test (Q/702J0301-
2008). The water temperature is set at 28◦ C and test system degassed
for more than one hour before testing. The water velocity can change
from 2m/s to 10m/s. The water velocity of each test continuously
enhances until the skins are washed off.
The test skins are divided into two groups. The first group includes
smooth skins (G-1) and biomimetic micro-groove drag reduction
surface (G-2) made from silicone rubber, whose sizes are all 170 mm ×
97 mm×2 mm (length × width × height). They need to be pasted upon
the surface of test model 1. The second group is made from waterborne
epoxy resin, including smooth skins (H-1), nano-long chain skins (H-
2) and synthetic bio-replicated hierarchical shark skin (H-3). The drag-
reduction performances of nano-long chain surface and synthetic bio-
replicated hierarchical shark skin can be achieved by experiments on
the second group. The test surface of the second group with size
170 mm×97 mm×0.3 mm (length × width × height) should be pasted
to test model 2. A total of 704 silicone sealants are used as adhesive
because of their merits such as high water resistance, high adhesive
strength and long curing time, which provide enough time to adjust the
position of skins after pasting. It should be noted that the height of skins
should be kept consistent during pasting. The seam should be sealed
and smoothed using of 704 silicone sealants to avoid infiltration of
water and forming of disturbed flow. Moreover, the direction of micro-
groove of skins should be consistent with the direction of flow.32−34
The experimental results are listed in Tables 4.4 and 4.5. The drag-
reduction rate shown in Table 4.6 is calculated on basis of literature.
Table 4.4. Resistance Value of Silicone Rubber Skins.

Flow
velocity
(m/s) 2 2.5 3 3.3 3.6 3.9 4.2 4.5 5 5.5 6 6.5 7 7.5 8
G-1 0.6 1.1 1.8 2.4 3.1 3.9 4.9 5.8 7.6 9.8 12.2 14.8
G-2 0.3 0.6 1.6 2.1 2.8 3.5 4.3 5.2 6.9 8.8 11 13.4 16 18.9 21.9
Table 4.5. Resistance Value of Waterborne Epoxy Resin Skins.
Flow
velocity
(m/s) 3 3.3 3.6 3.9 4.2 4.5 5.0 5.5 6 6.5 7 7.5 8 9
H-1 0.7 1.5 2.4 3.3 4.2 5.2 7.0 8.9 11.0 13.2 15.5 17.9 20.6 26.4
H-2 0.8 1.5 2.2 3.0 3.8 4.7 6.2 8.0 9.9 11.9 14.1 16.5 19.0 24.5
H-3 0.8 1.4 2.1 2.8 3.6 4.4 5.7 7.2 8.7 10.3 11.9 13.7 15.5
Table 4.6. Drag-Reduction Rate of Test Skins.
Flow
velocity
(m/s) 3 3.3 3.6 3.9 4.2 4.5 5.0 5.5 6 6.5 7 7.5 8 9
G-2 15.3 12.8 11.3 10.0 9.6 9.4 9.2 9.3 9.5 9.8
H-2 −20 0.7 6.8 9.2 10.2 10.7 10.9 10.4 10.0 9.5 8.9 8.3 7.7 7.5
H-3 −7.1 6 10.6 13.2 15.0 16.3 18.2 19.6 20.9 22.0 22.9 23.8 24.6
Except for the smooth silicone rubber skins (G-1) which were washed
off beyond 6.5 m/s, the flow velocity of the other test skins are up
to 8 m/s. The resistance value curves based on Tables 4.4 and 4.5
are shown in Fig. 4.27, and the drag-reduction rate curve based on
Table 4.6 is shown in Fig. 4.28.
From Table 4.6 and Fig. 4.28, we can see that the drag-reduction
rate of synthetic bio-replicated hierarchical shark skin increases rapidly
along with the velocity of flow rising as the velocity is less than
3.5 m/s and increases gently as the flow velocity exceeds 3.5 m/s.
The drag-reduction rate is up to 24.6% when the test terminates at
flow velocity up to 8 m/s. The drag-reduction rate of synthetic bio-
replicated hierarchical shark skin is higher throughout than that of
the nano-long chain skins, while it surpasses the biomimetic micro-
groove drag-reduction surface just after the flow velocity up to 3.9 m/s.
Fig. 4.27. Resistance at different flow velocities.
Fig. 4.28. Drag-reduction rate at different flow velocities.
Especially when flow velocity exceeds 5.5 m/s, the drag-reduction rate
of synthetic bio-replicated hierarchical shark skin is larger than the sum
of micro-groove and nano-long chain skins. It can be forecasted that the
synthetic drag-reduction effect should become more remarkable with
the flow velocity increasing. From Table 4.6 and Fig. 4.28, we can see
that the resistance of synthetic bio-replicated hierarchical shark skin is

higher than that of the smooth skins (H-1) when flow velocity is low. It
was because the long-chain of DRA stretching in water has not begun
to reduce drag when the water flow was laminar flow or the turbulence
intensity was weak. Even the hydrophilic groups of long-chain of DRA
hamper the water flow, which will lead to the resistance increasing. Only
with the velocity of flow increasing does the drag-reduction effect of
long-chain gradually become remarkable.35−38
4.7. Conclusion
In this chapter, different characters of artificial shark skin surfaces are

illustrated in detail, and the main conclusions are as following:
1. The marking key point method is an effective way to analyze the

replicating precision of biological shark skin, and the results tell us
that the replicating precision can surpass 90%, which validates the
reliability of the bio-replicated method.
2. The chemical property of artificial shark skin surfaces is investigated
systematically, and the results tell us that the grafting step has been
completed successfully.
3. The mechanical properties of artificial shark skin surface are explored
in detail. The experimental results tell us that when the amount
of dibutylo-phthalate is 0.75g (which means that the mass ratio is
0.75%), the composite shark skin has the best mechanical perfor-
mance.
4. The water tunnel and wind tunnel are built, and the drag-reduction
efficiency of shark skin surfaces are tested. The conclusion can be
reached that the drag-reducing efficiency of common artificial shark
skin is greater than 8%, and that of composite shark skin surface is
higher than 20%.
References
[1] Luo YH, Liu YF, Zhang DY. (2015) Hydrodynamic Testing of a Biological
Sharkskin Replica Manufactured Using the Vacuum Casting Method. Surf Rev
Lett 22(2): 1550030.
[2] Motta P, Habegger ML, Lang A, et al. (2012) Scale Morphology and Flexibility
in the Shortfin Mako Isurus oxyrinchus and the Blacktip Shark Carcharhinus
limbatus. J Morphology 273: 1096–1110.
[3] Zhang DY, Luo YH, Li X, et al. (2011) Numerical simulation and experimental
[4] Luo YH, Zhang DY. (2011). Study on the Micro-replication Precision of Shark
Skin. Applied Mechanics and Materials 44–47: 1151–1157.
[5] Luo YH, Zhang DY, Liu YF, et al. (2015). Chemical, Mechanical and Hydro-
dynamic Properties Research on Composite Drag Reduction Surface Based
on Biological Sharkskin Morphology and Mucus Nano-long Chain. Journal of
Mechanics in Medicine and Biology 15(5): 1530084.
[6] Agrim S, Robert WD, Steven PH, Michael SS. (2014) Drag Reduction Using
Riblet Film Applied to Airfoils for Wind Turbines. J Sol Energ-T ASME 136:
021007.
157–171.
[9] Zhang DY, Li YY, Han XX, et al. (2011d). High-precision bio-replication of
synthetic drag reduction shark skin, Chinese Science Bulletin 56(9): 938–944.
[10] Chen HW, Che D, Zhang X, et al. (2015) Large-proportional shrunken bio-
replication of shark skin based on UV-curing shrinkage. J Micromech Microeng
25: 017002.
[11] Han X, Zhang D. (2008b) Study on the micro-replication of shark skin. Sci
China Technol Sci 51(7): 890–896.
[12] Luo YH, Liu YF, Anderson J, et al. (2015d) Improvement of water repellent and
hydrodynamic drag reduction property on bio-inspired surface and exploring
sharkskin effect mechanism. Appl Phys A 120: 369–377.
[13] Cong Q, Feng Y, Ren LQ. (2006). Affecting of riblets shape of non-smooth
surface on drag reduction. Journal of Hydrodynamics, Ser. A 21(2): 232–238.
(In Chinese.)
[14] Daniel TL. (1981) Fish mucus: In situ measurement of polymer drag reduction.
Biol Bull 160(3): 376–382.
[15] Bhushan B, June YC, Koch K. (2009b) Micro-, nano- and hierarchical structures
for superhydrophobicity, self-cleaning and low adhesion. Philos Trans R Soc A
367: 1631–1672.
[16] Han X, Wang J. (2011) A Novel Method for Fabrication of the Biomimetic
Shark-Skin Coating. Adv Mater Res 239–242: 3014–3017.
[17] Kerry L. (1994) Functions for fish mucus, Rev Fish Biol Fisher 4: 401–429.
[18] Wen L, Weaver JC, Lauder GV. (2014) Biomimetic shark skin: design, fabrication

[20] Choi KS, Gadd GE, Pearcey HH, et al. (1989) Tests of drag-reducing polymer
coated on a riblet surface. Appl Sci Res 46: 209–216.
[22] Luo YH, Zhang DY, Chen HW. (2012b) Research on Manufacturing Vivid
Trans-scale Shark Skin Surface and Drag-Reducing Effect Simulation. Adv Sci
Lett 5(1): 49–55.
[23] Koury E, Virk PS. (1995) Drag reduction by polymer solutions in a riblet-lined
pipe. Appl Sci Res 54: 323–347.
[24] Matthias S, Stanislav G. (1991). Biological Micro- and Nanotribology: Nature’s
Solutions. Berlin, Springer-Verlag Berlin Heidelberg.
[26] Singh RP, Karmakar GP, Rath SK, et al. (2000) Biodegradable drag reducing
agents and flocculants based on polysaccharides: Materials and applications.
Polym Eng Sci 40(1): 46–60.
[27] Luo YH, Liu YF, Zhang DY. (2015b). Advanced Progress in Nature Gas
Pipelining Applying different Drag Reduction/Energy Saving Technologies: A
Review. Eur J Environ Civil Eng 19(8): 931–949.
[28] Pan JF, Chen HW, Zhang DY, et al. (2013) Large-scale solvent-swelling-
based amplification of microstructured sharkskin, Journal of Micromechanics and
Microengineering 23: 075018.
[29] Pratapkumar N, Donggang Y. (2007) Rapid pattern transfer of biomimetic
surface structures onto thermoplastic polymers. Mater Sci Eng C 27(4):
794–797.
[30] Springer VG, Gold JP. (1989) Sharks in Question: the Smithsonian Answer Book.
Smithsonian Institution Press, Washington and London.
[31] Toms BA. (1948) Some observations on the flow of linear polymer solutions
through straight tubes at large Reynolds numbers. Conference: First Interna-
tional Congress on Rheology 2: 135–141.
[32] Martin S, Bhushan B. (2014). Fluid flow analysis of a shark-inspired microstruc-
ture. J Fluid Mech 756: 5–29.
[33] Viswanath PR (2002) Aircraft viscous drag reduction using riblets. Prog Aerosp
Sci 38: 571–600.
[34] Walsh MJ. (1983) Riblets as viscous drag reduction technique. AIAA J 21(4):
485–486.
[35] Luo YH, Zhang DY, Liu YF, Ng EYK. (2015c) Bio/Micro-Rolling Fabrication
of Biological Sharkskin Morphology on Semi-Cured Coating and Drag Force.
Experimental Research 16: 1650016.
[36] Zhang DY, Luo YH, Chen HW. (2011b) Application and numerical simulation
research on biomimetic drag-reducing technology for gas pipelining. Oil Gas-
Eur Mag 37(2): 85–90.
[37] Zhang DY, Luo YH, Chen HW, Jiang XG. (2011c) Exploring drag-reducing
grooved internal coating for gas pipelines. Pipeline Gas J 238(3): 58–60.
[38] Christodoulou C, Liu KN, Joseph DD. (1991) Combined effects of riblets and
polymers on drag reduction in pipes. Phys Fluids 3(5): 995–996.
Chapter 5
Application of Biomimetic Shark Skin

Surface in Natural Gas Pipelining
J. Wang,∗ Y.T. Zhao,∗ L. Zhang,∗ Y. Luo,† E.Y.K. Ng‡


For purpose of increasing the transportation capacity of natural gas

pipelines and protecting them from corrosion, internal coating tech-
nology has been widely used, with remarkable benefit. However, with
the reduction of inner wall roughness, small convex regions are all
completely submerged within the viscous sublayer, and the pipeline can
be termed a “hydraulic smooth pipe,” which implies that wall friction
is difficult to lower further even with continuous smoothing of the
coated surface. Therefore, how to reduce flow friction and increase
transmission capacity on the basis of the internal smooth coating is
gradually becoming an urgent problem to be resolved; perhaps bio-
inspired/biomimetic drag-reducing technology is a good approach.
In this chapter, according to the actual working circumstances of the
West-to-East Project in China, the relevant parameters are calculated
to justify the existence of the “hydraulic smooth pipe.” And then, the
143
forming property of AW-01 Epoxy Resin which is widely adopted in

China is analyzed comprehensively via the light-sectioning method, and
the relationship of forming precision and curing degree of epoxy resin
is made certain, laying the foundation for the manufacturing process
of bio-inspired natural gas pipelines. For completing the bio-inspired
surface in a large area, the Pre-Cured Micro-Rolling Technology
(PCMRT) is advanced. To verify the drag-reduction and forming effect,
experiments with air are carried out, followed by the experiment with
actual natural gas in the field transmitting pipeline. The results of the
experiments are consistent with that in numerical simulation, which
proved the accuracy and reliability of applying the bio-inspired drag-
reducing technology in natural gas pipelining.
5.1. Internal Coating Technology
The flow state of natural gas in a pipeline is almost in turbulence, but not
all the fluid in the flow section is in the same turbulent state. Near the
pipe wall, the transverse pulsating movement is constrained by the wall,
so the turbulent shear stress generated by the pulsation is very small,
and viscous drag friction plays the dominant role. Therefore, the layer
of fluid is basically in the laminar flow state, and the thin layer is called
the viscous sublayer. When the natural gas flows through the pipeline,
if the absolute roughness of the pipeline is greater than the thickness of
the viscous sublayer, the roughness of pipeline plays the primary role in
the energy loss of natural gas transmission. Under these conditions,
further lowering the roughness of the inner wall can contribute to
increasing the transmission rate of natural gas pipelining. According
to the afore-mentioned theory, internal coating technology has been
widely applied, marking a milestone in the history of natural gas
pipelining. Its advantages are as follows: (1) improving the flowing
properties; (2) anticorrosion provided before construction; (3) helpful
for detecting faults in the inner wall of the pipeline.
The internal coating technique was mainly adopted for water and
crude oil in the early phase worldwide. It is first used in the pipelines of
sulphur-containing crude oil and natural gas in 1947, and for the trunk
gas transmission pipeline in 1953. The Transcontinental Pipeline Co. in
the US carried out the internal coating on the two natural gas pipelines
Application of Biomimetic Shark Skin Surface in Natural Gas Pipelining 145
from Nov. 1958 to Dec. 1959. Through several decades’ effort, internal
coating technology has undergone great improvement. Canada, Italy,
France, the former Soviet Union and the Netherlands all adopted drag-
reduction internal coating technology, and the transmission rate was
increased by 3.8%, 6.2%, 10%, 12% and 30% respectively.1,2 China’s
scientific researchers carried out field tests in Dagang City — the results
indicated that the throughput of the natural gas pipeline with the AW-
01 coating was improved by 20.9% over the untreated pipeline.3,4
The steps of internal coating technology include heating and dehu-
midification, de-rusting, wind-cleaning, coating spraying and curing;
the schematic drawing is shown in Fig. 5.1. The roughness of the
naked pipe internal wall is more than 45 µm, which will become about
20 µm after de-rusting and 4.5 µm covered by the coating, as shown
in Fig. 5.2. In order to enhance the adhesive force and effectively avoid
drop-out of coating, the characteristics and properties of coating were
investigated and explored, and the flexible rolling method during the
curing process was proposed. The result showed that the adhesion effect
was enhanced by about 10% by rolling without introducing additional
materials.
Heating and De-

humidification
De-rusting
Coating spraying
Wind-cleaning
Curing
Fig. 5.1. Schematic drawing of internal coating technology.

Internal coating wall
Wall after de-rusting
Naked pipe wall
Fig. 5.2. Variation of internal wall roughness.
At present, internal coating technology has been widely applied

in the first and second pipelines of the West-to-East gas transmission
projects and other projects in China, bringing enormous economic
benefit. Since the first pipeline of the West-to-East gas transmission
project was completed and put into operation in August 2004, gas from
the Traim gas-field has supplied 12 provinces, and huge economic and
environmental benefits have been received since then. However, with
the rapid development of all of China, the demand for natural gas in
areas along the first pipeline, especially in the middle and southeast
coastal areas, is growing very fast — the first pipeline has been clearly
overwhelmed. Even taking into account some other methods, such
as increasing pressure and so on, the gas boost is still limited, and it
cannot meet the growing demands of these areas. It is predicted that
the gap of gas supply and demand will be about 80 billion m3 a year
in 2020, therefore the necessity of constructing the second pipeline
of the West-to-East gas transmission project is self-evident. A result of
Alashankon
Tuha Oilfield Beijing
Dushanzi
Urumchi
Shanshan Shijiazhuang
Zhongwei Jingbian
Traim Gasfield
(Lunnan) Huaiyang
Chaklamud Nanjing
Gasfield Xi’an
Lanzhou
Wuhan
Shanghai
Chongqing
Nanchang
Available Lines
Nanning Guangzhou
Second Pipelines of West-to-East
Fig. 5.3. Preliminary pipeline alignments of West-to-East project in China.
Table 5.1. Parameters of Second Pipeline of West-to-East Gas Transmission

Project.
Diameter/ Working
Starting Ending thickness pressure Throughput
points points (mm) (MPa) (108 × m 3 /yr)
Trunk Line 1 Alashankou Nanchang 1219/18.4 12 300

Trunk Line 2 Nanchang Guangzhou 1016/15.3 10 130
Branch Line 3 Lunnan Shanshan 1016/15.3 10 120
Branch Line 4 Tuha Shanshan 610/7.9 10 30
Branch Line 5 Jingbian Xi’an 1016/15.3 10 120
Branch Line 6 Nanchang Shanghai 914/11.9 10 70
Branch Line 7 Guangzhou Nanning 711/11 10 30
lengthy discussion and analysis, the preliminary pipeline alignments and

parameters of the second pipeline are shown in Fig. 5.3 and Table 5.1,
respectively.
5.2. Coated Pipeline Becoming “Hydraulic

Smooth Pipe”
For the natural gas pipeline, the thickness of the viscous sublayer δ0
is changing with the Reynolds numbers and the inner diameters of
Fig. 5.4. Hydraulic rough pipe.
pipeline, and the semi-empirical formula of computing δ0 is as follows:

58.31d
δ0 = (5.1)
Re0.875
Q
Re = 1.536 , (5.2)
dµ
where d is the inner diameter of the pipeline, Re is the Reynolds number
of the pipeline, Q is volume flow rate under the standard status, and
is relative density of natural gas.
When the natural gas flows through the pipeline, if the thickness
of the viscous sublayer of nature gas is less than the absolute roughness
of pipeline, that is δ0 < , as Fig. 5.4 shows, the pipeline is called a
“hydraulic rough pipe,” and the roughness of the pipeline plays the
primary role in the energy loss of natural gas transmission. Under
these conditions, further lowering the roughness of the inner wall can
contribute to increasing the transmission rate of natural gas pipelining.
If the thickness of the viscous sublayer of natural gas is more than
the absolute roughness of pipeline, that is δ0 > , as Fig. 5.5 shows,
the pipeline is called a “hydraulic smooth pipe” — the small convex
parts of the pipeline inner wall are all completely submerged in the
viscous sublayer, the friction between the gas flow and pipe wall is the
viscous resistance of the viscous sublayer, and the natural gas likely
flows through the absolutely smooth pipeline, so the friction cannot
be reduced even if the roughness of the inner wall of the nature gas
pipeline is further lowered.
According to the actual situations and computation, the relevant
parameters in the process of calculation are as follows: = 0.6,
Fig. 5.5. Hydraulic smooth pipe.
µ = 1.55e-5Pa·s (12 MPa) or µ = 1.43e-5 Pa·s (10 MPa). In

accordance with the data in Table 5.1 and Formula (7.2), the Reynolds
numbers of different pipelines are approximately available as follows:
Re1 = 4.785e7, Re2 = 2.696e7, Re3 = 2.489e7,

Re4 = 1.032e7, Re5 = Re3 = 2.489e7,
Re6 = 1.607e7, Re7 = 8.898e6
In accordance with the data in Table 5.1 and Formula (5.1), the
results of calculating the thickness of the viscous sublayer are as shown
below:
58.31d1 58.31d2
δ01 = = 13.14 µm, δ02 = = 18.09 µm,
Re0.875
1 Re0.875
2
58.31d3 58.31d4
δ03 = = 19.40 µm, δ04 = = 25.28 µm,
Re0.875
3 Re0.875
4
58.31d6
δ05 = δ03 = 19.40 µm, δ06 = = 25.70 µm,
Re0.875
6
58.31d7
δ07 = = 33.37 µm
Re0.875
7
At present, the AW-01 epoxy resin coating has reduced the average
absolute roughness of the inner wall from 45 µm to 5.5 µm, so
the internal coating technology has enabled gas pipelines become
“hydraulic smooth pipes.” How to further reduce wall resistance and
improve the transportation capacity of nature gas pipeline under the
same circumstances has developed into a new problem to be resolved.

Fortunately, bio-inspired/biomimetic technology is a good approach.
Biomimetic drag-reducing technology has gained tremendous suc-
cess in the field of fluid engineering so far. Walsh at the NASA Langley
Research Center5 found that bio-inspired shark skin micro-grooved
surfaces can effectively reduce wall friction in turbulent flowing stations.
Since then, many researchers have carried out lots of work and explored
the drag-reduction mechanism in this field. The standpoint was put
forward by Bechert in late the 1990s6,7 that the flowing vortex could
enhance the slow brands gathered in the spreading direction and permit
the small momentum fluid to burst, and the effect of micro grooves is
to form the “second vortex” to curb the gathered phenomena, the
result of which is to inhibit the momentum exchange and reduce
the wall resistance. Liu et al.8 performed an experiment with pipes
lined with a film of micro-grooved equilateral triangles, and the results
showed that the drag-reduction efficiency could reach 5%–7% in the
completely developed turbulent flow. Koeltzsch et al.9 explored the
velocity field over convergent and divergent riblet patterns by hot-
wire measurements in turbulent pipe flow. Significant changes in the
near wall velocity field were discovered, and drag reductions up to
about 10% were measured. The research achievements of Viswanath10
indicated that riblets using symmetric v-grooves with adhesive-backed
film manufactured by the 3M company in the US had been widely
investigated and the results had revealed enormous consistency with
regard to the degree of drag reduction as well as certain aspects of
flow structure, and the maximum viscous drag-reduction in the range
of 4%–8% had been measured on a variety of two-dimensional flows
with zero or mild pressure gradients. Choi et al.11 studied the direct
numerical simulation (DNS) of micro flow field over the grooved
surface for the first time in 1993, and the drag-reduction mechanism
was investigated and explored in depth. Lang et al.12,13 explored the
influence of bristled shark skin on boundary layer control, and it was
concluded that the scales of some fast-swimming sharks were pliable
and might erect passively, which could affect the drag-reduction effect.
Therefore, it has developed into an indisputable fact that this kind
of bio-inspired surface has the obvious effect of drag-reduction in
Fig. 5.6. Schematic drawing of pipe with smooth coat and micro-grooved coat.
fluid engineering. The flow of natural gas in the pipeline is mostly

in a turbulent flowing state, the application prospects of the micro-
grooved drag-reducing technology on the pipeline in order to increase
the transmission rate on the basis of internal coating are inevitably very
bright. The schematic diagram of smooth coating inner wall pipeline
and groove coating inner wall pipeline are compared in Fig. 5.6.
For grooved drag-reducing technology, the uppermost factor
affecting the drag-reducing efficiency is the non-dimension groove
width s + and the non-dimension groove height h + . According to
the relevant literature, when s + is 16∼19, h + is 12∼14, the grooved
surface has the best drag-reducing effect. Of all grooves, the easiest
to manufacture is the triangle straight groove. Therefore, taking the
triangle straight groove as an example, the analysis and calculations are
illustrated as follows.
Non-dimension groove width and height:

+ su ∗ s τ sU∞ λ + hu ∗ h τ hU∞ λ
s = = = , h = = = ,
v v ρ v 8 v v ρ v 8
(5.3)
where s -Width of the groove, unit: m
h-Height of the groove, unit: m
v-Kinematic viscosity of fluid, unit: m 2 /s
U∞ -Free stream velocity of fluid, unit: m/s .
For the calculation of λ (give name), the Brill formula2 is adopted
in this paper, whose expression is as follows:
1 Re
√ = 2 lg . (5.4)
λ 4.5223 lg Re − 3.8215
Table 5.2. Different Grooves on Different Pipelines.

H
H Parameters s -width h-height s + -non- h + -non-
H
HH of of dimension dimension
HH U∞ groove groove groove groove
HH (m/s ) (µm) (µm) width height
Pipelines H
Trunk Line 1 10∼12 55 40 16.80∼20.16 12.22∼14.66
Trunk Line 2 8∼10 65 45 16.30∼20.37 11.28∼14.10
Brunch Line 3 6∼7 85 65 16.70∼18.75 12.29∼14.34
Brunch Line 4 4∼5 120 90 16.09∼20.11 12.07∼15.09
Brunch Line 5 6∼7 85 65 16.70∼18.75 12.29∼14.34
Brunch Line 6 6∼7 85 65 16.57∼19.34 12.67∼14.78
Brunch Line 7 4∼5 120 90 16.27∼20.33 12.20∼15.25
The velocity and viscosity of flowing natural gas change with the
working pressure and other factors. In the trunk lines (working pressure
is 12 MPa), the kinematic viscosity is about 0.935e-6 m2 /s and the
velocity is about 8∼12m/s; in the branch lines (working pressure
is 10 MPa), the kinematic viscosity is about 0.947e-6 m2 /s and the
velocity is about 4∼7m/s. To achieve the best drag-reducing effect,
according to the above-mentioned formulas, the results can be obtained
as tabulated in Table 5.2.
5.3. Biomimetic Shark Skin Drag-Reduction

Technology
For AW-01 epoxy resin, the appropriate temperature fit for curing
is 40◦ C–50◦ C; SEM images of coating at different temperatures are
shown in Fig. 5.7. They indicate the small and fine lattice, closed packed
internal structure and good forming quality using the light-sectioning
method.
The capacity of epoxy resin plastic deformation in its curing is
analyzed and discussed in this chapter, and the relationship of plastic
deformation and curing degree of epoxy resin (AW-01) in the curing
process is investigated and analyzed by experimental research. The
curing process of epoxy resin is the crossing course of physical and
chemistry change: When the curing reaction reaches a certain level,
the system will see a sudden increase in its hardness, and transfer
(a) Curing at 40˚C (b) Curing at 45˚C (c) Curing at 50˚C
Fig. 5.7. SEM of AW-01 epoxy resin curing at different temperatures.
into insoluble, non-melting, with cross-linked structure of the gel,

which does not have the capacity of flowing, but of plastic deformation
under external pressure. Its characteristics have been applied in micro-
impressing technology; for the traditional micro-imprinting method,
the pressing zone-time is broad, but the keeping time is also very
long, which is limited to be applied on actual engineering, and how
to reduce the keeping time ultimately is an urgent problem to be
solved. And if the keeping time can be reduced into the time of padding
mold, the forming efficiency can be improved greatly. However, how to
determine the capacity of plastic deformation in the curing process is an
urgent issue. The common feature of complex fluid flow is the flowing
complexity and the difficulty of analysis of flow behavior, so it is difficult
to get a conclusion via theoretical calculations. Therefore, aiming the
particular material of epoxy resin (AW-01 epoxy resin, which has been
widely applied in the first and second pipelines of the West-to-East gas
transmission project in China), the relationship of plastic deformation
capacity and curing degree in the curing process is investigated by
experimental research, in which the capacity of plastic deformation is
measured by the instantaneous shaping precision.
To determine the ability of plastic deformation of epoxy resin at
different times, the experiment research of light-sectioning method is
used in this chapter; the schematic diagram is shown in Fig. 5.8, which
is mainly composed of parallel light, rigid probe, magnifying system,
screen and so on. The instantaneous feeding displacement of rigid
probe is performed, when the coating is fully cured, through testing
the morphology of coating surface, the capacity of plastic deformation
rigid probe
parallel light
coating of epoxy resin

video display
signal processing
amplification system
Fig. 5.8. Schematic diagram of light-sectioning method.
120
60 L: 2.75 mm L: 1.81mm
0
–60
–120
0.8 1.6 2.4 3.2 4.0 4.8 5.6 6.4
Position [mm]
(a) 3-D morphology (b) Cross-section curve of corresponding curve

of coating surface
Fig. 5.9. 3-D morphology of epoxy resin and the cross-section curve.
is investigated by testing the shaping precision of the epoxy resin

coating.14
In the process of light-sectioning method, the feeding displacement
of the rigid probe is exerted by screw rotation, whose accuracy can be
controlled in 1 µm. The solution of the naked human eye is 0.1mm,
and the magnification of the system is about 50 times. The feeding
displacement is 200 µm, and the testing system error is 2 µm in
the experiment process, the probe feed is 200 µm, so test results of
the visual impact of the largest error is 1%, in the permissible error
range, hence this method can be used for testing the epoxy-coated
plastic deformation. The morphology of the coating surface and the
corresponding cross-section curve is shown in Fig. 5.9. The depth
of liquid film is 300 µm, which can be measured by ruler, the curing
temperature is set as 45◦ C, and the keeping time is 147 minutes; the
experiment cross is about 1 second. Until the coating is fully cured, the
morphology of the surface can be received, as presented in Fig. 5.9;
the depth of the corresponding is 175.8 µm, and the shaping precision
is about 87.9%.
When the curing time is set as 40◦ C, 45◦ C and 50◦ C respectively,
and the keeping time is set as different values, the depth of concave
morphology and instantaneous shaping precision can be received as
achieved in Table 5.3.
As shown in Fig. 5.10, for the curing temperature of 40◦ C, for
example, the instantaneous shaping precision can reach more than 90%,
and in the corresponding time-zone, the capacity of plastic deformation
is the best.
For studying the properties and characteristics of AW-01 epoxy
resin system further, the kinetics of the cure reaction is monitored
by differential scanning calorimetry (DSC) and investigated using
constant heating methods, which is the important basis of composite
materials forming technology and can supply the theoretical principle
and foundation. In the process of imprinting, one needs to analyse the
characteristic and shaping precisions of AW-01 epoxy resin coating.
The instantaneously shaping precision of epoxy resin coating and
the relationship of curing degree and shaping precision are shown
in Fig. 5.10. Based on Table 5.3 and Fig. 5.10, the conclusion can
be reached: The best plastic deformation is corresponding with the
curing degree (α) of between 0.82–0.86 to achieve the best plastic
deformation. Therefore, the shaping precision is before descending
and after downing. In summary, the relationship of curing degree and
shaping precision was investigated, and when the curing degree is from
0.82 to 0.86, the best plastic deformation can be obtained, which has
great significance for epoxy resin shaping.
5.4. Numerical Simulation of Drag-Reducing

Effect
In validating the drag-reduction effect of grooved coating, direct
numerical simulation is carried out in this chapter. To ensure high
accuracy of computation, the domain of simulation and computation is
not the whole pipeline, but for a particular angle with symmetry planes,
as shown in Fig. 5.11.
At the same time, in order to ensure the high accuracy of compu-
tation and reducing the total number of grid, the mesh distribution
May 24, 2016
13:32
Table 5.3. Relationship of Shaping Precision and Curing Circumstances.
Keeping Depth/ Shaping Keeping Depth/ Shaping Keeping Depth/ Shaping

T/◦ C Time/min um Precision T/◦ C Time/min um Precision T/◦ C Time/min um Precision
Bio-Inspired Surfaces and Applications

40 150 151.0 75.5% 45 130 156.0 78.0% 50 100 145.2 72.6%
40 152 155.0 77.5% 45 132 158.2 79.1% 50 102 156.2 78.1%
40 155 156.4 78.2% 45 135 160.2 80.1% 50 105 172.2 86.1%
40 157 160.2 80.1% 45 137 170.0 85.0% 50 107 174.2 87.1%
40 160 165.4 82.7% 45 140 172.0 86.0% 50 110 179.8 89.9%
40 162 175.8 87.9% 45 142 175.4 87.7% 50 111 180.4 90.2%
40 166 176.4 88.2% 45 143 178.4 89.2% 50 112 181.2 90.6%
40 168 179.4 89.7% 45 144 179.4 89.7% 50 113 182.2 91.1%
40 170 180.4 90.2% 45 145 180.0 90.0% 50 114 180.8 90.4%
40 173 181.2 90.6% 45 146 181.0 90.5% 50 115 180.6 90.3%
40 175 182.0 91.0% 45 147 181.4 90.7% 50 117 180.4 90.2%
40 176 181.6 90.8% 45 149 183.0 91.5% 50 118 180.2 90.1%
9in x 6in
40 178 181.2 90.6% 45 150 183.2 91.6% 50 119 180.0 90.0%
40 180 180.4 90.2% 45 151 181.6 90.8% 50 120 179.0 89.5%
40 181 180.2 90.1% 45 155 179.6 89.8% 50 121 176.6 88.3%
40 182 177.0 88.5% 45 157 176.0 88.0% 50 122 176.0 88.0%
b2378-ch05
40 187 175.0 87.5% 45 160 164.2 82.1% 50 123 174.4 87.2%
40 190 172.0 86.0% 45 163 159.0 79.5% 50 100 145.2 72.6%
page 156
Fig. 5.10. Appropriate time-zone to exert pressure (0.82 < α < 0.86).
Smooth surface
Slip air water
Outlet
Symmetry Flowing direction
Inlet
Grooved surface
Fig. 5.11. Computational fluid zone.

should comply with the principles as follows15−18 : (1) To capture the

features of flow field near the wall, the grid near the upper wall and
lower wall location in the turbulent boundary layer should be refined.
(2) Limited by computer capacity, the total number of grids should
be constrained, so the grid should be sparse in the region far from the
wall. In the vertical direction, the height is divided into 40 line-grids by
the ratio of 1.25:1, and the parameters of meshing whole volume are
set as following: Elements-Tet/Hybrid, Type-TGrid; the total number
of grids is about 1,470,000. In the course of numerical simulation,
the RNG k-ε turbulence model and Enhanced Wall Treatment near
the wall are selected, others parameters are set as default, which can
exactly simulate the flow field situation near the tiny grooves. In
addition, choosing of the boundary conditions should follow the below
principles: (1) Because of the fully developed turbulence, the flow at
inlet and flow outlet of the computational flow field should be set as
periodical boundary condition. (2) Among the span wise of the flow
field, the flow is not constrained. In order to ensure the flow field
without other interference, the boundary condition along the span wise
of the computational flow field is set as symmetry condition. (3) The
upper smooth wall and the lower grooved surface wall are set as non-
slip and impenetrable boundary condition. The mesh of computational
fluid zone is shown in Fig. 5.12.
In the process of numerical simulation, the number of iterations
is set as 150, when the average velocity of gas in the pipeline is set
as 4.03m/s; the contours of shear stress are shown in Fig. 5.13. The
shear stress on the tips and the valleys is about 269.56 pa∼285.45 pa
and 31.72 pa∼47.58 pa, respectively, and for the smooth surface, the
shear stress is between 158.58 pa∼174.44 pa, and the integral of shear
stress on the smooth surface and grooved surface is 0.0010972N and
0.0010019N, so the drag-reduction efficiency of the grooved surface
is 8.68%.
Figure 5.13 shows the contours of velocity on a particular cross
section of the flow field — when the region is farther from the wall,
the velocity is higher, and the velocity near the grooved surface is also
less than that near the smooth wall. Moreover, the flow on the groove
valley is relatively smooth, and due to the influence of the grooves, the
velocity vortex is generated near the tips, so the turbulence is inhibited.
Fig. 5.12. Meshing of grids.
Fig. 5.13. Contours of Z-Wall shear stress (Pa).

Fig. 5.14. Contours of velocity (m/s).
In the process of manufacturing and employing the gas pipelines

with a grooved inner surface, the following bugs can be produced, such
as heaves and hollows, seams, irregularity, damage of grooves, deposits
of dust and so on. Figure 5.14 shows the contours of shear stress and
velocity vectors on flowing direction on grooved surface with heaves
and hollows. The sheer stress on top of the heaves is higher than others,
which is because of the higher velocity there, however, the shear stress
on the bottom of the hollow is low, even negative due to the negative
velocity. And the back flow can lead to the low shear stress. From the
computation, the heaves can increase the wall friction more than the
hollows, and this is to be avoided.
Because the rolling organization is composed of several rollers, the
rolling seams on the coating surface are inevitable, and the existing
seams have the two phenomena, which are the accumulation of material
and gaps of grooves. The shear stress on the grooved surface with rolling
seams is shown in Fig. 5.16. The accumulation of material can increase
the wall friction, which should be prevented, and the small gaps of
grooves have little influence on the drag-reduction effect, hence in the
manufacture of machining rollers, some minus allowance should be
guaranteed.
Fig. 5.15. Influence of heaves and hollows on drag-reduction effect. (a) Contours
of shear stress (Pa), (b) Velocity vectors (m/s) on grooved surface.
(a) (b)
Fig. 5.16. Contours of z-wall shear stress on grooved surface with rolling seams.
(a) Grooved surface with material accumulated, (b) Grooved surface with gaps.
In the process of rolling the grooved inner wall, because of the

influence of roundness of pipelines, the pressure exerted on the inner
wall coating of pipeline is not well-proportioned, so irregularities in
the grooved surface will be produced. The shear stress on the irregular
grooved surface is shown in Fig. 5.17. By the relevant data analysis,
when the irregularity of the grooved surface is not apparent, it can
produce only little negative influence on drag reduction, even as
positive influence.
In the process of applying the grooved surface pipelines, dust can
deposit in the valley of the grooves, and the shear stress on the grooved
surface is presented in Fig. 5.18. Due to the height reduction of the
groove, the effect of inhabit turbulence is decreased. And when the
height becomes half of the original, and if the dust continues to be
Fig. 5.17. Shear stress on the irregular grooved surface.
Fig. 5.18. Shear stress on the snapped grooved surface.
deposited, the height is less than the thickness of viscous sublayer, the
effect of the grooved surface is diminished, and the drag reduction
is less apparent. Thus the gas that enters the pipelines should be
cleaned.
In the process of employing the grooved surface, the pigging
process is operated, and the peaks of the grooves can be snapped or
fretted, where the influence of the broken grooved surface on drag
reduction is obvious. The shear stress on the surface is shown in
Fig. 5.19. When the height of groove is becoming the 3/4, the drag
reduction effect will become 1/2, and when the height is becoming
(a) (b)
Fig. 5.19. Shear stress on the snapped grooved surface. (a) Micro-grooved surface
with wreckage (b) Micro-grooved surface with abrasion
Dehumidification Derusting Cleaning Spraying Curing PCMRT Cured
Fig. 5.20. Process of machining biomimetic drag-reducing pipe.
2/3, the drag reduction effect will disappear. Therefore, the method
of reinforcing the hardness of the grooves should be researched and
adopted.
5.5. Fabrication of Micro-Grooved Elastic Rollers
The length of natural gas pipes various from 6 m to 12 m, so in order to

complete the manufacture of grooved surface in such a large area, the
rolling process is the first choice. One is aiming for the characteristics
and properties of AW-01 epoxy resin, which has been widely used in
the West-to-East project in China. The step of pre-cured micro-rolling
technology (PCMRT) can be easily added in the available production
line of internal coating technology while other steps do not need to be
changed. The process of machining biomimetic drag-reducing pipes is
shown in Fig. 5.20.
In the process of instantaneous imprinting technology, if the
keeping time is not enough, the polymer will not fill the mold
completely. At the same time, if the keeping time is too long, the
imprinting efficiency will be affected. For purpose of high precision
and efficiency, Heyderman et al.2,3 obtained the formula of marinating
Fig. 5.21. PCMT process and the image of rolling.
Fig. 5.22. Error analysis of pipe roundness.
time,

η0 S 2 1 1
tf = − , (5.5)
2P hf2 h02
in which, tf is the keeping time, S is the viscosity of polymer, P is the
pressure, h0 is the height of the pattern, and hf is the height of pattern.
In order to verify the forming effect, the PCMRT with rigid rolling
is used. The image of rolling is presented in Fig. 5.21. In the rolling
process, the velocity is about 0.2 m/s–0.3 m/s.
For the rolling system with steel pipe, rigid rolling will lead to
the failure, so a new method and process should be explored. For
the PCMRT, due to the existence of roller’s manufacturing error and
roundness error of pipe, a small clearance will be produced, which can
reach to 0.5 mm, as shown in Fig. 5.22, so the rigidity-contact rolling
will lead to the failure, and the elastic rolling is required. Additionally,
if the fixed supporting way is adopted with the gravity and roundness
error of pipe and beam, the rolling system will produce large deflection
Fig. 5.23. Error analysis of gravity with long pipe.
for pipe length of 6 m to 12 m of more than 10 mm, as illustrated in

Fig. 5.23. The deformation of the elastic roller cannot however adapt to
the total errors, therefore, a new supporting method should be explored
further to eliminate the influence of such a large error.
In the curing process of internal coating (AW-01 epoxy resin), its
rigidity will increase gradually, so harder outward surface of the roller
is demanded to realize the manufacturing biomimetic drag-reducing
surface on the coating. At the same time, to enable the outward surface
of roller and coating surface of pipe to contact with each other com-
pletely, the whole roller should express its elasticity and deformability to
some extent. Therefore, an outward-hard and inner-soft roller should
be fabricated. The important steps are as follows: (1) machining the
positioning plate; (2) positioning of central shaft and mold with smooth
inner wall; (3) casting the rubber silicon and curing; (4) demolding
of elastic roller until it is cured completely; (5) positioning of elastic
roller and mold with grooved inner wall; (6) casting the epoxy resin of
QS-125 and curing (produced by Beijing Qingda Qishi New Material
Technology Corporation); (7) Gaining the outward-hard and inner-
soft elastic roller by demolding. The thickness of epoxy resin is 1.5 mm;
the width of grooves on the outward surface is 200 µm. The schematic
of fabricating the elastic roller is shown in Fig. 5.24.
In ensuring the fluency of rolling process and eliminating the series
of errors more easily, the angle of each roller should be limited, the
whole rolling mechanism is composed of 18 rollers, and each machined
20◦ of the whole circle; the diagram of the rolling mechanism is shown
in Fig. 5.25. In addition, the diameter of the rolling mechanism is
designed to be 3% more than that of the natural gas pipe in order
to exert pressure on the coating surface. The micro-grooved roller
mechanism is shown in Fig. 5.26.
Positioning Plate Positioning of central shaft Positioning of mould Casting silicon rubber
with smooth inner wall and curing
Positioning of mould with Demoulding of outward-hard

Demoulding of elastic roller Positioning of elastic roller
grooved inner wall, casting and inner-soft elastic roller
epoxy resin and curing
Fig. 5.24. Diagram of elastic roller fabrication.
Fig. 5.25. Diagram of roller mechanism.
5.6. Design of Rolling Equipment
(1) Rolling design scheme — Adding-weight method

In order to eliminate the influence of deformation and gravity, the
adding-weight method is applied, as shown in Fig. 5.27. The rolling
system mainly contains elastic roller equipment, pumping equipment,
parallel guide, guiding rod, adding weight and so on.
Fig. 5.26. Elastic roller mechanism.
Nature gas pipe
Elastic roller equipment
Supporting bar
Pumping Equipment
Enlarged
Adding-weight
Parallel guide rail
Fig. 5.27. Rolling scheme of nature gas pipes (Adding-weight method).
In the rolling process, if the guiding rod is connected with the

roller equipment rigidly, due to the length of guiding rod and gravity,
a large deformation will be produced, so the rolling equipment should
be connected with flexible rope, and the gravity of roller equipment
will be offset by the adding-weight, and it likes floating without gravity
in the pipe, and it can move under the drove by the force.
Fig. 5.28. Schematic diagram of overhanging method.
(2) Rolling design scheme — Overhanging method

To avoid the influence of gravity and straightness error of pipe and
beam, which has a length of 6–12 m, the overhanging method is
investigated and explored. In this method, the roller mechanism
completes the movement vertically in the pipe drove by the flexible
rope; the schematic diagram is shown in Fig. 5.28. The process of
machining biomimetic drag-reducing pipes by overhanging method are
as following: (1) The pipe with pre-cured coating and pulley mechanism
are lifted using the electric hoist device — the steel pipe is drooping
vertically under gravity; (2) The hook descends out of the bottom
of pipe for hanging the roller mechanism using fixed pulley, which
can change the direction of force; (3) Connecting roller mechanism
with hook, the PCMRT with outward-hard and inner-soft elastic roller
mechanism is performed under force; (4) Put down the pipe and take
out the roller mechanism, and the next pipe can be machined with the
same steps. In the actual processing, the length of the steel pipe is 6 m,
Fig. 5.29. Image of overhanging method.
the inner diameter is 105 mm, the thickness is 4.5 mm, the total mass
is about 70 kg, and rated value of the electric device is about 100 kg.
The filed machining image is shown in Fig. 5.29, and the total time
of machining is about 30s. For the trunk pipeline of the West-to-East
project in China, the maximum diameter is 1,219 mm, the length is
12 m, the thickness is 12.8 mm, the whole mass is about 4,600 kg, and
for the electric hoist device, the maximum of rated value can reach more
than 40 × 103 kg, so it can also be machined by the same method.
5.7. Experiments on Drag-Reduction Effect

The forming effect of the biomimetic drag-reducing surface is one of
the key evaluating targets, and it is very necessary to analyze the 3-D
morphology on the internal coating surface. Small samples of pipes are
Fig. 5.30. Samples of biomimetic drag-reducing pipe.
shown in Fig. 5.30, and the 3-D morphology of the biomimetic drag-
reducing surface is included in Fig. 5.31. The width of the grooves is
about 200 µm, and the width of the micro-grooves on the outward
surface of the roller is also 200 µm, so the forming accuracy in the
width direction is about 100%, and the forming effect is satisfactory.
Based on the theory that “the pressure potential on the inner
wall of pipelines is decreasing along with the axial direction; the shear
stress (frictional resistance) is proportional to the pressure drop at the
same length of pipe,” so the frictional efficiency of pipes with different
Reynolds numbers can be obtained through measuring the pressure
drop and the fluid velocity of pipes. Considering the feasibility and
convenience of experiment in the lab, the inner diameter of the steel
pipe is chosen as 105 mm, and the total length of testing pipe is 36 m,
with six 6 m long pipes. The diagram of the testing and distribution of
pressure tapping holes are indicated in Fig. 5.32. To obtain the fully
developed turbulent flow, the first testing point should be far away from
the inlet of gas where the distance is 40 times that of the inner diameter.
Moreover, the gas tightness of the system should be ensured to allow
reliability of test results.
As shown in Figs. 5.33 and 5.34, the testing system of the
drag-reduction experiment is mainly composed of the coated pipes,
Fig. 5.31. Biomimetic drag-reducing surface.
Fig. 5.32. Diagram of pipes testing and distribution of pressure tapping holes.
high-speed impeller fan, frequency converter, multiple pressure gauge,

pitot tube and inclined micro-pressure gauge. The high-speed impeller
fan provides air flow supply, the frequency converter is for regulating
the air velocity, the multiple pressure gauge is used to test the pressure
at various lengths, the pitot tube and inclined micro-pressure gauge
are used to measure the air flow velocity in the pipes. In the testing
system, the testing errors of pressure and wind speed mainly arise from
the naked eye’s reading error of the alcohol column of the micro-
manometer. In the process of experiments, the tilt angle (β) of the
micro-manometer is set as 45◦ , and the working medium is industrial
alcohol, whose density is 0.8 × 103 kg/m3 , the gravity acceleration
9.8 m/s2 . The reading error of the micro-manometer is 0.5 mm, so
Fig. 5.33. Frictional coefficient testing system of pipes.
the absolute error is 2.8 Pa, which is calculated as ρg l sin(α) =

0.806 × 9.8 × 0.5 × sin(π/4) ≈ 2.8 Pa.
In the drag-reduction experiments with coated pipes, the seam
between the different pipes should be calibrated exactly, as shown in
Fig. 5.34 (a), then the sealant is smeared on the joining seam and the
rigid hoop is used to encapsulate the pipes, as presented in Fig. 5.34
(b), for the purpose of ensuring airtightness. The adjusted stents are
necessary to ensure a proper alignment, as illustrated in Fig. 5.34 (c).
For the purpose of clean visual sense, the industrial alcohol in the micro-
pressure gauge is colored, as shown in Fig. 5.34 (d).
Based on the formulas and the actual experimental conditions,
when the width of the micro-grooves is set as 200 µm in the drag-
reduction experiments with air, a promising drag-reducing effect can
be obtained, as given in Tables 5.4 and 5.5. “T” is the temperature
of the experimental environment, “α” is the slope of the pitot tube,
“β” is the tilt angle of micro-manometer, “H” is the frequency of
frequency converter (with the increasing of H, the output of the high-
speed impeller fan will rise gradually), “hi” is the length of the industrial
alcohol column in the glass tubes as illustrated in Fig. 5.34 (d), “l” is
the length of the industrial alcohol column in the Pitot tube.
(a) (b)
(c) (d)
Fig. 5.34. Details of testing system.
The pressure distribution of smooth and bio-inspired drag-

reducing coating pipes with air were obtained and plotted in Figs. 5.35
and 5.36. It can be seen that the pressure on different testing points is
on the same line, which indicate that the influence of seams is negligible.
The drag-reducing (DR) efficiency with different Reynolds num-
bers of bio-inspired micro-grooved pipes can also be calculated as
shown in Table 5.6.
(2) check where is this from?? Drag Reduction Experiment with Natural
Gas
The drag-reducing effect of bio-inspired micro-grooved pipes has been
validated with air, and in order to complete the industrialization,
experimental research using natural gas must be further carried out.
The field testing experiment is performed in Dongying City, Shandong
Province, the People’s Republic of China. For convenience of testing
and calculations, the pipeline is paved in a straight line without branch
and elbow. The schematic diagram of field testing experiment is shown
May 24, 2016
13:32
Table 5.4. Experimental Results of Smooth Coating Pipes with Air.

T β H h0 h1 h2 h3 h4 h5 h6 h7 h8 h9 h10 h11 l
(/◦ C) α (◦ ) (Hz) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm)
−4.5 0.4 45 7.5 381.1 362.7 363.8 366.4 367.6 370.1 371.3 373.9 375.0 377.6 378.8 380.6 6.1
−4.5 0.4 45 10 381.1 349.8 351.1 356.1 357.4 362.4 363.7 368.7 370.0 375.1 376.3 380.1 11.4
−4.5 0.4 45 12.5 381.1 334.7 336.7 344.0 346.1 353.3 355.4 362.6 364.7 372.1 374.1 379.6 18.5
−4.5 0.4 45 15 381.1 313.5 316.1 326.8 329.3 340.1 342.7 353.4 356.0 366.7 369.3 377.6 27.5
−4.5 0.4 45 17.5 381.1 289.2 292.1 307.2 310.0 325.2 328.1 343.2 346.0 361.2 364.1 375.1 38.3
−4.5 0.4 45 20 381.1 263.8 266.9 286.9 290.1 310.1 313.2 333.2 336.0 356.3 359.4 374.2 50.3
−4.5 0.4 45 22.5 381.1 234.8 238.4 263.7 267.3 292.7 296.3 321.6 325.3 350.6 354.2 373.6 64.1
−4.5 0.4 45 25 381.1 201.4 206.4 237.2 242.2 273.1 278.1 308.7 313.7 344.5 349.5 371.7 81.0
−4.5 0.4 45 27.5 381.1 164.4 169.7 207.2 212.5 250.1 255.3 292.8 298.1 335.6 340.9 369.5 97.7
−4.5 0.4 45 30 381.1 123.7 129.3 174.6 180.2 225.6 231.2 276.6 282.1 327.5 333.1 367.3 119.5
9in x 6in
−4.5 0.4 72 32.5 374.1 148.1 153.1 192.1 197 236.1 241.1 280.1 284.9 324.1 328.9 359.9 139.8
−4.5 0.4 72 35 374.1 112.2 117.2 163.4 168.5 214.6 219.7 265.1 270.9 317.1 322.1 358.1 164.9
−4.5 0.4 72 37.5 418.0 121.3 126.7 179.7 185.1 238.2 243.6 296.6 302.1 355.1 360.5 401.2 188.3
−4.5 0.6 72 40 427.0 86.7 93.8 153.9 160.9 221.1 228.1 288.1 295.2 355.3 362.3 409.3 145.2
b2378-ch05
page 174
May 24, 2016
Application of Biomimetic Shark Skin Surface in Natural Gas Pipelining
13:32
Table 5.5. Experimental Results of Bio-Inspired Micro-Grooved Coating Pipes with Air.

T β H h0 h1 h2 h3 h4 h5 h6 h7 h8 h9 h10 h11 l
(/◦ C) α (◦ ) (Hz) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm)
−4.5 0.4 45 7.5 384.9 365.8 367.2 369.8 371.1 373.7 374.8 377.6 378.7 381.4 382.6 384.4 6.1
−4.5 0.4 45 10 384.9 353 354.8 359.8 361.1 366.1 367.5 372.5 373.8 378.8 380.1 383.8 11.5
−4.5 0.4 45 12.5 384.9 339 341.2 348.3 350.4 357.5 359.5 366.7 368.7 375.8 377.8 383.4 18.5
−4.5 0.4 45 15 384.9 319.1 321.6 332.1 334.5 345 347.5 358 360.4 370.8 373.3 381.4 27.5
−4.5 0.4 45 17.5 384.9 296.2 299.3 313.8 316.5 331 333.7 348.3 351 365.5 368.2 378.9 38.2
−4.5 0.4 45 20 384.9 272 274.8 294.8 297.8 316.8 319.8 339 341.8 361 363.8 378 50.3
−4.5 0.4 45 22.5 384.9 243.5 248.8 272.8 276.4 300.4 303.8 328 331.4 355.5 358.8 377.4 64.5
−4.5 0.4 45 25 384.9 211.3 218 247.3 252.1 281.3 286.1 315.4 320.2 349.5 354.2 375.5 81.7
−4.5 0.4 45 27.5 384.9 175.6 181.6 217.6 222.6 258.6 263.6 299.6 304.6 340.6 345.6 373.3 97.7
−4.5 0.4 45 30 384.9 134.8 140.2 184.2 189.5 233.5 239.0 282.8 288.3 332.3 337.7 371.1 119.7
9in x 6in
−4.5 0.4 72 32.5 377.9 154.5 159.3 198.7 203 242 246.2 284.7 289.5 327.5 332.7 364.9 141.1
−4.5 0.4 72 35 377.9 114.6 120 165.6 171.6 217.3 223 269.1 274.2 321.5 325.5 362.4 165.5
−4.5 0.4 72 37.5 408.3 105.2 111.5 164.1 171.1 224.2 230.6 284.2 290.1 344.2 349.7 391.5 189
−4.5 0.6 72 40 385.8 31.8 39.4 100.9 109.2 170.8 178.9 241.6 248.6 311.6 318.4 367.6 145.4
b2378-ch05
175
page 175
Fig. 5.35. Pressure distribution of smooth coating pipe (air).
Fig. 5.36. Pressure distribution of bio-inspired drag-reducing pipe (air).
in Fig. 5.37. The total length of the smooth internal coating pipeline
and bio-inspired drag-reducing pipeline is set as 198 m (6 m × 33), and
the inner diameter of pipeline is 105 mm. All of the steel pipes are
connected by the welding method.
Table 5.6. Drag-Reduction Efficiency of Grooved Surface with Different

Reynolds Numbers.
Re 31250 46875 62500 78125 93750 109375 125000 140625

DR/% 3.25 4.85 5.61 6.47 6.94 7.30 8.23 8.61
Re 156250 171875 187500 203125 218750 234375 250000
DR/% 8.90 8.32 7.51 6.52 3.92 3.21 1.21
Flowing direction of natural gas
Testing Point 1 Testing Point 2 Testing Point 3
Internal coating pipeline Biomimetic drag-reducing pipeline
Fig. 5.37. Diagram of field testing experiment.
Fig. 5.38. Image of fabricated biomimetic drag-reducing pipes.
For the same pipeline and the same length, the friction coefficient
is proportional to the pressure drop. Therefore, the drag-reducing
efficiency can be obtained by measuring the pressure of natural gas. The
field images of pipeline construction are shown in Figs. 5.38 and 5.39.
The pipeline under construction is shown in Fig. 5.40. For the actual
field testing situation, the pressure of the natural gas is about 0.3 MPa;
Fig. 5.39. Sealing the pipes.
Fig. 5.40. Natural gas pipeline under construction.

Fig. 5.41. Pressure gauge for field testing.
the pressure of testing points can be measured by electric pressure

gauge, as presented in Fig. 5.41. Its measuring range is 0–0.4 MPa
and the measuring error is less than 0.02%.
In obtaining a good drag-reducing effect, the width of the micro-
grooves on the internal coating is set as 150 µm. Based on the theory
that the friction coefficient is ratio to the pressure drop (KPa) for the
same pipeline, the drag-reducing efficiency of bio-inspired pipe can be
calculated as follows:
fsmooth − fbiomimetic Psmooth − Pbiomimetic
η= × 100% = × 100%
fsmooth psmooth
(P1 − P2 ) − (P2 − P3 )
= × 100%,
P1 − P 2
(5.6)
where P1 is the pressure of testing point 1, P2 is the pressure of testing

point 2, and P3 is the pressure of testing point 3.
In the process of field testing, the flow rate (Q ) of natural gas is
controlled by throttle valve. The pressure of different testing points
with flow rate set at different values is tabulated in Table 5.7.
The relationship between drag-reducing efficiency and the
Reynolds number of bio-inspired micro-grooved pipelines with air and
natural gas is plotted in Fig. 5.42.
Table 5.7. Results of Field Testing Experiment with Natural Gas.
Q/m3 /h Re P1 /KPa P2 / KPa P3 / KPa DR /%
1 94.875 57518 299.953 299.624 299.304 2.74

2 97.625 59185 300.093 299.752 299.418 2.05
3 100.93 61189 300.113 299.750 299.397 2.75
4 106.98 64857 300.001 299.604 299.219 3.02
5 202.13 122542 295.945 294.822 293.744 4.01
6 205.98 124876 295.964 294.806 293.700 4.49
7 206.53 125210 295.857 294.695 293.589 4.82
8 241.18 146216 303.736 302.213 300.770 5.25
9 317.63 192565 302.970 300.439 298.045 5.41
10 377.58 228910 300.663 297.253 294.054 6.19
11 379.23 229910 300.616 297.169 293.947 6.53
12 434.23 263254 299.470 295.223 291.253 6.52
13 489.23 296598 299.248 294.018 289.241 8.66
14 506.28 306935 297.635 292.033 286.917 8.68
15 539.28 326941 296.782 290.604 284.962 8.67
16 573.25 347536 296.018 289.202 282.970 8.56
17 605.83 367288 296.183 288.708 281.858 8.36
18 682.83 413969 293.032 283.652 275.029 8.07
19 685.08 415333 293.007 283.561 274.870 7.99
20 722.85 438232 291.817 281.375 271.635 6.72
21 738.93 447980 291.740 280.799 270.509 5.95
22 741.55 449569 291.739 280.787 270.460 5.71
23 799.15 484489 289.531 277.176 265.409 4.76
24 820.47 497414 289.842 276.794 264.309 4.31
25 891.53 540495 286.086 270.723 255.986 4.07
26 926.68 561805 284.837 268.278 252.366 3.91
27 928.18 562714 284.944 268.459 252.545 3.46
28 984.28 596725 283.071 265.039 247.602 3.30
29 986.15 597859 277.090 258.948 241.361 3.06
30 988.02 598993 281.679 263.677 246.226 3.06
It can be seen that the maximum drag-reducing efficiency with both

air and natural gas can surpass 8% (about 8.90% and 8.68%, respec-
tively). Moreover, the drag-reducing efficiency according to Reynolds
numbers is first ascending to the maximum, and then descending.
The experimental results are consistent with the conclusions in the
relevant literature to some extent, which proves the feasibility of
applying the bio-inspired drag-reducing technology to natural gas
pipelining.
Fig. 5.42. Drag-reducing efficiency with air and natural gas.
The total length of the second pipeline in the West-to-East gas

transmission project is 7,256 km, and the designed output is 300 ×
108 km. The total number of compressor stations is 195, the total
power is more than 1,000 MW, and the gas consumption is about
10×108 m3 /yr. If the drag-reduction efficiency is calculated as 7%, the
relevant economic benefit can be achieved as follows: (1) The number of
compressor stations can be reduced to 180, which can lead to lowering
the total cost of 800 million yuan. (2) The total saved gas consumption
is about 70 million m3 . If the market price is calculated as 3 yuan/m3 ,
the savings in natural gas are worth about 210 million yuan.
Due to the addition of the new working procedure in manufac-
turing micro-groove coating, the manufacturing costs will increase
by about 10 yuan/m2 , so the total additional cost is calculated as
following:
10 yuan/m2 × 3.14 × 1.219 m × 7256 × 1000 m = 2.8 × 108 yuan.
Therefore, if the gas pipeline with micro-grooved coating can be
put into application, the total cost of investment will be reduced about
500 million RMB yuan, and the saved fee is about 210 million yuan
every year, so the economic benefit is very considerable.
5.8. Conclusion
In this section, the biomimetic drag-reducing technology is put forward

to be applied in natural gas transportation. The characteristics of
AW-01 epoxy resin, the processing steps of the internal coating
process and pre-cured micro-rolling technology are presented, and the
rolling equipment is designed. In addition, the simulation of the drag-
reduction effect of the new type of gas pipeline is performed, and the
conclusions are as follows:
(1) The convergence of biomimetic drag-reducing technology and
internal coating technology is put forward on gas pipelining, and
the pre-cured rolling technology, which is fit for the forming of the
thermosetting materials, is advanced — the effect is very good.
(2) In order to manufacture the grooved surface, rolling technology is
advanced, and the effective groove surface is machined.
(3) In order to manufacture the bio-inspired micro-grooved surface
and eliminate the errors of the rolling system, the outward-hard
and inner-soft rollers are machined, and the vertically overhanging
method is explored.
(4) For verifying the drag effect of the new pipeline, numerical simu-
lation is carried out, and it is concluded that the new application
can reduce friction and increase the transmission capacity on the
basis of smooth internal coating.
(5) The influence of bugs on grooved surface on the drag-reduction
effect is analyzed, which built the drag reduction foundation of the
biomimetic grooved surface.
(6) Drag-reduction experiments with air and natural gas were carried
out — the results show that the drag-reducing efficiency can
surpass 8%, which is consistent with results in the relevant literature.
This confirms the feasibility of applying bio-inspired drag-reducing
technology to natural gas pipelining.
(7) The economic benefit of adopting the new technology is in
comprehensive calculation and analysis, and the economic benefit
is enormous.
References
[1] Zhang DY, Luo YH, Chen HW. (2011) Application and numerical simulation
research on biomimetic drag-reducing technology for gas pipelining. Oil Gas-
Eur Mag 37(2): 85–90.
[2] Zhang DY, Luo YH, Chen HW, Jiang XG. (2011) Exploring drag-reducing
grooved internal coating for gas pipelines, Pipeline Gas J 238(3): 58–60.
[4] Luo YH, Zhang DY. (2012) Experimental research on biomimetic drag-
reducing surface application in natural gas pipelines. Oil Gas-Eur Mag 38(4):
213–214.
[5] Walsh MJ. (1983) Riblets as viscous drag reduction technique. AIAA J 21(4):
485–486.
157–171.
59–87.
[8] Liu KN, Christodoulout C, Ricciust O, Joseph DD. (1999) Drag reduction in
pipes lined with riblets. AIAA J 28(10): 1697–1698.
[10] Viswanath PR. (2002) Aircraft viscous drag reduction using riblets. Prog Aerosp
Sci 38: 571–600.
[11] Haecheon C, Parviz M, John K. (1993) Direct numerical simulation of turbulent
flow over riblets. J Fluid Mech 255: 503–539.
[14] Luo YH, Zhang DY, Liu YF, Ng EYK. (2015c) Bio/Micro-Rolling Fabrication
of Biological Sharkskin Morphology on Semi-Cured Coating and Drag Force.
Experimental Research 16: 1650016.
[15] Zhang DY, Luo YH, Li X, et al. (2011a) Numerical simulation and experimental
[16] Cong Q, Feng Y, Ren LQ. (2006) Affecting of riblets shape of non-smooth
surface on drag reduction. Journal of Hydrodynamics, Ser. A 21(2): 232–238.
(In Chinese.)
[17] Luo YH, Liu YF. (2014b) Numerical simulation of micro flow field on
biomimetic sharkskin micro-grooved surface. Adv Mater Res 884–885:
378–381.
[18] Luo YH, Liu YF, Zhang D, Ng EYK. (2014c) Influence of morphology for drag
reduction effect of sharkskin surface. J Mech Med Biol 14: 1450029.
Chapter 6
Biomimetic Surfaces for Enhanced

Dropwise Condensation Heat Transfer:
Mimic Nature and Transcend Nature
Youmin Hou,∗,‡ Zuankai Wang,† Shuhuai Yao∗

∗ Department of Mechanical and Aerospace Engineering,
The Hong Kong University of Science and Technology,
Hong Kong, China
† Department of Mechanical and Biomedical Engineering,
City University of Hong Kong, Hong Kong, China

‡ yhou@connect.ust.hk
Water vapor condensation plays a key role in a wide range of industrial

applications involving phase change heat transfer process, such as
power generation, thermal management, water harvesting and desali-
nation. When water vapor is cooled below its equilibrium saturation
temperature at the specified pressure, it can transition to the liquid
phase with a latent energy release during this phase change process.
Compared to the homogeneous nucleation taking place completely
within a supersaturated vapor, the heterogeneous nucleation is more
ubiquitous, and is characterized by a low reduced energy barrier owing
to the presence of solid interfaces. Thus, the heterogeneous nucleation
is intricately reliant on the physicochemical properties of solid surfaces.
For a hydrophilic surface with high excess energy, the condensed water
tends to form a liquid film covering the entire surface, which is termed
as filmwise condensation. In contrast, on a hydrophobic surface, liquid
185
condensates ball up, and it has been shown that the condensation heat
transfer coefficient associated with this so-called dropwise condensation
was tenfold higher than that of filmwise condensation.1−13 Yet, despite
exciting progress, the state of the art condensation materials still could
not satisfy the demanding requirements on the thermal management
in high heat-flux electronic devices,14−16 thermal management in
aerospace and chemical processing, etc.17,18
Recently there has been an exciting revival in engineering bio-
inspired robust dropwise condensation surfaces. This is partially due to
the progress in the fundamental understanding of previous unidentified
physical phenomena during the phase change process, as well as
advances made in the preparation of various structured surfaces. In
particular, over the course of evolution of several million years, many
natural biological organisms have developed exquisite strategies to sur-
vive in and adapt to a dynamically changing environment.19−22 Thus,
nature offers many remarkable inspirations for engineers to develop
novel surfaces to achieve enhanced properties which are impossible
using conventional approaches.23,24 In recent years, inspired by lotus
leaves,25 desert beetles26 and Nepenthes pitcher plants,27 a variety
of functional surfaces have been created to improve condensation
performances.
Fig. 6.1. (a) Filmwise condensation on a smooth copper tube and (b) dropwise
condensation with jumping droplet on a superhydrophobic copper tube.
Biomimetic Surfaces for Enhanced Dropwise Condensation Heat Transfer 187
In this chapter, our objectives are to develop an appreciation for

recent bio-inspired strategies for condensation enhancement as well
as to provide a basis for controlling or predicting the condensation
performance of the biomimetic surfaces.
6.1. Fundamentals of Dropwise Condensation

Understanding the fundamental mechanisms of the interfacial phase
change phenomena at these functional surfaces is crucial for the
optimization of surface design and equipment operation conditions.
The mechanism of dropwise condensation on smooth condensing
surfaces has been well studied over the past 50 years. In an early
study in 1966, Le Fevre and Rose4,28 proposed a model that derived
the heat transfer of dropwise condensation by combining thermal
resistance analysis of a single droplet with effective mean droplet size
distribution. As an extension of such a theoretical modeling concept,
precise calculations of dropwise condensation heat transfer have been
gradually developed by many investigations.29−31
In heterogeneous condensation, the condensed water acts as an
additional thermal resistance to the phase change heat transfer between
the vapor and surface, whether it is in the form of droplet or liquid film.
The thermal resistance increases with the condensed droplet size or
liquid film thickness, thus it is desirable to remove the condensate in a
rapid manner. For dropwise condensation, the condensed droplets can
roll off a vertical hydrophobic surface upon reaching a characteristic
capillary length (∼2 mm) due to the action of gravity. These rolling
droplets can merge and sweep the others in their falling path, thereby
refreshing the condensing surface for water re-nucleation. Because of
the sweeping and renewal of the droplet growth process, dropwise con-
densation has been considered as the preferred mode of condensation.
6.1.1. Individual droplet nucleation and growth

Interfacial water condensation starts from the heterogeneous nucle-
ation of embryo liquid droplet at the solid–vapor interface. According
to the classical nucleation theory, in general, the energy barrier G for
the formation of a liquid nucleus on a surface strongly depends on the
intrinsic wettability of the surface (i.e. the droplet contact angle).

πσlv rcr2
G = (2 − 3 cos θ + cos3 θ) (6.1)
3
where σlv is the liquid–vapor surface energy and rcr is the critical radius,
respectively.
The critical radius is obtained by Kelvin’s classical equation,
ln(p/p∞ ) = 2σlv /nl kTrcr , where p is the vapor pressure over a curved
interface of radius rcr , p∞ is the equilibrium vapor pressure above a flat
surface of the condensed phase at temperature T , nl is the number of a
water molecules per unit volume of the liquid, and k is the Boltzmann
constant. Accordingly, the nucleation rate J a water nucleus is given by
J = J0 exp (−G/KT ), where J0 is a kinetic constant.
It indicates that the nucleation rate of an embryo liquid droplet on a
hydrophilic surface is much higher than that on a hydrophobic surface.
In other words, the vapor can condense on hydrophilic surfaces at a
lower subcooling or supersaturation as compared to hydrophobic
surfaces. It is important to note that although the small contact
angle of hydrophilic materials benefits the initial water nucleation,
the consequent large pinning force greatly hinders the heat transfer
coefficient after the initial period due to water accumulation on the
condensing surface. Therefore, dropwise condensation has long been
recognized as the more efficient condensation mode for phase change
heat transfer.
For an individual droplet, a thermal circuit model is established
to account for different thermal resistances between the vapor and
condensing surface, including the resistance due to the curvature of
the droplet (Tc ), the liquid–vapor interfacial resistance due to direct
vapor molecule accommodation at the droplet interface (Ti ), the
conduction resistance through the droplet (Td ) and the hydrophobic
coating resistance (Thc ).
The temperature drop is due to droplet curvature (Tc ) given by
2Tsat σ
Tc = (6.2)
rhfg ρw
where Tsat is the water vapor saturation temperature, σ is the water
surface tension, hfg is the latent heat of vaporization, and ρw is the
liquid water density.
Since the droplet surface curvature, along with condensate sur-

face tension and surface subcooling (supersaturation) determines the
minimum thermodynamically-viable droplet radius rmin ,
2Tsat σ
rmin = (6.3)
hfg ρw T
Equation (6.2) is reduced to
rmin
Tc = (Tsat − Ts ) (6.4)
r
where rmin is the minimum viable droplet radius and Ts is the
condensing surface temperature.
The liquid-vapor interfacial resistance is responsible for the tem-
perature drop Ti between the vapor and the droplet surface which is
given by
q
Ti = Tsat − Ti = (6.5)
hi 2πr (1 − cos θ)
2
where q is the heat transfer rate through the droplet and hi is the
condensation interfacial heat transfer coefficient given by
1/2 2
2α M hfg
hi = (6.6)
2 − α 2πRTS vg T S
where α is the condensation coefficient; it is taken as unity (ranging
from 0 to 1) in the calculations which represent the ratio of vapor
molecules that will be captured by the liquid phase to the total number
of vapor molecules reaching the liquid surface. For the pure vapor
environment without noncondensable gas, it is appropriate to assume
α = 0.9. M is the gas molecular weight, R is the gas constant and νg is
the water vapor specific volume.
In general, when heat is transferred through the droplet to the
contact base, the conduction is the primary mechanism because the
condensed droplets are sufficiently small. Therefore, in the calculation,
the internal convection of the droplet can be neglected, and this
resistance is modeled as a pure conduction resistance, leading to the
droplet conduction temperature drop (Td ),
qθ
Td = Ti − Tb1 = (6.7)
4πrkw sin θ
where Tb1 is the liquid temperature of the droplet base and kw is the
condensed water thermal conductivity.
The temperature drop due to the resistance of the hydrophobic

coating material on the condensing surface is given by
qδhc
Thc = Tb1 − Ts = (6.8)
πr 2 khc sin θ
2
where Ts is the temperature of the condensing surface beneath the

hydrophobic coating, δhc is the hydrophobic coating thickness, and khc
is the coating thermal conductivity.
The total temperature difference between the vapor and the
condensing surface is the sum of the four temperature drops calculated
in Eqs. 6.2–6.8.
T = Tc + T i + T d + T hc

1 1 1 rθ δhc
=q 2 + +
πr 1 − rmin /r 2hi (1 − cos θ) 4kw sin θ khc sin2 θ
(6.9)
Thus, the heat transfer rate through a droplet with the radius of r
is represented as
T πr 2 (1 − rmin /r)
q= δ
(6.10)
1
2hi (1−cos θ)
+ 4kwrθsin θ + k sin 2θ
hc
According to the single droplet growth model, the condensation

heat flux through a droplet highly depends on the droplet radius and
contact angle. For dropwise condensation with a constant contact
Fig. 6.2. (a) Schematic of dropwise condensation on a hydrophobic surface

(b) Droplet thermal resistance diagram showing the liquid–vapor interface (Ri ),
droplet conduction (Rd ), hydrophobic coating (Rhc ) thermal resistances.
angle, the smaller droplet promises a larger heat flux on the condensing
surface. It suggests that a condensing surface covered by small droplets
enables better condensation performance when compared to that
covered by large droplets.
6.1.2. Droplet size distribution in dropwise

condensation
The analysis of single droplet growth indicates that the size distribution
of condensed droplets has significant impact on overall dropwise con-
densation performance since more condensation heat transfer occurs
through small droplets and thus the surface occupied by large droplets
is considered to be an ineffective heat transfer area. To estimate the
steady state droplet size distribution during condensation, Tanaka
introduced the population balance theory by taking into account the
two mechanisms for droplet growth: direct vapor accommodation onto
the droplet surface and the coalescence with neighboring droplets. The
derivation of the condensed droplet size distribution is based on the
droplet number conservation in a certain size range, i.e., the number
of droplets entering a given size range equals the number of droplets
leaving the same size range.
Assuming the droplet contact angle remains constant from the
initial nucleation to the departure from the surface, the droplet growth
rate (G) is given by
dr
G= (6.11)
dt
The population density of droplets n(r) is defined as the number
of droplets of radius r per unit area. For an arbitrary droplet size range
r1 − r2 , the number of droplets entering or leaving the range by growth
can be expressed as An 1 G1 dt or An 2 G2 dt , respectively. In the size
range, the number of droplets swept by the falling droplets is defined
as Sn 1−2 rdt , where S is the sweeping rate at which the substrate
surface is renewed by falling droplets, n1−2 is the average population
density in the size range r1 and r2 , and r equals r2 −r1 . By considering
the population balance for condensed droplets, the number of droplets
entering by growth must equal the sum of the number of droplets
leaving by growth and the number of droplets swept away,
An1 G1 dt = An2 G2 dt + Sn1−2 dt (6.12)
Equation 6.12 can be simplified as
A(G2 n2 − G1 n1 ) = −Sn1−2 r (6.13)
As r approaches zero, n1−2 becomes a point value, and the

equation can be written as
d(Gn) n
=− (6.14)
dr τ
where τ is the sweeping period defined as τ = A/S.
The heat transfer rate through a droplet, q, is related to the
droplet growth rate G = dr/dt by calculating the differential enthalpy
increment of the newly condensed vapor:
dV
q = ṁhfg = ρw hfg = πr 2 ρw hfg (1 − cos θ)2 (2 + cos θ)G (6.15)
dt
Substituting Eq. 6.15 into Eq. 6.10, the droplet growth rate can
be calculated as
T 1 − rmin /r
G= × δhc
ρw hfg (1 − cos θ)2 (2 + cos θ) 1
2hi (1−cos θ)
+ rθ
4kw sin θ
+ khc sin2 θ
1 − rmin /r
= A1 (6.16)
A2 r + A 3
where
T
A1 = (6.17)
ρw hfg (1 − cos θ)2 (2 + cos θ)
θ
A2 = (6.18)
4kw sin θ
1 δ
A3 = + (6.19)
2hi (1 − cos θ) khc sin2 θ
Integrating Eq. 6.14 with respect to r,

Gn r
d(Gn) dr
= − (6.20)
Gnmin Gn rmin τ
Solving Eq. 6.20, the size distribution for small droplets growing
by direct condensation is given by
Gn min
n(r) =
Gn

A2 (r − rmin
2
)
× exp +2rmin (r − rmin ) + rmin ln (r − rmin )
2
τA 1 2

A3
+ [r − rmin + rmin ln(r − rmin )] (6.21)
τA 1
For large droplets grown by coalescence, the size distribution N (r)

was established by Le Fevre and Rose,
−2/3
1 r
N (r) = (6.22)
3πr 2 rmax rmax
where rmax is the effective maximum droplet radius (droplet departure

radius) which is estimated based on the force balance between droplet
adhesion Fc and gravity Fg ,
Fc = c2r sin θσ (cos θr − cos θa ) (6.23)

2 − 3 cos θ + cos3 θ 3
Fg = πr ρg (6.24)
3
where θr and θa are the receding and advancing contact angles,
respectively. c refers to a numerical constant depending on the droplet
shape and the substrate’s stiffness. From the force balance between Fc
and Fg , the radius of sliding droplets can be deduced as
12
6c (cos θr − cos θa ) sin θ σ
rmax = (6.25)
π 2 − 3 cos θ + cos3 θ ρg
As previously mentioned, re is defined as the effective radius when

droplets growing by direct vapor condensation begin to merge and
grow by coalescence. Because this radius is equal to half of the mean
spacing between the active nucleation sites on the condensing surface,
by assuming that the nucleation sites form a square array, re can be
written as,
re = (4Ns )−1/2 (6.26)
where Ns is the number of nucleation sites on a unit area of condensing
surface. Because the population density n(r) and N (r) should keep the
continuity at the boundary of effective radius, the boundary condition is
set as n(r) = N (r) at r = re . Thus, the governing equation in Eq. 6.21
with this boundary condition is solved analytically, resulting in

1 re −2/3 r(re − rmin ) A2 r + A3
n(r) = exp(B1 + B2 )
3πre3 rmax rmax r − rmin A2 re + A3
(6.27)
where

A2 re2 − r 2 r − rmin
B1 = + rmin (re − r) − rmin ln
2
(6.28)
τA1 2 re − rmin

A3 r − rmin
B2 = re − r − rmin ln (6.29)
τA1 re − rmin
Deduced from the boundary condition,
d ln n(r) d ln N (r) 8
= =− (6.30)
d ln r d ln r 3
The sweeping period τ, is expressed as
3re2 (A2 re + A3 )2
τ= (6.31)
A1 (11A2 re2 − 14A2 re rmin + 8A3 re − 11A3 rmin )
Therefore, the total surface steady state condensation heat flux q
was obtained by incorporating the individual droplet heat transfer rate
(Eq. 6.10) with the droplet size distributions (Eqs. 6.22 and 6.27)
re rmax
q = qd (r)n(r)dr + qd (r)N (r)dr (6.32)
rmin re
The theoretical analysis of dropwise condensation heat transfer

evidently indicated that the reduced droplet average size, lower heat
transfer resistance, as well as higher droplet density are crucial for the
efficient dropwise condensation heat transfer. Recent advancements
in fabrication techniques have allowed for the development of many
biomimetic functional surfaces, where nearly spherical water droplets
form with high mobility and minimal droplet adhesion. In addition, the
role of surface structuring on wetting characteristics has been studied in
detail32−35 to enhance condensation performance by reducing droplet
departure sizes and enabling faster clearing of the surface for re-
nucleation.
6.2. Jumping Droplet Condensation

on Superhydrophobic Surfaces
When vapor condenses on non-wetting flora and fauna, such as a lotus
leaf23 and a cicada wing,36,37 an interesting physical phenomenon of
droplet spontaneous jumping associated with condensation on super-
hydrophobic surface occurs. This has attracted considerable attention
from scientists and engineers. When the small condensed droplets
(∼10–100 um) merge on a superhydrophobic nanostructured surface,
the coalesced droplets can be self-propelled due to the release of
excess surface energy.38−42 This finding provides an exciting approach
to enhance the dropwise condensation efficiency since the droplet
departure size is significantly reduced by the coalescence-jumping
behavior.43
With the rapid development of surface engineering in the past few
years, artificial superhydrophobic surfaces could be simply reproduced
by using a combination of low surface energy materials and nanoscale
surface roughness.44 Generally, in static wetting conditions, deposited
sessile droplets on these superhydrophobic surfaces can achieve the
apparent contact angles exceeding 150◦ and contact angle hysteresis
approaching 0◦ , showing a strong and steady non-wetting property.
However, the dynamic droplet wetting is much more complicated
during the phase change process, that is, the condensed droplet may
remain either in suspended state or penetrate into the nanostructures
due to the slight variation in condensation conditions or length scale of

surface roughness. As shown in Fig. 6.4, the superhydrophobic surface
loses its non-wetting property when surface temperature gradually
decreases to ∼0◦ C.45 This degradation of static droplet contact angle
can be attributed to the condensed liquid bridges within the nanos-
tructures before the droplet deposition. The coalescence of deposited
droplet base and underlying condensate results in Wenzel droplet
formation. The distinct droplet wetting features would profoundly
impact the droplet adhesion and individual droplet growth during
condensation. As a result, recent studies have focused on understanding
the dependency of droplet wetting morphology on the surface structure
length scale and seeking the optimal surface design for the efficient heat
transfer applications.
Fig. 6.3. Natural superhydrophobic surfaces and corresponding droplet jumping

behavior. Scanning electron micrograph of lotus leaf (a) and cicada wing.36 (b) show-
ing the surface structures. (c) Time lapse images of condensation showing the
spontaneously droplet self-propelling triggered by coalescence.38
Fig. 6.4. (a) Scanning electron micrograph of the artificial superhydrophobic

nanostructured copper oxide surface fabricated by immersion in 2.5 M NaOH and
0.1 M (NH4 )2 S2 O8 at 4◦ C for 5 min. (b) Apparent contact angle variation on the
copper superhydrophobic surfaces (a) in different surface temperatures. The significant
decrease in contact angle is caused by the coalescence of deposited sessile droplet with
underlying tiny Wenzel-state droplets condensed earlier on the as-prepared surface.45
6.2.1. Dynamic condensation morphologies

On an ideal, smooth surface, the droplet wetting at equilibrium static
state is usually described by the intrinsic contact angle (θ), which is
calculated by Young’s equation46 :

−1 γsv − γsl
θ = cos (6.33)
γlv
where γsv , γsl and γlv are the solid–vapor, solid–liquid, and liquid–vapor
interfacial tensions, respectively. On this basis, Cassie47 and Wenzel48
proposed two different mechanisms to describe the surface roughness
effect of equilibrium droplet wetting on superhydrophobic surfaces,
cos θcapp = ϕ(cos θ + 1) − 1 (6.34)
cos θwapp = r cos θ (6.35)
app app
where θc and θw denote the apparent Cassie and Wenzel contact
angle, respectively. ϕ is the ratio of the solid area contacting the droplet
to the projected area. The surface roughness r is the ratio of the actual
total surface area to the projected area, θ is the intrinsic Young’s contact
angle.
In the case of condensation on superhydrophobic surfaces, how-

ever, the droplet growth presents a dynamic wetting process, which
cannot be simply described by Eqs. 6.34 and 6.35. From the observa-
tion of water condensation by using environmental scanning electron
microscopy (ESEM), the nanoscale condensate initially nucleates on
the top of nanostructures as well as within the interspace on the
superhydrophobic surfaces.49 The subsequent droplet growth, how-
ever, shows distinct wetting modes due to the different length scale
of surface roughness. Specifically, three general droplet morphologies
have been identified from a number of experimental studies, as shown
in Fig. 6.5.
Fig. 6.5. Environmental scanning electron microscopy (ESEM) images of droplets

condensed with the (a) suspended Cassie, (b) partial wetting, and (c) highly pinned
Wenzel morphology on a nanostructured surface, scale bar is 10 µm. The time lapse
schematics of (d) Cassie, (e) partial wetting and (f) Wenzel droplets showing the
different growth process during condensation on nanostructured surface (schematics
not to scale).
For the droplet nucleating on top of the nanostructures, it is

unconditionally stable at the suspended Cassie state when it grows in the
absence of interactions with other droplets (Fig. 6.5a). For the droplet
nucleating within the nanostructure, it grows to fill the interspace, and
then either grows above the nanostructures to form a partial-wetting
Cassie droplet with a liquid bridge at the base (Fig. 6.5b), or laterally
spreads inside the nanostructures to form a Wenzel droplet with a
completely wetting base (Fig. 6.5c). Enright et al. proposed an energy
criterion to conservatively estimate the formation of partial wetting and
Wenzel droplets on superhydrophobic surfaces,50
cos θcadv 1
E∗ = =− (6.36)
cos θwadv r cos θ
where θcadv and θwadv denote the droplet advancing contact angles in
Cassie and Wenzel states, respectively. The dimensionless ratio E ∗
represents the energetic comparison of non-equilibrium advancing
Cassie and Wenzel states. When E ∗ < 1, the partial-wetting Cassie
droplets should emerge, while when E ∗ > 1, the nanodroplet is apt
to grow as highly pinned Wenzel morphology. It is important to note
that the nucleation density also impacts the wetting mode transitions
during droplet growth. If the mean spacing between the droplet nuclei
(active nucleation sites) is separated by 2∼5 times larger than the
characteristic pitch spacing of surface structures, the emergence of
partial-wetting or Wenzel droplets should follow the estimation of E ∗ . If
droplets nucleate too close to each other, the droplet interactions would
occur on a similar length scale as the nanostructure interspace, leading
to Wenzel droplet formation without regard to the energy criterion
in Eq. 6.36.
6.2.2. Enhanced heat transfer via jumping-assisted

droplet departure
When the droplets grow and begin to merge with others, self-removal
via coalescence-induced droplet jumping can be observed for both
suspended and partial-wetting Cassie droplets on superhydrophobic
surfaces. This suggests that such two droplet wetting morphologies
are in favor of the spontaneous jumping in condensation because

the droplet contact pinning is below the critical threshold of self-
removal. In contrast, the Wenzel droplets demonstrate the large droplet
departure diameters owing to the significant adhesion of complete
wetting droplet base. Unlike the conventional droplet shedding by
gravity on hydrophobic surfaces, the spontaneous jumping behavior
greatly decreases the droplet average size during condensation. The
experimental measured data showed that in condensation on the copper
oxide superhydrophobic surface with high nucleation density, N ≈
5 × 109 m−2 , the mean spacing of droplet coalescence was ∼8 µm and
the efficient droplet self-removal enabled more than 75% of condensed
droplets with diameter smaller than 10 µm.51 As a result, according to
the theoretical thermal analysis of dropwise condensation in Eqs. 6.10
and 6.32, the overall heat transfer performance can be significantly
improved due to the highly reduced average size of condensed
droplets.
Miljkovic first underscored the importance of droplet wetting mor-
phology on condensation thermal resistance via an in situ ESEM study
of the suspended and partial-wetting Cassie droplet morphologies on
superhydrophobic nanostructured surfaces (Fig. 6.5b, e). Although the
suspended state of condensed droplets was considered to be essential
for the droplet self-removal, the gas trapped within the nanostructures
underneath droplets may form an additional thermal barrier and greatly
hinder the overall heat and mass transfer performance.35,52,53 Their
experimental results showed that if both suspended and partial-wetting
Cassie droplets were self-propelled at identical size, the heat transfer of
partial-wetting droplets was 4∼6 times higher than that of suspended
Cassie droplets. Accordingly, the previous dropwise condensation heat
transfer model in Eq. 6.10 was further extended by considering the
influences of droplet wetting morphology associated with the thermal
resistances of surface structures (Fig. 6.6).
T πr 2 (1 − rmin /r)
qpw =
−1 (6.37)
kP ϕ kw (1−ϕ)
1
2hi (1−cos θ)
+ 4kwrθsin θ + k sin
1
2 θ δ k +hk
hc P hc
+ δhc kw +hkhc
hc
where ϕ is the surface solid fraction, and kp and h denote the thermal
conductivity and height of nanostructures, respectively.
Fig. 6.6. Schematic of thermal circuits in a partial-wetting droplet showing the heat
transferred through the droplet conduction resistance (Rd ) and the parallel paths
including (i) the hydrophobic coating (Rhc ) followed by the pillar (Rp ) and (ii) the
liquid bridge (Rlb ) followed by the hydrophobic coating (Rhc ). (Schematic is not to
scale.)
Based on the updated theoretical analysis, the superhydrophobic

nanostructured surfaces had 56% condensation heat flux enhancement
for partial-wetting droplet morphology and 71% heat flux degradation
for suspended morphology in comparison to smooth hydrophobic
surfaces.51 The experimental study of condensation on a copper
oxide superhydrophobic nanostructured surface verified the developed
theoretical prediction. In the thermal characterization, the tested
superhydrophobic surface was energetically favorable for the formation
of partial-wetting Cassie droplet during condensation at low supersat-
urations (S < 1.12), and thus can achieve 25% higher overall heat flux
and 30% higher condensation heat transfer coefficients compared to a
smooth hydrophobic surface. Nevertheless, when supersaturations are
higher (S > 1.12), the excessive droplet nucleation density resulted in
the liquid film spreading within the nanostructures. Thus, the droplet
wetting morphology transitioned from mobile jumping droplets to the
highly pinned Wenzel droplets, as shown in Fig. 6.7. This surface
flooding led to a 40% degradation of condensation heat transfer
coefficient as compared to the convention dropwise condensation on
smooth surface.
Fig. 6.7. (a) Jumping droplet condensation on a superhydrophobic nanostructured

CuO tube with supersaturation of S = 1.08 (inset: magnified view of the jumping
phenomena, scale bar is 500 µm). (b) Wenzel droplet formation resulted from flooded
condensation on the same CuO tube with supersaturation of S = 1.54.51
6.2.3. Hierarchical superhydrophobic structures

for continuous droplet condensation
In order to avoid the surface flooding and Wenzel droplet formation,
some hierarchical superhydrophobic surfaces with multiscale roughness
were designed by mimicking the natural two-tier architectures of
the lotus leaf, which consists of microscale papillae and nanoscale
surface roughness. The three-dimensional multiscale surface architec-
tures result in a mixed Wenzel–Cassie droplet mode during water
condensation, in which the microstructures are preferentially wetted
while nanostructures remain dry. When the rapid growing droplets
on microstructures merge with each other, the hierarchical structures
enable the wetted droplet base to recede from the micro-patch and
transition to the mobile Cassie state. As a result, this hierarchical
superhydrophobic surface can achieve the continuous jumping droplet
condensation with a long period of time as compared to these
conventional superhydrophobic surfaces.49,54−58
Chen et al. first reported a significant promotion of droplet self-
removal via a well-designed micropyramid superhydrophobic surface
with multiscale roughness.56 The microstructures were wisely fabri-
cated in a pyramid configuration with vertical smooth base for droplet
fast growth and inclined nanograssed sidewalls for efficient droplet
departure (Fig. 6.8a, b). In order to study the impact of microstructure
Fig. 6.8. (a) Scanning electron micrograph of the novel hierarchical nanograssed
micropyramid arrays showing the micropyramids with smooth base, nanograssed
sidewalls and interspace. (b) Schematic of condensation dynamics on the hierarchical
surface. The synergy between the micro and nanoscale roughness features results in
the stable Cassie droplet morphology and upwards surface tension force, which assist
in droplet departure. (Schematic is not to scale.) (c) Time-lapse ESEM images of
condensation on the hierarchical surface showing the spontaneous droplet jumping
and surface renewal. Red dashed circles highlight areas of the surface with droplets just
prior to coalescence and subsequent jumping, while blue dashed circles highlight areas
of the surface right after droplet jumping. The center-to-center spacing of nanograssed
micropyramid arrays is 20 µm.56
design on the overall condensation performance, the base dimension

of the micropyramids is ∼14 µm × 14 µm and the center-to-center
spacing of the micropyramid arrays are patterned as 20 µm, 30 µm,
and 40 µm (S20, S30 and S40), respectively. In the condensation
experiment under ambient conditions, the hierarchical surface with
Fig. 6.9. Comparison of dropwise condensation dynamics on surfaces with various

structure designs. Nano, S20, S30, and S40 denote nanograssed surface and hier-
archical surfaces with the center-to-center spacing of 20, 30, 40 µm, respectively.
(a) Measured droplet number densities on the various surfaces. Error bars are from
multiple measurements at different areas of the same sample. (b) Quantification
of the cumulative droplet departure volumes on four different surfaces. (c) His-
togram of droplet diameters, averaged in 8-min condensation, on four different
surfaces.56
spacing of 20 µm not only had the highest percentage of droplets with

diameter smaller than 10 µm (∼60%), but also achieved 450% increase
in the cumulative droplet departure volumes when compared to the
superhydrophobic nanograssed surface. These results highlight the
importance of hierarchical surface structures on enhanced condensation
performance, and also imply that the microscale features should be
appropriately arranged to benefit the droplet growth and coalescence-
induced droplet jumping simultaneously.
The enhanced self-propelled droplet jumping by hierarchical

surface opens up new horizons for biomimetic surface design. The
synergistic cooperation between hierarchical roughness structures con-
tributes directly to long-term dropwise condensation, which is essential
for practical implementations. With further understanding of the
inherent relation of multiscale structure length scales, the optimized
scalable hierarchical superhydrophobic surface can be widely used in
various industrial applications associated with condensate transport,
such as vapor chamber, water harvesting and dehumidifiers, etc.
6.3. Spatially Controlled Condensation

via Heterogeneous Wettability
Although superhydrophobic surfaces exhibit superior droplet depar-
ture efficiency, the higher nucleation energy barrier of hydrophobic
materials/coatings increases the subcooling temperature (supersatu-
rations) required for the initial water condensation on surfaces, as
mentioned in Section 6.1.1. Such a large temperature difference
between vapor and condensing surface, to some extent, would limit
the overall heat transfer coefficient during the dropwise condensation.
Interestingly, a Stenocara beetle surviving in the Namib Desert of
southern Africa inspired the scientists to solve this issue. The Namib
Desert is one of the most arid areas of the world, but this beetle
is able to live there by collecting condensed water on its bumpy
back surface (elytra) from fog-laden wind every early morning. In
2001, Parker first revealed the mechanism of water capture by the
unique surface structures with hybrid (biphilic) wettability on Stenocara
beetle’s elytra.59 On a macroscopic scale, the elytra are covered
with random bump arrays with diameter of ∼0.5 mm and spacing
of ∼0.5–1.5 mm. (Fig. 6.10a). On the microscopic scale, the peaks
of the bumps are smooth without covering wax, providing a strong
hydrophilicity during the condensation. Yet, the sloping sides and inter
valleys are covered in microstructures with a coating of hydrophobic
wax (Fig. 6.10b). The scanning electron micrograph (SEM) showed
that the natural microstructure consists of flattened hemispheres with
Fig. 6.10. (a) Image of over-wings (elytra) of adult female desert beetle Stenocara
showing the hybrid (biphilic) bump structures. (b) Macroscopic image showing
the waxes test on bump of the elytra. Underlying areas of the elytra are stained
positively (waxy, colored), whereas the peaks of the bumps remain unstained (wax-free,
black). (c) Scanning electron micrograph of the surface structures on the underlying
areas.59
diameter of 10 µm which are patterned in regular hexagonal arrays

(Fig. 6.10c).
When fog contacts the elytra surface, water preferentially condenses
on the peak of bumps because of the extreme low nucleation energy
barrier on these hydrophilic areas. The higher wettability of the bump
peak prevents the condensed water from being blown by wind and
also benefits the water accumulation by merging the water striking
on the surrounding hydrophobic slopes. Until the droplet contact
pinning is overcome by gravity or other external forces, the condensed
droplets detach from the bumps and roll down the beetle’s back to its
mouthparts.
6.3.1. Increased nucleation rate on hybrid surfaces

The incredible water capture ability of Namibian beetles indicates that
droplet nucleation depends strongly on the intrinsic wettability of the
surface. According to the classical nucleation theory in Eq. 6.1, the
nucleation energy barrier (G) continuously increases with Young’s
contact angle of the condensing surface, suggesting that the vapor
has much more difficultly forming a nucleus on hydrophobic surfaces
as compared to hydrophilic surfaces under identical conditions.60
Therefore, by patterning the condensing surface with hydrophobic
and hydrophilic regions (biphilic areas), the droplet nucleation sites
Fig. 6.11. Optical (a, c) and environmental scanning electron microscope (b, d)
images of water condensation on blade surface. (a–b) Random water nucleation on
the homogeneous hydrophobic blade surfaces resulting in the liquid bridge within
the interspaces. (c–d) Spatially controlled water nucleation on the hydrophilic top of
PVA-coated blade surface resulting in simultaneous and uniform growth of suspended
droplets.62
and wetting morphology can be manipulated during the condensa-

tion process. Varanasi et al.61 and Mishchenko et al.62 reported the
spatial preference for water condensation, as shown in Fig. 6.11.
The experiment results indicated that such top-hydrophilic designs of
microstructures initially promote the growth of Cassie droplets instead
of forming the liquid bridges in surface cavities due to the spatially
controlled water nucleation.
With the aim of creating wettability contrast for spatially con-
trolled water nucleation, the hybrid surfaces were developed to effi-
ciently prevent the droplet wetting within the surface roughness at
the initial condensation stage. However, the condensed droplets on
the structured hybrid surface show a significant pinning force because
of the higher interfacial free energy between the droplet base and
hydrophilic areas.26,63,64 As a result, these droplets cannot depart
from the hydrophilic patches via coalescence-jumping during the
merging process. This immobile coalescence would disrupt the droplet
suspension state and lead to filmwise condensation on the hybrid
surfaces.63
6.3.2. Mutual reinforcement of “Namibian beetles”

and “lotus leaves”
The top-hydrophilic hybrid microstructured surfaces enable the con-
densates to form Cassie droplets on the top of surface roughness in a
metastable state. Yet these suspended droplets would gradually fall into
the cavities of microstructures due to immobile droplet coalescence.
Thus, an efficient approach to promote condensate departure from
hydrophilic areas is to maintain the enhanced droplet nucleation rate
and spatial control of condensed droplets for long-term condensation
on hybrid surfaces. Recently, Hou et al. developed a novel hybrid
surface with patterned high-contrast wettability, which intelligently
introduced the superhydrophobic nanostructures of the “lotus leaf”
to cooperate with the hybrid pattern of “Namibian beetles,” as shown
in Fig. 6.12a.65 By confining the hydrophilic patches on the top of
micropillars surrounded by superhydrophobic nanograss, such a novel
hybrid surface enables not only higher density water nucleation, but also
droplet depinning on the extreme hydrophilic areas. Particularly, the
unique morphological and wetting features achieve a natural recurrent
transition from filmwise condensation to dropwise condensation, which
is distinct from conventional superhydrophobic surfaces (Fig. 6.12b).
Unlike immobile coalescence on microstructured hybrid surfaces,
the filmwise-to-dropwise transition on the hybrid surface enabled the
droplet to coalesce with a large contact angle (>165◦ ). It provided
sufficient driven energy to overcome the contact adhesion on the
hydrophilic patterns, naturally resulting in a droplet self-propelling. As
a result, the active nucleation sites of hydrophilic areas can be exposed
for a new round of filmwise condensation and following morpho-
logical transitions. Such a remarkable recurrent process on a hybrid
surface allows for the mutual reinforcement of filmwise and dropwise
condensation by judiciously optimizing the heterogeneous structure
length scale. The experimental comparison of condensation dynamics
on various condensing surfaces demonstrated that the proper surface
design of hydrophilic and superhydrophobic structures can achieve
intensive droplet nucleation, rapid growth and efficient departure
simultaneously. The thermal measurement also showed that the heat
transfer coefficient on the optimized hybrid surface demonstrated a 63%
Fig. 6.12. (a) Scanning electron micrograph of the hybrid nanostructured silicon
surface consisting of micropillar arrays and surrounding nanograss. (b) Schematics of
the hybrid surface showing the heterogeneous wettability. The top of the micropillar
(dark red area) is hydrophilic, which is energetically favored for droplet nucleation and
filmwise growth. A hydrophobic layer is selectively coated on the nanograss (green
area), providing a global superhydrophobicity for dropwise condensation. (c) Time-
lapse ESEM images of water vapor condensation on hybrid nanostructured silicon
surface. The droplets (e.g., A, B, C) on the nanograss grow in the CCA (constant
contact angle) mode, while droplets (e.g., D, E, F) on the hybrid surface grow in
the CCL (constant contact line) mode. The distinct morphology of CCL droplets
demonstrates the condensation transition from the filmwise mode to dropwise mode.65
enhancement as compared to the conventional dropwise condensation

on the flat hydrophobic surface, as shown in Fig. 6.13.
The similar phase change phenomenon was observed on bottom-
hydrophilic hybrid surfaces with combination of superhydrophobic
nanoroughness. He et al. developed a hierarchical superhydropho-
bic porous surface comprising micro-pores with hydrophilic base.66
Fig. 6.13. Condensation morphologies and corresponding heat transfer measure-

ments on various condensing surfaces, including (a) flat hydrophilic (b) flat hydropho-
bic (c) nanograssed superhydrophobic surface, and (d) hybrid surface. (e) Steady-state
overall surface heat flux q as a function of subcooling T for different condensing
surfaces, undergoing filmwise, dropwise, and hybrid-wetting modes. (f) Steady-state
condensation coefficient h as a function of the partial pressure of saturated vapor Pv for
different condensing surfaces. Error bars indicate the propagation of the measurement
errors associated with the temperatures of bulk gas, substrate back, and heat sink, as
well as the humidity measurement.65
Fig. 6.14. Optical microscopy images of condensation process on (a) the micro/
nanoporous superhydrophobic surface, (b) the micro/nanoporous superhydrophobic
surface modified with PVA, (c) the micro/nanoporous superhydrophobic surface
modified with PAM; scale bar for all photo-micrographs represents 20 µm. The time
scale is given in the images. The optical micrographs show controllable condensation,
coalescence, and self-removal on the micro/nanoporous superhydrophobic surface
modified with hydrophilic polymers.67
By properly controlling the wettability of micro-pores, the bottom-

hydrophilic hybrid surface demonstrated a significant increase in the
droplet condensation rate and self-removal efficiency (Fig. 6.14). This
combinatorial advantage of nucleation spatial control and droplet self-
departure is promising for enhancing the condensation performance
on the basis of superhydrophobic surfaces.67 Nevertheless, the sys-
tematic thermal characterizations and unified heat transfer models of
these hybrid surfaces are still limited. Therefore, further investigation
on hybrid surfaces is needed for the future development of hybrid
surfaces.
6.4. Directional Condensate Departure

on Gradient Surfaces
Besides jumping droplet condensation, many plants and animals pro-
vide some other effective ways to promote condensate motion. The
Cactaceae species living in deserts have developed their unique surface
structures, including spines, barbs, grooves, and trichomes, for water
collection and transportation.68,69 As shown by Fig. 6.15, small droplets
start to condense on the barb and at the tip of the spine, then move
directionally along the grooved spine and coalesce into big droplets.
The big droplets then further move along the spine until reaching the
bottom and are finally absorbed by the belt-structured trichomes. The
similar phenomena of directional water movement have also been found
on spider silks,21 butterfly wings,70 lotus leaf margin,71 rice and bamboo
leaves.72,73
6.4.1. Rapid droplet movement induced by Laplace

pressure difference
Such spontaneous condensate transport can be achieved by a gradient of
Laplace pressure due to gradually varied surface structure or surface free
energy. As the driving force of droplet movement, the Laplace pressure
Fig. 6.15. (a) SEM image of cactus O. microdasys, showing a gradient surface
structures. Scale bar: 100 µm. (b) Schematic image showing the spontaneous water-
collection process by the cactus stem.
difference p can be calculated using the Laplace–Young equation:

1 1
p = γlv + (6.38)
r1 r2
where r1 is the curvature radius at the left side of the surface, and r2
is the curvature radius at the right side of the surface. γlv is the surface
tension of the liquid–vapor interface.
Contact-angle hysteresis is the resistance for droplet movement,
which is obtained by
F = γlv l (cos θr − cos θa ) (6.39)
where l is the droplet characteristic length, θa is the advancing contact
angle, and θr is the receding contact angle.
Yang et al. designed a hydrophobic surface consisting of a structural
gradient that was similar to the gradient in spider silk and cactus,
and observed the spontaneous motion of droplet with diameter of
500 µm.74 The Laplace pressure driven droplet movement can also
be achieved by creating surface energy gradient. Chaudhury and
Whitesides fabricated silicon surfaces with a surface energy gradient
through the control of the chemical composition of silanes on the
surface.75 Such a surface energy gradient generated a force gradient
around the contact line, and drove the droplet towards the direction of
increasing wettability. Khoo et al. further revealed that the velocity of
droplet self-motion reached up to 0.5 m/s via a well-designed wedge
shaped wettability gradient.76
6.4.2. Enhanced condensation heat transfer

on gradient surfaces
From condensation experiments on the surface with energy gradients,
Chaudhury et al. found the merging droplets moved toward more
hydrophilic regions, which was similar to the droplet motion on a
cactus.75 Guided by the surface energy gradient and propelled by the
released energy during coalescence, the small droplets (0.1∼0.3 mm)
reached a speed of 20∼40 µm/s, much faster than movement in
Marangoni flows. As compared to the filmwise condensation mode,
the heat transfer coefficient of heat exchangers can be increased
by a factor of 3∼10 depending on different surface subcooling.
Fig. 6.16. Selected images showing the directed motion of water droplets (as streaks
of light) from hydrophobic to hydrophilic stripes when a mist of water drops is
produced on the surface at room temperature. The arrows indicate the direction of
the drops when they get pulled into the hydrophilic regions.77
Most recently, Chaudhury et al. presented an approach not only

to simplify the gradient surface fabrication but to also achieve the
same condensate autonomous transportation from hydrophobic to
hydrophilic regions.77 Such a condensing surface consists of alternate
stripes of hydrophobic/hydrophilic zones, as shown in Fig. 6.16.
During the condensation experiments, the vapor condensed in a thin
film on the hydrophilic stripes while in droplets on the hydropho-
bic stripes. As the coalesced droplets reached near the hydrophilic
stripe, they merged with the liquid film and were collected on
the stripe.
Several research groups demonstrated that alternative surface pat-
terns could be optimized for condensation enhancement.78−80 Peng
et al. carried out systematic heat transfer measurements on the copper
condensing surfaces with vertically patterned hydrophobic/hydrophilic
stripes.79 The experiment results indicated that the maximum con-
densed droplet size and the thermal performance were controlled by
adjusting the characteristic length scale of hydrophobic and hydrophilic
stripes. The optimum hydrophobic region width was ∼0.55 mm and
the corresponding optimum maximum droplet radius was ∼0.25 mm.
When the surface subcooling was set at 2.0 K, the heat transfer
performance of the optimum hybrid surface demonstrated a ∼23%
increase compared to that of dropwise condensation on a flat surface.
Fig. 6.17. Selected images showing the droplet size distribution on different
hydrophobic–hydrophilic hybrid surfaces with (a) hydrophobic stripe width of
∼0.55 mm and (b) ∼0.95 mm. (c) The comparison of heat transfer characteristics
of steam condensation on different hydrophobic–hydrophilic hybrid surfaces to that
of homogeneous hydrophobic surface.79
Although the hydrophilic region acted as the drainage path on the

surface, the heat transfer performance decreased as the hydrophilic
region width increased. Thus, the hydrophilic stripes should be
designed as narrow as possible if efficient condensate absorption is
enabled.
Ghosh et al. pointed out that the condensation performance
of alternative hydrophobic/hydrophilic surfaces might be further
improved by mimicking the natural surface patterns of banana leaves.80
Figure 6.18a shows a wedge-shaped hydrophilic track on the condens-
ing surface transporting the collected condensates from the narrow
end to the wider one. By creating a liquid film curvature gradient,
the wedge shaped hydrophilic track allowed efficient movement of
water along the tapered track via capillary pumping effect. Inspired
by the vein network of banana leaves, the wedge-shaped hydrophilic
patterns were arranged to form an optimal 2-D liquid transport
system. The condensation experiments indicated that the bio-inspired
interdigitated pattern achieved ∼9% enhancement in water collection
over the conventional vertically patterned condensing surface under
environmental conditions with temperature of 35◦ C and humidity
of 80%.
Fig. 6.18. (a) Wedge-shaped hydrophilic track aiding the autonomous transport of
condensate from the narrow end to the wider end. (b) Typical vein structure of a banana
leaf. (c) Condensation on an interdigitated wettability-patterned design inspired from
the leaf vein network layout, scale bar is 20 mm. (d) Water collection rate for two
different ambient environments for the unpatterned control case (bare Aluminum),
straight vertical patterns, and bio-inspired interdigitated wettability patterns.80
Furthermore, the wetting heterogeneity can also lead to

autonomous water transport within the surface roughness. Anderson
et al. proposed a special gradient surface design consisting of dense
nanowires with a hydrophilic base and hydrophobic tips.81 This
vertical surface energy gradient enabled a long-range collective droplet
coalescence, resulting in the periodic refreshment of nucleation sites
for continuous condensation. As shown in Fig. 6.19, the experimental
visualization indicated that this novel condensate movement strategy
also relied on the Laplace pressure-driven coalescence of Wenzel
droplets through the wetted substrate, rather than requiring condensed
droplets to remain in Cassie stable state for coalescence-induced self-
propelling.
Fig. 6.19. (a) Selected optical images of long range droplet coalescence showing the
Laplace pressure-driven droplet movement. Encircled droplets are in the Cassie state
and therefore do not interact with the surrounding coalescing droplets. (b) Schematics
showing the coalescence process of Wenzel droplets through wetted substrate.81
6.5. Suspended Condensation

on Lubricant-Infused Surfaces
Microstructured substrate with infused lubricating fluid, an interesting
natural phenomenon of Nepenthes pitcher plants, provides a remark-
ably simple alternative idea for dropwise condensation enhancement in
recent years. In pitcher plants, the infused liquid is effectively locked
within the microscale roughness of the surface and forms a continuous
overlying film which significantly decreases the contact adhesion force.
When insects land on this lubricant-infused surface, they slide from the
rim into the digestive juices at the bottom of the Nepenthes pitcher
plants.27
Fig. 6.20. (a) Image of Nepenthes pitcher plant. (b) Cross-sectional shot of the
Nepenthes, which shows the capture of insects by creating an extremely slippery surface
using microscopic ridges bathed in nectar.
6.5.1. Self-repairing slippery property

on lubricant-infused surfaces
Wong et al. successfully reproduced the slippery property of Nepenthes
pitcher plants on the artificial lubricant-infused porous surface.27,82,83
Two types of porous solids were used as experimental substrates, includ-
ing uniformly patterned epoxy-resin-based nanostructured surfaces and
Teflon nanofibrous membranes with random network. The lubricating
fluids used for the surface fabrication were perfluorinated fluids, which
were added onto the porous substrates to form the over-coated layer.
With well-matched solid and liquid surface energies and roughness
length scale, the infused lubricants spread spontaneously onto the
whole substrate through capillary wicking.
The lubricant-infused surface displayed a very low droplet adhesion
against liquids of surface tension ranging from ∼17.2 ± 0.5 mN/m for
n-pentane to 72.4 ± 0.1 mN/m for water due to the extreme liquid
repellency as shown in Fig. 6.21b. For the low-surface-tension liquids
(γ < 25 mN/m), the measured droplet contact angle hysteresis (<2.5◦ )
and volume (∼4.5 µl) suggested the droplet pinning force was 0.83 ±
0.22 µN on the lubricant-infused surface, which was nearly one order
of magnitude lower than on the nanostructured superhydrophobic
Fig. 6.21. (a) Schematics showing the fabrication of a SLIPS by infiltrating a

functionalized porous/textured solid with a low-surface energy, chemically inert liquid
to form a physically smooth and chemically homogeneous lubricating film on the
surface of the substrate. (b) Comparison of contact angle hysteresis as a function of
surface tension of test liquids (indicated) on SLIPS and on an omniphobic surface.
The advancing and receding contact angles of a liquid droplet are denoted as θadv ,
and θrec , respectively. SLIPS 1, 2 and 3 refer to the surfaces made of Teflon porous
membrane (SLIPS 1), an array of epoxy posts of geometry 1 (pitch, 2 mm, height,
5 mm, post diameter, 300 nm) (SLIPS 2) and an array of epoxy posts of geometry
2 (pitch, 900 nm, height, 500 nm–2 mm, post diameter, 300 nm) (SLIPS 3). Error
bars indicate standard deviations from three independent measurements. (c) Time-
lapse images showing the capability of a SLIPS to self-heal from physical damage with
∼50 µm wide on a timescale of the order of 100 ms.27
surfaces. In particular, unlike the superhydrophobic surface, the contact

angle hysteresis on the lubricant-infused surface was independent of the
liquid surface tension. This finding greatly broadens the application
areas of condensation heat transfer for various working liquids. In
addition, the lubricating liquid film can act as a self-healing coating for
rapidly restoring the liquid-repellent function even when the porous
substrate was damaged by abrasion or impact. Such self-repairing

mechanism was attributed to the fluidic nature of the lubricating layer,
that is, the liquid spontaneously flowed towards the damaged area and
refilled the voids due to the capillary action (Fig. 6.21c).
6.5.2. Heat transfer enhancement of immersion

condensation
Since the lubricant-infused surfaces exhibited significant reduced
contact pinning, Anand et al. attempted to test the condensation
dynamics on such surfaces to analyze the feasibility of heat transfer
enhancement.84 The experimental data showed that the lubricant-
infused surface could increase the droplet mobility when they reached
100 µm in diameter, resulting in the improved water sweeping motion
during condensation. The experiment results showed that the gravity-
induced droplet departure diameter on lubricant-infused surface was
reduced to ∼1.5 mm while the droplet shedding size on superhy-
drophobic surface was ∼3.5 mm in diameter for Wenzel droplet
at same condition. The lubricant-infused surface serves as a viable
surface engineering approach to enhance condensation heat transfer
by preventing the condensate penetration.
However, the reduced droplet departure size does not necessarily
address the other aspects influencing condensation heat transfer rates,
such as droplet nucleation density. To solve this issue, Xiao et al.
fabricated a novel lubricant-infused copper surface and carried out
Fig. 6.22. (a) Schematic of condensation on a lubricant-infused surface showing

droplets remain afloat on a lubricant film without penetrating into the surface
roughness. (b) ESEM image of water condensed on a lubricant-infused surface.84
Fig. 6.23. (a) Schematic showing water vapor diffusing through the thin lubricant
film and forming immersed droplets on the tips of micropillars. (b) Magnified schematic
showing the nuclei formation on high-surface-energy sites of TFTS nano-agglomerates
on micropillar tips in the lubricant film. (c) Image of dropwise condensation on a
hydrophobic copper tube. (d) Image of condensation on an lubricant-infused TFTS-
coated CuO tube. Significantly higher droplet density was observed on the lubricant-
infused surface while a low departure radius of 0.98 ± 0.13 mm was maintained.85
systematic condensation heat transfer experiments to study the cor-

responding thermal efficacy.85 By introducing the nanoscale high
surface energy domains (TFTS nano-agglomerates) on the Krytox
lubricant-infused nanostructured CuO surface, a unique “immersion
condensation” was observed during the phase change process, where
the nucleating water droplets immersed within the infused lubricant
(Fig. 6.23a, b). The combination of surface energy heterogeneity,
reduced oil–water interfacial energy, and surface structuring enabled
the droplet nucleation density to increase by over one order of
magnitude while maintaining the ultra-low droplet adhesion. The
visualization data showed that the average droplet departure radius was
reduced from 1.83 ± 0.31 mm for the typical hydrophobic surfaces to
0.98 ± 0.13 mm for the immersion condensation surfaces. As a result,
the thermal measurement demonstrated that the condensation heat
transfer coefficient on such lubricant-infused heterogeneous surfaces
was enhanced by ∼100% compared to the dropwise condensation on
flat hydrophobic surface in the presence of non-condensable gases.
6.6. Outlook of Bio-Inspired Strategies

for Condensation Enhancement
Despite much recent research focused on the condensation phe-
nomenon on biomimetic functional surfaces, the durability of
hydrophobic functionalization and micro/nanoscale surface structures
for long-term dropwise condensation remains a primary issue in practi-
cal implementations. In recent years, self-assembly monolayers (SAMs)
of fluorosilanes have been widely used in laboratory studies because
the covalent bonding of silane-based SAMs provides a stable thermal
property, which significantly increases the longevity of hydrophobic
coating. However, some scientists also pointed out the fluorosilane
coating may degrade when exposed to steam at 100◦ C.86 As upgraded
materials, the fluoropolymer thin films, rare earth oxides and graphene
have been recently developed aiming for lower thermal resistance and
robust coating.86−88 Nevertheless, further investigation is still needed
to explore the advanced materials for surface coating with superior
thermal conductivity and conformity.
The research has demonstrated the potential for condensation
heat transfer on hybrid surfaces with wetting contrast and lubricant
infused surfaces. Particularly, the cooperation of different bio-inspired
strategies offers a large design space for advanced condensing surfaces.
Theoretical studies along with systematic thermal characterization will
be required to investigate the underlying physics and optimum surface
design. The key issue regarding the hybrid surface is the optimal
range of wetting contrast and length scale of surface structures. The
distinct condensation process on hybrid surfaces leaves a lot of unsolved
questions for future development. In comparison, lubricant-infused
surfaces provide a promising approach to improve the condensation
heat transfer for liquids with low surface tension. However, the
progressive lubricant loss during condensate departure remains an
important problem for long-term operation using the lubricant-infused
surfaces. The related studies should be quite useful for the optimization
of industrial applications associated with refrigeration cycle.
For jumping droplet condensation, the surface flooding phe-
nomenon greatly limited the utilization in many applications with
high condensation heat flux. Miljkovic et al. recently demonstrated

that the charge in condensed droplets can be separated during self-
jumping on superhydrophobic surfaces.89 The condensing surfaces
absorbed the negative charge and formed an electric double layer at
the liquid/coating interface. As a result, these jumping droplets were
positively charged upon departure and may retract to the surface due
to the electrostatic force. A transverse electric field applied on the
superhydrophobic surface may facilitate the droplet jumping away from
the surface. Thus the overall heat transfer coefficient can be further
enhanced by 50% compared to that without the applied electric field
for low supersaturations.90 Their work has inspired more research
on applying external approaches such as Coulomb force,89,90 electro
wetting,91−93 vibration94,95 and ionic wind,96,97 in combination with
the biomimetic surfaces for enhanced dropwise condensation heat
transfer.
References
[1] Nusselt W. (1916) Die Oberflachenkondesation des Wasserdamffes. The surface
condensation of water. Zetrschr Ver Deutch Ing 60: 541–546.
[2] Schmidt E, Schurig W, Sellschopp W. (1930) Versuche über die Kondensation
von Wasserdampf in Film-und Tropfenform. Technische Mechanik und Thermo-
dynamik 1: 53–63.
[3] Rose JW. (1981) Dropwise Condensation Theory. Int J Heat Mass Tran 24:
191–194.
[4] LeFevre, Rose J. (1966) A theory of heat transfer by dropwise condensation.
In: Chemical Engineering Progress, pp. 86.
[5] Rose J. (1988) Some aspects of condensation heat transfer theory. International
communications in heat and mass transfer 15: 449–473.
[6] Rose JW. (1998) Condensation Heat Transfer Fundamentals. Chemical Engi-
neering Research and Design 76: 143–152.
[7] Ma X, Rose JW, Xu D, et al. (2000) Advances in dropwise condensation heat
transfer: Chinese research. Chem Eng J 78: 87–93.
[8] Enright R, Miljkovic N, Alvarado J, et al. (2014) Dropwise Condensation on
Micro-and Nanostructured Surfaces. Nanoscale and Microscale Thermophysical
Engineering 18: 223–250.
[9] Rose J. (1980) Approximate equations for forced-convection condensation in
the presence of a non-condensing gas on a flat plate and horizontal tube. Int J
Heat Mass Tran 23: 539–546.
[10] Rose J. (1976) Further aspects of dropwise condensation theory. Int J Heat Mass
Tran 19: 1363–1370.
[11] Rose J. (1999) Condensation heat transfer. Heat and Mass Transfer 35:
479–485.
[12] Stylianou S, Rose J. (1983) Drop-to-filmwise condensation transition: heat
transfer measurements for ethanediol. Int J Heat Mass Tran 26: 747–760.
[13] Tanner D, Pope D, Potter C, West D. (1968) Heat transfer in dropwise
condensation at low steam pressures in the absence and presence of non-
condensable gas. Int J Heat Mass Tran 11: 181–190.
[14] Boreyko JB, Zhao Y, Chen C-H. (2011) Planar jumping-drop thermal diodes.
Appl Phys Lett 99: 234105.
[15] Roadmap I. (2009) International technology roadmap for semiconductors,
2009 edn. Executive Summary. Semiconductor Industry Association.
[16] Thome JR. (2006) The new frontier in heat transfer: Microscale and nanoscale
technologies. Heat Transfer Engineering 27: 1–3.
[17] Peters TB, McCarthy M, Allison J, et al. (2012) Design of an Integrated
Loop Heat Pipe Air-Cooled Heat Exchanger for High Performance Electronics.
Components, Packaging and Manufacturing Technology, IEEE Transactions on 2:
1637–1648.
[18] Miljkovic N, Wang EN. (2011) Modeling and optimization of hybrid solar
thermoelectric systems with thermosyphons. Solar Energy 85: 2843–2855.
[19] Feng L, Li S, Li Y, et al. (2002) Super-Hydrophobic Surfaces: From Natural to
Artificial. Adv Mater 14: 1857–1860.
[20] Su YW, Ji BH, Huang Y, Hwang KC. (2010) Nature’s Design of Hierarchical
Superhydrophobic Surfaces of a Water Strider for Low Adhesion and Low-
Energy Dissipation. Langmuir 26: 18926–18937.
[21] Zheng Y, Bai H, Huang Z, et al. (2010) Directional water collection on wetted
spider silk. Nature 463: 640–643.
[22] Hou YP, Chen Y, Xue Y, et al. (2012) Water Collection Behavior and Hanging
Ability of Bioinspired Fiber. Langmuir 28: 4737–4743.
[23] Zorba, V, Stratakis, E, Barberoglou, et al. (2008) Biomimetic Artificial Surfaces
Quantitatively Reproduce the Water Repellency of a Lotus Leaf. Adv. Mater. 20:
4049—4054.
[24] Nosonovsky M, Bhushan B. (2007) Biomimetic superhydrophobic surfaces:
multiscale approach. Nano Lett 7: 2633–2637.
[25] Patankar NA. (2004) Mimicking the lotus effect: influence of double roughness
structures and slender pillars. Langmuir 20: 8209–8213.
[26] Zhai L, Berg MC, Cebeci FC, et al. (2006) Patterned superhydrophobic surfaces:
toward a synthetic mimic of the Namib Desert beetle. Nano Lett 6: 1213–1217.
[27] Wong T-S, Kang SH, Tang SK, et al. (2011) Bioinspired self-repairing slippery
surfaces with pressure-stable omniphobicity. Nature 477: 443–447.
[28] Rose J, Glicksman L. (1973) Dropwise condensation — the distribution of drop

sizes. Int J Heat Mass Tran 16: 411–425.
[29] Umur A, Griffith P. (1965) Mechanism of dropwise condensation. J Heat Trans
87: 275–282.
[30] Kim S, Kim KJ. (2011) Dropwise condensation modeling suitable for superhy-
drophobic surfaces. J Heat Trans 133: 081502-1–081502-7.
[31] Abu-Orabi M. (1998) Modeling of heat transfer in dropwise condensation. Int
J Heat Mass Tran 41: 81–87.
[32] Patankar NA. (2010) Supernucleating surfaces for nucleate boiling and dropwise
condensation heat transfer. Soft Matter 6: 1613–1620.
[33] Bocquet L, Lauga E. (2011) A smooth future? Nature Materials 10: 334–337.
[34] Kang SH, Wu N, Grinthal A, Aizenberg J. (2011) Meniscus lithography:
Evaporation-induced self-organization of pillar arrays into Moireì patterns. Phys
Rev Lett 107: 177802.
[35] Miljkovic N, Enright R, Wang EN. (2012) Effect of droplet morphology on
growth dynamics and heat transfer during condensation on superhydrophobic
nanostructured surfaces. Acs Nano 6: 1776–1785.
[36] Zhang G, Zhang J, Xie G, et al. (2006) Cicada Wings: A Stamp from Nature
for Nanoimprint Lithography. Small 2: 1440–1443.
[37] Wisdom KM, Watson JA, Qu X, et al. (2013) Self-cleaning of superhydrophobic
surfaces by self-propelled jumping condensate. Proc Natl Acad Sci 110: 7992–
7997.
[38] Boreyko JB, Chen CH. (2009) Self-Propelled Dropwise Condensate on Super-
hydrophobic Surfaces. Phys Rev Lett 103: 184501.
[39] Boreyko JB, Chen C-H. (2010) Self-propelled jumping drops on superhy-
drophobic surfaces. Physics of Fluids (1994–present) 22.
[40] Lv C, Hao P, Yao Z, Niu F. (2015) Departure of Condensation Droplets on
[41] Tian J, Zhu J, Guo H-Y, et al. (2014) Efficient Self-Propelling of Small-Scale
Condensed Microdrops by Closely Packed ZnO Nanoneedles. The Journal of
Physical Chemistry Letters 5: 2084–2088.
[42] Wang F-C, Yang F, Zhao Y-P. (2011) Size effect on the coalescence-induced
self-propelled droplet. Appl Phys Lett 98: 053112.
[43] Dietz C, Rykaczewski K, Fedorov AG, Joshi Y. (2010) Visualization of droplet
departure on a superhydrophobic surface and implications to heat transfer
enhancement during dropwise condensation. Appl Phys Lett 97: 033104.
[44] Luo Y, Li J, Zhu J, et al. (2015) Fabrication of Condensate Microdrop Self-
Propelling Porous Films of Cerium Oxide Nanoparticles on Copper Surfaces.
Angew Chem Int Ed 54: 4876–4879.
[45] Feng J, Qin Z, Yao S. (2012) Factors affecting the spontaneous motion of
condensate drops on superhydrophobic copper surfaces. Langmuir 28: 6067–
6075.
[46] Young T. (1805) An Essay on the Cohesion of Fluids. Philos T Roy Soc 95:
65–87.
[47] Cassie ABD, Baxter S. (1944) Wettability of porous surfaces. Transactions of the
Faraday Society 40: 546–551.
[48] Wenzel RN. (1936) Resistance of Solid Surfaces to Wetting by Water. Indus-
trial & Engineering Chemistry 28: 988–994.
[49] Rykaczewski K, Paxson AT, Anand S, et al. (2012) Multimode multidrop
serial coalescence effects during condensation on hierarchical superhydrophobic
surfaces. Langmuir 29: 881–891.
[50] Enright R, Miljkovic N, Al-Obeidi A, et al. Condensation on superhydrophobic
surfaces: the role of local energy barriers and structure length scale. Langmuir
28: 14424–14432.
[51] Miljkovic N, Enright R, Nam Y, et al. (2012) Jumping-droplet-enhanced
condensation on scalable superhydrophobic nanostructured surfaces. Nano Lett
13: 179–187.
[52] Miljkovic N, Wang EN. (2013) Condensation heat transfer on superhydropho-
bic surfaces. MRS Bulletin 38: 397–406.
[53] Miljkovic N, Enright R, Wang EN. (2013) Modeling and Optimization of
Superhydrophobic Condensation, J Heat Trans 135: 111004–111004.
[54] Chen C-H, Cai Q, Tsai C, et al. (2007) Dropwise condensation on superhy-
drophobic surfaces with two-tier roughness. Appl Phys Lett 90: 173108.
[55] Cheng J, Vandadi A, Chen C-L. (2012) Condensation heat transfer on two-tier
superhydrophobic surfaces. Appl Phys Lett 101: 131909.
[56] Chen X, Wu J, Ma R, et al. (2011) Nanograssed micropyramidal architectures for
continuous dropwise condensation. Advanced Functional Materials 21: 4617–
4623.
[57] Lo CW, Wang CC, Lu MC. (2013) Spatial control of heterogeneous nucleation
on the superhydrophobic nanowire array. Advanced Functional Materials 24:
1211–1217.
[58] He M, Zhou X, Zeng XP, et al. (2012) Hierarchically structured porous
aluminum surfaces for high-efficient removal of condensed water. Soft Matter
8: 6680–6683.
[59] Parker AR, Lawrence CR. (2001) Water capture by a desert beetle. Nature 414:
33–34.
[60] Yamada Y, Ikuta T, Nishiyama T, et al. (2014) Droplet Nucleation on a
Well-Defined Hydrophilic–Hydrophobic Surface of 10 nm Order Resolution.
Langmuir 30: 14532–14537.
[61] Varanasi KK, Hsu M, Bhate N, et al. (2009) Spatial control in the heterogeneous
nucleation of water. Appl Phys Lett 95: 094101.
[62] Mishchenko L, Khan M, Aizenberg J, Hatton BD. (2013) Spatial control
of condensation and freezing on superhydrophobic surfaces with hydrophilic
patches. Advanced Functional Materials 23: 4577–4584.
[63] Yao C-W, Alvarado JL, Marsh CP, et al. (2014) Wetting behavior on hybrid
surfaces with hydrophobic and hydrophilic properties, Appl Surf Sci 290: 59–65.
[64] Yao CW, Garvin TP, Alvarado JL, et al. (2012) Droplet contact angle behavior
on a hybrid surface with hydrophobic and hydrophilic properties. Appl Phys Lett
101: 111605–111606.
[65] Hou Y, Yu M, Chen X, et al. (2014) Recurrent Filmwise and Dropwise
Condensation on a Beetle Mimetic Surface. ACS Nano 9: 71–81.
[66] He M, Zhang Q, Zeng X, et al. (2013) Hierarchical porous surface for efficiently
controlling microdroplets’ self-removal. Adv Mater 25: 2291–2295.
[67] Hou Y, Yu M, Chen X, et al. (2015) Filmwise-to-dropwise condensation
transition enabled by patterned high wetting contrast. Journal of Heat Transfer
137(8): 080907–080907.
[68] Ju J, Bai H, Zheng Y, et al. (2012) A multi-structural and multi-functional
integrated fog collection system in cactus. Nature Communications 3: 1247.
[69] Guo L, Tang GH. (2015) Experimental study on directional motion of a single
droplet on cactus spines. Int J Heat Mass Tran 84: 198–202.
[70] Liu C, Ju J, Zheng Y, Jiang L. (2014) Asymmetric Ratchet Effect for Directional
Transport of Fog Drops on Static and Dynamic Butterfly Wings. ACS Nano
8(2): 1321–1329.
[71] Zhang J, Wang J, Zhao Y, et al. (2008) How does the leaf margin make the
lotus surface dry as the lotus leaf floats on water? Soft Matter 4: 2232–2237.
[72] Wu H, Zhang R, Sun Y, et al. (2008) Biomimetic nanofiber patterns with
controlled wettability. Soft Matter 4: 2429–2433.
[73] Wu D, Wang J-N, Wu S-Z, et al. (2011) Three-Level Biomimetic Rice-Leaf
Surfaces with Controllable Anisotropic Sliding. Advanced Functional Materials
21: 2927–2932.
[74] Jing-Tang Y, Chen JC, Ker-Jer H, Yeh JA. (2006) Droplet manipulation on
a hydrophobic textured surface with roughened patterns. Journal of Microelec-
tromechanical Systems 15: 697–707.
[75] Daniel S, Chaudhury MK, Chen JC. (2001) Fast Drop Movements Resulting
from the Phase Change on a Gradient Surface. Science 291: 633–636.
[76] Khoo HS, Tseng F-G. (2009) Spontaneous high-speed transport of subnanoliter
water droplet on gradient nanotextured surfaces. Appl Phys Lett 95: 063108.
[77] Chaudhury MK, Chakrabarti A, Tibrewal T. (2014) Coalescence of drops near
a hydrophilic boundary leads to long range directed motion. Extreme Mechanics
Letters 1: 104–113.
[78] Chatterjee A, Derby MM, Peles Y, Jensen MK. (2014) Enhancement of
condensation heat transfer with patterned surfaces. Int J Heat Mass Tran 71:
675–681.
[79] Peng B, Ma X, Lan Z, et al. (2015) Experimental investigation on steam
condensation heat transfer enhancement with vertically patterned hydrophobic–
hydrophilic hybrid surfaces. Int J Heat Mass Tran 83: 27–38.
[80] Ghosh A, Beaini S, Zhang BJ, et al. (2014) Enhancing Dropwise Condensation
through Bioinspired Wettability Patterning. Langmuir 30: 13103–13115.
[81] Anderson DM, Gupta MK, Voevodin AA, et al. (2012) Using Amphiphilic
Nanostructures To Enable Long-Range Ensemble Coalescence and Surface
Rejuvenation in Dropwise Condensation. ACS Nano 6: 3262–3268.
[82] Epstein AK, Wong T-S, Belisle RA, et al. (2012) Liquid-infused structured
surfaces with exceptional anti-biofouling performance. Proc Natl Acad Sci 109:
13182–13187.
[83] Vogel N, Belisle RA, Hatton B, et al. (2013) Transparency and damage toler-
ance of patternable omniphobic lubricated surfaces based on inverse colloidal
monolayers. Nature Communications 4: 2176.
[84] Anand S, Paxson AT, Dhiman R, et al. (2012) Enhanced condensation on
lubricant-impregnated nanotextured surfaces. ACS Nano 6: 10122–10129.
[85] Xiao R, Miljkovic N, Enright R, Wang EN. (2013) Immersion condensation on
oil-Infused heterogeneous surfaces for enhanced heat transfer. Scientific Reports
3: 1988.
[86] Paxson AT, Yagüe JL, Gleason KK, Varanasi KK. (2014) Stable dropwise
condensation for enhancing heat transfer via the initiated chemical vapor
deposition (iCVD) of grafted polymer films. Adv Mater 26: 418–423.
[87] Azimi G, Dhiman R, Kwon H-M, et al. (2013) Hydrophobicity of rare-earth
oxide ceramics. Nature Materials 12: 315–320.
[88] Preston DJ, Mafra DL, Miljkovic N, et al. (2015) Scalable Graphene Coatings
for Enhanced Condensation Heat Transfer. Nano Lett 15(5): 2902–2909.
[89] Miljkovic N, Preston DJ, Enright R, Wang EN. (2013) Electrostatic charging
of jumping droplets. Nature Communications 4: 2517.
[90] Miljkovic N, Preston DJ, Enright R, Wang EN. (2013) Electric-Field-Enhanced
Condensation on Superhydrophobic Nanostructured Surfaces. ACS Nano 7:
11043–11054.
[91] Kim J, Kaviany M. (2007) Purging of dropwise condensate by electrowetting.
J Appl Phys 101: 103520.
[92] Kim J, Kaviany M. (2010) Electrowetting Purged Surface Condensate in
Evaporators. Heat Transfer Engineering 31: 101–107.
[93] Lee SJ, Hong J, Kang KH, et al. (2014) Electrowetting-Induced Droplet
Detachment from Hydrophobic Surfaces. Langmuir 30: 1805–1811.
[94] Boreyko JB, Chen C-H. (2009) Restoring Superhydrophobicity of Lotus Leaves
with Vibration-Induced Dewetting. Phys Rev Lett 103: 174502.
[95] Boreyko JB, Baker CH, Poley CR, Chen C-H. (2011) Wetting and Dewetting
Transitions on Hierarchical Superhydrophobic Surfaces. Langmuir 27: 7502–
7509.
[96] Go DB, Maturana RA, Fisher TS, Garimella SV. (2008) Enhancement of external
forced convection by ionic wind. Int J Heat Mass Tran 51: 6047–6053.
[97] Go DB, Garimella SV, Fisher TS, Mongia RK. (2007) Ionic winds for locally
enhanced cooling. J Appl Phys 102: 053302.
Chapter 7
Large-Scale Fabrication of Biomimetic

Drag-Reduction Surface via
Huawei Chen, Deyuan Zhang, Xin Zhang,

Da Che
School of Mechanical Engineering,
Beihang University, Beijing, China
Shark skin possess unique hierarchical structure and superior drag

reduction performance. Recently, bio-replication approach has been
proposed and developed to directly take use of its surface function.
However, optimum drag reduction performance of real shark skin just
appear at his living conditions. If the application conditions is great
different from its living conditions, the drag reduction performance will
be declined. In order to widen its application, large-scale adjustment
should be developed for its hierarchical structure. Then in this chapter,
we will introduce swelling and shrunken process to adjust the vivid
microstructure.
7.1. Introduction
With the improvement of micro/nano fabrication capability, surface
structure has become more sophisticated than ever to meet serious
demands on surface functions. Natural evolution provides natural
surface treasures with marvelous surface function to inspire innovative
design, such as the drag reduction of shark skin. Recently the shark skin
229
effect has drawn worldwide attention because of its inherent superior

drag reduction.1−5 Although the biomimetic drag-reduction skin such
as the microgroove drag-reduction riblet has opened the door for
various industrial applications, e.g. oil pipeline, the drag-reduction
performance is greatly lowered as compared with natural shark skin.6−10
Recently, the bio-replication forming approach has been put forward
to take advantage of the natural surface as much as possible. The
natural surface is directly taken as the molding template to achieve
1:1 replication.11−13
As the classic product of natural selection, sharks have evolved
unique hierarchical surface morphology on their skins for survival in
their living environment. Since adaptation to the living environment
is the sole driving force of evolution, sharks show the best drag
reduction function only in their living environment. Especially, the
natural surface function is impaired or even reversed as staying far from
the living environment. Bio-replicated shark skin performs maximum
drag reduction about 12% only when the relative fluid flow is around
5m/s, which agrees well with normal swimming speed of shark [14].
The maximum drag reduction commonly declines with improvement
of fluid flow speed beyond 5m/s. Therefore, adjusting the surface
morphology of the shark skin to accommodate different fluid flow
environment is necessary for spreading industrial applications of shark
skin effect. According to Walsh’s observation that maximum drag
reduction of micro-riblets is dependent on optimum spacing s + and is
affected by depth-to-width ratio h/s , equal-proportional adjustment of
surface morphology is one way to guarantee the optimal drag reduction
function of shark skin preserving at different speed fluid flow such as
ship, submarine [15].
7.2. Bio-Replication of Vivid Shark Skin
Almost all of biomimetic drag-reduction riblet surfaces were mainly

focused on two-dimensional (2-D) riblets with simple cross sections,
such as sawtooth, scalloped and blade, and their maximum benefits in
reduction of fluid drag have not reached 10%.7−10
Although the design of the riblet surface for drag reduction took
inspiration from nature, the drag-reducing efficiency in general is much
Large-Scale Fabrication of Biomimetic Drag-Reduction Surface 231
lower than that of biological prototypes such as real shark skin. The
swimming performance of real shark skin has been explored, and it
has been experimentally or theoretically clarified that its drag-reducing
efficiency can reach more than 12%, which is superior to that of
simple 2-D riblets. Lang et al. investigated the influence of bristled
shark skin on control of boundary layer, and found that the scales of
some fast swimming sharks were pliable to adjust the drag-reduction
effect.11,12 Oeffner and Lauder compared the self-propelled swimming
speed of different surfaces and found that real shark skin foils actually
had benefit in swimming performance of an average of 12.3% (with
a maximal improvement of almost 20%) as compared with the same
foils with the denticles sanded off.13 Numerous studies have indicated
that real shark skin is more advanced than simple 2-D riblet surfaces
in drag reduction. The forming of biomimetic shark skins with surface
morphology close to the biological prototype is still hard work. By
simplifying or magnifying the shark scales, many investigations have
fabricated some kinds of biomimetic shark skins.17−19 Recently, Zhang
et al. directly took real shark skin as a replica template to fabricate vivid
shark skin, and the testing results in water tunnel demonstrated that
the wall drag was reduced about 12%.20−22 However, this method still
has difficulty with fabricating continuous large-area shark skin surface.
In order to resolve the problems in large area fabrication of
hierarchical drag-reduction surface structure, a novel bio-replication
coating approach is proposed on basis of rapid UV curable polymer, in
which the real shark skin is directly taken as a replica template.
7.2.1. Materials and bio-replication coating device

Micro-embossing or -imprinting has become more prevalent for low-
cost and mass-production of micro- or nano-structures. However
micro-embossing or -imprinting was generally limited to regular 2-
D micro-patterns, and is not suitable for complex 3-D micro-patterns
as like shark skin due to considerable difficulty of demolding. In
order to make transfer of natural shark skin onto a free-form surface
possible, flexible bio-replication coating approach of shark skin is
proposed as shown in Fig. 7.1 (reproduced from Fig. 5.20 of same
research group to ease reading), whose mechanism is based upon
micro-embossing.
Fig. 7.1. Schematics of natural function surface morphology coating technology.
Materials
The skin of Carcharhinus brachyurus (1.4 m in length, Beijing
Fishery Corp.), which is a typical fast swimming shark, was taken
as the template for bio-replication coating. As the material for the
UV transparent embossing belt, Polydimethylsiloxane (PDMS) was
applied and the Sylgard 184 elastomer kit (Dow. Corning, US) was
used. The thickness of the soft negative mold was set at 5mm, as
thin as possible to ensure sufficient UV transmission. A commercial
water solution of an ionomeric polyurethane diacrylate (PUC, trade
name UCECOAT 6558, MW = 10,000 g/mol) was used as the
base UV-curable resin. Darocur 1173 (Ciba Specialties Chemical)
was employed as photoinitiator.23,24 A standard protein sample and
soybean lecithin were provided by Shanghai Hufeng Biotechnology
Co., Ltd. Fluorescein isothiocyanate (FITC) was used to label the
protein/liposome, and FITC-protein conjugates were prepared by us.
Pretreatment of shark skin template

Dermis and subcutaneous tissue was scraped clean from a fresh shark
skin template. The shark skin template was rinsed five to eight times
with de-ionized water to clean off blood and other impurities. After
being flattened and nailed to a rigid plate, the shark skin template was
placed in 2.5% glutaraldehyde (C5H8O2) for three hours to maintain
its original micromorphology via chemical fixation. The chemically
fixed shark skin template was rinsed for one hour with three to four
changes of a 0.1mol/L phosphate buffer solution (pH 7.2). Any
residual solution on the template was rinsed off three to four times
with clear and de-ionized water. Gradient dehydration was conducted
to prevent template deformation resulting from excessive water loss in
drying. The chemically fixed shark skin samples were immersed in 30%,
50%, 75%, 80%, 95% and 100% ethanol solution to dehydrate them step
by step, each for 15 to 30 min. Finally, the samples were kept inside a
drying oven at 60◦ C for 12 hours, which completed the pretreatment
of the shark skin template.
Bio-replication coating device based on UV curable paint

All the processes of bio-replication coating involve pre-treatment of
shark skin, soft negative mold preparation and rapid micro-embossing
of UV curable paint. A long strip of the soft negative mold is connected
to a flexible ring to meet coating requirements of various free-form
surfaces.
Pretreatment of shark skin is important to maintain its surface
micro-morphology and improve strength as the molding template.
Here pre-treatment of shark skin consists of following steps: rigid
fastening, cleaning, chemical fixation, rinse, dehydration and drying.
To replicate shark skin in large area, a piece of fresh shark skin
(200 mm × 550 mm) with good morphology was selected to be the
shark skin template and pretreated as follows. Shark skin template is
rinsed three to five times in clear water, and then two to three times
in de-ionized water to remove the blood or other impurities. After
being flattened and nailed to a rigid plate, the shark skin template is
put into 2.5% glutaraldehyde (C5H8O2) over three hours to maintain
its original micromorphology via chemical fixation. The chemical fixed
shark skin template is rinsed one hour with three to four changes of
0.1 mol/L pH7.2 phosphate buffer solution. The residual solution is
rinsed off three to four times in clear and deionized water. Gradient
dehydration process is conducted to prevent template deformation
from excessive water loss in drying. Chemical fixed shark skin samples
are immersed in a concentration of 30%, 50%, 75%, 80%, 95% and 100%
ethanol solution to dehydrate step by step, each for 15–30 minutes.
Shark skin template Pouring PDMS
Soft negative mold Demolding
Fig. 7.2. Preparation of soft negative mold.
After keeping the samples in a drying oven for over 12 hours at 60◦ C,
the pretreatment of the shark skin template is completed. The SEM of
pretreated shark skin is shown in Fig. 7.3 (a) or Fig. 5.52.
The soft negative mold is obtained via soft lithography, in which
the pretreated shark skin is taken as a template of the soft lithography
as shown in Fig. 7.2. Pouring of PDMS upon the template and de-
molding of the soft negative mold from template are the predominant
procedures of soft lithography. In order to ensure PDMS is fully poured
into the micro-cavities of the shark skin template, all the processes of
soft negative mold preparation are conducted in a vacuum chamber to
squeeze air from the cavities of the shark skin. The SEM photo of the
soft negative mold obtained by soft lithography is shown in Fig. 7.3(b).
The mechanism of large-area bio-replication coating instrument
is illustrated as shown in Fig. 7.4 or Fig. 5.51. A long strip of
the soft negative mold with micro-morphology of shark skin tightly
enwraps three plastic wheels, and an ultraviolet lamp is set inside. The
soft negative mold strip is 500 mm in length and 185 mm in width.
The distance between the strip and the UV lamp (power 300 W and
wavelength 395 nm LED UV lamp, Shanghai Machine Optoelectronic
(a) Shark skin template (b) Soft negative mold (c) Bio-replication coating
Fig. 7.3. Micro-morphology of shark skin, soft negative mold and bio-replication
coating.
Fig. 7.4. Natural function surface morphology bio-replication coating instrument.
Technology Co. Ltd) is 20 cm. An ultrasonic vibration PZT is mounted

on the wheel close to the spraying side of UV curable paint. High-
frequency ultrasonic vibration (frequency 20 kHz and amplitude 1 µm)
is generated throughout the coating processes to assist the UV curable
paint in filling the soft negative mold. This is because high frequency
vibration has a significant liquefaction effect. The curing rate of UV
curable paint is greatly dependent on the power of the UV lamp,
thickness of the soft negative mold and depth of UV curable paint.
After minimum curing time Tmin , which is examined by various coating
experiments, the relation between the distance L between two wheels
and the speed of coating V is decided as
L
V = .
Tmin
7.2.2. Analysis of surface morphology accuracy

of bio-replication coating
Bio-replicated accuracy is assessed by comparing the surface micro-
morphology between the shark skin template and the replicated coating
surface. In general, the placoid shape of shark skin is dependent not only
on its position but also the species of shark. One hole is drilled in the
shark skin before preparation of the soft negative mold as a mark to find
the same placoids on the shark skin template and the coating surface.
The surface morphologies of the same placoid are shown in Fig. 7.5.
The contour at the middle of the placoid in a shark skin surface
is a crucial curved shape. Such contours of the shark skin template
and replicated coating surface are indicated in Figs. 7.6 and Fig. 7.7,
(a) Shark skin template (b) Coating surface
Fig. 7.5. SEM of shark skin and replicated coating.
Fig. 7.6. Horizontal cross-section contour of shark skin and replicated coating.
Fig. 7.7. Vertical cross-section contour of shark skin and replicated coating.
Fig. 7.8. Cross section of bio-replication coating surface.
respectively. From Fig. 7.7, we can see that the placoid ridge of the
shark skin has a slant, about 11◦ . Moreover, it is obvious that the cross-
section contour curves almost coincide, which indicates the accuracy
of bio-replication coating is high enough. Figure 7.8 shows the cross
section photo of the bio-replication coating surface, through which we
can see that the wedge angle on the back of the scale can be perfectly
transferred onto the coating surface.
The surface morphological results of bio-replicated shark skin are
listed in Table 7.1, where S is the average width of a groove and
H is the average height of a riblet. In Table 7.1, compared with
Table 7.1. Replication Coating Comparison of Riblet, L = Left Side;

R = Right Side.
Shark skin Bio-replication Replication

Type template coating error
Average height H(µm) 26.9 ± 0.2 26.2 ± 0.2 −2.60%

Average width S(µm) L 68.44 ± 0.3 68.36 ± 0.3 −0.10%
R 61.66 ± 0.3 61.41 ± 0.3 −0.40%
shark skin template, replicated coating errors of S and H are −0.4%

and −2.6%, respectively. Negative values indicate shrinkage of the
coating surface, which is caused primarily by heat from the UV curing
process. However, the errors are minor enough to achieve a satisfactory
replication accuracy of micro-riblets.
7.2.3. Drag reduction test

Drag reduction tests are conducted in vacuole water tunnel, whose
diagram is shown in Fig. 7.9 or Fig. 6.15. The length of the test
pipe is 0.6 m and the distance between the two orifices of pressure
drop measurement is 0.5 m. Buffer pipe section is set longer than 20 m
to stabilize water flow at front of test pipe section. The inner wall of
test pipe section is painted with the same thickness coating. The flow
velocity in the test section is adjustable between 0∼20 m/s. The power
of water pump is 18.5 kW.
The shape of the test pipe is square. All the skins on each side
of the square are fabricated using large-area bio-replication coating.
Three types of coating surface involving smooth, V-type microgroove
riblet and bio-replicated vivid shark skin are prepared to compare drag
reduction, in which the smooth coating surface (surface roughness
Ra < 0.5 µm) is taken as the reference surface. The surface material
is the same, and the V-type microgroove riblet height is 35 µm, peak
spacing is 80 µm. The experiments are conducted according to the
rules for vacuole water tunnel test (Q/702J0301-2008). The water
temperature is set 28◦ C and test system degassed for more than one
hour before testing.
Fig. 7.9. Experiment system of water tunnel.
The drag-reduction rates (DR) of the microgroove riblet and bio-

replication coating surface were calculated by
Pt − Pr
DR = × 100%, (7.1)
Pr
where Pr indicates pressure drop of smooth reference surface, and

Pt is that of the tested coating surface with microgroove riblet and
bio-replicated vivid shark skin.
The drag reduction rates of microgroove riblet and bio-replicated
vivid shark skin are shown in Fig. 7.10. The drag reduction rate of
microgroove riblet is about 8%, which is in great accordance with
the previous literature.4 Moreover, the bio-replication coating surface
shows drag reduction of about 12%, higher than with microgroove
riblets. This is because bio-replicated shark skin preserves the hier-
archical structure of the natural surface as closely as possible, and is
more complex than regular microgrooves. Unlike the U/V-shaped
microgroove structure, the longitudinal section of the shark skin
scale is not parallel to the direction of water flow — in particular,
there exists an attack angle. The attack angle is observed in a range
of 10–35◦ . The attack angle is one of the important factors for
producing the high drag-reduction efficiency, which can decrease the
turbulence intensity on the basis of grooves, and can also lead to the
Fig. 7.10. Drag reduction test of microgroove riblet and bio-replicated vivid shark
skin.
occurrence of tiny back flow on the valley of the scale. In addition,

due to the fine and sophisticated morphology, the shark skin has
become a kind of super-hydrophobic surface. The effective boundary
slipping phenomenon on the fluid–solid interface may be produced
when the fluid moves, which can decrease the velocity gradient and
viscous resistance significantly; perhaps it is another important factor
to produce “shark skin effect.”14−16
7.2.4. Anti-fouling test

The anti-fouling test is performed by immersing the bio-replication
coating samples in an open algae pond. During the tests, all the
samples are kept still to speed up the propagation of algae upon their
surface. Three types of samples consisting of smooth, microgroove
and bio-replicated vivid shark skin are also prepared, with the same
surface material. Since diatoms in general have strong absorbability,
1mg/mL protein/FITC solution is used to make diatom display clear
Fig. 7.11. Static antifouling test (a) smooth surface (b) microgroove riblet surface
(c) bio-replicated shark skin (left: Microscopic image, right: Fluorescence image).
at fluorescence spectroscopy. The static and dynamic anti-fouling tests

were conducted respectively. By rotating the samples in the algae pond,
the dynamic test was also carried out for two weeks.
Figure 7.11 shows the bio-fouling status of surface samples in the
static test after immersion for two weeks. The dynamic test results
which also lasted two weeks are presented in Fig. 7.12. We find
that the anti-fouling results for the dynamic test were better than
for the static test, indicating the diatoms more easily propagate in
the static test. In the static test, the smooth surface was colonized
and finally entirely covered by a great number of diatoms. Compared
with the smooth surface, diatoms propagated upon the microgroove
riblet surface declined remarkably. It should be noted that the diatoms
Fig. 7.12. Dynamic antifouling test (a) smooth surface (b) microgroove riblet
surface (c) bio-replicated shark skin (left: Microscopic image, right: Fluorescence
image).
distribute unevenly upon the microgroove surface, primarily growing

along the microgroove, as indicated in Fig. 7.11 (b). However, with
respect to the bio-replication coating surface, diatoms were hardly
found except for sporadic colonization at the foot of the placoid, which
implyies that the shark skin has very significant anti-fouling properties.
A Fourier transform infrared spectrometer AVATAR 360 FT-IR was
applied to obtain the infrared spectrum of the samples before and
after immersion. Figure 7.13 shows the infrared spectrums of the test
samples. Because all the test samples were fabricated with the same UV
curable material, the spectrums of the samples are almost same. After
immersion for two weeks, the spectrum of test samples hardly varies.
Fig. 7.13. Infrared spectrum of the samples.
This indicates that the difference in the anti-fouling is a result of the

surface morphology of each test samples.
The anti-fouling coefficient Ai_j was defined as follows:
Ni−surface_j
Ai_j = × 100%, with i = ‘G’, ‘S’ or ‘R’, (7.2)
Nsmooth_static
where Ni−surface_j indicates the number of algae growing upon

microgroove ‘G,’ smooth surface ‘S’ or replicated shark skin surface ‘R’
in j-state. The j is ‘s’ to indicate static test and ‘d’ to indicate dynamic
test, and Nsmooth_static is that of the smooth surface in static test. The
diatoms propagated upon the surface samples are statistically counted
as listed in Table 7.2. Compared to about 100 diatoms per square
millimeter of smooth surface in the static test, about 50 and 4 diatoms
grow upon microgroove and bio-replicated vivid shark skin in the static
test, respectively. And the diatoms decline in the dynamic test — in
particular only about 2 diatoms can be found on the bio-replicated
shark skin. Anti-fouling coefficients of microgroove and bio-replicated
Table 7.2. Diatoms Propagated upon Surface Samples.
Static test Dynamic test

Number Anti-fouling Number Anti-fouling
of ditoms coefficient of ditoms coefficient
Surface type (/mm2 ) Ai_s (%) (/mm2 ) Ai_d (%)
Smooth surface i = S 100 — 65 65%

Microgroove surface i = G 50 50.0% 35 35%
Bio-replicated shark skin i = R 4 4.0% 2 2%
vivid shark skin related to smooth surface in the static test are 50%
and 4%, which indicate bio-replicated shark skin marvelously protects
the surface from bio-fouling. Previous studies have made clear that the
effects of surface microtexture are clearly related to the scale of the
test organisms, and fouling organisms that are larger than the scale of
the microtexture would have reduced adhesion strength due to fewer
attachment points, i.e. attachment point theory. Hierarchical surface
characteristics of bio-replicated vivid shark skin will lower the number of
attachment points, which reduces adhesion strength more remarkably
than regular microtexture.
7.3. Large Proportional Amplification

Bio-replicated shark skin shows the best drag reduction of about
8% only when the water speed is around 5m/s as compared with
micro groove riblets,4−6 which implies that optimal velocity range
of drag reduction is around 5m/s for shark skin. Adjusting optimal
velocity range of drag reduction via controllable modification of
micromorphology is very necessary to widen its industrial applica-
tions. The microgroove drag-reduction mechanism has illustrated that
equal-proportional scaling is one way to guarantee the optimal drag-
reduction function of shark skin being retained under different ambient
conditions.
The bio-replication forming approach is based on the micro
molding process, thus large-scale amplified bio-replicated shark skin
can be fabricated if the female mold material has equal-proportional

swelling property, so-called bio-scaling forming. To ensure the excellent
function of the natural surface, mold material with superior microscopic
structure and large-scale swelling is of great importance in bio-scaling
forming. Nevertheless, few materials meet the demands so far, for exam-
ple some metals and ceramics25,26 can be amplified to some degree due
to thermal expansion effect, but the expansion coefficient is too low.
Other materials such as copolymer hydrogels27 and sponges28−30 can be
large-scale swollen in water, but the micro-pores swell simultaneously
to destroy the original microscopic structure.
Over past decades, polymer science has made great progress thanks
to worldwide R&D, especially in the development of novel polymer
material. Among their specific properties, large-scale swelling is a long-
term pursued one. Polydimethylsiloxane (PDMS), as one type of cross-
linked polymer, generally swells to large scale in response to some
organic solvent like n-hexane.31−33 Moreover, its surface microscale
structure remains undamaged even in the case of large-scale swelling,
which meets the demands of bio-scaling forming. The swelling effect
of PDMS in general is to be avoided in micro fluidic chip fabrication.34
Only a few papers make positive use of its swelling effect. X. Li et al.
applied PDMS as stamp material to adjust 2-D lattice spacing of
colloidal crystals by about 1.49 times large-scale swelling.35,36 Daniel
P. Brennan et al. synthesized a special solvent-swelling PDMS to
form micropore arrays for filling other particles.37 Bjorn Samel et al.
proposed one approach to control fluid flow in micro channel by use
of thermal-swelling PDMS.38 K.L. Lai used PDMS absorbing poor
solvent in resist to deswell the resist of swollen PMMA gel during
imprinting, which resulted in easy demolding.39 Almost all of them are
related only to simple application of the PDMS swelling effect on a 2-D
surface, or even 1-D channel. There is still no literature about accurate
control of the PDMS swelling effect and modulation of complex 3-D
micro-structure.
In an attempt to solve the geometric unadjustable problem of
function surface bio-replicated in bio-replication forming, one novel
bio-scaling forming technology is proposed by taking advantage of the
PDMS swelling effect, especially for accurate swelling control of micro-
scale surface structure.
7.3.1. Experimental section

The skin of Carcharhinus brachyurus was chosen as the bio-scaling
template. Polydimethylsiloxane (PDMS) elastomer kits (Sylgard 184)
were purchased from Dow Corning (Midland, MI). Simethicone
(Xingye Silicone Material Corp, Tsingtao, China) and n-hexane (Tian-
jin Guangfu Fine Chemical Research Institute) were used as mold-
releasing agent and swelling solvent respectively in this experiment.
Epoxide resin (Model: XY-2) and deionized water (Juyuan Corp,
Beijing) were used as received.
7.3.2. Controllable bio-scaling forming process

based on PDMS swelling
Controlling the swelling of PDMS is very important for accurate
amplification of natural functional surface micromorphology. As shown
in Fig. 7.14, the proposed bio-scaling forming process is divided
Epoxide
Shark skin or
resin
amplified mould
Moulding
of PDMS
Liquid or different Degassing
PDMS concentration of
gaseous n-hexane
Swelling of
Airproof Gaseous
PDMS mould
solifying n-hexane
Degassing
& solifying
Flexible
Flexible
demoulding
demoulding
Flexible
demoulding
PDMS Amplified
mould bio-replicate
shark skin
Step 2: Swelling of PDMS mould Step 3: Forming of amplified

Step 1: Moulding of PDMS
bio-replicated shark skin
Fig. 7.14. The process of amplifying bio-replicated shark skin with gaseous n-hexane.
into three steps: molding of PDMS, large-proportional swelling of

PDMS (in gaseous or liquid n-hexane), and forming of amplified bio-
replicated shark skin. The second step, i.e. large-proportional swelling
of PDMS mold, is the most important to achieving accurate bio-scaling
forming. The PDMS mold is placed in an airtight container filled with
liquid or different concentration gaseous n-hexane for a certain time.
The swelling ratio and rate of the PDMS mold is dependent on the
concentration of the gaseous n-hexane. Afterwards, the swollen PDMS
mold is taken out as template in third step, upon which epoxide resin
is poured and degassed for three minutes. The poured mold is placed
in the airtight container full of gaseous n-hexane for six hours. The
concentration of the gaseous n-hexane is same as that used in the second
step. The amplified bio-replicated shark skin is obtained after epoxide
resin is solidified upon the swelling PDMS mold.
7.3.3. Swelling ratio control of PDMS

The size of the PDMS mold sample in this experiment is 15 mm
× 10 mm × 5 mm. The PDMS mold swells in large scale in both
gaseous and liquid n-hexane, and the optimum swelling ratio of PDMS
reaches about 34% in gaseous and 39% in liquid n-hexane, as shown
in Fig. 7.15.
Fig. 7.15. (a) The original PDMS mold (b) The PDMS mold swells in gaseous
n-hexane.
Fig. 7.16. The swelling trend of PDMS in gaseous n-hexane of different concentra-
tion (D is the original size of the PDMS mold, D0 is the amplified size of the PDMS
mold, S = D/D0).
In the gaseous n-hexane, the concentration of n-hexane plays great

role in swelling ratio and swelling rate. Figure 7.16 shows the PDMS
sample volume change along with maintaining time in gaseous n-
hexane. The trend of PDMS swelling in gaseous n-hexane is similar, i.e.
the size of PDMS mold swelling slowly as maintained inside gaseous n-
hexane. Because the swollen PDMS is taken as the bio-replication mold
in step 3, swelling to a stable state is inevitable for accurate formation
of the function surface. The PDMS sample takes at least 10 hours
to reach the highest point of swelling behavior and retains stability
in volume afterwards, which point is called the best swelling ratio.
Figure 7.17 shows the best swelling ratio at different gaseous n-hexane
concentrations. It can be seen that the best swelling ratio of PDMS
increases with improvement in the gaseous n-hexane concentration.
But when the concentration is above 1%, the swelling stops and stays at
around 34%, which is the optimum swelling ratio in all concentrations
Fig. 7.17. The best swelling ratios at different gaseous n-hexane concentrations.
of gaseous n-hexane. Thus, the swelling ratio of bio-scaling forming can

be controlled by elaborative choice of n-hexane gaseous concentration.
The swelling trend of PDMS mold in the liquid n-hexane is similar
to that in the gaseous n-hexane as shown in Fig. 7.18. The PDMS
mold swells and reaches best swelling ratio, 39%, after 400 minutes
(Fig. 7.19). When compared to Fig. 7.16, it is obvious that both the
swelling rate and best swelling ratio in liquid n-hexane is much higher
than in gaseous n-hexane. However, the PDMS mold cannot swell in
equal proportion as when partially immersed in liquid n-hexane. The
swelling ratio of the partially immersed region is higher than other
region, resulting in each surface of the PDMS mold becoming concave,
as shown in Fig. 7.20. To ensure high-precision bio-replication, the
PDMS mold must be prevented from being partially immerse in liquid
n-hexane.
If the fully swollen PDMS mold is transferred from liquid n-hexane
to gaseous n-hexane at over 1% concentration, the PDMS mold slowly
shrinks from 39% to 34%, which will take one hour. The shape of the
PDMS mold is maintained in the process of shrinking.
Fig. 7.18. (a) The original PDMS mold (b) The PDMS mold swells in liquid
n-hexane.
Fig. 7.19. The swelling trend of PDMS in liquid n-hexane.

Fig. 7.20. The PDMS mold swelling when partially immersed in liquid n-hexane.
7.3.4. Control of swelling rate of PDMS

The swelling rate of PDMS is greatly affected by gaseous n-hexane
concentration. Generally, the PDMS mold takes less time to reach
the best swelling ratio when the gaseous n-hexane concentration is
higher. However, when the gaseous n-hexane concentration is higher
than 1%, the time to the best swelling ratio (34%) will keep stable at
about 2,000 min, as shown in Fig. 7.21. The n-hexane concentration
needs to be chosen according to the required swelling ratio of bio-
scaling forming. And the PDMS mold must be immersed for sufficient
time to reach the best swelling ratio. The PDMS mold swells more
rapidly in liquid n-hexane, taking just 400 minutes to reach 39% best
swelling ratio. Moreover, the full-swollen PDMS mold immersed in
liquid can slowly shrink to best swelling ratio of gaseous n-hexane
as keeping it inside gaseous n-hexane for just 1h. Consequently,
the PDMS swelling efficiency can be improved by initially swelling
in liquid n-hexane and then transferring it into air-tight gaseous
n-hexane.
Fig. 7.21. The time PDMS spends to reach the best swelling ratio at different gaseous
n-hexane concentrations.
7.3.5. Micromorphology accuracy of bio-scaling

forming
The biomimetic shark skin is fabrication in step 3 (Fig. 7.14) of bio-
scaling forming by pouring epoxide resin upon the swollen PDMS
mold. Figure 7.22 shows the results of the bio-scaling formed shark
skin at 1% gaseous n-hexane. By comparison between original and
amplified shark skin, we can see that the template is amplified in equal
proportion in the macro scale, and the amplification ratio is about
34%. More importantly, the micro morphology of amplified shark skin
surface is also kept satisfactorily. The single scale marked with a circle in
Figs. 7.23a and b separately shows the original shark skin scale and its
corresponding amplified scale. Comparing the shape and size of each
corresponding scale, it can be seen that the micro structures of shark
skin are well preserved and amplified in equal proportion.
Fig. 7.22. (a) The original real shark skin and (b) its corresponding amplified repli-
cation (c) The SEM image of the amplified replication.
Fig. 7.23. SEM images of (a) the real original shark skin and (b) its corresponding
amplified replication.
To quantitatively analyze bio-scaling forming accuracy, micromor-

phology of the amplified biomimetic shark skin whose PDMS female
mold swollen in 1% gaseous n-hexane was scanned by means of 3-D
white light interference profilometer. The 3-D white light interference
image of the original shark skin scale and the corresponding amplified
one are separately shown in Fig. 7.24. As the main structure of a
shark skin scale is made up of three ridges and two grooves, such
relating characteristical parameters are obtained to assess amplification
accuracy. The locations of surface curves are labeled by black lines as
shown in Fig. 7.24, in which ri(i=1,2,3) and Ri(i=1,2,3) separately
indicate scale ridges of real shark skin and corresponding amplified
replication, and gi(i=1,2) and G1i(i=1,2) separately indicate grooves.
Fig. 7.24. 3-D white light interference images of (a) the real shark skin scale and
(b) its corresponding amplified replication.
The amplified replication is contracted 1.34 times in every direction to

make them uniform in scale. If bio-scaling forming accuracy is higher,
surface micromorphology of the real shark skin will be closer to that of
1.34 times contracted bio-scaled shark skin. The characteristic curves
of real shark skin and its corresponding contracted bio-scaled shark
skin are separately shown as solid and dot dash lines in Fig. 7.25. The
value difference between solid line and dot dash line represent the bio-
scaling error which can be statistically accumulated from 90 points over
lines. Error statistic shows that the average error is 0.0268 µm and the
maximum error is 0.0352 µm, i.e. the bio-scaling forming accuracy is
higher than 95%.
To further study the bio-scaling forming accuracy, nine more
surface curves perpendicular to the middle ridge are obtained at
intervals of 18 µm between two adjacent lines. These curves were built
in Matlab as shown in Fig. 7.26, in which the original shark skin and
bio-scaled shark skin are separately shown as black and blue lines, and
bio-scaled shark skin was contracted 1.34 times in every direction. The
z-value difference between real and bio-scaled shark skin represents
bio-scaling forming error. From Fig. 7.26, it is clear that the average
error is 0.0185 µm and maximum error is 0.0502 µm.
Fig. 7.25. The surface curves of the real shark skin and its corresponding amplified
replication which has been zoomed out 1.34 times. The curve is located at (a) r1 and
R1 (b) r2 and R2 (c) r3 and R3 (d) g1 and G1 (e) g2 and G2.
Fig. 7.25. (Continued)
7.4. Large-Proportional Shrunken

Bio-Replication of Shark Skin Based
on UV-Curing Shrinkage
The bio-replication forming approach is based on the replica mold
process, thus large equal-proportional shrunken bio-replication can
be conducted if negative mold materials possess shrinking properties.
Fig. 7.26. Modeling of (a) the original shark skin scale in black line (b) its amplified
replication in blue line and (c) their combination.
High-quality microscopic surface structure and equal-proportional

shrinkage are two important demands for shrinking material to ensure
translation of excellent natural surface function. Nevertheless, only
few materials meet such demands so far. Fortunately, over the past
few decades, material science has made great progress thanks to
worldwide R&D. Previous studies have proposed a large-scale equal-
proportional amplification bio-replication approach on basis of solvent-
swelling effect of Polydimethylsiloxane,40−42 which can adjust the
micro-riblets of shark skin to achieve about 140% solvent-swelling ratio
at one-time amplification. But achieving large-proportional shrinkage
in bio-replication of shark skin is still a bottleneck problem.
As is well known, UV-curable material can shrink at equal propor-
tions after curing, and its microscale surface structure remains intact. It
has become a hot research area to study the shrinkage of UV-curable
material, however there are only few papers dealing with positive use
of the shrinking effect. Almost all current studies focus on how to
reduce the shrinkage of UV-curable materials, such as reducing the
shrinkage by low-temperature,43 more illumination time,44 low light
intensity,45 changing the composition of the UV-curable material,46
and other methods.47 There is still no literature about controlling of
UV-curable material shrinkage to adjust complex 3-D microstructure.
One novel large-proportional shrunken bio-replication is proposed

by taking advantage of the shrinking effect of UV-curable material.
The shrinking effect of UV-curable material is investigated, and the
shark skin is taken as a template for large-proportional shrunken bio-
replication. The shrunken bio-replicated accuracy is tested by the
comparison between the original shark skin and the large-proportional
shrunken bio-replicated shark skin. Finally, the drag reduction rate
of the bio-replicated shark skin is tested in the water tunnel,
and the efficiency of large proportional shrunken bio-replication is
validated.
7.4.1. Materials and method

The skin of Carcharhinus brachyurus was chosen as the original
template. Bisphenol-A epoxy acrylate (621A-80, Eternal Chemical Co.,
Ltd., Taiwan) was used as the pre-polymer of UV curable material.
Triethylene glycol diacylate (EM223, Eternal Chemical Co., Ltd., Tai-
wan) was taken as the diluent. 2-Benzyl-2-(dimethylamino)-1-[4-(4-
morpholinyl)phenyl]-1-butanone was purchased as the photo-initiator
(Insight High Technology Co., Ltd., Beijing). The 621A-80 and
EM223 was mixed at 1:1 in weight, and then poured into 1% solu-
tion of 2-Benzyl-2-(dimethylamino)-1-[4-(4-morpholinyl)phenyl]-1-
butanone solution. After full mixing, the solution is used for large-
proportional shrunken bio-replication.
7.4.2. Large-proportional shrunken bio-replication

process
The proposed large-proportional shrunken bio-replication process is
divided into three sub-steps, as shown in Fig. 7.27 preparation of shark
skin negative mold by use of UV-curable material, shrinking of UV-
curable mold, and bio-replication forming of shark skin from shrunken
negative mold. Accurate control of the shrinkage, especially of the
UV-curable material mold, is essential to achieving large-proportional
shrunken bio-replication of shark skin. Degassing is conducted by
a vacuum chamber (DZF-6020, Yiheng Corp., Shanghai). All the
processes are conducted at 20◦ C.
Shark skin
UV-curable Shrunken mould

material
Epoxide resin
Standing
at 20
Degassing &
UV-curing After
20 hours Degassing
Flexible
demoulding Shrunken
shark skin
UV-curable mould
Step 1: Moulding of Step 2: Shrinking of Step 3: Forming of
UV-curable material UV-curable mould shrunken shark skin
Fig. 7.27. The process of equal-proportional shrunken bio-replication.
7.4.3. Shrinking ratio and rate

of large-proportional shrinkage
of UV-curable mold
The UV-curable mold generally has an inherent shrinkage property,
i.e. size of curable mold shrinks with prolonged placing time in air. The
shrinking process of the UV-curable cylindrical mold sample (diameter
11 mm; height 10 mm) is shown in Fig. 7.28. It is clear that size
of UV-curable mold changes remarkably with increase in placing time,
and gradually becomes stable in size after 20 hours. The size shrinkage
along height (Z) and radius direction (R) of the mold are approximately
same, which indicates large-scale equal-proportional shrinkage can be
achieved. Moreover, the optimum shrinking ratio generally increases
with ambient temperature because the volatilization of the diluents
improves with temperature. The shrinking ratio S is presented to
Fig. 7.28. The shrunking process of UV-curable mold.
assess the shrinking degree by (D0-D)/D0, where D0 is the side

length of original UV-curable mold and D is the side length of the
shrunken mold.
The volume change of UV-curable mold along with placing time in
air is presented in Fig. 7.29a. The size of mold slowly shrinks after full
curing of UV-curable materials, and the speed of shrinkage gradually
declines until a constant shrinking ratio is reached. The shrunken
mold is directly taken as the bio-replication template, whose volume
should be kept in the stable state to ensure accurate bio-replication of
the natural surface. Evaporation of diluent determines the shrinking
of UV-curable mold, as shown in Fig. 7.30. The shrinking ratio of the
UV-curable mold can be adjusted by changing the diluent content of
UV-curable material. The volume of the UV-curable mold uniformly
shrinks, and gradually reaches a stable state after at least 20 hours.
The shrinking behaviors of UV-curable mold in height and radius are
almost the same. More importantly, the shrinking ratio hardly varies
with change of thickness (Fig. 7.29b). The maximum shrinking ratio
of UV-curable mold at 20◦ C is around 23%, which ensures large-
proportional shrunken bio-replication.
Fig. 7.29. The shrinking behavior of UV-curable mold at 20◦ C.
Evaporation of diluent
Fig. 7.30. The shrinking mechanism of photo polymerization.
7.4.4. Micromorphology accuracy of large

proportional shrunken bio-replication
The quality of surface morphology is another important factor for bio-
replicated shark skin. SEM photos of the original shark skin template
and large-proportional shrunken bio-replicated shark skin are shown in
Fig. 7.31. In order to obtain the same scale, a micro-hole was drilled
on the original shark skin template as the anchor point. We can see
that the bio-replicated shark skin scale from the UV-curable material
is remarkably shrunken as compared with the corresponding original
shark skin scale, whose shrinking ratio is also about 23%. Moreover,
the microstructures of the shark skin scale are well preserved even
Fig. 7.31. SEM images of (a) the original shark skin and (b) its corresponding
shrunken replication.
though large-proportional shrinkage of bio-replicated shark skin occurs

after curing, indicating the large-proportional shrunken bio-replication
based on UV-curable materials can provide desirable micro-scale surface
quality.
To quantitatively analyze large-proportional shrunken bio-
replication accuracy, micro-scale surface morphologies of test samples
are scanned by means of 3-D white light interference profilometer.
The 3-D scanned surface morphologies of the original shark skin and
the corresponding large-proportional shrunken bio-replicated shark
skin are separately included in Fig. 7.32. As the main structural
characteristics, this shark skin scale is made up of three ridges and two
grooves. The relating characteristic parameters of ridges and grooves
are obtained to assess the large-proportional shrunken bio-replication
accuracy. The position of the ridge and groove curves are labeled
with black lines as shown in Figs. 7.32a and b, in which Ri(i=1,2,3)
and ri(i=1,2,3) separately indicate ridges of original shark skin and
corresponding shrunken bio-replicated shark skin, Gi(i=1,2) and
gi(i = 1,2) separately indicate grooves, and Di(i=1,2) and di(i=1,2)
are the vertical curves over the scale. The scanned data are used
to rebuild these curves as shown in Fig. 7.33, in which the value
of shrunken bio-replicated shark skin is divided by 0.77 in every
direction to make them uniform in scale. The closer the curves of
shrunken bio-replicated shark skin superimpose the original shark skin,
Fig. 7.32. 3-D white light interference images of (a) the real shark skin scale and
(b) its corresponding shrunken replication.
the higher the large-proportional shrunken bio-replication forming

accuracy.
The shrunken bio-replication accuracy is assessed by use of statisti-
cal average on difference of 80 points over lines. Statistical assessment
shows that the average error is 0.01921 µm and the maximum error
is 0.51077 µm, indicating the forming accuracy of large-proportional
shrunken bio-replication is higher than 95%.
7.4.5. Drag-reduction test of shrunken

bio-replicated shark skin
In order to validate translation of optimal drag-reduction fluid velocity
region of large-proportional shrunken bio-replicated shark skin, the
drag-reduction test was conducted in vacuole water tunnel at the China
Ship Scientific Research Center (CSSRC). The length of the test section
in vacuole water tunnel is 3.2 m with diameter 0.8 m. The measurement
range of strain gauge balances is 300 N and its resolution is 0.1 N. Test
model with hollow elliptical sphere at front end and hollow cylinder at
posterior segments is made from aluminum alloy LY12, whose size is
90 mm in external diameter and 500 mm in cylinder length as shown in
Fig. 7.34 or Fig. 5.13 or Fig. 6.26. The size of UV-curable mold used
to fabricate test skins is 83 mm × 3200 mm × 35 mm in width, length
Fig. 7.33. Surface curves of the real shark skin and its corresponding shrunken
replication (a) ridge curves (b) groove curves (c) vertical curves.
Fig. 7.34. Drag-reduction test model.
Fig. 7.35. Drag-reduction test.
and thickness. All the test skins are made from waterborne epoxy resin,
including smooth skin, original bio-replicated shark skin, and shrunken
bio-replicated shark skin. The test skins are strongly pasted to the test
model, and drag-reduction experiments are conducted according to
the rules for vacuole water tunnel test (Q/702J0301-2008). The water
temperature is set 20◦ C and test system degassed for more than one
hour before testing.
Taking the smooth skin as reference, the drag-reduction ratio of
the original bio-replicated shark skin and the shrunken bio-replicated
shark skin are shown in Fig. 7.35. It is obvious that the optimal
drag-reduction velocity region can be adjusted by large-proportional
shrunken bio-replication, i.e., the peak of drag-reduction ratio transla-
tional moves by control of bio-replication shrinkage. In particular, the
peak of drag reduction remains identical, about 11%, which agrees well
with our previous studies. The adjustability of the drag-reduction peak
indicates that large-proportional shrunken bio-replication can obtain
excellent drag reduction of nature surface from living environment to
various industrial environments.
References
[1] Walsh M, Weinstein L. (1978) Drag and heat transfer on surfaces with small
longitudinal fins. In: 11th Fluid and Plasma Dynamics Conference.
[2] Liu K, Christodoulou C, Riccius O, Joseph DD. (1990) Drag reduction in pipes
lined with riblets, AIAA J 28: 1697–1699.
[3] Choi H, Moin P, Kim J. (1993) Direct numerical simulation of turbulent flow
over riblets. J Fluid Mech 255: 503–539.
[4] Bechert D, Bruse M, Hage W, et al. (1997) Experiments on drag-reducing
59–87.
[5] Walsh MJ. (1983) Riblets as a viscous drag reduction technique. AIAA J 21:
485–486.
[6] Dean B, Bhushan B. (2010) Shark-skin surfaces for fluid-drag reduction in
turbulent flow: a review. Philos T Roy Soc A 368: 4775–4806.
[7] Wilkinson S, Anders J, Lazos B, Bushnell D. (1988) Turbulent drag reduction
research at NASA Langley: progress and plans. Int J Heat Fluid Fl 9: 266–277.
[8] Lee S-J, Lee S-H. (2001) Flow field analysis of a turbulent boundary layer over
a riblet surface. Exp Fluids 30: 153–166.
[9] Dean B, Bhushan B. (2012) The effect of riblets in rectangular duct flow. Appl
Surf Sci 258: 3936–3947.
[10] Viswanath P. (2002) Aircraft viscous drag reduction using riblets. Prog Aerosp
Sci 38: 571–600.
[11] Lang A, Motta P, Habegger ML, et al. (2011) Shark skin separation control
mechanisms. Marine Technol Soc J 45: 208–215.
[12] Lang A, Motta P, Hidalgo P, Westcott M. (2008) Bristled shark skin: a

microgeometry for boundary layer control? Bioinspiration & Biomimetics 3:
046005.
[13] Oeffner J, Lauder GV. (2012) The hydrodynamic function of shark skin and two
biomimetic applications. J Exp Biol 215: 785–795.
[14] Zhang D, Luo Y, Chen H. (2011) Application and numerical simulation research
on biomimetic drag-reducing technology for gas pipelining. Oil Gas-European
Magazine 37: 85–90.
[15] Zhang D, Luo Y-H, Li X, Chen H-W. (2011) Numerical simulation and
experimental study of drag-reducing surface of a real shark skin. Journal of
Hydrodynamics, Ser. B 23: 204–211.
[16] Yuehao L, Deyuan Z. (2013) Investigation on fabricating continuous vivid
sharkskin surface by bio-replicated rolling method. Appl Surf Sci 282: 370–
375.
[17] Ball P. (1999) Engineering shark skin and other solutions. Nature 400: 507–
509.
[18] Bechert D, Bruse M, Hage W, Meyer R. (1997) Biological surfaces and their
technological application—laboratory and flight experiments on drag reduction
and separation control. AIAA paper 97-1960.
divergent wall riblets. Exp Fluids 33: 346–350.
[20] Zhang D, Li Y, Han X, et al. (2011) High-precision bio-replication of synthetic
drag reduction shark skin. Chinese Sci Bull 56: 938–944.
[21] Chen H, Zhang X, Zhang X, et al. (2013) Large-scale equal-proportional
amplification bio-replication of shark skin Based on solvent-swelling PDMS.
J Appl Polym Sci 130: 2383–2389.
[22] Chen H, Zhang X, Che D, et al. (2014) Synthetic Effect of Vivid Shark Skin and
Polymer Additive on Drag Reduction Reinforcement. Advances in Mechanical
Engineering 6: 425701.
[23] Stenzel V, Wilke Y, Hage W. (2011) Drag-reducing paints for the reduction of
fuel consumption in aviation and shipping. Prog Org Coat 70: 224–229.
[24] Wang J-Y, Lv M-X, Kou Y, et al. (2012) UV-curable phthalazinone-based epoxy
methacrylate antifouling coatings with N-vanillylnonanamide as antifoulant.
J Coating Technol Res 9: 309–315.
[25] Jakubinek MB, Whitman CA, White MA. (2010) Negative thermal expansion
materials. J Therm Anal Calorim 99(1): 165–172.
[26] Roy R, Agrawal DK, McKinstry HA. (1989) Very Low Thermal Expansion
Coefficient Materials. Annu Rev Mater Sci 19(1): 59–81.
[27] Shim WS, Yoo JS, Bae YH, Lee DS. (2005) Novel injectable pH and temperature
sensitive block copolymer hydrogel. Biomacromolecules 6(6): 2930–2934.
[28] Friedrich T, Tieke B, Stadler FJ, Bailly C. (2011) Copolymer Hydrogels of
Acrylic Acid and a Nonionic Surfmer: pH-Induced Switching of Transparency
and Volume and Improved Mechanical Stability. Langmuir 27(6): 2997–3005.

[29] Mabilleau G, Basle MF, Chappard D. (2006) Evaluation of surface roughness of
hydrogels by fractal texture analysis during swelling. Langmuir 22(10): 4843–
4845.
[30] Gui XC, Cao AY, Wei JQ, et al. (2010) Soft, Highly Conductive Nanotube
Sponges and Composites with Controlled Compressibility. ACS Nano 4(4):
2320–2326.
[31] Yoo SH, Cohen C, Hui CY. (2006) Mechanical and swelling properties of PDMS
interpenetrating polymer networks. Polymer 47(17): 6226–6235.
[32] Lee JN, Park C, Whitesides GM. (2003) Solvent compatibility of
poly(dimethylsiloxane)-based microfluidic devices. Anal Chem 75(23): 6544–
6554.
[33] Favre E. (1996) Swelling of crosslinked polydimethylsiloxane networks by pure
solvents: Influence of temperature. Eur Polym J 32(10): 1183–1188.
[34] Dangla R, Gallaire F, Baroud CN. (2010) Microchannel deformations due to
solvent-induced PDMS swelling. Lab on a Chip 10(21): 2972–2978.
[35] Li X, Wang TQ, Zhang JH, et al. (2010) Modulating Two-Dimensional
Non-Close-Packed Colloidal Crystal Arrays by Deformable Soft Lithography.
Langmuir 26(4): 2930–2936.
[36] Ren ZY, Li X, Zhang JH, et al. (2007) Tunable two-dimensional non-close-
packed microwell arrays using colloidal crystals as templates. Langmuir 23(15):
8272–8276.
[37] Brennan DP, Dobley A, Sideris PJ, Oliver SRJ. (2005) Swollen poly (dimethyl-
siloxane) (PDMS) as a template for inorganic morphologies. Langmuir 21(25):
11994–11998.
[38] Samel B, Sandstrom N, Griss P, Stemme G. (2008) Liquid Aspiration and
Dispensing Based on an Expanding PDMS Composite. J Microelectromech S
17(5): 1254–1262.
[39] Lai KL, Leu IC, Hon MH. (2009) Soft imprint lithography using swelling/
deswelling characteristics of a polymer mold and a resist induced by a poor
solvent. J Micromech Microeng 19: 037001.
[40] Chen H, Zhang X, Zhang D, et al. (2013) Large-Scale Equal-Proportional
Amplification Bio-Replication of Shark Skin Based on Solvent-Swelling PDMS.
J Appl Polym Sci 130: 2383.
[41] Pan J, Chen H, Zhang D, et al. (2013) Large-scale solvent-swelling-based
ampli?cation of microstructured sharkskin. J Micromech Microeng 23: 075018.
[42] Pan J-F, Chen H-W, Zhang D-Y, et al. (2013) Accurate Bio-scaling Processing
of Micro Structured Shark Skin Based on Swelling of Polydimethylsiloxane. Int
Polym Proc 28: 341.
[43] Lu B, Xiao P, Sun M, Nie J. (2007) Reducing volume shrinkage by low-
temperature photopolymerization. J Appl Polym Sci 104: 1126.
[44] Uyama S, Irokawa A, Iwasa M, et al. (2007) Influence of irradiation time on

volumetric shrinkage and flexural properties of flowable resins. Dent Mater J 26:
892.
[45] Felipe L, Schneider J, Cavalcante LM, Silikas N. (2010) Shrinkage Stresses
Generated during Resin-composite Applications: A Review. J Dent Biomech
2010: 131630.
[46] Spinell T, Schedle A, Watts DC. (2009) Polymerization shrinkage kinetics of
dimethacrylate resin-cements. Dent Mater 25: 1058.
[47] Micheal M, El-Zaher N. (2003) Efficiency of ultraviolet/ozone treatments in
the improvement of the dyeability and light fastness of wool. J Appl Polym Sci
90: 3668.
Chapter 8
Study of Flow over Dimpled Cylinder

for Drag Reduction
Tan S.P.,∗ Koh J.H. and Ng Y.K. Eddie
School of Mechanical & Aerospace Engineering, Nanyang

Technological University,
50 Nanyang Avenue, Singapore 639798
∗ shupeitan91@gmail.com
Drag reduction of flow over a circular cylinder is done by having a

dimpled surface. Computational fluid dynamics (CFD) simulations are
done to investigate the degree of drag reduction by dimpled cylinder
as compared to smooth cylinder, and the results are to be validated by
wind tunnel experimentations.
The simulations were conducted for Reynolds number (Re) ranging
from 20,000 to 160,000, which corresponds to the critical Re range
of a golf ball. The aerodynamics coefficients such as drag coefficient
and pressure coefficient were compared between both dimpled and
smooth cylinders and the results showed effective drag reduction by an
average of 9.16%. On the other hand, experimentations were carried
out for Re 60,000 to 140,000 due to wind speed limitations, and the
differences were significant as compared to the simulation results. These
were mainly due to the interference of flow during the experiment and
the inaccuracies of the CFD model. Nonetheless, drag reduction was
also achieved through wind tunnel experimentations where an average
value of 8.39% was recorded.
271
8.1. Introduction
Over the years, there have been many attempts made in rationalizing
the definitions and terminology behind drag. Total drag is the force
corresponding to the rate of decrease in momentum. It is unfavourable
as it causes structural vibrations, and hence it is extremely detrimental
to many engineering structures. Therefore, drag reduction is of par-
ticular interest to many industries as it leads to higher durability of
equipment.1−5
Drag reduction is done through extensive research on controlling
the fluid flow over bluff bodies. It can be achieved through two
main approaches: passive and active flow controls. Passive flow control
relies on the structural components of the bodies such as dimples and
riblets,3−5 and active flow control makes use of additional installation of
devices to direct wake flow structures such as an upstream installation
of a small control rod.
The focus of this chapter is to investigate the effects of dimples
on the surface of a circular cylinder. It is to induce transition and thus
create a turbulent boundary layer to delay the point of separation in
order to reduce drag.
8.2. Flow over a Circular Cylinder
Due to the viscosity of real fluid, a thin layer called the boundary
layer will form. When the upstream flow reaches the leading edge of
a flat surface, fluid starts to slow down and a laminar boundary layer
is created. However, the fluid climbs up over smooth rounded surface
of the cylinder due to favorable pressure gradient and may accelerate
if it is a suction surface. Due to the no-slip condition imposed by
the viscosity of fluid, the fluid slows down as it moves downstream.
Hence, the boundary layer begins to grow. As the laminar boundary
layer continues to grow, it will become unstable and transition to a
turbulent boundary layer. Turbulent flow is highly irregular and eddies
occur. When a viscous fluid flows over a cylinder, there will be vortices
in a thin boundary layer adjacent to the cylinder where occurrence
of boundary layer separation and a trailing wake will form eventually
behind the cylinder.1,2
Study of Flow over Dimpled Cylinder for Drag Reduction 273
Fig. 8.1. Flow comparison between laminar and turbulent boundary layer.6
Separation of flow from the surface will occur, which is the result of
a rising external pressure along the surface (adverse pressure gradient)
that resists the flow. In order to oppose this rising pressure, the fluid
particles lose their kinetic energy. Separation of flow is able to occur
for both laminar and turbulent boundary layers. However, for the
turbulent boundary layer, the greater momentum transfer due to eddies
will be able to raise the velocity near the surface. Therefore, it has more
kinetic energy to oppose the adverse pressure gradient. With this, the
point of separation is delayed (refer to Fig. 8.1 for laminar and turbulent
flow comparison), and hence drag is reduced.6
Therefore, some methods of drag reduction are to induce the
transition to turbulent boundary layer. In this chapter, having dimples
on a circular cylinder produces turbulence in the slow-moving air, and
the more energetic layer separates much later, thus leading to a smaller
wake and drag. This method was inspired by the case of golf balls, where
further distance is achieved, encouraging the incorporation of dimples
on circular cylinders.
Fig. 8.2. Variation of CD with Re (2-D circular cylinder).7
Drag is usually characterized by its own dimensionless coefficient,

CD , which is a function of the Reynolds number (Re); their relationship
(based on experimental data) is shown in Fig. 8.2. From the figure,
it can be observed that CD is very large when Re is less than 100.
However, it decreases monotonically until Re ≈ 3 × 105 . From this Re
onwards, there is significant drop of CD from about 1 to 0.3 and then
recovery to 0.6 for Re = 107 . This is due to the sudden transition from
laminar to turbulent boundary layer.
However, the drag coefficient, CD , does not depend only on Re.
The surface roughness of the cylinder may affect it as well. This is
because the drag force is made up of two major components: pressure
drag and skin friction drag. As shown in Fig. 8.3, for a smooth cylinder,
the transition occurs at Re ≈ 5 × 105 , whereas for a rough cylinder, the
transition occurs at a lower Re of around 105 .
Vortex shedding can be observed at Re of around 105 where
the flow has become unstable and the vortices will be alternately
be shed from the body and flow downstream. It becomes turbulent
and transform into a distinct wake with the laminar boundary layer
separating from the surface.
There have been many methods to investigate drag reduction. In
this chapter, drag reduction of a dimpled cylinder is observed by means
of wind tunnel experimentations and CFD simulations.
Fig. 8.3. Comparison of CD between smooth and rough cylinders.7
8.3. Experimental Method

The smooth and dimpled cylinders to be used in this study were
250 mm in length, with diameter of 50 mm. The dimples were arranged
in 29 rows and each dimple was 5 mm apart in each row. The rows were
alternated by 15 degrees from the previous row. The dimple array will
give a dimple density of 1.392 dimples per mm of the length of cylinder.
The dimple diameter was set to 6 mm. The depth of the dimples was
set to 1 mm, thus the dimple depth-to-dimple diameter ratio was 0.17,
which was between 0.15 to 0.35.8 These dimples were then machined
through use of a computer numerical controlled milling machine.
The cylinders were then further modified to incorporate the use of
a pressure transducer to record the pressure readings on the surface
of the cylinders. These pressure readings were converted to their own
dimensionless parameter, pressure coefficient CP . The modification of
the cylinders involved milling and drilling operations. Material 30 mm
in diameter and 80 mm in depth was removed to allow the installation of
pressure tappings from both sides of the cylinders. Figure 8.4 illustrates
the changes.
Next, the cylinders were drilled with holes as pressure tappings
on the surfaces. There were two pressure tappings on the smooth
Fig. 8.4. Original and modified cylinder.
Fig. 8.5. Pressure tappings on smooth cylinder.
cylinder, one on each side, and they were aligned along the length
of the cylinder, as shown in Fig. 8.5. On the other hand, there were
four pressure tappings on the dimpled cylinder, two on each side (one
will be inside a dimple, another outside a dimple) and aligned along
the length, as shown in Fig. 8.6. The extra two tappings as compared
to the smooth cylinder were used to investigate if there are pressure
differences inside a dimple and outside a dimple. Due to this drilling
operation, imperfections such as scratches and chafing were left on the
surfaces, affecting the relative surface roughness, which may lead to
higher drag force acting on the cylinder.
The tappings were connected to the thin-walled stainless steel pipe
of 20 mm in length and to the pressure transducer via flexible tubes as
shown in Fig. 8.7.
The pressure transducer which was used to measure the pressure
readings around the cylinder’s surface was with a digital indicator as
Fig. 8.6. Pressure tappings on dimpled cylinder.
Fig. 8.7. Thin-walled pipe and flexible tube.
shown in Fig. 8.8. The model is Flotech Setra 230 with a pressure
range of ±17,000 Pa.
The wind tunnel experiment was carried out in the Nanyang
Technological University (NTU) closed loop wind tunnel facility. It
is the AF6407 Wind Tunnel with a test section 720 mm height ×
2000 mm length × 780 mm width. The wind tunnel experimentation
set up is shown in Fig. 8.9. The wind tunnel is able to provide an air
Fig. 8.8. Pressure transducer and digital indicator.
Fig. 8.9. Wind tunnel experimentation set up.
velocity range from 3 ms−1 to 90 ms−1 , with a turbulence intensity of

approximately 0.1 %.
The cylinder was mounted on the force sensor using an additional
attachment piece. A level gauge was used to make sure that the cylinders
were leveled horizontally, and a force sensor was used to record the drag
force. This arrangement is shown in Fig. 8.10.
Both the experiments and simulations were originally planned for
Re of 20,000 to 160,000 to capture the critical region of extreme drag
reduction shown in Fig. 8.2. However, results for Re below 60,000
were ignored due to the limitations of the force sensor in measuring
smaller forces. Also, due to limitation of the wind speed provided by
the wind tunnel, cases with Re of 160,000 were also not conducted.
Fig. 8.10. Mounting of cylinder with the help of level gauge.
Table 8.1. Experimental CD vs Re Results for Both Cylin-

ders.
Re Smooth cylinder Dimpled cylinder % Difference
60,000 0.955 0.896 6.58

80,000 0.951 0.869 9.44
100,000 0.945 0.863 9.50
120,000 0.938 0.855 9.71
140,000 0.934 0.875 6.74
Therefore, the experiments were only carried out with Re range of

60,000 to 140,000 in intervals of 20,000.
For both smooth and dimpled cylinders, the drag forces were
compiled and represented by CD as tabulated in Table 8.1.
From the experimental results, drag reduction was observed with
the dimpled cylinder as compared to its smooth counterpart, as shown
in Fig. 8.11. The drag reduction in this experiment is an average
of 8.39%.
In this experiment, the authors are also interested in observing the
possibility of a critical region where the drag dips drastically. From the
comparison of CD vs Re results of the smooth and dimpled cylinders,
it shows a seemingly lowest CD value at around Re 120,000 before
increasing at Re of 140,000.
The CD results of the dimpled cylinder were compared to a
study done by Bearman and Harvey9 — their critical region is much
Fig. 8.11. Comparison of experimental CD vs Re results for smooth and dimpled

cylinders.
Fig. 8.12. Comparison of CD for dimpled cylinder with Bearman and Harvey.8
more significant and it occurs at around Re 100,000, as shown

in Fig. 8.12.
The comparison made with Bearman and Harvey9 shows that the
current experiment setup is unable to replicate the drop in CD at around
Re of 100,000. This may be due to several factors. First, cylinder
parameters such as surface roughness and dimple density were different.

The authors’ dimple density was lower than that in the experiment
conducted by Bearman and Harvey, which had a dimple density of
1.81 dimples per mm. Second, the experiment setups of both studies are
different, and the authors’ experimental set-up was not isolated enough
due to the presence of the mounting system and other apparatus.
Therefore, it is deduced that the critical region of the cylinder has not
occurred in the study, and it is likely to be delayed to Re > 140,000.
As for the analysis of the pressure readings, they were compiled and
calculated into their respective pressure coefficients, CP . The pressure
readings were taken from 0◦ to 180◦ with an interval of 15◦ . For the
smooth cylinder, two readings were taken for each angle and averaged.
For the dimpled cylinder, two readings were taken from inside the
dimples and two from outside the dimples. The CP is then plotted
against the angle θ (for the case of Re 60,000) as presented in Fig. 8.13.
From the graph, it can be noticed that the curve corresponding
to the smooth cylinder is rather far away from the theoretical curve
calculated using potential flow theory. On the other hand, both inside
and outside dimple curves do not differ much from each other and
have a lower-pressure region between 45◦ and 120◦ . This shows that
Fig. 8.13. Experimental CP against θ for Re of 60,000.

the dimple cylinder is capable of reducing drag when compared with

the smooth cylinder. The graphs for the other Re have shown similar
trends as compared to Re of 60,000 and will be included with the
simulated pressure results later.
8.4. Simulation Method
Three-Dimensional Computational Fluid Dynamics (CFD) simulations

were done using ANSYS Workbench 14.5.7 (FLUENT) and the CD
results were extracted. All the model geometries were done in Design-
Modeler found within the Workbench instead of SolidWorks to prevent
any geometrical errors in the importing process. The size of the fluid
domain takes in the exact dimensions of the wind tunnel test section.
The cylinder was placed at 750 mm (15D; D refers to the diameter
of the cylinder) from the inlet boundary and 1250 mm (25D) from the
outlet boundary, as shown in Fig. 8.14. The inlet was set as velocity-in,
the outlet was set as pressure-out, and the rest as no-slip walls.
As the ANSYS Workbench available in the NTU Computer-Aided
Engineering (CAE) Laboratory is an educational version, the total
number of elements is restricted to 512,000. For the smooth cylinder,
a quadrilateral structured mesh was used and the number of elements
used was 472,740. However, due to the complex geometry of the
dimpled cylinder, a fully structured mesh was not achievable. Hence,
a hybrid mesh was used where tetrahedral elements were used near
the surface of the dimples and quadrilateral structured mesh was used
for the remaining domain. The number of elements was 508,866. The
meshes of both cylinders are included in Fig. 8.15.
There are many different approaches available to solve turbulence
problems computationally and these are normally selected based on the
problem application and computation resources available. For turbu-
lence modelling, there are three main approaches: Direct Numerical
Simulation (DNS), Large Eddy Simulation (LES) and Reynolds Aver-
aged Navier–Stokes Simulation (RANS). Even though the accuracy of
results using DNS and LES increased due to the higher mesh density, for
the practicality of this study, RANS is used. However, the limitation of
the RANS is that this approach solves the time-averaged Navier–Stokes
equations, thus giving the time-averaged velocity contours instead of
Fig. 8.14. Geometry setup of the CFD simulation.
Fig. 8.15. Meshes of smooth (left) and dimpled cylinders.
the instantaneous ones. This may cause the vortex shedding to be less
observable.
Under the RANS turbulence modeling, there are many models to
choose from in FLUENT. In industrial CFD, the two most widely used
models are k-epsilon (k-ε) and k-omega (k-ω) models.
According to Jasmina et al,10 the k-ε turbulence model has proven

to be unsuitable for flows with curved surfaces and with boundary
layer separation. The k-ε model used often in flow simulation has been
shown to poorly represent flow behavior in highly curved regions.11
On the other hand, the k-ω models are better for flows with adverse
pressure gradient and the Shear Stress Transport (SST) turbulence
models showed good agreement for both smooth and dimpled surfaces.
Hence, in this simulation, the SST k-ω model is used.
The Pressure-Based Solver (PBS) and the Pressure-Implicit with
Splitting of Operators (PISO) scheme were selected as the method of
solving. The PISO scheme is used for transient flow problems, which is
applicable in this simulation. The discretization schemes are of second
order as they provide better accuracy than first order ones. However,
when the PISO scheme was selected for the dimpled cylinder simu-
lation, numerous errors were detected and simulation was not able to
proceed smoothly. The errors might be caused by the poor mesh quality
of the dimpled cylinder with partially structured mesh being used.
Due to these errors, the pressure-velocity coupling scheme was
changed to Coupled and simulations were able to continue without
further errors. The FLUENT input parameters applied in the problem
setup were tabulated in Table 8.2.
Appropriate time step sizing is required to achieve the best results
possible without sacrificing valuable computational resources. The
smaller the time step size, the longer the computational time despite
its ability to provide more accurate results. The required time step
size can be calculated by using the Strouhal (St) number, which is
Table 8.2. FLUENT Input Parameters.
Solver Pressure-Based, Transient

Unsteady Formulation Second Order Implicit
Turbulence Model SST k-ω
Discretization Pressure: Standard
Momentum: 2nd Order Upwind
Turbulent Kinetic Energy: 2nd Order Upwind
Turbulent Dissipation Rate: 2nd Order Upwind
Pressure-Velocity Coupling PISO (Smooth cylinder)
Coupled (Dimpled cylinder)
approximately 0.2 for Re 10,000 to 160,000. At least 20 to 25 time steps

are required to capture the vortex shedding and by choosing the fewest
time steps (i.e. 20), computational time can be reduced as compared to
25 time steps.
Each simulation was allowed to have a flow time of 5s to reach
convergence, hence the calculations for the respective time step size
and number of time step can be computed for each Re, and they are
tabulated in Table 8.3.
The simulations were allowed to run for 5s and the CD results were
averaged over the number of time steps. For both smooth and dimpled
cylinders, the averaged drag coefficients, CD are shown in Table 8.4.
From the numerical simulated results, there is observed drag reduc-
tion by the dimpled cylinder as compared to its smooth counterpart,
as presented in Fig. 8.16. The drag reduction is an average of 9.16%,
with greater reductions as Re number decreases.
Table 8.3. Time Step Size Used for Each Re.
Re number Velocity (ms−1 ) Time stepping size (s) No. of time steps
20,000 5.84 0.00214 2,337

40,000 11.7 0.00107 4,673
60,000 17.5 0.000714 7,003
80,000 23.4 0.000534 9,364
100,000 29.2 0.000428 11,683
120,000 35.1 0.000356 14,045
140,000 40.9 0.000306 16,340
160,000 46.7 0.000268 18,657
Table 8.4. Simulated CD Results for Both Cylinders.
Re number Smooth cylinder Dimpled cylinder % Decrease
20,000 0.706 0.562 20.4

40,000 0.659 0.555 15.8
60,000 0.614 0.548 10.7
80,000 0.596 0.549 7.89
100,000 0.584 0.541 7.36
120,000 0.573 0.539 5.93
140,000 0.565 0.549 2.83
160,000 0.557 0.544 2.33
Fig. 8.16. Simulated CD vs Re results with Fluent RANS.
Fig. 8.17. Graph of CP against θ with simulated and experimental results for Re of
60,000.
As for the simulated pressure readings, the CP values were extracted

directly from FLUENT and arranged with an interval of 15◦ . The CP
values were then plotted against the angle θ (in the case of Re 60,000),
as shown in Fig. 8.17, which is similar to Fig. 8.13 shown previously
with the addition of simulated values.
Table 8.5. Measured Pressure Readings for Smooth

Cylinder at Re 140,000.
Pressure Pressure Average

Angle (◦ ) reading 1 (Pa) reading 2 (Pa) reading (Pa)
0 940 940 940

15 730 750 740
30 220 300 260
45 −580 −370 −475
60 −1110 −770 −940
75 −1170 −500 −835
90 −820 −520 −670
105 −830 −530 −680
120 −830 −530 −680
135 −810 −510 −660
150 −840 −530 −685
165 −830 −550 −690
180 −820 −540 −680
From Fig. 8.17, it can be seen that the CP values are close for the
dimpled cylinder (blue, yellow and orange curves) before 60◦ . There is
a slight discrepancy at 60◦ and 75◦ for the smooth cylinder where the
experiment captured a higher CP . There are much larger discrepancies
at angles after 105◦ between the experimental and simulated CP values.
The experimental CP values are much lower than the simulated values.
This is because during the wind tunnel experiment, the recorded
pressure readings were observed to be extremely low from one side
of the cylinder as illustrated in Table 8.5 for the smooth cylinder at Re
140,000.
From Table 8.5, it can be seen that the measured pressure readings
at 75◦ differ by as much as −670 Pa. This resulted in a far less accurate
average pressure reading. This problem arised for the smooth cylinder
for all Re (60,000 to 140,000) from 30◦ onwards, but was not so
obvious for the dimpled cylinder. Initially, the authors suspected that
the extreme low pressure detected (on the right side from the direction
of air flow) may be due to the pressure tappings. Hence, the cylinder
was flipped over and the experiment repeated. However, the extreme
low pressure was again detected at the right side. Thus, it may be other
Table 8.6. Validation of CD Results.
Smooth Dimpled
% %
Re Experiment Simulated Difference Experiment Simulated Difference
60,000 0.955 0.614 35.7 0.896 0.548 38.8

80,000 0.951 0.596 37.3 0.869 0.549 36.8
100,000 0.945 0.584 38.2 0.863 0.541 37.3
120,000 0.938 0.573 38.9 0.855 0.539 37.0
140,000 0.934 0.565 39.5 0.875 0.549 37.3
disturbances within the wind tunnel which affected both pressure and
drag force readings.
The overall trends for the graphs are noticeable for all Re and
the respective graphs were plotted from Figs. 8.A.1 to 8.A.8 in the
Appendix.
Both the current experimental and simulated results were compared
for validation of the accuracy of the simulation. Only comparisons
between Re 60,000 to 140,000 were made as this was the Re range
for the experiment, and the CD results are shown in Table 8.6.
The differences observed in all results were approximately 35 to
40%. These deviations can be attributed to the following reasons. The
high experimental CD results may be due to the interference in the
air flow in the wind tunnel set up due to the pressure transducer and
tubings as well as the scratches made on the cylinders. The simulations
conducted have many limitations such as element count restriction and
selection of the RANS turbulence modeling. The surface roughness of
the material was also not accounted for in the CFD simulations.
8.5. Conclusion
From both the experimental and simulated results, drag reductions
were achieved by the dimpled cylinder as compared to its smooth
counterpart which corresponded to 8.39% and 9.16%, respectively.
The differences between the experimental and simulated results
are significant. It may be the insufficiently fine mesh used due to the
limitation on the number of elements. Also, the experiment set-up is
not isolated enough to prevent the interference of flow. Nonetheless,

the key hypothesis that dimpled cylinders reduce the drag forces on the
cylinders has been shown.
Clearly, more work can be done to further improve this study.
For the experimental work, better experimental setup can help to
produce more accurate and useful results, such as (1) high-precision
manufacturing of cylinder with specific surface roughness, (2) more
precise force and sensors, (3) adding end-plates to cylinder to reduce
end-gap losses and (4) particle image velocimetry of the flow of
cylinder. For the simulation study, increasing the number of elements
and improving mesh quality can also yield more representative results.
Using more computational intensive turbulence modeling such as LES
can also help in capturing and understanding the difference in the
vortex shedding of the cylinder.
Opportunities to expand this study are aplenty. With a reliable and
accurate CFD model, the work can be expanded to optimization of the
sizing, shape and arrangement of dimples on the cylinder. Studies on
the effects of dimpled cylinder on vortex shedding and vortex-induced-
vibration are also of great interest. Low-Re analysis on dimples on micro
tubes may be of interest to the biomedical industry.
While these challenges are addressed, the usage of dimpled cylinders
can be far ranging in many different industry and applications.
References
[1] Pang S, Ng EYK, Chiu WS. (2013) Comparison of Turbulence Models in Near
Wake of Transport Plane C-130H Fuselage. AIAA J 50(3): 847–852. [DOI:
10.2514/1.C032031]
[2] Ng EYK, Pang S, Chiu WS. (2013) Computational Aerodynamics of the C-130
Predicting Airflow Effect on Airdrop. International Journal of Computational
Methods 10(5): 1350025–1350033. [DOI: 10.1142/S0219876213500254]
[3] Luo Y, Liu YF, Zhang D, Ng EYK. (2014) Influence of Mor-
phology for Drag Reduction Effect of Sharkskin Surface, Journal of
Mechanics in Medicine and Biology 14(2): 1430001-1–14300016. [DOI:
10.1142/S0219519414500298]
[4] Luo Y, et al. (2015) Chemical, Mechanical and Hydrodynamic Properties
Research on Composite Drag Reduction Surface Based on Biological Sharkskin
Morphology and Mucus Nano-long Chain. Journal of Mechanics in Medicine
and Biology, accepted.
[5] Luo Y, D Zhang, Y Liu, et al. (2016) Bio/Micro-Rolling Fabrication of

Biological Sharkskin Morphology on Semi-Cured Coating and Drag Force
Experimental Research, Journal of Mechanics in Medicine and Biology, 16(02):
1650016 [15 pages] (DOI:10.1142/S0219519416500160)
[6] Drag of a Sphere. (2010) Retrieved 27 Sept 2013 from http://www.grc.nasa.
gov/WWW/k-12/airplane/dragsphere.html
[7] Anderson J. (2011) Fundamentals of Aerodynamics, 5th ed. McGraw-Hill,
New York.
[8] Yoon CH, Park JM, Park YC. (2010) Dimples Lifting Pipe for Mining Deep-Sea
Mineral Resources. United States Patent.
[9] Bearman P, Harvey J. (1993) Control of Circular Cylinder Flow by the Use of
Dimples. AIAA J 31(10): 1753–1756.
[10] Jasmina B, Zivojin M, Milos M. (2012). Experimental and Numerical Investiga-
tion of Flow around a Sphere with Dimples for Various Flow Regimes. Thermal
Science 16(4): 1013–1026.
[11] Daniel W. (2010) Analysis of Curvature Effects on Boundary Layer Separation
and Turbulence Model Accuracy for Circulation Control Applications.
Appendix
Fig. 8.A.1. Pressure coefficient plot against angle (◦ ) at Re 20,000.





Chapter 9
Fluid Flow in Biomimetics Simulated

Vessel Having a Grooved Surface:
An Investigation of the Effect
of Riblets in Drag Reduction
Guangming Hu
Department of Mechanical and Electrical Engineering

Guangdong Engineering Polytechnic, No. 18,
Yuxing Road, Tian he District, Guangzhou, 510520, China
hgm92122@163.com; 178601711@qq.com
Research in the 1970s has shown that small grooves on surfaces,

which are known as riblets, are inspired by biometric structures that
occur naturally and are capable of reducing drag that resulted from
the fluid flow. These structures form a constraint against Reynolds
stresses when aligned with the flow, which is related to the growth
of eddies in low speed region. Such a region usually occurs along the
internal surface of a duct. Through many years of research, various
forms of riblets have been created to optimize drag reduction in fluid
flows. We hypothesized that a set of riblets that are spaced reasonably
far enough and of sufficiently small size, as well as having the least
cross sectional area can be optimal in drag reduction. Therefore, in
our experiments, we aim to research the most effective type of riblet
geometry (i.e. L, U, and V). To achieve this, we carried out series of
experiments to test the effectiveness of riblets in reducing air flow in
pipes using the wind tunnel. Besides working on the type of riblets,
we also attempt to discover the optimal spacing between riblets for
295
highest drag reduction. Our results show that 3mm V-shaped riblets
spaced 13 mm apart gives the optimal design. A nozzle pertaining to
a unique curvature was constructed, and air flow through it resulted
in different speeds at different region of the nozzle. In conclusion, the
findings in this experimental study were based on the nozzle design
such that the best riblets with the optimal spacing between them were
used. Our research proved to be a success when we demonstrate how
riblets when applied on the nozzle can reduce drag significantly.
9.1. Introduction
In nature, we can find riblet structures on the body surfaces of animals

such as fish, birds, snakes, etc. For maximal drag reduction, research on
optimization of riblets is often carried out, and biomimetics scientists
have explored the drag-reduction level relating to the shape and size
of riblets using experiments and simulations based on examination of
natural skin structures of animals (as shown by the photorealistic images
in Fig. 9.1).
In particular, the drag-reduction effect of the riblets based on bionic
study of shark skin (such as the one illustrated in Fig. 9.1(a)) has
attracted extensive attention worldwide. The riblets-like bionic non-
smooth surface provides a new kind of low-drag surface microscopic
topological structure for the structural design of ships and varieties of
oceanic vehicles.1 The feathers of birds are composed of the pinnule,
barbule, and pinna rachis. Flight and tail feathers possess three typical
Fig. 9.1. Microstructure of animals’ body surfaces based on microstructures pertain-

ing to (a) shark skin, and (b) flight feather. These images were adapted from Refs. 1
and 2, respectively.
Fluid Flow in Biomimetics Simulated Vessel Having a Grooved Surface 297
barbules: nodular, hooked, and nonhooked. Hooked barbules and

nonhooked barbules engage with each other and form a massy surface
to reduce flight drag. Characteristics of the drag-reducing surface of a
feather are illustrated in Fig. 9.1(b). A similar microstructure can be
seen in nearly every flying bird. The grooved structure of the feature
may be chosen as a biomimetics sample for studying the effect of feather
riblet structure2 on flight drag reduction.
The riblet technology can be widely used and has huge potential
for growth, depending on its satisfactory reliability and applicability.
It has been put into trial applications in various fields,3 namely, the
surfaces of watercraft, aircraft, inner wall of pipeline, athletic swimsuit,
etc. Additionally, some of the correlational research has forecasted and
reported the drag-reduction rate of these riblets. Reports claimed that
bionic riblets with a specific shape could have a drag-reduction rate
over 30%.4,5 The Coustols’ experiment got a drag reduction rate of
10∼15%.6 Walsh gained a drag-reduction rate up to 8%.7
Experimental studies in the 1970s showed that small grooves
(riblets) aligned with the flow had the property of modifying the near-
wall structure of the boundary layer. In particular, the riblets proved
to work as a constraint to the production of the Reynolds stresses
associated with the growth and eruption of the eddies in the low-
speed regions of the boundary layers. Riblets have been employed in
microsurface design such that the geometrical variations can change
the near-wall structure of the flow and help to reduce drag.8,9 This
study explores the relationship between the dimensional configuration
of riblets and reduced drag flow. The optimal dimension and shape of
the riblets in particular to the least drag is applied onto a nozzle. The
variation of the riblet dimension is based on the velocity at every cross-
section of the structure perpendicular to the flow. The main objective is
to optimize the flow speed of the air at the nozzle output by minimizing
the drag as the air flows through the nozzle inlet surface.
Li et al. have investigated the time series of velocity vector field and
statistics in the turbulent boundary layer (TBL) over riblet surfaces
utilizing time-resolved particle image velocimetry (TRPIV). They
concluded that the streamwise turbulent intensity and the Reynolds
shear stress of the TBL over riblet surfaces are both smaller than
the ones over smooth surfaces at the same wall-normal position,
indicating that riblet surfaces weaken the flow turbulence reducing the
momentum exchange and energy loss in the flow.10
Riblets are also effective in the presence of developed turbulent air
flow.11,12 The skin-friction coefficient of turbulent boundary layer flow
over a smooth wall with transverse square grooves is investigated for
four grooved-wall cases. The four grooved-wall configurations pertain
to a type of rough wall, which is characterized by regularly spaced two-
dimensional square cavities (grooves) placed normal to the flow. In this
study, the use of riblets to reduce air drag on a flat plane is investigated.
A riblet is effectively an indented groove line that is aligned parallel to
the flow of air across a surface. Different dimensional configurations
of the riblets are experimented with on a flat plane and tested in a
wind tunnel. Then, four such planes are constructed into a rectangular
duct. The difference in velocity of air flow at the inlet and outlet duct
is measured for variation in air speed and dimension of the grooves.
Next, the rectangular planes are replaced by trapezoidal planes and
inclined at an angle to form a rectangular nozzle. Based on the velocity
profile along the longitudinal flow along the nozzle, the indentation
of the planes will be varied to give the least drag along the surface.
Finally, we consider complex geometrical nozzle structure based on
trigonometrical functions to improve upon the flow.
As is well known, riblets dampen turbulence intensities and
Reynolds shear stress near the wall for the drag-reducing configuration.
Surface drag is the friction between the fluid and the surface it is
moving along. Basically, our research is concerned with applying riblets
in nozzles to reduce the drag of the fluid passing through them.
Specifically, the riblets lessen surface friction between the fluid passing
through and the walls of the nozzle due to the presence of small grooves
on the surface of the nozzles. Research in the 1970s has shown that
riblets are capable of reducing drag on fluid flow. These riblets when
aligned with the flow, work as a constraint to stresses caused by the
fluid flow, caused by the slowing down of the fluids from the friction
at the sides of the pipe they are passing through, otherwise known
as Reynolds stresses. Through many years of research, many forms of
riblets have been created to optimize drag reduction in fluid flows. Our
research attempts to mimic this effect in nozzles, which are used to
speed up fluid flow in rockets, in order to maximize this desired effect.
Scientific concepts based on fluid dynamics are employed to provide

an explanation for how drag is reduced by altered flow pattern and
Reynolds stress caused by riblets.
At a given Reynolds number, drag reduction is found to increase
with increasing feature size and spacing, as in laminar flows. The
main purpose of this research is to develop new technologies that
can aid efficient conduct of air flow through ducts and pipes. In this
case, we aim to promote faster flowing fluid, which can help increase
the efficiency of fluid-related mechanisms. In a way, this also raises
awareness in the scientific community as well as in the public of the
existence of riblets, and of their usefulness. This also establishes the
wrong concept that fluids travel faster only along an extremely smooth
surface, a direct opposite of the rough surfaces produced by the riblets
to reduce drag. Also we want to test out a new type of C-shaped riblets
which have never been tried before.
Our hypothesis is that the riblets have a potential drag reduction
function when applied in nozzles, and it seems logical to think that
a small sized triangular riblet configuration can produce a larger
surface area for more effect. Additionally, the riblets should be spaced
reasonably far enough to prevent more drag from being created by
excessive turbulence in the fluid flow. Ultimately, the riblets should
provide the requisite surface to minimize Reynolds stresses and drag
without drastically reducing the amount of cross sectional area.
Fluid dynamics is basically the study of the physics involving
interactions with the surface of the moving gas or liquid, based on the
principle of energy conversation. The understanding of this concept can
help us to better comprehend our research into how riblets can work
to produce our hypothesized results. Surface drag is simply, in layman
terms, the friction between the fluid and the surface it is moving along.
Our research aims to reduce this friction.
The link between these two flows is that laminar flows basically
transform into turbulent flows at high velocities. The nature of laminar
flow in pipes is that the layers of fluids flow parallel to each other
causing the flow in the middle to move faster, while fluid layers close
to the surface flow slower due to the surface drag. Turbulent flows
are basically when the fluid molecules flow at random with vigorous
collisions; basically, the flows are all roughly at the same speed in a pipe
due to their random nature throughout the fluid flow. This effect is
easily observed when one simply lets water flow at high and low speeds
from household taps.
In general, riblets are grooves on the surface that force turbulence
to occur so that it acts as a cushion for the above laminar flow. These
riblets when aligned with the flow work as a constraint to stresses of
the fluid flow (otherwise known as Reynolds stresses) caused by the
slowing down of the fluids due to friction at the sides of the pipe it
is passing through, Although the flow is seemingly slowed down, on
average the liquids flow at a higher average velocity.
Chamorro et al. investigated the drag reduction on a wind turbine
airfoil partially or fully covered with riblets in a wind tunnel based on
mean drag that was measured via wake survey (momentum deficit) and
lift via a sensitive force balance. They discovered that the most efficient
riblet configuration for a completely covered airfoil was the V-groove
shape of 100 µm height.13 Anderson et al. studied the combined effect
of a polymer with 3M riblets and the comparison with sand roughened
and commercially rough surfaces was made, which showed that the
absolute drag reduction of the 3M riblets appears to be independent
of the polymer presence.14
Viscosity is one property of the flow and a main factor that affects
the drag. Viscosity of flows does have an impact on the efficiency of
riblets; this is important, as we would need to use the wind tunnel
setup for our upcoming experiments to test for the effect of riblets on
flows. It gives us a simple understanding of how rough surfaces alter
turbulent flows and allows us to conceptualize how such rough surfaces
are akin to riblets on the surface of pipes. In an online article, NASA
equipped a racing yacht with a “riblet” skin that helped the craft slide
through the sea more smoothly.15 This interesting article has helped in
the understanding of riblets as an essential item in aerodynamics.
Nozzles are basically used to accelerate or decelerate fluids. There
are two main types of nozzles: divergent and convergent; the former
has a big opening and a small exit, and the latter a big exit and small
opening (seen commonly in the tail of the rockets where the air is blown
out from combustion). The fluids speed up at the divergent end due to
the law that pressure is force over surface area. With a smaller surface
area, should the pressure remain constant, the force naturally increases
with increasing speed. We will also be using nozzles to test our riblet
design.
The various possible shape designs of riblets are triangular and
semicircular riblets. There are basically three types of riblets, L, U and
V. In this experimental study, we have used V-shaped riblets as they
have been proven to be more effective in most cases, as well a new type
of C-shaped riblet to research alternative riblet shapes.
9.2. Methodology
9.2.1. Experimentation measurements and rigs

This is basically an experimental method that we have adopted to
make our research much more straightforward to handle. We used
cylindrical ducts, commonly known as household PVC pipes, that are
three inches in diameter, and we tested out the riblets of all sizes and
shapes at different spacing with different wind speeds of the wind tunnel
in the cylindrical duct to find which is the optimum configuration
of these factors to reduce drag the most, before applying it to our
experimental model, the nozzle, which would be further discussed in
Section 6.15, therefore eliminating all other configurations of riblets in
the experimental process.
Liu et al. used pipes lined with a film of grooved equilateral triangles
of base 0.11 mm to show that the maximum reduction occurs when the
height of the riblets is 11–16 wall units, and correlates well with the
Taylor microscale of the fluctuating velocity gradient.16 Microscopic
riblets that are aligned parallel to the flow are an effective means to
reduce the turbulent skin friction by up to 10% in comparison with a
smooth surface, and may be investigated experimentally by direct wall
shear stress measurements.17,18
Throughout our entire experimental process, we have used several
pieces of apparatus to conduct our experiments. For the first stage of our
experimentation, we used a cylindrical duct and pasted Perspex strips
inside to simulate a real pipe (Fig. 9.2). A digital manometer was used to
measure the pressure and air speed at the entrance of the pipe and at the
end of the pipe. The second stage of our experimentation process was
the use of riblets in a self-constructed nozzle. We worked with different
Fig. 9.2. The production process of the pipe with riblets on its internal surface. We
show the follow step procedures: (a) Design the cross section of triangular and circular
riblets; (b) paste the riblets together; (c) insert riblets into the cylindrical duct; and
(d) secure the assembled pipe.
riblet shape and spacing to determine the optimal configuration for

minimizing the drag.
9.2.1.1. Digital manometer

A digital manometer was used to measure the air pressure and speed
of air coming in and out of the cylindrical duct, to measure the drag
reduction, to the accuracy of 1 Pa for measuring pressure and 0.1 ms-1
for measuring wind speed.
9.2.1.2. Pitot tube

A pitot tube was connected to the digital manometer, and its end placed
into the wind tunnel to allow the wind to enter the manometer and
have its pressure and speed measured.
9.2.1.3. Constructing the riblets

Koury et al. showed that a pipe that is lined with 0.15 mm V-groove
riblets, with flow at Reynolds numbers from 300 to 150,000, can
be used for drag reduction.19,20 Riblet design is a useful means of
reducing viscous drag on a body and has potential aerodynamic and
hydrodynamic applications.21 Perspex strips were procured and cut up
into 40 cm strips to fit into our pipe. There were two main types of
riblets used, triangular and circular. We used 3 mm and 5 mm versions
of each. To make our research more effective, we also tested out the
optimal spacing between each riblet. The different types of spacing used
were 3, 5, 7, 9, 11 and 13 mm for each of the four variations of riblets.
A 500 mm long pipe was implemented. We shortened the pipe
to 400 mm eventually, as it would be more economical with regards
to the number of riblets that had to be purchased. This cylindrical
duct mimicked a pipe, and by pasting riblets along its circumference,
we could thus simulate a real life pipe with the addition of riblets. It
was roughly one-third the size of the wind tunnel. This is because any
larger size than this might have caused excessive turbulence against the
cylindrical ducts from blockage, which affects the overall wind speed
and pressure and thus affects the results.22−24
9.2.2. Nozzle
A nozzle that is constructed based on trigonometrical function is also
implemented to verify the test results by the straight pipe with riblets
at its inner lining. Here, the cross-sectional area of the flow, A0, will be
proportional to the height of the profile, H. Since velocity is inversely
proportional to A0, it is also inversely proportional to H. Therefore for
different H values, specific configurations of riblets will apply.
For a defined nozzle shape profile (Fig. 9.3), it is necessary to
produce a shell design of specific thickness along the profile such that
groove lines can be engraved onto the shell wall. This will enhance the
flow laminar properties by reducing turbulence as much as possible.
The equation describing the geometry of the wall for a semi-circular
nozzle constructed using trigonometrical functions is:
H h s π
=1+ sin (z) , for α ≤ z ≤ β, s ≥ 0, (9.1)
H0 H0 2l
Fig. 9.3. The design of the nozzle was based on (a) schematic diagram of nozzle
based on curvatures generated from different shape parameters range; and (b) assembly
and checking the sectional view of nozzle.
where s is a parameter determining the shape of the nozzle, H0 is the

radius of the semi-circular duct or height of the channel, h and l are the
height and length of the nozzle respectively. The flow of fluid is axially
along the z-axis. By detecting the air velocity flow along the internal
nozzle surface, it is possible to implement groove lines of variable depth
such that generation of vortices can be reduced optimally.
At the widest end, the nozzle was 8 by 8 inches long, while at the
smaller end it was 4 by 4 inches. The y = 5 tan x graph was used as it
gives a uniform curvature to the nozzle and nicely showed a curve that
had its endpoints horizontally two units apart; this is useful as knowing
that the nozzle was 4 inches at the convergent end and 8 inches at
the divergent end, there would be 2 inches of horizontal difference
in lengths at both sides. Taking two units on the graph as 2 inches,
we found the coordinates of the different points along the graph and
plotted these values in inches on the plastic board, and plotted a mirror
image 4 inches apart at the convergent end and cut out the shape to
get the board. The most effective drag-reducing riblets are then pasted
on all sides of the nozzle for testing.
9.2.3. Wind Tunnel

Our project required the use of a wind tunnel at our wind tunnel facility
(Fig. 9.4). This wind tunnel can generate different wind speeds. The fan
Fig. 9.4. Experimentation performed after setting up (1) Wind creation equipment
(2) Desk (3) Brackets (4) Plastic cover (5) Air inlet pipe (6) Nozzle (7) Air outlet pipe
(8) Lock rings.
blade rotation in the wind tunnel is activated by the motor frequency

control. This controls the suction speed of the air through the nozzle
that is placed in the wind tunnel.
9.2.4. Verification experimentation

The verification experimental procedure is meant to verify whether
the effect of drag reduction achieved with the optimum configuration
of riblets in the cylindrical duct is mirrored when applied to the
experimental nozzle model. This gives a convincing and objective
conclusion as to whether the application of riblets in nozzles is possible
and effective. The verification serves to verify the fact that the riblet
configuration in optimum drag reduction in the cylindrical ducts also
has the same effect when used in a nozzle.
• Control Experiment
The control experiment was to find the drag caused by the nozzle when
placed in a convergent way and divergent way without riblets and was
carried out by just running the experimental procedure without riblets
in the nozzle.
• Actual Experiment
The 3 mm triangular 13 mm spacing riblets were placed in the nozzle
parallel to the wind, at wind speeds controlled by the fan blade
rotation frequencies of 10 Hz, 15 Hz, 20 Hz, 25 Hz, 30 Hz and 35 Hz,
and tested when the nozzle was orientated both the convergent and
divergent way.
• Data Recording
The data recorded were mainly the pressure and speed of exiting wind
with the digital manometer.
9.2.5. Elimination experiment

The elimination experiment basically involved recording of drag reduc-
tion in cylindrical ducts or PVC pipes when a variety of riblets were
attached on its interior surfaces in the direction parallel to the wind of
the wind tunnel (Fig. 9.4). These pipes were placed in the wind tunnel
at different wind speeds. The manometer was used to find the pressure
of the wind in the wind tunnel at the entrance point and exit point of
the pipe. The difference in force was then calculated to find the drag
force. The riblet configuration that cause the most drag reduction was
then be used in the verification experiment. This saves time and money
as the riblets would not be need to be cut in a shape that fits the riblets.
Instead they are placed uniformly in a pipe; wind passing through will
also be more stable and easy to record, and also only the most effective
riblets would then be needed to be purchased in abundance to fill up
the larger nozzle.
• Control Experiment
The control experiment was to find the drag caused by the pipe in the
first place without riblets and was done by just running the experimental
procedure without riblets in the pipe.
• Actual Experiment
Riblets of sizes 5 mm and 8 mm, shapes circular and triangular, spacing

between the riblets of 0 mm, 5 mm, 7 mm, 9 mm, 11 mm and 13 mm,
were attached to the surface of the pipe and run through the wind
tunnel at fan blade rotation frequencies of 10 Hz, 15 Hz, 20 Hz, 25 Hz,
30 Hz, and 35 Hz.
• Data Recording
The data recorded were mainly the pressure and speed of exiting wind
with the digital manometer.
9.2.6. Preliminary calculations

The ultimate aim of the experiments was of course to find out the drag
reduction of the various experimental set-ups. However, these values
were not obtained immediately and required certain calculations and
formulas which are stated below.
9.2.6.1. Cross-sectional area of wind tunnel

The wind tunnel is 13 by 13 inches, and 1 inch is equal to 0.0254 m.
Therefore to calculate the cross-sectional area of the wind tunnel,
0.0254 must by multiplied by 13, and the product squared; which
is equal to 0.10903204 m2 .
9.2.6.2. Cross-sectional area of cylindrical duct

The cross-sectional area of a cylindrical duct is a circle, which has an
area of πr2 where r is the radius. The internal diameter of the pipe is
7.62 cm when measured with a vernier caliper, and thus its radius is
3.81 cm, which is 0.0381 m. Therefore, the cross-sectional area of the
cylindrical duct is equal to 0.004560367 m2 .
9.2.6.3. Cross-sectional area of riblets

Riblets in this experiment come in either triangular shapes or circular
shapes — the triangular shape to simulate the V-shaped riblets and
circular to test out the effectiveness of a new type of riblet, the C-
shaped ones. In the case of triangular riblets, the 3 mm types have a
base length of 3 mm or 0.003 m and a height of 4 mm or 0.004 m and
thus a cross sectional area of 1/2 × 0.004 × 0.003, which equates to
0.000006 m2 ; the 5 mm types have a base length of 5 mm or 0.005 m
and a height of 6 mm or 1/150m and thus an cross sectional area of
1/2 × 1/150 × 0.005, which equates to 0.0000167 m2 . In the case of
circular riblets, the 3 mm types have a diameter of 3 mm or 0.003 m
and thus a radius of 1.5 mm or 0.0015 m and thus a cross-sectional
area of π(0.0015)2 , which equates to 0.0000071 m2 ; the 5 mm types
have a diameter of 5 mm or 0.005 m and thus a radius of 2.5 mm or
0.0025 m and thus a cross-sectional area of π(0.0025)2 , which equates
Table 9.1. Riblet Configuration.
Riblet configuration
Cross-sectional
Size (mm) Shape Length*height/ diameter (mm) Area (m2 )
3 Triangular 3*4 0.0000060

5 Triangular 5*62 /3 0.0000167
3 Circular 3 0.0000071
5 Circular 5 0.0000196
to 0.0000196 m2 . (Table 9.1, Riblet Configuration). Depending on the

number of riblets that can be fitted into the cylindrical duct or nozzle
with the given spacing in between, the total cross-sectional area of the
individual riblets is the cross-sectional area for one riblet multiplied by
its number.
9.2.6.4. Entrance and exit pressure at cylindrical duct

Pressure is measured in pascals and defined by force over area. There-
fore, naturally force is equal to pressure multiplied by area. This force
is measured in newtons. The entrance force is obtained by multiplying
the entrance pressure to the cross-sectional area of the wind tunnel, and
the exit force by multiplying the exit pressure to the cross-sectional area
of the cylindrical duct subtracted that of the riblets.
9.2.6.5. Frictionless and actual exit force of nozzle

As mentioned above, pressure is measured in pascals and defined by
force over area. Thus, it naturally follows that assuming that there is no
drag, the force would be constant, and as cross-sectional area decreases,
pressure would increase proportionately to it. The convergent side of
the nozzle is 4 by 4 inches, while the divergent side of the nozzle is
8 by 8 inches; by calculation we would know that 8 by 8 inches is
four times the area of 4 by 4 inches, and thus the divergent end of
the nozzle would have one quarter of the pressure of the convergent
end. The wind from the wind tunnel passes into the nozzle at the
same pressure as the entrance pressure, and if it passes from divergent
to convergent, the pressure would multiply by four, theoretically; if
it passes from convergent to divergent, the pressure would divide by
four, theoretically. This value obtained after multiplication and division

of the entrance pressure is the frictionless exit force. However, we know
that because of friction, this value can never be obtained. Therefore, to
obtain the actual exit force, we measure the pressure of the wind at the
exit point of the nozzle and multiply it by the surface area of the exit
end of the nozzle.
9.2.6.6. Drag force

Drag force through the cylindrical duct is found by subtracting the exit
force from the entrance force. Drag force through the nozzle is found
by subtracting the actual exit force from the expected exit force.
9.2.6.7. Drag reduction and reduction percentage

Drag reduction is found by subtracting the drag force through the
cylindrical duct with riblets from the drag force through the cylindrical
duct without riblets. To obtain the percentage of drag reduction, we
divide the drag reduction by the drag force through the cylindrical duct
without riblets and multiplied it by 100%.
9.2.6.8. Graphical data analysis

The data is first recorded in tabular form during the experimental
period for easy recording; all the data is then compiled into a unified
table where calculations involving drag reduction are carried out. These
values reflected in the table also allow easy reference when they are to
be used to plot graphs in a later stage. The graphical representation of
the data is meant to allow easier comprehension and interpretation of
the results obtained from experimentation. It shows clear trends and
patterns to aid the analysis. We have used MATLAB, which consists of
a 3D-graph plotting tool, to represent the data in the form of a 3D-
graph. This also allowed easier comparison of the various graphs unique
to the riblet configurations.
9.2.6.9. Off-design performance considerations

Some considerations will need to be taken into account when designing
the experimental procedures. Flow alignment, surface quality and
pressure gradients are the parameters that affect the performance of

riblets as explained below:
(1) Angle of flow: The effect of deterioration of riblet performance

is negligible for flow misalignment up to 15 degrees from the
horizontal flow. For angles above 40 degrees, the riblets become
ineffective. Although the surface of the nozzle varies in inclination
with respect to the axial flow through it, the expansion of the air
from the throat enables a flow that will be parallel to the surface.
Therefore it is still possible to install riblets to streamline the flow
and reduce vortices.
(2) Surface contamination: Contamination on the flow surface that is
due to deposition of dust, combustion particulate or atmospheric
aggression may render the riblet effect to be void.
(3) Pressure gradients: Pressure gradients have a minor effect of only
1% to 2% on the total skin friction drag reduction.25,26
9.3. Results and Discussion
This section explains the experimental procedures that are to be carried

out for the study of riblet effect on the flow over a surface. The results
are ultimately applied to the design of an effective nozzle.
In this study, we perform laminar flow analysis of a smooth surface
duct of height 0.5 m and length 3 m that is implemented with L-shaped
riblets. The parameters involved in drag reduction in this study are
riblet height h, width w, width of spacing d, speed of air flow and
angle of inclination. The performance indicator is the difference in the
inlet and outlet air speed labelled δ. This is a measure of how much air
momentum is lost due to the frictional drag inside a duct.27
The depth and width of groove lines on a smooth surface positioned
parallel to the air flow is varied in steps of 2 mm respectively for 10
measurements with each air flow speed increment of 20 mh-1 starting
from 10 mh-1. The spacing between adjacent grooves is standardized as
5 mm. The response surface curves corresponding to the 10 different
inlet speeds for the variation of δ with respect to the [h, w] grid are
presented below.
In a similar fashion, the width and the spacing between adjacent

groove lines are now varied in steps of 2 and 3 mm respectively for a
fixed depth given 10 different air inlet flow speed. The response surface
curves corresponding to the 10 different inlet speeds for the variation
of δ with respect to the [w, d] grid are presented below.
9.3.1. Graphical and tabular data analysis

9.3.1.1. 3 mm circular riblets
Here, 3 mm circular riblets show that there is no real optimum
frequency at which they perform. However, their peak performance is
found at the middle range spacing, as the smaller spacing might narrow
the pipe too much and cause too much turbulence, clogging the pipe
and the larger spacing to result in too little turbulence generated in
order to cushion the flow. Overall, drag is created as shown by the
data points on the graph that lie in the negative region (Appendix A:
Table 9.2, 3 mm Circular Riblets Tabular Data).
9.3.1.2. 3 mm triangular riblets

Here, 3 mm triangular riblets show that there is an optimum frequency
at lower frequencies at which they perform, and the peak performance
is found in the largest range spacing. And overall, they reduce drag as
more than half of the data points on the graph lie in the positive region.
Here we see that the riblets are completely useless with small spacing
areas and high frequency areas as they probably cause too much of
turbulence and clog the pipe (Appendix A: Table 9.3, 3 mm Triangular
Riblets Tabular DataRiblet Configuration).
9.3.1.3. 5 mm circular riblets

Here, 5 mm circular riblets show that there is no real optimum
frequency at which they perform, and peak performance is found in
the largest range spacing. And overall, this configuration increases drag
as the data points on the graph lie in the negative region. Again, this can
show how smaller spacing areas cause excessive turbulence and constrict
the pipe flow (Appendix A: Table 9.4, 5mm Circular Riblets Tabular
Data).
9.3.1.4. 5 mm triangular riblets

Here, 5 mm triangular riblets show that there is no real optimum
frequency at which they perform, and peak performance is found in
no distinct areas. And overall, the configuration increases drag as the
data points on the graph lie in the negative region. Again the same trend
is shown at smaller spacing areas, with lower drag reduction (Appendix
A: Table 9.5, 5 mm Triangular Riblets Tabular Data).
9.3.1.5. 3 mm circular riblets versus 3 mm triangular

riblets
Here, we see that triangular riblets are more effective than circular
riblets in almost all areas and are far more efficient at the large spacing
low frequency areas. Here, with regard to riblet shape, the triangular
design far exceeds the circular one in terms of efficiency.
9.3.1.6. 5 mm circular riblets versus 5 mm triangular

riblets
Here, we see that triangular riblets are more effective than circular
riblets. But yet again, we realise that the 5 mm riblets give an even
reading throughout the graph and the data points lie in the negative
zone. This suggests their effectiveness in all areas.
9.3.1.7. 3 mm circular riblets versus 5 mm circular riblets

Here, we see that 3 mm circular riblets are more effective than 5 mm
circular riblets, and for the factor of size 3 mm far exceeds 5 mm.
9.3.1.8. 3 mm triangular riblets versus 5 mm triangular

riblets
Here, we see that again, the 3 mm is the better choice in the size.
As shown on the graph, the drag-reduction percentages for 3 mm
triangular riblets are much above those of the 5 mm triangular riblets.
9.3.1.9. Overall analysis

From the data (Fig. 9.5), we can see that the 3 mm triangular riblets
with 13 mm spacing are the most effective. The C-shaped riblets or
circular riblets that we have advocated have also shown to be promising,
where the 3 mm circular with 7 mm spacing riblets are the second most
effective. The third most effective is in fact not using any riblets at
all, thus showing that all other configuration of riblets that we have
tested in this experiment are incompetent in terms of drag reduction
(Appendix A: Table 9.6, Different Shape Riblets Tabular Analysis).
9.3.2. Verification experiment graphical

and tabular data analysis
9.3.2.1. Convergent nozzle
The results are rather promising as the graph lies completely in the
positive region, meaning that riblets do help in drag reduction in
convergent nozzles. A puzzling thing is that the trends have all been
reversed, whereby higher frequencies bring much more drag reduction
than lower frequencies, as compared to previously in the elimination
experiments. Also, there is a sudden low region at frequency 20 Hz.
Probably the nozzles with larger area allow the forming of more
turbulence at the surfaces as compared to pipes, which are narrow;
this allows steady flow at high velocities hence the trend, especially
in a convergent nozzle where fluids are accelerated. The low point at
20 Hz is probably the transition point between stable flow and chaotic
turbulence flow which achieves overall acceleration as seen from the
non-stopping and unstable variation of the wind speeds and pressure
while recording, the instability and the low drag reduction.
9.3.2.2. Divergent nozzle

Again, the results are rather promising as the graph lies completely in
the positive region, meaning that riblets do help in drag reduction in
divergent nozzles. A startling fact is that the drag reduction values now
go up to range from 10% to 40%. There is a region of negligence from
fan blade rotation frequency of 10 Hz to 15 Hz as the wind speed and
Fig. 9.5. Different shape riblets graphical analysis based on (a) 3 mm circular riblets;
(b) 3 mm triangular riblets; (c) 5 mm circular riblets; (d) 5 mm triangular riblets;
(e) 3 mm circular riblets versus 3 mm triangular riblets; (f) 5 mm circular riblets versus
5 mm triangular riblets; (g) 3 mm circular riblets versus 5 mm circular riblets; and
(h) 3 mm triangular riblets versus 5 mm triangular riblets.
Fig. 9.6. Different shape nozzle graphical analysis based on (a) divergent nozzle
with 13 mm spacing, 3 mm triangular riblets; and (b) convergent Nozzle with 13 mm
spacing 3 mm triangular riblets.
pressure are far too low to be computed. The most effective region has
been brought forward to 20 Hz and there is a low region at 25 Hz
(Fig. 9.6). The reduction is also quite surprisingly at high frequency
ends. All this is probably due to the fact that the fluids decelerate so
much in the divergent nozzle, and a single turbulence caused by the
riblets accelerates the fluids so much more than its original without
riblets. Non-stopping and unstable variation of the wind speeds and
pressure while recording is also experienced especially at high velocities.
The low point at 25Hz is probably also a transition point between two
stages of fluid flow.
A few percentage points of drag reduction is actually a significant
amount when nozzles are applied in industries where great amounts
of fluid flow are concerned. As such, every quantity of efficiency saves
cost and is attractive. Furthermore, at high or even supersonic speeds,
these forces are mostly amplified.
In the nozzles of rockets, acceleration of the combusting jet air
that is released can propel the rocket upwards. All sorts of spray, such
as fire extinguisher nozzles and sprinkler nozzles, need to be designed
efficiently. In times of emergency, the increase in efficiency of fluid
flows makes a big difference in saving lives. The question is if our
nozzle were made with different dimensions, such as the size of a rocket
nozzle, would the same layout of riblets still be the optimum layout?
Not exactly, very large nozzles like rocket nozzles would allow more
turbulence and fluids in a rocket nozzle move at supersonic velocities.

Thus, it would probably work better with a slightly different range of
riblets. However, the concepts applied and the trends observed in this
study remain the same.
At certain frequencies of fluid flow, riblets do not seem to be as
effective. At lower air speeds, there seemed to be not enough turbulence
being produced to form an air cushion. Yet, for high speeds, excessive
turbulence is produced and results in clogging.
9.4. Conclusion
Riblets are groove lines or indentations that are protruded into a
structure such as an aerofoil or duct walls such that flow over it can
be made as laminar (Blasius) as possible. The reduction of turbulence
is enhanced by channeling air through the grooves and allowing the air
region above the flat face to stick more closely to the surface instead
of diverting from it. We investigate the optimum ratio for fin-height
versus riblet spacing, physical dimensions of the riblet, along with the
optimum shape using the (L, U, or V) riblet design configuration.
In this study, the main objective is to understand how riblets will
reduce the turbulence across the flow over a flat surface and how
to improve the technique such that the velocity of moving air is
optimized. We devised different riblet configurations in particular to
their dimensional specification on the inner surface of a rectangular
duct and vary the speed of flow through it. The optimization of the
flow velocity is obtained based on trial and error as we vary the width
and depth of the groove lines in incremental steps based on different
inlet air speed. A reference chart is constructed from the experimental
results, which is able to predict the required width and depth of the
grooves based on the air flow properties.
From the experiments we proved that our hypothesis correct for the
3 mm triangular 13 mm spacing riblets. This is by far the least cross-
sectional area reducing, and this thus prevents clogging and blockage
of fluids from travelling through. Being small in size, it could space out
maximally and cause smooth turbulences, instead of abrupt and sharp
ones that would in turn slow down the flow. From the tests in the
pipes and nozzle, the 3 mm triangular 13 mm spacing riblets turn out
to decrease drag all the same effectively. Although values and trends
are roughly different for both elimination experiments and verification
experiments, the fact is that there is indeed effective drag reduction
from the riblets. Also, we have found that C-shaped riblets do have some
extent of effectiveness, and might turn out to work better in fluid flows
in certain scenarios. This experiment brought us to see both nearby
application of our research as well as long-term ones; mainly application
in everyday sprays and household pipes. And in the future, when the
research on riblets becomes more mature and rocket technology has
advanced to a higher level, we may even apply it to rocket nozzle
designs.
References
[1] Xu C, Wang J, Luan S, et al. (2010) Analysis of drag reduction mechanism of
the bionic microscopic riblets surface, 2010 3rd International Conference on
Biomedical Engineering and Informatics, pp. 2394–2398.
[2] Feng B, Chen D, Wang J, Yang X. (2014) Bionic Research on Bird Feather for
Drag Reduction, Advances in Mechanical Engineering. Article ID 849294.
[3] Chao X, Wang J, Liu Y, Jiang L. (2009) A review: the research and application
of drag reduction by biomimetics riblets technique. Journal of Dalian Fisheries
University 24: 235–237.
[4] Wu M. (2009) Application and development of the imitation shark skin
swimsuits, Progress in Textile Science & Technology 2: 90–91.
[5] Mohammadi A, Floryan J. (2013) Pressure losses in grooved channels. Journal
of Fluid Mechanics 725: 23–54.
[6] Coustols E. (1989) Behavior of internal manipulators: ‘Riblet’ models in
subsonic and transonic flows. AIAA 89–0963.
[7] Walsh MJ. (1982) Turbulent boundary layer drag reduction using riblets. AIAA
82–0169.
[8] Walsh MJ. (1983) Riblets as a Viscous Drag Reduction Technique. AIAA
Journal 21(4): 485–486.
[9] Campitelli G, Krastev VK, Huebsch W. (2014) Effects of Moving Surface Riblets
on a Transitional Flow affected by Adverse Pressure Gradient. AIAA 2014–
2651.
[10] Li S, Yang S, Jiang N. (2013) Trpiv measurement of drag-reduction in the
turbulent boundary layer over riblets plate. Chinese Journal of Theoretical and
Applied Mechanics 45(2): 183–192.
[11] Enyutin GV, Lashkov YA, Samoilova NV. (1995) Drag reduction in riblet-lined
pipes. Fluid Dynamics 30(1): 45–48.
[12] Hemmasian MM, Rad M. (2008) Drag-Reduction Characteristics in Air Flow

over the Iced Surface at Re > 104, International Journal of Advanced Design
and Manufacturing Technology 1(2): 35–41.
[13] Chamorro LP. (2013) Drag reduction of large wind turbine blades through
riblets: Evaluation of riblet geometry and application strategies. Renewable
Energy 50: 1095–1105.
[14] Anderson GW, Rohr JJ, Stanley SD. (1993) The Combined Drag Effects of
Riblets and Polymers in Pipe Flow. Journal of Fluids Engineering 115(2):
213–221.
[15] Dunbar B. (2008) NASA riblets for Stars & Stripes (April 22, 2008). Retrieved
from http://www.nasa.gov/centers/langley/news/factsheets/Riblets.html
[16] Liu KN, Joseph DD, Riccius O, Christodoulou C. (1990) Drag reduction in
pipes lined with riblets. AIAA Journal 28(10): 1697–1698.
[17] Grüneberger R, Hage W. (2011) Drag characteristics of longitudinal and
transverse riblets at low dimensionless spacings. Experiments in Fluids 50(2):
363–373.
surfaces and their optimization with an adjustable geometry. Journal of Fluid
Mechanics 338(5): 59–87.
[19] Koury E, Virk PS. (1995) Drag reduction by polymer solutions in a riblet-lined
pipe. Flow Turbulence and Combustion 54(4): 323–347.
[20] El-Samna OA, Chun HH, Yoon HS. (2007) Drag reduction of turbulent flow
over thin rectangular riblets. International Journal of Engineering Science 45:
436–454.
[21] Brima A, Atmani R, Achour B. (2008) Theoretical Approach on the Study of
Turbulent Water Flow over Smooth and L-shaped Riblets Surfaces. Interna-
tional Journal of Fluid Mechanics Research 35(2): 161–171.
[22] Vukoslavcevic, Wallace JM, Balint JL. (1992) Viscous drag reduction using
streamwise- aligned riblets. AIAA Journal 350(30): 1119–1122.
[23] Beghidja A, Gouidmi H, Benderradji R. (2006) Study of the interaction shock
wave boundary layer with the K-ω turbulence model. WSEAS Trans Inf Sci Appl
3(5): 921–926.
[24] Nasrin S, Milad B, Asghar BR. (2009) Analyzing the cross section effect of hyper-
sonic flow past a conical body via perturbation method. Proc IASME/WSEAS
Int Conf Fluid Mech Aerodyn, pp. 182–187.
[25] Liu J-H, Jiang N. (2008) Frequency Response of Near-Wall Coherent Structures
to Localized Periodic Blowing and Suction in Turbulent Boundary Layer.
Chinese Physics Letters 25: 1738–1741.
[26] Bandyopadhyay PR, Ahmed, A. (1993) Turbulent boundary layers subjected to
multiple curvatures and pressure gradients. Fluid Mech 246: 503–527.
[27] Park SR, Wallace JM. (1994) Flow Alteration and Drag Reduction by Riblets in
a Turbulent Boundary Layer. AIAA 32(1): 31–38.
Appendix A
9.5. Drag Reduction Percentage Analysis

in Different Size, Shape, Spacing
and Frequency, Tabular Data Below
were Came from Experimental Process
Table 9.2. 3 mm Circular Riblets Tabular Data.
Riblet Configuration
Drag-Reduction
Size (mm) Shape Spacing (mm) Frequency/Hz Percentage (%)
3 Circular 0 10 −0.541282305
3 Circular 0 15 −0.328844036
3 Circular 0 20 −0.424708473
3 Circular 0 25 −0.477004206
3 Circular 0 30 −0.565239188
3 Circular 0 35 −0.472926961
3 Circular 5 10 0.17540007
3 Circular 5 15 0.028251929
3 Circular 5 20 −0.210558439
3 Circular 5 25 −0.213089559
3 Circular 5 30 −0.318340925
3 Circular 5 35 −0.266533425
3 Circular 7 10 0.219686804
3 Circular 7 15 0.071103445
3 Circular 7 20 0.132221019
3 Circular 7 25 −0.070196867
3 Circular 7 30 −0.111488004
(Continued )
319
Table 9.2. (Continued )
Drag-Reduction
3 Circular 7 35 −0.170984614
3 Circular 9 10 0.249211294
3 Circular 9 15 −0.135320576
3 Circular 9 20 −0.140372293
3 Circular 9 25 −0.142059706
3 Circular 9 30 −0.159481709
3 Circular 9 35 −0.142487178
3 Circular 11 10 −0.11502249
3 Circular 11 15 0.12109688
3 Circular 11 20 −0.119316449
3 Circular 11 25 −0.019345949
3 Circular 11 30 −0.135559453
3 Circular 11 35 −0.121114101
3 Circular 13 10 −0.101490432
3 Circular 13 15 −0.101490432
3 Circular 13 20 −0.10527922
3 Circular 13 25 −0.106544779
3 Circular 13 30 −0.119611282
3 Circular 13 35 −0.160955576
Table 9.3. 3 mm Triangular Riblets Tabular Data Riblet Configuration.
Drag-Reduction
3 Triangular 0 10 −0.45945469
3 Triangular 0 15 −0.242470443
3 Triangular 0 20 −0.336899135
3 Triangular 0 25 −0.389149142
3 Triangular 0 30 −0.393002927
3 Triangular 0 35 −0.43423853
3 Triangular 5 10 0.208875004
3 Triangular 5 15 0.060642027
3 Triangular 5 20 −0.178727554
3 Triangular 5 25 −0.080837718
3 Triangular 5 30 −0.04365511
(Continued )
Drag-Reduction
3 Triangular 5 35 −0.23461225
3 Triangular 7 10 −0.137836407
3 Triangular 7 15 0.097015523
3 Triangular 7 20 0.008229935
3 Triangular 7 25 −0.144700828
3 Triangular 7 30 −0.082090054
3 Triangular 7 35 −0.09150715
3 Triangular 9 10 −0.114863673
3 Triangular 9 15 0.12126452
3 Triangular 9 20 0.032882012
3 Triangular 9 25 −0.120584024
3 Triangular 9 30 −0.135372279
3 Triangular 9 35 −0.120946873
3 Triangular 11 10 −0.097634122
3 Triangular 11 15 −0.097634122
3 Triangular 11 20 0.051371069
3 Triangular 11 25 −0.204316092
3 Triangular 11 30 −0.115066437
3 Triangular 11 35 −0.102804842
3 Triangular 13 10 0.302854372
3 Triangular 13 15 0.389299289
3 Triangular 13 20 0.216757992
3 Triangular 13 25 0.011655709
3 Triangular 13 30 −0.101529209
3 Triangular 13 35 −0.036425321
Table 9.4. 5 mm Circular Riblets Tabular Data.
Drag-Reduction
5 Circular 0 10 −0.902137174
5 Circular 0 15 −0.709746398
5 Circular 0 20 −0.811942384
5 Circular 0 25 −0.864439768
5 Circular 0 30 −0.997383173
(Continued )
Drag-Reduction
5 Circular 0 35 −0.862048571
5 Circular 5 10 −0.095241082
5 Circular 5 15 −0.233618445
5 Circular 5 20 −0.327899995
5 Circular 5 25 −0.473532353
5 Circular 5 30 −0.531605696
5 Circular 5 35 −0.524612611
5 Circular 7 10 −0.375890489
5 Circular 7 15 −0.154263786
5 Circular 7 20 −0.104529492
5 Circular 7 25 −0.204248853
5 Circular 7 30 −0.215510091
5 Circular 7 35 −0.193361395
5 Circular 9 10 −0.319506916
5 Circular 9 15 −0.094747792
5 Circular 9 20 −0.33143458
5 Circular 9 25 −0.238892767
5 Circular 9 30 −0.299650697
5 Circular 9 35 −0.336428059
5 Circular 11 10 −0.281917867
5 Circular 11 15 −0.281917867
5 Circular 11 20 −0.292442277
5 Circular 11 25 −0.393380545
5 Circular 11 30 −0.33225356
5 Circular 11 35 −0.400450772
5 Circular 13 10 −0.244328818
5 Circular 13 15 −0.015393133
5 Circular 13 20 −0.106047234
5 Circular 13 25 −0.256496691
5 Circular 13 30 −0.287953085
5 Circular 13 35 −0.257268516
Table 9.5. 5 mm Triangular Riblets Tabular Data.
Drag-Reduction
5 Triangular 0 10 −0.765757817
5 Triangular 0 15 −0.765757817
5 Triangular 0 20 −0.794344685
5 Triangular 0 25 −0.80389349
5 Triangular 0 30 −1.039323121
5 Triangular 0 35 −0.851975606
5 Triangular 5 10 0.341174226
5 Triangular 5 15 0.059598007
5 Triangular 5 20 −0.112278751
5 Triangular 5 25 −0.306932762
5 Triangular 5 30 −0.375542219
5 Triangular 5 35 −0.35263582
5 Triangular 7 10 −0.319065757
5 Triangular 7 15 −0.094282124
5 Triangular 7 20 −0.186247557
5 Triangular 7 25 −0.334955621
5 Triangular 7 30 −0.376034108
5 Triangular 7 35 −0.284633562
5 Triangular 9 10 0.100972766
5 Triangular 9 15 −0.043763379
5 Triangular 9 20 −0.281330409
5 Triangular 9 25 −0.284712278
5 Triangular 9 30 −0.319628992
5 Triangular 9 35 −0.285569006
5 Triangular 11 10 −0.239299318
5 Triangular 11 15 −0.010084216
5 Triangular 11 20 −0.248232714
5 Triangular 11 25 −0.251216716
5 Triangular 11 30 −0.282025581
5 Triangular 11 35 −0.251972652
5 Triangular 13 10 −0.207392742
5 Triangular 13 15 0.023594947
5 Triangular 13 20 −0.215135019
5 Triangular 13 25 −0.316922082
5 Triangular 13 30 −0.24442217
5 Triangular 13 35 −0.271122986
Chapter 10
3-D Modelling of Biological Systems

for Biomimetics
Shujun Zhang,∗,† Donghui Chen,† Kevin Hapeshi ∗

and Xu Zhang‡
∗ Schoolof Computing and Technology,
University of Gloucestershire, UK
† Key Laboratory of Bionics Engineering,
Ministry of Education, Jilin University, China

‡ CED Group of Rolls-Royce UTC,
Faculty of Engineering and the Environment,

University of Southampton, UK
With the advanced development of computer-based enabling technolo-

gies, many engineering, medical, biology, chemistry, physics and food
science etc have developed to the unprecedented levels, which lead to
many research and development interests in various multi-discipline
areas. Among them, biomimetics is one of the most promising and
attractive branches of study. Biomimetics is a branch of study that
uses biological systems as a model to develop synthetic systems. To
learn from nature, one of the fundamental issues is to understand
the natural systems such animals, insects, plants and human beings
etc. The geometrical characterisation and representation of natural
systems is an important fundamental work for biomimetics research.
3D modeling plays a key role in the geometrical characterisation and
representation, especially in computer graphical visualization. This
chapter firstly presents the typical procedure of 3D modelling methods
325
and then reviews the previous work of 3D geometrical modelling

techniques and systems developed for industrial, medical and animation
applications. Especially the chapter discusses the problems associated
with the existing techniques and systems when they are applied to 3D
modelling of biological systems. In addition, the chapter also presents
two case studies of authors’ own work. Based upon the discussions, the
chapter proposes some areas of research interests in 3D modelling of
biological systems and for Biomimetics.
10.1. Introduction
Biomimetics is a branch of study that uses natural biological systems

as a model to develop synthetic systems. While approaches vary, the
principle is the same in that using biomimetics, human beings learn
from nature. Biomimetics is also known by several names — bionics
and biognosis, etc. Basically it is the concept of taking ideas from nature
and implementing them in another technology such as engineering,
design and computing. The concept is very old (the Chinese wanted
to make artificial silk 3,000 years ago; in Greek mythology Daedalus’
wings could be regarded as an early design failure). Biomimetics is
gathering momentum because of recent advances in technology and
the ever increasing need for sympathetic and sustainable technology.
For details of the definition and discussion of biomimetics, please refer
to Refs. 1–3.
To learn from nature, one of the fundamental issues is to understand
the biological systems. The understanding consists of (1) the descrip-
tion of the natural system as well as its behaviors and functions; (2) the
specification of the mechanisms of how natural systems perform their
functions; (3) the identification of the physical, chemical and biological
properties of natural systems; (4) the geometrical characterization of
natural systems including morphological surface features and internal
structures and (5) any other related areas which are useful for the
design and development of synthetic systems. Among the above areas,
the geometrical characterization of natural systems is the important
fundamental work for biomimetics research. Since human beings
looked to nature for inspiration more than 3,000 years when the
3-D Modelling of Biological Systems for Biomimetics 327
Chinese hankered after an artificial silk,2 people have been using various
methods to obtain the geometrical information of natural systems.
Many people have published their work on geometrical characterization
of animals, insects, plants and human beings.4−39 The work includes
techniques for acquisition of geometrical information, data processing
algorithms, 3-D modelling algorithms, 3-D geometrical computation
theory and methods for computer graphical visualizations. However,
most of these technologies and methods have been designed and
developed for engineering, medical-related application and animation
purposes. When they are applied to 3-D modelling of biological sys-
tems for biomimetics research and applications, many problems often
emerge. These problems include: (1) unsuitable accuracy, resolution
and workable areas of 3-D scanners for raw data capturing of biological
objects; (2) low efficiency algorithms for raw data pre-processing, point
cloud/image edge detection and segmentation; (3) 3-D geometrical
mathematical models for 3-D modelling, representation and computer
graphical visualization of biological objects with special features such as
sharp corners of a biological object, insect tarsus, field mouse hairs and
tine features of flowers of some plants, etc.; (4) as far as the authors
are aware, there is no dedicated 3-D modelling system for biolog-
ical objects, including scanners, associated software and stand-alone
software; and (5) there is no well-accepted methodology/protocol in
the biomimetics community for people to follow in enhancing quick
information sharing, efficient knowledge generation and promotion.
This chapter tries to discuss those problems, present some typical
applications and propose some areas of research interests with the
purpose of presenting problems, and hopefully attracting more people’s
interests in efforts for design and development of systems suitable for
3-D modelling of biological objects.
To avoid the repetition, in this chapter, for convenience, the term
“biological objects” will be used to refer to all complex biological
systems, namely animals, insects, plants and human beings. The size
of a biological object can be as big as a mammal animal,or as small as
an insect and as tiny as a cell.
Following the introduction, Section 10.2 will present the dis-
cussion of the methods for 3-D geometrical modelling used in
practice. The discussion will be focused on the methods for raw
data capturing, pre-processing, edge detection and segmentation,

including the problems of the existing 3-D modelling technology
and methods, especially the problems when they are applied to 3-D
modelling of biological objects. Section 10.3 will present the typical 3-
D digitizing technologies/scanners, their characteristics and suitability
for digitizing biological objects. Section 10.4 will discuss various 3-D
geometrical mathematical modelling methods and their characteristics.
Then, Section 10.5 will present three applications of 3-D modelling to
biomimetics study. Based upon the above sections, Section 10.6 will
propose some areas of interest for 3-D modelling of biological objects,
and finally Section 10.7 tries to draw conclusions.
10.2. 3-D Geometrical Modelling
10.2.1. Introduction
3-D geometrical models of a physical object can be defined as a data set
in an organized numerical form. They can be used to generate various
3-D models of different formats to graphically represent the object.
Computer graphical visualization is the most popular and powerful
method. There are many 3-D computer models used in engineering,
science, medical, military, arts and many other areas. The list can
grow much longer easily. The methods for generation of the 3-D
computer models can be classified into two main types: one is reverse
engineering (RE), which is currently used widely for industrial and
animation applications. RE is the process of duplicating an existing
component or subassembly without the aid of drawings, documenta-
tion or computer model data by using 3-D geometry computation,
computer graphics, engineering analysis and measurement of existing
parts to develop technical data required for successful commercial
reproduction.40,41 In this chapter, the RE-based method is called
Method 1 (M1); the other method is direct generation on computer
through 3-D computer software such as Solid Works and Auto CAD
etc., called Method 2 (M2). The difference between the two methods
is that the RE method employs 3-D digitizing techniques to obtain
the surface information of a physical object in the form of a set of
geometrical points of three position components,42 and sometimes
three surface color information,43 while the direct generation of a
3-D model is mainly based on the human–computer interface of 3-

D software to input the geometrical information. That is, almost all
geometrical information is obtained using 3-D digitizing scanners in
the RE method while the direct 3-D generation method employs
limited geometrical information mainly from manual measurement of
an existing object. Figure 10.1 shows the 3-D models of engineering
components, animals, insects, and plants generated using M2.44,45
These models are certainly useful for biomimetics study. However,
they lack detailed geometrical information about the morphological
surfaces; so many small but very important surface features have been
ignored. Comparatively, there are more problems associated with M1;
this is especially true when the method is applied in 3-D modelling
of a biological object. Besides, M2 can be seen loosely as a sub-set of
M1. So this section will mainly present M1 procedures and discuss the
associated problems.
10.2.2. Procedure of M1
10.2.2.1. Procedure
The procedure of typical M1 consists of five steps for CAD model
creation. It involves data acquisition, data processing, segmentation,
curve and surface model generation and solid model creation, as shown
in Fig. 10.2.
Fig. 10.1. 3-D models of engineering component, animals and plants generated
using Method 1. Some jpg files of this figure were from 3D Café and 3D Model Works
(www.3dcafestore.com/3dmodels1.html, http://3dlenta.com/en/plants.html).
Fig. 10.2. The five steps of a typical M1.

10.2.2.2. Data acquisition

Data acquisition is the first step in M1. A target component (e.g. a
morphological surface of an animal, insect or plant) is normally scanned
using an appropriate scanning strategy and 3-D digitizing equipment.
Several settings may be required for a full scanning of a complex surface.
At each scanning, the geometrical positions of many points on the
surface and surface color information are captured. The position of
each geometrical point is described by three positional co-ordinate
components, x, y, and z. The color of the surface is described by three
color components, r (red), g (green) and b (blue). All points together
are called the point cloud.
The surface information of an object can also be captured by normal
video or CCD cameras used for Photogrammetry, a technology which
has been used for more than 150 years,46−50 or for Photogeometry, a
process in which a series of still photos of an object are taken and a
combination of proprietary and off-the-shelf software is used to derive
a 3-D representation of an object in space.
The above scanning methods can only be used to acquire the geo-
metrical and color information of the surface of an object without slic-
ing it. No information of internal features can be obtained. So various
other technologies have been developed to address this problem. The
typical methods are to obtain the cross section information by either
directly scanning or cutting the object into a serial of cross sections51
for medical-related applications. Medical imaging has come a long
way since 1895, when German physicist Wilhelm Conrad Röntgen
observed strange flickers cast by his cathode-ray instruments. Within
months, Röntgen had used the mysterious “X-rays,” as he called them,
to produce an image of the bones of his wife’s hand, revolutionizing
medicine. For the first time, physicians could peek inside the body
without cutting it open or probing an orifice.
3-D scanners for medical-related applications are normally serial
cross-section based. Capturing technologies include (1) sections
obtained directly (e.g. serial-section microscopy, confocal microscopy,
or mechanical sectioning of an object); (2) serial sections obtained
from MRI (magnetic resonance imaging), CT (computer tomographic
reconstruction), ultrasound and X-ray; (3) 3-D holography; and
(4) direct calculation from mathematical models, etc. In this chapter, we

refer these methods as serial cross-section based 3-D scanning methods.
In engineering applications, some efforts have been made to build
3-D models of porous materials using microscope to capture the images
of sliced surfaces of the porous materials and then processing these
images using various software and algorithms.52 The advantage of this
method is that the accuracy of the model can be as high as a few microns
though the proper selection of the magnifying factor of the microscope.
However, this is a very time-consuming process. For example, to build
a 3-D model of 2 mm by 9 mm by 9 mm porous metal could take up
to two to three months.
10.2.2.3. Data processing

Data processing mainly consists of two tasks: (1) point clouds alignment
to form a single point cloud to represent the whole surface and (2) point
cloud sampling.
(a) Point clouds alignment

As discussed above, it is very often required to have several settings to
obtain a full scanning of a surface, so more than one point clouds
are collected. They should be assigned together to form a single
point cloud. There are normally two ways to do it: (1) Physical
method — alignment through a single physical scanning reference
point or multi physical scanning reference points and a homogenous
matrix describing the positional and orientation relationships between
the reference points; (2) Soft computing method — alignment by
picking several (at least three) non-collinear points (at least one point
should not be co-planar to the other two) of each cloud to generate a
homogenous matrix. The physical method is easy to use, but it takes
long time and it is more expensive to make additional fixtures for
setting up the object. The errors of this method are dependent on
both the quality of the fixtures and the experience of the operators.
The soft computing method is relative cheap and easy to use. but
it demands good mathematical models for the efficient computation
of the homogeneous matrices. Some scanners and associated software
provide this feature. But it is often found that the performance of
this feature is not as good as claimed by the marketing promotion

activities.41,53
(b) Sampling methods

Before or after alignment of the point clouds, due to the nature of three-
dimensional scanners, the number of scanned points is often large,
and accuracy varies. The number can be from millions to billions. A
large number of data points cannot be manipulated efficiently in late
processing and application. So, it is desirable to appropriately reduce the
number of data points. However, it needs to be determined precisely
how many (or at what interval) points will be sampled. This amounts
to saying that original parts cannot be reconstructed correctly if the
scanned data is sampled beyond the tolerance.
In this chapter, the two sampling methods will be discussed: vector
sampling and rate sampling.54,55
The algorithm for the vector sampling method can be described as
follows:
Suppose there are three consecutive scanned points (P a ,P b ,P c )
from a point set. Then, two directional vectors can be computed with
the points: V 1 = P b − P a and V 2 = P c − P b . The idea of the
vector sampling method is to remove the center point (P b ) if the
angle (θ) between two direction vectors, V 1 and V 2 , is less than the
predefined angle tolerance (θt ). At the same time, the distance between
first two points (P a and P b ) must be taken into account. If the distance
(l = |P a − P b |) from a considered point to the last retained point is
larger than the distance tolerance (lt ), the considered point will be
preserved.
Sampling by rate is a simple but powerful technique to reduce
highly redundant points within a given tolerance. This method removes
every point whose sequence number is a multiple of “α” along a scan
line, where α is the value of a user-input integer. Both methods are
often employed together for sampling point clouds.
For images stack captured using serial cross-section based 3-D
scanning methods, some pre-processing work is required for image
enhancement prior to 3-D reconstruction. The pre-processing usu-
ally involves application of image filters (mathematical algorithms
implemented in software) to the entire data set to remove noise and

artifacts, smooth or sharpen the images, or to correct for problems with
contrast and/or brightness. While these filters are generally performed
as pre-processing steps, they can also be carried out after a 3-D model
has been reconstructed from the image stack.
Median and Gaussian filters have the general effect of smoothing
images. These are used to eliminate noise and background artifacts and
to smooth sharp edges, but also tend to remove some of the details
in small objects. This kind of filter is useful if we want a smooth 3-D
surface model of biological objects.
Sharpening filters can be used to emphasize details in the image
stack, but also have the effect of highlighting noise and other small
artifacts. The application of sharpening filters is most useful when the
image stack consists of fine structural components of a specimen, or
when edge enhancement is desired. This kind of filter is useful if we
want a 3-D model with small features of biological objects. However,
this function sometimes cannot be applied when modelling a biological
object for the applications of design and development of a tool that
needs a sharper edge.
The contrast and brightness of the image stack can be adjusted to
enhance perception of the sampled specimen. This is usually done by
changing the ramping of the greyscale values for the dataset. Histogram
equalization can be used to improve contrast by a non-linear mapping
of the grey levels in an image. This technique is most commonly used
when the grey levels are concentrated in a small portion of the range
of possible values.
It is important to realize that the application of filters to the data set
can ultimately affect quantitative measurements of 3-D reconstructions
produced from it. As such, the application of filters in some instances
is used only for display purposes.
10.2.2.4. Segmentation and edge detection

(a) Edge detection
Edge points are either the outer boundary among constituent regions
with relatively distinct discontinuities, or internal boundaries in a
region. Edge detection is a crucial step since geometrical features are

recognized and extracted from the information of edge points.
There are normally two kinds of edge points: sharp edges and
smooth edges. Sharp edge features can be often found on the surfaces
of animals and plants. Sharp edges can be detected by the existence
of tangency discontinuities. The process of sharp edge detection is
straightforward since the existence of tangency discontinuities can be
easily checked along the scan lines. But there are some errors here, since
the points on the true edge could be omitted during the scanning. This
error is dependent on the resolution of the scanner. This is often found
to be a barrier to generating a satisfactory 3-D model of a biological
object.
The variation of curvature values can be used to detect smooth
edges. One characteristic provides a reasonable solution for detecting
smooth edges in a curvature plot: “spike.” These spikes in the curvature
plot give a clear indication of the location at which the points are to be
partitioned into different sets. Hence, the presence of spikes can then
be used to detect the presence of smooth edges.
(b) Segmentation
An ordinary engineering component or a biological object consists
of many distinct geometric shapes. The whole scanned point cloud
sometimes needs to be divided into several regions according to its
constituent shapes. This process is called segmentation. The whole
segmentation process needs to be successfully accomplished, since it
is normally a prerequisite step to recognizing features from the discrete
scanned data. It has been manually carried out by RE operators. The
mouse is normally used to select appropriate points on the displayed
image. The point data captured in the mouse “window” are ready to be
mapped to a feature. Segmentation must be continued until the whole
regions of an object are isolated completely. The level of subdivision
depends upon the complexity of the object.
It is the trend that more work has been done for automatic
segmentation though it is still in the early stage for the practical
applications of fully automatic segmentation.56 It is especially true for
the method based RE with normal industrial 3-D digitizing scanners.
As for image stack obtained using serial cross-section based 3-D

scanning methods, segmentation refers to the process of extracting the
desired object (or objects) of interest from the background in an image
or data volume. There are a variety of techniques that are used to do
this,57−59 ranging from the simple (such as thresh holding and masking)
to the complex (such as edge/boundary detection, region growing
and clustering algorithms). Segmentation can be aided through manual
intervention or handled automatically through software algorithms. It
can be performed before building the 3-D reconstruction by processing
of images in the image stack, or after the 3-D model has been formed.
10.2.2.5. Curve/surface model creation

The point clouds generated from segmentation can be used to generate
curve/surface models using commercial software such as Imageware’s
Surfacer60 and Raindrop Geomagic Studio,61 etc. Various curve and
surface models created at this stage are useful for biomimetics research
and application for analyzing and understanding the functions and
geometrical features of biological systems.
10.2.2.6. Solid model creation

Solid models can be directly created from curve or surface models
generated in Section 10.2.2.5 by using commercial CAD software
such as Solid Edge or Solid Works, etc. Solid models are useful for
simulation, animation performance analysis of biological systems, man-
made system design and manufacturing in biomimetics research and
application.
After pre-processing of the image stack captured using serial cross-
section based 3-D scanning methods, it can then be reconstructed into
a 3-D volumetric dataset. This is usually achieved using either volume
or surface rendering techniques.
The resulting 3-D image from cross-section based methods such as
NMR and confocal microscopy is normally a discrete scalar field, that
is, it is a 3-D grid with values given at each point of the grid. But this
kind of 3-D image is not very useful for biomimetics study.
Volume rendering is a computer graphics technique whereby the
object or phenomenon of interest is sampled or subdivided into many
cubic building blocks, called voxels (or volume elements/pexel). A

voxel is the 3-D counterpart of the 2-D pixel and is a measure of unit
volume. Each voxel carries one or more values for some measured or
calculated property of the volume and is typically represented by a unit
cube. The 3-D voxel sets are assembled from multiple 2-D images and
are displayed by projecting these images into 2-D pixel space where
they are stored in a frame buffer.
In surface rendering, the volumetric data must first be converted
into geometric primitives by a process such as isosurfacing, isocon-
touring, surface extraction or border following. These primitives (such
as polygon meshes or contours) are then rendered for display using
conventional geometric rendering techniques.
Both techniques have advantages and pitfalls. A major advantage
of the volume rendering technique is that the 3-D volume can be
displayed without any knowledge of the geometry of the dataset and
hence without intermediate conversion to a surface representation. This
conversion step in surface rendering can sometimes be quite complex,
especially if surfaces are not well defined (i.e. noisy 2-D images) and
can require a lot of user intervention (such as manual contour tracing).
On the other hand, because the 3-D dataset is reduced to a set of
geometric primitives in surface rendering, this can result in a significant
reduction in the amount of data to be stored, and can provide fast
display and manipulation of the 3-D reconstructions produced by this
method. By contrast, since all of the image stack data is used for volume
rendering, computers with lots of memory and processing power are
required to handle volumes rendered in this manner. Because the
entire dataset is preserved in volume rendering, any part, including
internal structures and details (which may be lost when reducing to
geometric structures with surface rendering) may be viewed. This could
prove to be a powerful method for animation of biological systems
to understand their principles and mechanisms at the early stage of
biomimetics research.
10.3. 3-D Digitizing Technologies/Scanners
As discussed above, the first step of creating 3-D models of an

existing physical object is to collect geometrical, and sometimes color,
information of the morphological surface and internal features of an

object. In this step, 3-D digitizing technologies/scanners play the
central role. There are a variety of commercially available technolo-
gies/scanners that can be used to digitize objects from the molecular
scale on up to multi-story buildings. Many of the commercial products
are available on the market. Based on the applications, 3-D scanners
can generally be classified into three main categories: 3-D scanners
for industrial applications, for medical applications and for animation
applications.
10.3.1. 3-D scanners for industrial applications

The main technologies used in 3-D scanners are sensing technologies
including tracking, imaging, range finding or a combination of these.
Based on the sensing technologies, 3-D scanners for industrial appli-
cations can be further divided into two sub-categories: contact and
non-contact scanners.
10.3.1.1. Contact 3-D scanners

The sensing technology in contact 3-D scanners is tracking systems that
digitize by positioning a probe on the surface of an object and triggering
the computer to record the location, as shown in Fig. 10.3. The
simplest tracker is a mechanical linkage or pantograph. Computerized
Measuring Machines (CMM), such as Cyclone made by Renishaw42
and X330 by Faro Technologies62 are robust 3-D mechanical trackers
for manufacturing applications and landscape documentation and
survey. There are manual and automated contact-tracking systems.
Manual tracking systems requiring a large amount of patient, skilled
labor like Faro Arm, but they can sometimes digitize an object directly
into polygonal models, eliminating the need for the reconstruction
phase. Automated probe tracking systems produce point clouds that
will require fully reconstruction line Renishaw Cyclone. Most contact
3-D scanners are designed for the accuracy from 10 µm to 200 µm,
which is suitable for industrial applications such as reverse engineering
of an existing component and error analysis.41,53 But some automated
trackers such as the Scanning Probe Microscope (SPM) can be used to
create 3-D models of molecular scale objects.
Fig. 10.3. (a) Part being scanned. (b) Point cloud collected by Cyclone 3D scanner.
Fig. 10.4. A contact 3-D scanner was used to find out the leg and body layout design
of various insects.
3-D contact scanners can be used for biomimetics research in areas

that do not need detailed geometrical features such as scanning the
overall shape of a dolphin for CFD analysis to learn about its wonderful
fluid dynamics performance63−64 or to find out the leg and body layout
design of various insects to study their incredible ability to carry loads
or study their walking efficiency to advance man-made limbed machine
innovation,65−67 as shown in Fig. 10.4.
But most 3-D scanners for industrial applications are not suitable for
biomimetics research that requires understanding and capturing small
geometrical features of a biological object. For example, to learn from
the female dung beetle’s unique surface features to design and develop
low resistance and efficient earth-moving equipment such as bulldozers

and ploughs,4,7 it is necessary to scan its forehead, which requires
scanners with much higher accuracy and resolution than normal 3-D
contact scanners for industrial applications.
It is worth pointing out that a Japanese company, Mitutoyo, mar-
keted a contact 3-D scanner which can capture geometrical information
with the accuracy as high as ± (0.5 + L/100) µm, resolution as high
as 0.1 µm and measuring force as small as 0.01 N,68 that is suitable
for scanning small geometrical features. But the scanner’s range is only
50 mm and there is no confirmed report that size of the probe is small
enough to reach the small features or that there is associated software
that is suitable for biomimetics research.
10.3.1.2. Non-contact 3-D scanners

The sensing technology in non-contact 3-D scanners involves tracking
systems that digitize by emitting light (laser or normal) onto the object
to capture the geometrical information in the form of three positional
components (x, y, z) or three positional components (x, y, z) and three
color components (r, g, b). The typical examples of this kind of 3-D
scanner are Hymarc’s Hyscan and Wicks Wilson’s human body and
head scanners.
The Hyscan 45 is a high-performance 3-D scanning laser digitizer
which rapidly acquires precise XYZ data points from an object. Hyscan
systems map surface information in a continuous high-speed non-
contact fashion with ±0.001 (±0.025 mm) ±3 sigma accuracy, which
satisfies demanding applications such as reverse engineering, design,
rapid prototyping and inspection in industry. The Hyscan laser digitizer
is designed to fit to any Coordinate Measuring Machine (CMM), Com-
puter Numerically Controlled (CNC) Machine, or other translation
device. The integrated design yields a powerful yet user friendly system
which can be quickly and easily be interchanged with existing probes
or tools.
Figure 10.5 shows the principle behind a 3-D laser scanner.69−71
The TriForm Head Scanner from Wicks and Wilson is designed
specifically to capture the human head in monochrome or full color.
The capture technology is that the scanner projects striped patterns of
Fig. 10.5. Schematic diagram of laser scanning system.69 .
normal white light onto the subject, using no lasers or other radiation.
The patterns are stored by a digital camera and image software is used
to process 3-D geometrical and color information to create a full color
3-D image of head, including complex capturing of hair. The main
advantage for this scanner is the light source is not lasers or other
radiation that could damage the living object. The other advantages
include very short scanning time (from about two seconds for HS1
SINGLE-VIEW to about 10 seconds for HS2 Double View) and color
information of the surface is captured simultaneously. These features
make this scanner suitable for scanning biological bodies (animals,
insects and plants) for some biomimetics research that does not need
high accuracy and resolution, since the TriForm scanners’ accuracy is
as low as ±1 mm.43 Figure 10.6 shows the TriForm HS2 Double View
Head Scanner.
There are many other 3-D scanners available on the market. For
more scanner products and their working principles, please see Refs. 70
and 72. Most of them are not suitable for biomimetics research
that requires the capture of small geometrical features of biological
bodies.8,67
10.3.2. 3-D scanners for medical applications

This section briefly discusses 3-D scanners mainly for medical-related
applications from the biomimetics point of view. For more details, see
Refs. 73 and 74. 3-D scanners for medical applications mainly rely
Control
Computer Light
Source
Camera
Frame
Store
Projection Grating
Pizzo electric Lens
Translator Lens
Object
(a) (b)
Fig. 10.6. HS2 Double View Head Scanner and diagram of Auto-MATE System.41
on an image approach using a series of slices through the object. The

slice image can be obtained by actually cutting the object and taking
optical photographs of the ends or by using advanced sensors such
as ultrasound, Magnetic Resonance Imaging (MRI), X-Ray Computed
Tomography (CT), serial-section microscopy and confocal microscopy.
The slices can be used to produce volumetric data (voxels) or feature
extraction might be used on the images to produce contour lines. Both
forms of data can be readily converted to polygonal and surface models.
In serial-section microscopy, the tissue being studied is sectioned
into a number of slices and each slice is put into a microscope. Then,
images are captured of each slice. To recreate how the tissue looked
before we sectioned it, we must put all the images of all these slices
back together again, just as if we were putting the real slices of tissue
back together again.
In confocal microscopy, the microscope can obtain a single plane
of image data without having to slice the tissue. In this case, we don’t
need to realign the images of the slices much, but just stack them back
together and then visualize the result.
In MRI, the imaging device acquires a number of cross-sectional
planes of data through the tissue being studied. All of these planes must
be stacked back together to obtain a complete picture of what the tissue
was like.
Problems associated with image acquisition using serial-cross-

section-based 3-D scanners are discussed as follows:
(1) It is usually very expensive and requires a tightly controlled opera-

tional environment.
(2) The file size of the image stack is very large. For example,
when using confocal microscopy, an image stack consisting of 50,
512*512 8-bit images occupies 12.8 MB of storage space. The
3-D size of the image stack not only has implications for storage
space, but also has adverse effects on the processing, display and
manipulation of this data. That could be a serious problem for 3-D
modelling of biological objects.
(3) It is difficult to select thickness of the slice. If the thickness is too
small, the number of slices is too high, which leads to large file
size; if thickness is too big, some small features (like the boundary
location between two distinct features) would be omitted.
(4) Confocal microscopy is highly destructive. As the laser scans the
specimen, it is selectively bleached in the X/Y plane and non-
selectively bleached in the Z plane. If the specimen is scanned one
plane at a time, not only is that plane bleached, but all other cross-
sectional planes above and below that plane are also bleached. If
the laser power is set to its maximum, the bottom slice in an image
stack would have received as much irradiation as if all of the images
were captured at the same plane. It has been suggested that this
non-selective bleaching effect in the Z plane can be reduced by a
process called Z-axis distributed averaging.75
(5) It is usually more difficult to use serial-section microscopy and
put the slices back together than to use confocal microscopy,
MRI, or CT. However, it is more difficult to use an imaging
technique like MRI or confocal microscopy to “see” a small object
than to use serial-section microscopy. Because the information
from surrounding areas blurs out what you’re looking for, the
smallest thing that an MRI can “see” is about 1mm cubed. For
confocal microscopes, the smallest object that is detectable is
about 1/10 µm (1/10,000 mm). But once you slice the object
up, you can use other forms of microscopy such as an elec-
tron microscope to be able to see objects almost as small as
1/, µm (1/100,000,000 mm). There are even newer “atomic

force” microscopes that even let you detect individual atoms, but
not many people (as yet) have done 3-D reconstructions at this
minute level. The problem is, the smaller one goes, the more
artifacts can be introduced, so the reconstruction process gets
much more difficult.
From the above discussion, it can be noted that serial cross-section

based 3-D scanners can be a good choice for some biomimetics research
such as understanding of a biological system’s mechanisms at the early
stage of biomimetics research. However, we are far from the situation
where we can rely on them to meet all the requirements of biomimetics
research and application.
10.3.3. 3-D scanners for animation applications

Animation applications include art, fashion, clothing, film, e-Business,
engineering system concept design, medical treatment such as pros-
thetics, orthopedic fittings, biopsy, radiation therapy, and robotic hip
replacement, etc.
3-D scanners for animation applications are different from the
others in terms of their relatively low cost and high efficiency of data
acquisition. But the accuracy is not as high as that for industrial and
medical applications. The typical accuracy is from 200 µm to 2 mm.
There are also two kinds of scanners: contact and non-contact.
For example, the MicroScribe G276 is an affordable 3-D digitizing
system marketed more towards the animation industry. MicroScribe
products capture the physical properties of three-dimensional objects
and translate them into complete 3-D models. Wicks Wilson’s normal
white light scanners are also ideal for animation applications. Other
scanners include the EOS Systems PhotoModeler, which uses man-
ual feature digitizing in disparate viewpoint images to create object
models. Geometrix introduced a product at Siggraph 1998 that applies
automated photogrammetry to video images. This technology holds
significant promise for rapid creation of 3-D models using conventional
and digital video cameras. Figure 10.7 shows a Photogrammetry system
and its principle.77
Fig. 10.7. Principle of Photogrammetry.77
Due to the nature of function performances such as low cost

and high efficiency of data acquisition, 3-D scanners for animation
applications can be used in biomimetics study that requires geometrical
models of large size and low accuracy.
10.4. 3-D Geometrical Mathematical Modelling

Methods
As for the 3-D modelling of cars, airplanes and engineering equipment,
etc., Computer Aided Geometric Design (CAGD) plays a major role in
3-D modelling of animals, insects and plants for biomimetics research
and development. The mathematics behind CAGD is also indispensable
in computer graphical visualization of the models. There are many
methods such as polynomial and parametric interpolation, least squares
and Bezier curves, Hermite and natural cubic splines, tensor product
surfaces, lofting and Coons surfaces, B-splines, B-spline curves and
surfaces, interpolation with B-spline curves, least squares B-spline
methods, and NURBS (non-uniform rational B-splines). This section
discusses some mathematical modelling methods and their suitability
for 3-D modelling of biological objects.
10.4.1. Polynomial interpolation for curves

The central idea of interpolation is to find a polynomial which goes
through prescribed data points (xi , yi , zi ), 0 ≤ i ≤ n. Some methods
(such as Newton interpolation) can be used to find the unique z =

f (x, y) of degree less than or equal to n — one less than the number
of points. The mathematics for finding this equation is elementary
and well known.78−80 The main problems associated with polynomial
interpolation for curve models are: (1) It cannot be trusted for
extrapolation purposes, which indicates that it is not reliable for analysis
and design of a product using geometrical information generated
by extrapolation of the curve model obtained by scanning small
area of biological objects for biomimetics research; (2) It inevitably
generates wiggling of high degree polynomials within the range of
interest. The wiggling problem can be solved using mathematical
treatment.80
10.4.2. Bezier curves

Polynomial interpolation is suitable for a “best fit” polynomial of lower
degree. If the number of points, N , is limited and a higher degree of
polynomial is required, it is normally difficult to get an exact fit. The
method of least squares is one popular solution for such circumstances.
However, Bezier curve is another alternative. It starts with points
P0 , P1 , . . . , Pn and ends up with a parametric curve that is polynomial
of degree less than or equal to n in each slot.
The Bezier curve associated with control points P0 , P1 , . . . , Pn is
defined for 0 ≤ t ≤ 1 by
n
n i
B(t ) = t (1 − t )n−i Pi . (10.1)
i
i=0
The functions attached to the points here are known as the Bernstein
basis functions. Cubic Bezier curves use four control points, and the
basis functions are (1 − t )3 , 3(1 − t )2 , 3(1 − t )t 2 , t 3 . These curves have
many applications, sometimes in the equivalent Hermite formation,
specified in terms of just two control points with specified tangent
vectors there.79 Large curves can be built up by stringing many such
segments together, with tangent continuity at all points. This property
can sometimes be found to be very useful in biomimetics. For example,
due to limitation of workable range of 3-D scanners, a large area of
biological surface can be divided into a matrix of small regular areas.
Then each small area can be scanned separately. The point cloud for
each scanning can be directly processed using software to generate the
desired Bezier curves based on the requirement of biomimetics research
and application. All Bezier curves generated for each small area can
be stringed together to form Bezier curves over the whole area. This
method will save the multi-point clouds alignment time and reduce
errors due to this alignment.
Bezier curves always lie within the convex hull of their control
points, since on [0; 1] the Bernstein basis functions are clearly non-
negative and sum to 1. Generally speaking, only the first and last control
points are interpolated. The intermediate control points influence the
curve’s shape in a different way, acting more like magnets. This feature
shows that Bezier curves cannot be used for the case that requires high
accuracy in biomimetics research. There are various ways to adjust the
influence of the control points. One could repeat some points, i.e.,
list them more than once, but increasing the number of points also
increases the degree of the resulting curve. Another restriction inherent
to the Bezier approach is the fact that the curves change totally as soon
as one control point is moved. This makes it difficult to build 3-D
models that require modifications at small local areas.
It can be noted that both polynomial interpolation and Bezier
curve methods are limited to polynomials. Now let us not restrict
to polynomials. Given points P 1 , P 2 , . . . , P n and corresponding
“weights” w0 , w1 , . . . , wn , the associated rational Bezier curve is
defined on [0, 1] by

n n i
i=0 t (1 − t )n−i wi Pi
i
R(t ) = . (10.2)
n n i
i=0 t (1 − t ) Pi
n−i
i i
If the weights are all equal then the rational Bezier curve reduces to the
ordinary Bezier curve, since the denominator simplifies to the common
weight. Hence only select control points can be “emphasized.” An
important application of rational quadratic Bezier curves is to the
construction of (bits of) conic sections — including circles, ellipses
and hyperbolas — without resorting to trigonometric or hyperbolic
functions.79,80
n
Now all curves can be in the form of C (t ) = i=0 fi (t )Pi ,
for control points P i , and basis functions fi (t ). The polynomial
interpolation functions can also be realized in this way where the fi (t )
are the well-known Lagrange functions.79 In the polynomial Bezier
case, the fi (t ) are the Bernstein basis functions. It will often happen

that ni=0 fi (t ) = 1 as well, at least on some parameter interval, which
ties in with the convex hull property mentioned above when the basis
functions are also non-negative.
10.4.3. Elementary surface patches

The interpolation and Bezier curves can be used to come up with two
broad classes of surface patches: tensor product surfaces and lofting
surfaces. A surface patch is defined as a function of two parameters u,
v, plotted over some rectangle in the u and v plane, taking values in
three spaces. Surfaces of the type z = f (x, y) can be realized this way,
as S (u, v) = [u, v, (f , u, v)].
A tensor product surface patch can be built up from cubic Beziers.
Let start with a grid of 16 control points Pi,j (0 ≤ i, j ≤ 3), and then
consider the Bernstein functions 3!/i!(3 − i)!u i (1 − u)3−i and 3!/
j !(3 − j )!v j (1 − v)3−j , used in the definitions of the cubic Bezier curves
controlled by P0,0 , P1,0 , P2,0 , P3,0 and P0,0 , P0,1 , P0,2 , P0,3 , respectively.
We use the set of all possible products of these one variable functions
as a basis for our surface: for u, v belong to [0, 1], define S(u, v)
to be
3
3 3 i
u (1 − u)3−i v i (1 − v)3−i Pi,j . (10.3)
i j
i,j =0
This surface interpolates the corner points P0,0 , P0,3 , P3,0 , and P3,3 .
Its precise shape may be varied by altering these and the other control
points.
A simpler example of a tensor product surface patch is the bilinear
surface patch, which also turns out to be the easiest example of the
second class of surface patches, lofting surface. Let start with linear
Beziers, i.e., straight lines, connecting four points Q0,0 , Q1,0 , Q1,1 ,
and Q0,1 (in that order), yielding:
Bill (u, v) = (1 − u)(1 − v)Q0,0 + (1 − u)vQ0,1

+ u(1 − v)Q1,0 + uvQ1,1 (10.4)
for u, v ∈ [0, 1]. This is the simplest surface joining four points. It is a
ruled surface, and interpolates the lines which connect the given points.
Replacing the four straight lines in the bilinear surface patch with
four curves yields a general curved surface patch. For example, use
curves Cu,0 and Cu,1 joining points Q0,0 to Q1,0 and Q0,1 to Q1,1 and
curves C0,v and C1,v joining points Q1,0 to Q1,1 and Q0,0 to Q0,1 ,
respectively. A hammock between the curves Cu,0 and Cu,1 can be
obtained by lofting in the v direction. That is
Loftv (u, v) = (1 − v)Cu,0 + vCu,1 (10.5)
for u, v ∈ [0, 1]. This linearly interpolates between corresponding

points on each curve.
Similarly, lofting in u direction yields
Loftu (u, v) = (1 − u)C0,v + uC1,v . (10.6)
The Coons surface construction gives a surface patch which interpolates

all four of the space curves. Its parametric equation is given by:
Coons (u, v) = Loftu (u, v) + Loftv (u, v) − Bil (u, v). (10.7)
The Coons surface can be modified by multiplying three component

surfaces with three weight factors to form a general form of 3-D surface.
This can be used to build 3-D surface of biological morphological
surfaces with small features for biomimetics research and development.
It is also possible to build various 3-D models by stitching together
lofting, Coons, and Bezier patches — or the more general B-spline
surfaces or NURBS patches discussed in Sections 10.4.4 and 10.4.5.
10.4.4. B-splines
B-splines are a class of functions made up of pieces of polynomials,
joined together in some fashion. We start by choosing an m+1-tuple
T = [t0 , t1 , . . . , tm ] of non-decreasing real numbers, which is called the
knot vector. We then define the B-splines Bi,j of order j recursively, as

follows. Set

1 on [ti , ti+1 ]
Bi,1 (t ) = (10.8)
0 elsewhere
for 0 ≤ i ≤ m − 1 (with Bm−1,1 (t ) = 1 on [tm−1 , tm ]), and for any
j − m, we define
Bi,j (t ) = wi,j (t )Bi,j −1 (t ) + (1 − wi+1,j (t ))Bi+1,j −1 (t ) (10.9)
for 0 ≤ i ≤ m − j , where the wi,j is given by

 (t − ti ) if ti+j −1 = ti
wi,j (t ) = (ti+j −1 − ti ) . (10.10)

0 otherwise
Given T and k ≤ m, we thus get m −k +1 piecewise polynomials Bi,k (t )
of degree at most k − 1. The pieces which constitute each function join
up at the knots, where they exhibit varying degrees of smoothness.
Order one B-splines Bi,1 (t ) are step functions, order two B-splines
Bi2k (t ) are zig-zag linears, order three B-splines Bi,3 (t ) are piecewise
quadratics, and so on.
The multiplicity of a given knot (i.e., how often it is repeated in
the knot vector) tells us a lot about the way the pieces of the B-splines
join up there: at a knot of multiplicity l , each Bi,k (t ) is at least k − l − 1
times continuously differentiable.79,80 If k − l − 1 = −1, this is to be
interpreted as a potential discontinuity.
If k − l − 1 = 0, this means that the B-splines are continuous but
not differentiable, i.e. B-splines have spikes. This feature of B-splines is
very useful for building 3-D models of special geometrical features such
as tarsus of an insect leg, insect cerci, horn of an ox and filed mouse
hair, as shown in Fig. 10.8.
Fig. 10.8. Biological surfaces can be 3-D-modelled with B-spline methods by letting
k − l − 1 = 0. Pictures from http://beetles.source.at/english/anim.htm
10.4.4.1. B-spline curves

Given a knot vector T = [t0 , t1 , . . . , tm ], and the corresponding m −
k + 1 B-splines Bi,k (t ) of order k, for some fixed k ≤ m, then if we also
have m − k + 1 control points P o , P 1 , . . . , P m−k , we can put everything
together to get a B-spline curve of degree k−1 defined on [tk−1 , tm−k+1 ]:

m−k+1
C (t ) = Bi,k (t )Pi . (10.11)
i=0
If T = [0, 0, 0, 1, 1, 1, 1] and k = 4, we see that C(t ) is just a cubic

Bezier curve; indeed any Bezier curve can be realized in a similar way.
10.4.4.2. B-Spine Surfaces

The tensor product bi-cubic Bezier surface patch can be generalized to
get B-spline surface patches. Let knot vectors U = [u0 , u1 , . . . , umu ],
V = [v0 , v1 , . . . , vmu ], of possibly different lengths, and two curve
orders ku ; kv . Given a rectangular mesh of (mu − ku + 1)_(mv − kv + 1)
control points Pi.j , we can then define
−kv +1
−ku +1 mv
mu
S(u, v) = Bi,ku (u)Bj ,kv (v)Pi,j . (10.12)
i=0 j =0
When both knots vectors are [0; 0; 0; 0; 1; 1; 1; 1] and both orders are
4, we get a bi-cubic Bezier surface patch as seen before.
10.4.5. Non-uniform rational B-splines

(NURBS) curve
Non-uniform rational B-Splines (NURBS)81 refers to curve and surface
generation using not-necessarily uniform (equally spaced) knots in
conjunction with the rational function approach already encountered
in above discussion of Bezier curves. The NURBS curve is explained
in detail by Piegl and Tiller.82 A pth-degree NURBS curve is deter-
mined by its control points {Pi }, weights {wi }, and knots {ui }. The
polygon formed by the {Pi } is called the control polygon. The knots
form a sequence of non-decreasing real numbers while ui ≤ ui+1
(i = 0, . . . , m − 1) and normally u0 = · · · = up = 0, um−p =

· · · = um = 1. The number of knots is m + 1. The U , U =
{u0 , . . . , up , up+1 , . . . , um−p−1 , um−p , . . . , um } is called the knot vector.
With the B-spline basis functions {Ni,p (u)} recursively defined on U as

1 if ui ≤ u ≤ ui+1
Ni,0 (u) =
0 otherwise
u − ui ui+p−1 − u
Ni,p (u) = Ni,p−1 (u) + Ni+1,p (u),
ui+p − ui ui+p−1 − ui+1
(10.13)
the NURBS curve is a function of the parameter u
n
Ni,p (u)wi Pi
C (u) = i=0 n . (10.14)
i=0 Ni,p (u)wi
It is normally assumed that wi > 0 and u ∈ [0, 1]. The number of
control points is n + 1.
Setting the rational basis functions as
Ni,p (u)wi
Ri,p (u) = n , (10.15)
j =0 Nj ,p (u)wj
where the {Ri,p (u)} are piecewise rational functions, Eq. (10.14) can
be rewritten as
n
C (u) = Ri,p (u)Pi . (10.16)
i=0
The NURBS curve has many useful and important characteristics. The
following presents some characteristics that are related biomimetics
research.
(1) Ri,p (u) = 0 for u ∈
/ [ui , ui+p+1 ]. Therefore, moving a single control
point Pi or changing the weight wi affects only the segment of the
NURBS curve for u∈ / [ui , ui+p+1 ], and outside this interval the
control point Pi has no effect. The control points of the NURBS
curve are said to exert the property of localness. A complementary
fact is that each segment of such curves is controlled by at most
k of the Pi ’s: the segment between tj and tj +1 is completely
determined by Pj −k+1 , . . . , Pj . This is the important property for
3-D-modelling biological surface that requires local modification.
(2) If u ∈ [ui , ui+1 ], then C(u) lies within the convex hull formed by
the control points from Pi−p to Pi . That is the control polygon
determines the general shape of the NURBS curve.
(3) All Ri,p (u) attain exactly one maximum for all u ∈ [0, 1] when
p > 0. When the knot vector reflects the geometric distribution
of the control points and all the weights have the same or close
values, the maximum values of Ri,p (u) generally occur at the
curve segment closest to the control point Pi . That is, under the
circumstance, the closer a curve point C (ū) is to a control point
Pi , the more substantially the modification of the control point Pi
affects the curve point C (ū).
(4) Weight modification has a perspective effect. If a weight wi changes,
each affected point C (ū) will move along a straight line defined
by the original point C (ū) and the control point Pi . C (ū) will
be pulled toward Pi if wi increases, and C (ū) will be pushed
away from Pi if wi decreases. On the other hand, control point
repositioning has a translational effect. That is, when a control
point Pi is repositioned to a new position P’i , all the affected
curve points move in a parallel direction. This property is useful
for studying biomimetics limbed robots. For example, for the
design and development of jumping robots, it is very important
to study the jumping principle and muscle motion of some insects’
legs. The 3-D modelling will play an important role. During the
3-D modelling, the control points of the NURBS curves are
positioned near the locations where muscles are attached to bones
or some locations along the muscles, then the NURBS curves are
geometrically associated with a generic polygon mesh representing
the muscles. The weights can then be changed or the control points
can be repositioned to simulate various motions. These 3-D models
can be used not only during the early stage performance analysis
and biomimetics design, but also for manufacturing at late stage.
(5) The two knots which are closer together seem to be dragging the
curve towards the control point. The double knot can be used to
obtain an interpolation of one of the control points. This property
can be used to build 3-D models of biological features with small
round corner by sneaking the two knots up on a knot from the
both sides.
The above features should be further exploited for the 3-D model
building of biological surfaces for biomimetics.
10.4.6. NURBS surfaces

If we extend one dimensional parameter space (u) in NURBS curve
to two dimensional parameter space (u, v), we get NURBS surface. A
non-uniform rational B-spline surface of degree (p, q) is defined by
m n
i=0 j =0 Ni,p (u)Nj ,q (v)wi,j Pi,j
S(u, v) = m n , (10.17)
i=0 j =0 Ni,p (u)Nj ,q (v)wi,j
Similarly, the property analysis for NURBS surface should be done for
the 3-D model building of biological surfaces for biomimetics.
10.5. Applications of 3-D Modelling

in Biomimetics
10.5.1. An innovative methodology of product

design from nature83
10.5.1.1. Introduction
Nature is an information sourcebook for behavior, function, color
and shape, which can inspire visual design and invention. Studying
the form and functional characteristics of a natural object can provide
inspiration for product design and help to improve the marketability
of manufactured products. The inspiration can be triggered either by
direct observation or captured with three-dimensional (3-D) digitizing
techniques to obtain superficial information (geometry and color). An
art designer often creates a concept in the form of a two-dimensional
(2-D) sketch while engineering methods lead to a point cloud in
3-D. Each has its limitations in that the art designer commonly lacks
the knowledge to build a final product from a 2-D sketch and the
engineering designer’s 3-D point clouds may not be very beautiful.
In this section, an innovative methodology of product design from
nature will be presented for Product Design from Nature (PDN),
coupling aesthetic intent and geometrical characteristics, exploring the
interactions between designers and nature’s systems in PDN. It was

believed that this approach would considerably reduce the lead time
and cost of product design and development from nature.
10.5.1.2. Discussion of bio inspiration of product design

In 1984 Nigel Cross published Developments in Design Methodology,
collecting influential articles about methods and procedures over the
past two decades. His book reflected developments in the design
research community. The approach was defined as “. . .the study of how
designers work and think; the establishment of appropriate structures
for the design process; the development and application of new design
methods, techniques and procedures; and reflection on the nature and
extent of design knowledge and its application to design problems.”84
Conceptual design is normally optimized by iteration due to the
lack of ideas at the early stage of design. The iteration process is
very time-consuming and expensive, returning to earlier stages to
modify the designers’ ideas, often resulting in long lead time and high
cost of introducing a new product onto the market. Moreover, the
design philosophy of “form follows function” is no longer sufficient;
the aesthetic aspect of a product has become a more and more
important element for success. In addition, the shift in manufacturing
paradigm will have a deep impact on design and operation of future
manufacturing systems.
The development of computer-based design has given rise to
beguiling but rather rigid images, and this in turn has caused researchers
such as Soufi and Edmonds to observe that “current computer aided
design (CAD) systems do not provide sufficient support to the early
conceptual stages of design.”85 For designers, it is important to
understand how invention works in design, and where and how it
breaks down. One of the inventive processes is conceptual design
where inspiration and observation can arise or be learned from nature.
This is called the conceptual design process “Product Design from
Nature” (PDN).
The inspiration from a natural system is called bio-inspiration.
Bio-inspiration can be triggered either by direct observation by an
art design professional or captured by an engineering designer using
3-D digitizing techniques to obtain surface information. An art design

professional often creates a conceptual design in the form of a 2-
D sketch while the engineering method leads to a point cloud in
3-D. Both methods are limited: the art design professional lacks
the knowledge to build a final product from a 2-D sketch and the
engineering designer’s 3-D point clouds are not very aesthetic. In
this methodology, the proposed method for product design integrates
reverse engineering, 3-D geometrical computation and inspiration from
a designer’s sketch of biological systems for conceptual and form design
and form aesthetics.
It is common to look to nature for inspiration,2,4,86 but there are
few instances of successful transfer of the technology into design. The
fundamental issues are to understand (1) the behavior and function
of the natural system; (2) how natural systems perform their functions
and their physical, chemical and biological properties; (3) the designer’s
observation captured in the freehand sketch; (4) geometrical character-
ization of natural systems including 2-D and 3-D models, a joint effort
of engineering and graphic design, which leads to the final conceptual
design based on functional requirements and form aesthetics; and (5)
the design of a prototype and its development by testing. Of these,
geometrical features, the designer’s sketched inspiration, and the form
aesthetics of natural systems are essential for designing a bionic product.
10.5.1.3. Proposition of a methodology for PDN

There are normally two methods for capturing geometric information
for bio-inspiration. First, art design, in which the direct observation
of a natural object is undertaken by art design professionals with their
special knowledge and experience. Second, the engineering method, in
which engineers measure a natural object using methods such as reverse
engineering. Both methods have their advantages and disadvantages.
The PDN method aims to combine the advantages of both methods
and overcome the disadvantages of each (see Fig. 10.9).
The procedure is as follows:
Stage 1 — Capturing surface geometrical information
A professional designer observes a biological object such as an apple, a
lotus leaf or a bird to get bio-inspiration. With their expert knowledge,
Fig. 10.9. The proposed methodology for PDN.
experience and talent, they use their imagination to design either a

pre-defined product or something totally new. The output is usually a
2-D sketch (freehand drawing) which can be scanned to get digitized
geometrical information in a format suitable for later processing
to generate the geometrical information required for 3-D models
using the algorithms in Stage 2. Alternatively, an engineer measures
geometrical and color data directly from a natural object. The outputs
are normally various sets of points in x-, y- and z-coordinates for
geometrical data and r, g and b for color. These are “point clouds”
which can be processed to generate information for geometrical design
(such as 3-D models) and aesthetic design using various algorithms in

Stage 2.
Stage 2 — Building a 3-D model
This consists of filtering, segmentation, edge detection, and initial 3-D
modelling, iterative analysis/optimization for the final 3-D models,
considering also the aesthetics. There should be sufficient collabo-
ration between professional designers, engineers and 3-D modelers.
The inputs from designers should be effectively integrated into 3-D
modelling to retain aesthetic beauty in the designed product. The
inputs from engineering design and the 3-D modeler are used to build
3-D models that are suitable for production in Stage 3. In this stage,
various algorithms have to be applied.
Stage 3 — Manufacturing a prototype
The aim of any product design including PDN is to provide a
description for manufacturing a physical product. This is important for
bionic engineering to be successfully integrated into the development
and manufacturing process. So the natural stage of the proposed
method should be to build some prototypes for testing, quality control
and production of the product for market. Various rapid prototyp-
ing techniques/methods are available commercially. Typical methods
include rapid prototyping, high-speed machining, laser cutting and
processing, etc.86
The innovation is in the integration of art design with engineering
design for PDN. A number of algorithms have to be used. Some of them
have been developed, but some of them still need to be developed. The
details of the procedures of art design direct observation — sketching
from a natural object — can be found in Ref. 83.
10.5.1.4. A case study of the design of light shade —

learning from lotus leaves
In this sub-section, a case study is presented to demonstrate how to
follow the proposed methodology to design a light shade by learning
from lotus leaves.
Figure 10.10 shows a 2-D sketch by an art design professional
based on bio-inspiration after observing a lotus plant. This sketch
Fig. 10.10. A 2-D sketch based on the bio-inspiration.83
Fig. 10.11. (a) sketch, (b) curves; (c) 3-D model.83
can be processed through scanning, sampling, edge detection and

segmentation to generate curves/surfaces (Fig. 10.11) which can be
used to build a 3-D wire frame model (Fig. 10.12). This 3-D model
can be further modified to generate the final 3-D model and design
specification — PDN. The final design can be used for a prototype,
as shown in Fig. 10.13. This prototype was produced using 3-D
printing.
Fig. 10.12. 3-D models.83
Fig. 10.13. Prototype.83
10.5.1.5. Summary
The proposed methodology is a method for innovative product design
from nature, coupling aesthetic intent and geometrical characteristics,
exploring the interaction between designers and nature systems in

product design from nature. From the case study, it can be seen that
the method is valid and useful for designers to do innovative design by
learning from nature.
However, the methodology needs to be further developed before
it can be used in real product design from nature. In particular, great
efforts have to be made on how to design and develop an effective
and efficient algorithm to build a 3-D model from a designer’s 2-D
sketches and how to design and develop a computer system that is either
workstation-based or web-based (though web-based is preferable) as
a platform for both art designers and design engineers to effectively
collaborate during the optimization of the design and aesthetics to
increase the marketability of the products.
10.5.2. A new parameterized feature-based generic

3-D human face model for emotional
bio-robots87
Emotional bio-robots are one of the important areas of bionic robots
applications. To represent human facial expressions is an essential
requirement for building emotional bio-robots because the expressions
can help bio-robots communicate with human beings emotionally.
To design and develop emotional robots, it is necessary to build a
generic 3-D human face model. While the geometrical features of
human faces are freeform surfaces with complex properties, it is the
fundamental requirement for the model to have the ability of repre-
senting both primitive and freeform surfaces. This requirement makes
Non-rational Uniform B-Spline (NURBS) suitable for 3-D human face
modelling.
In this sub-section, a new parameterized feature-based generic
3-D human face model is presented and implemented. Based on
observation of human face anatomy, the model defines 34 NURBS
curve features and 21 NURBS surface features to represent the human
facial components, such as eyebrows, eyes, nose and mouth, etc. These
curve models and surface models can be used to simulate different
facial expressions by manipulating the control points of those NURBS

features. Unlike the existing individual-based face modelling methods,
this parameterized 3-D face model also gives users the ability to use
the model to imitate any facial appearances. In addition, the potential
applications of the new proposed 3-D face model are also discussed.
Besides emotional bio-robots, it is believed that the proposed model
can also be applied in other fields such as aesthetic plastic surgery
simulation, film and computer game characters creation, and criminal
investigation and prevention.
10.5.2.2. Discussion of the methods for human face

3-D models
With the rapid development and application of 3-D modelling tech-
niques, a lot of research has been carried out to study human faces and
to build 3-D face models for various applications,88−90 especially for
emotional bio-robots, aesthetic surgery, crime detection and computer
games, etc. One intuitive approach to modelling human faces is using
scanned 3-D data. It is a type of active stereo vision method. The data
obtained using 3-D scanners — so-called “clouds” — is a complete set
of 3-D information including 3-D coordinates, colors, and textures
rather than the profile images of an object. The number of points
in “clouds” varies from hundreds of thousands to millions. Based
on these “clouds,” different algorithms are developed to build the
corresponding 3-D models.91−94
Another 3-D modelling method is the triangular patches method,
which is a surface approximation method. Although each triangle can be
expressed in a 2-D plane, numerous triangles in 3-D can be connected
together to approximate an arbitrary surface. A famous triangular
patches face model is CANDIDE,95 which uses hundreds of triangles
to represent a 3-D human face with several simple facial features. Since
the first CANDIDE model was proposed, several variations have been
proposed.96−98
The third approach of modelling 3-D human faces is the statistical
model-based method. The most successful statistical face model is the
morphable model established by Blanz and Vetter.99 A pixel-level 3-D
prototype face database was constructed to store the statistical data. A
morphable face model was derived by transforming shape and texture

into a vector space representation. Later, other researchers proposed
various modelling methods based on 3-D morphable model using
collected statistical scanning data sets.100−102
Even though there are several existing modelling methods, the
gap between these methods and the requirements of emotional bio-
robots applications still urges the creation of a parameterized generic
3-D human face model:
• First of all, human faces consist of a number of geometrical regions

which contain different muscles under the skin. The expressions are
too complicated to simulate with a single surface model because even
the minor expressions involve multiple muscles. Difficulties also exist
in transferring expressions of one bio-robot to another because there
is no parametric representation of the face models.
• Second, the 3-D face model required by emotional bio-robots
should be capable of generating faces with arbitrary appearances
through manipulating the facial features. The previous modelling
methods only provide mechanisms to represent a 3-D face from
data points.96,103 They were not mainly designed for representing
facial features and especially their boundary information. Another
approach for 3-D face modelling is for professional engineers to
design the model using 3-D modelling software. But the accuracy
will depend on the experience of the engineers. It is not easy for an
ordinary person to create a face model.
• Third, sometimes the target people cannot present for scanning —
even an image of the target person is difficult to obtain in some
cases. For example, the clear frontal images of terrorists are rarely
circulated in ordinary circumstances. So it is not feasible to build
3-D face models using existing modelling methods.
From the above discussions, it can be seen that it is necessary to develop

a parameterized generic 3-D face model that can be used to help the
engineers design and develop emotional bio-robots even when the
robots’ faces are not similar to any existing human beings, because
the appearances of the robots can be easily changed by morphing some
facial features of the generic model.
In this model, techniques of NURBS will be used. Details of

NURBS can be found in Section 10.4 or from Ref. 87.
10.5.2.3. Face model definition

To build up a parameterized generic 3-D face model, a human face
should be divided into a number of geometrical surfaces for the
purpose of expressing the whole face with features-based parameterized
geometrical models. Based on the features analysis (such as nose,
mouth, eyes and chin, etc.) on a human face104,105 and the convenience
and effectiveness of manipulating the generic 3-D model, 34 curve
features are defined on a human face (as shown in Fig. 10.14). These
curve features can be combined together to form 21 surface features
(as shown in Fig. 10.15).
10.5.2.4. Reverse computation and 3-D manipulation

of facial features
As soon as all facial features are defined, they need to be manipulated
to change the appearance of the model, and the adjacent features are
connected together, which means no holes and gaps on the surface
model. In practice, the degree of NURBS curve and surface is usually
chosen as 3 to get a balance between control flexibility and computation
complexity.106 All weights are set to 1 initially and can be adjusted
while manipulating the models. So the next step is to calculate NURBS
control points and knot vectors by reverse computation.
The problem of NURBS curve reverse computation can be
described as: Given a set of data points Q = {q0 q1 q2 . . . qm , to compute
the control points Pi , weightswi , and knot vector U with a specified
degree p which starts from q0 ends with qm , and passes through q2 to
qm−1 . The problem of NURBS surface reverse computation is similar
except it is described in a 3-D space.
Assume U = {u0 u1 u2 . . . un+1 un+2 , . . . un+p un+p+1 is the knot
vector of target NURBS curve, because the NURBS curve passes
through the first and last control points, such condition for knots holds:
u0 = u1 = u2 = · · · = up = 0, un+1 = nn+2 = un+3 = · · · =
un+p+1 = 1. Only the knots between up+1 and un are unknown. Because
Fig. 10.14. Features of NURBS curve face model.
Fig. 10.15. Features of NURBS surface model.

the clamped NURBS passes through the first and last control points,
these two points are set as duplicated control points in this paper. The
relationship between the number of control points n and the number
of data points m is n = m + 2. So the total number of knots in knot
vector is n + p + 1 = m + 2 + 3 + 1 = m + 6, and the total number of
control points is n + 1 = m + 3. The unknown knots in the knot vector
from up+1 to um+2 (un ) can be computed by parameterizing knots for
each data point qj . As expressed in Eq. (10.18), the cumulative chord
length method107 is used to calculate all knots:
u0 = u1 = u2 = u3 = 0
|qi − qi−1 |
ui+3 = ui+2 + m for i = 1, 2, . . . , m − 1
i=1 |qi − qi−1 |
um+3 = um+4 = um+5 = um+6 = 1. (10.18)
The target curve will pass through all m data point qi (0 ≤ i ≤ m),
which means m+1 equations can be obtained according to Eq. (10.19).
But according to the analysis above, the target NURBS curve has
m + 3 control points. The target NURBS curve is tangent to the
bounding polygon at the first and last control points. This introduces
two boundary conditions.
m
j = C (u) = N j ,p (ūj )P j
j =0

m
= N j ,3 (ūj )P j ( j = 0, 1, 2, . . . , m) (10.19)
j =0
Using these two boundary conditions and parameterized knot vector,

the linear equation of AP = D are generated as Eq. (10.20).
    
a0 b0 c0 ... ... ... P0 d0
a1 b1 c1 ... ... . . .    
   P1   d1 
. .   .   . 
 . ... ... ... ... . . .  . = . 
  .   . 
. . . . . . am+1 bm+1 cm+1 . . . Pm+1  dm+1 
... ... am+2 bm+2 cm+2 ... Pm+2 dm+2
(10.20)
Where d0 and dm+2 are the bounding conditions at the first and last
control point, dj = qj −1 for j = 1, 2, . . . , m + 1, and ai , bi , ci for
i = 1, 2, . . . , m + 1 are the basis functions of three degrees for each u i

previously computed using the cumulative chord length function. The
control points vector P can be computed by P = A −1 D.
As the degree, knot vector, and control points of the NURBS curve
are computed, the unique curve can be evaluated using the de Boor108
algorithm. It can also be considered as the NURBS curve whose weights
equal one. It can be changed freely by either control points or weights
factors. Figure 10.16 is an example of the reverse computation.
By extending the above method to 3-D space, it is easy to get the
NURBS surface reverse computation. First, all boundaries of the target
surface are computed using NURBS curve reverse computation. Next,
the knot insertion is performed to make sure each pair of opposite
boundaries has the same number of knots. Then, the “control net”
(matrix of control points) of the target NURBS surface is determined
by interpolating between each pair of control points on the opposite
boundaries. Finally, the surface can be evaluated according to computed
knots and control net. Figure 10.17 is an example of NURBS surface
reverse computation of the nose feature. By applying the reverse
Fig. 10.16. NURBS curve by reverse computation. Left: calculated control points;
Right: original data points.
Fig. 10.17. Nose surface reverse computation. Left: boundary NURBS curves;
Middle: computed control net; Right: evaluated NURBS surface.
computation for all defined features, the NURBS surface face model
can be constructed.
10.5.2.5. Experiment results

To verify the proposed model, a generic parameterized feature based 3-
D human face model is implemented in VC++.NET environment. The
constructed generic 3-D face models are vivid and intuitive with defined
adjustable features. Figures 10.18 to 10.21 illustrate the abilities of
morphing facial features of the proposed curve and surface face models.
According to the simulation, the features of the proposed models are
Fig. 10.18. Generic and morphed curve face models.

Fig. 10.19. Original surface model and model with morphed left eye.
Fig. 10.20. Curve face model and surface face model with wider nose.
Fig. 10.21. Morphed NURBS curve face model. Eyes and eye brows in different
shapes (left). Smiling mouth and morphed under jaw line (right).
Fig. 10.22. Morphed NURBS surface face model. Right brow in different shape
(left). Smiling mouth and morphed under jaw line (right).
flexible and can be manipulated to represent various appearances or

different facial expressions.
10.5.2.6. Summary
The method proposed can be used to build a parameterized generic
3-D human face model with NURBS curves and surface features by
applying reverse computation to data points. By comparing the new
proposed model with other existing 3-D modelling methods, this
parameterized face model can represent the real 3-D surface of a human
face mathematically to provide flexible control over all facial features.
The feature-based approach is very useful for performing manipulation
on all defined facial features to represent various facial expressions. By
combining with a face detection method,109 it is possible to generate
the 3-D face model from images of a person automatically, which can
potentially be applied in the field of face aesthetics surgery and crime
prevention.
10.5.3. 3-D geometry modelling of porous copper

materials
This sub-section presents a method for generating 3-D geometrical
modelling of porous materials. This method can be applied to build
3-D model of human bones. The procedure of the method will be

presented and some examples of 3-D models of porous copper will also
be presented.
10.5.3.2. The procedure of the suggested method

The geometrical characteristics of porous materials are that the size of
the pores is very small compared with the body size of the materials
(or components). For example, the pore size of a sandstone is a few
microns. There are a number of previous methods for modelling such
porous materials.110−111 However, these methods are not suitable for
building 3-D models of porous metals with pore size from 20 to 600
microns, especially when the distribution is not as even as that of natural
sandstones. So it is necessary to propose and develop a method for
building 3-D models of porous metals.
The principle of the proposed method is to slice the materials, get
the 2-D geometrical information (boundaries) of the pores found on
each slice and join all found boundaries of each pore in all slices to build
3-D models.
There are nine steps in the proposed method. The procedure is as
follows:
(1) Determine the number of slices and thickness of each slice.

(2) Mill the porous materials from the selected height to the required
height to obtain a number of slices. The number of slices can be
determined based on the application requirement. Due to the large
number of the pores in each slice, it is not necessary to have a large
number of slices. The height of each slice can be determined based
on the application requirement and the material production.
(3) Use a selected microscope to scan the surface of each slice to collect
the geometrical information of the cross-sections of all pores in
each slice. It should be pointed out that multi-scanning is required
to cover the whole surface area for each slice.
(4) Merge the images scanned in Step 3 to generate a single image for
each slice.
(5) Find the boundaries (x- y- and z-coordinates) of cross-sections of
all pores in each slice. Repeat the process to obtain the boundaries
of all pores in all slices.
(6) Input the identified cross-section boundaries into CAD software

to generate 3-D models for all pores in the materials.
(7) Build a 3-D solid model of the same size of the porous materials.
(8) Use Boolean operation functions in CAD software to generate a
3-D model of the porous materials.
(9) The 3-D model generated in Step 8 can be used as the basis
model for generating the 3-D model of a product of porous
materials. It should be pointed out that the product can be of
any required geometrical size and shape. Hence the method can
be used to generate any required 3-D model. So this study will not
provide a generic geometrical model of a given porous material,
but a generic method for building 3-D geometrical models of any
product design.
10.5.3.3. Case study of 3-D modelling of porous copper

A sample of porous copper was selected for 3-D modelling.
(1) Determine the number of slices and thickness of the slice
The pore distribution of the selected porous copper repeats every 200
microns in height. So in theory, it is reasonable to slice the material to a
200 micron height to obtain the geometrical information for building
the 3-D model. However, to increase the reliability and to balance the
amount of test work, it was decided that the number of slices of the first
copper sample is 60 and thickness of each slice is 100 microns. So the
total height of the test is 6,000 microns (6 mm). The images of these
60 slices can be evaluated to find high-quality ones for 3-D modelling.
(2) Mill the porous materials from the selected height to the
required height to obtain the slices
It is very important to use a milling machine with the required accuracy.
A CNC milling machine was used — a LEADWELL V-30 vertical CNC
made in Taiwan, and the controller is a FANUC Series 21-M and the
CNC system milling error is 0.002mm. A new cutter was used. The
cutter’s diameter is 2 cm to make sure the cross-section can be milled in
one go. From the observation of the milled surface, there are no clear
forced deformations found. That means the CNC milling method is
suitable for the test.
Fig. 10.23. Zeiss SteREO Discovery V20 microscope.
(3) Scan the surface of each slice to collect the geometrical

information of the cross-section of all pores for each slice
It was decided to use a microscope to capture the image of the cross-
sections. A Carl Zeiss SteREO Discovery V20 was used, as shown in
Fig. 10.23. As discussed above, multi-scanning is required to cover the
whole surface area for each slice. Based on the measurement of the sizes
of pores on each cross-section, the magnifier factor of 20 was selected
as a compromise between the clearness of the images and the number
of images. For the sample, 20 images are required to cover the whole
cross section area. A typical image is shown in Fig. 10.24.
(4) Merge the images scanned in Step 3 to generate a single image
for each slice
After a few trials, it was found that Microsoft painter is a convenient
and reliable software to merge images scanned in Step 3 to generate a
single image for each slice. Every effort has to be made to ensure that
all individual images of each slice are accurately merged into a single
one. There are some overlaps between images. The same pore features
in the overlaps of the images are identified as the reference for merging.
This work could be done by developing a Matlab program.
Fig. 10.24. A typical image captured using the microscope.
(5) Identify the boundaries (x- and y-coordinates) of cross-sections

of all pores in each slice
Considering the requirements of 3-D modelling and the time needed,
11 slices of the sample with high image quality were selected for further
geometrical information processing (the identification of boundaries of
pores, their 3-D modelling and pore sizes and distribution analysis).
A method has been developed using some functions of Matlab’s
artificial intelligent tool box and a few self-developed algorithms for
the boundary identification. Figure 10.25 shows the found boundaries
of pores for one typical slice. This kind of jpg file can be properly viewed
using free software such as Windows Photo Gallery.
During this process, the data of x- and y-components of all pores in
each slice was recorded in a single data file in a neutral file format. They
can be inputted into any software. The z-component of each point on
identified boundaries can be added using software like Excel.
(6) Input the identified cross-section boundaries into CAD soft-
ware to generate 3-D models for all pores in each slice
SolidWorks CAD was selected for building the 3-D models of the
porous materials. Since the x-, y- and z-components of all points of all
pores in each slice are saved in a single file, when this data file is loaded
into SolidWorks, the software treats the whole file as a single point
Fig. 10.25. Identified the boundaries of all pores.
Fig. 10.26. Point cloud in SolidWorks (whole slice).
cloud, as shown in Fig. 10.26. A macro was written in the SolidWorks

environment, and this macro can be used to load the data file for
each slice into the SolidWorks environment and automatically process
the data file to identify all points for each pore and to build curve
models of all pores in each slice. After running this macro, all curve
Fig. 10.27. Curve model of pores in a typical slice.
models for each slice are generated and saved in a single file, and these
curve models can be separately selected for solid modelling, as shown
in Fig. 10.27.
(7) Solid modelling of pores in each slice
The curve models generated in (6) were used for 3-D solid modelling
of pores, as shown in Fig. 10.28. It was noticed that due to the large
number of pores and irregularity of the pore shapes of a typical porous
copper, the file sizes of 3-D models of porous materials can be very
large. For example, the file size of a 3-D solid model of each slice of
this sample is about 100 MB. So it would be advisable to generate a
3-D model of a small piece. This model can be used to build the 3-D
model of the real product. The 3-D solid model of a typical slice is
shown in Fig. 10.29.
10.5.3.4. Summary
A method for generating 3-D models of porous materials has been
described. The experiment results show that the proposed method can
be used to build a 3-D model of the porous materials with a large range
of pore sizes and vey irregular pore shapes. However, this is a time-
consuming process. More work should be done to make the process as
automated as possible to reduce the time required for 3-D modelling
of porous metals.
Fig. 10.28. 3-D solid model of pores in a typical slice.
Fig. 10.29. 3-D solid model of a typical slice.

10.6. Discussion of Problems and Future

Research Interests Areas
As discussed above, there have been many different kinds of 3-D
modelling techniques and systems available on the market. They can
be applied to biomimetics research based on their accuracy, resolution,
operationability, cost and efficiency, etc. However, they have been
designed mainly for industrial, medical and animation applications.
So there are various limitations when they are directly applied to
biomimetics research and applications. This section will discuss these
limitations and try to list future research areas in 3-D geometrical
modelling of animal, insects and plants for biomimetics.
10.6.1. Limitations of existing 3-D systems

The main limitations of the most existing 3-D geometrical modelling
systems are:
10.6.1.1. 3-D Scanners

(1) The ranges of the accuracies and resolutions of most industrial
3-D scanners are not wide enough to meet all the requirements
of biomimetics research and applications. The typical accuracy
and resolution are within the range of 25 µm to 200 µm.
3-D geometrical modelling of many biological surfaces needs the
accuracy to be as high as from 0.5 µm to 10 µm and the resolution
as high as 0.2 µm to 1µm to secure the accurate geometrical
descriptions and acquisition of critical geometrical information at
local areas such as sharp tip points of insect tarsus, boundary of
convex spherical types of tiny features on the forehead of the female
dung beetle. The accuracy of some industrial CMM (co-ordinate
measuring machines) can be as high as ≤1 µm, but this accuracy
is for single-dimensional measurement, not for area scanning, and
the environment conditions are very strict.
(2) Though they have high accuracy and resolution, the workable
ranges/areas/spaces are not large enough for most scanners
developed for nano-technology study and medical applications.
The effective scanning area is typically within several square mm

for one setting, so it is very difficult to use this kind of scanner to
capture surface geometrical information. If doing multi-scanning,
the point clouds alignment or image connection will bring in
errors, which makes the whole process low accuracy.
(3) The measurement platform is not flexible enough for captur-
ing complex body geometry of biological objects. The sample
table/platform/bed of most scanners with high accuracy and
resolution are fixed horizontally or with very limited mobility
(Renishaw, Zeus, MicroScope, MRI, etc). To achieve complete
data acquisition, extra fixtures have to be used on the table to
allow the biological body to be moved to the required positions
and orientations. So, the fixtures have to be very accurate, which
leads to either long time delay or unbearable costs. This is especially
true for a single scanning of an individual surface, which is mostly
the case in biomimetics research.
10.6.1.2. Raw data pre-processing

(1) Noise filtering — 3-D modelling software provides modules
for noise filtering. However, the function is very limited and
manual operations have to be included to get a “clean” point
cloud suitable for the late processing and modelling operation.
This could be a big potential problem for 3-D biological body
modelling since morphological surfaces of biological bodies are
rarely regular primitives, even not smooth. Actually, one of the
promising areas in biomimetics is the study of geometrically non-
smooth features of biological systems to design man-made systems
with high performances such as the non-smooth surface of some
fighters, non-smooth earth-moving components and low-water-
resistance swimming suits. To design and produce morphological
non-smooth systems, it is necessary to have an accurate description
of geometrical non-smooth features of the biological body. These
features are normally in very small scale and in large number. The
noise filtering of such a point cloud is very challenging. As far as
the authors are aware, there isn’t any commercial software that is
suitable for this task.
(2) Sampling — Sampling techniques are relatively more mature

than those for edge detection and segmentation. However, most
sampling algorithms were developed for industrial and medical
and small-scale non-biotechnology applications. So they are not
functionally sufficient and efficiently high enough for 3-D geomet-
rical modelling in biomimetics research. For example, it is really
a challenge to sample a point cloud of a non-smooth surface of
a biological body, as discussed in 1), so that the sampled point
cloud has the desired point distribution of high density along
edges/boundaries areas and low density away from edges.
10.6.1.3. Segmentation and edge detection

(1) Edge detection is one of most desired techniques in reverse
engineering. Unfortunately, there have not been reports of fully
reliable software for autonomous edge detection, though efforts
have never been stopped.54−56
(2) There is also no reliable autonomous software segmentation,
especially for point clouds, though a considerable amount of work
has been done.56 It is very important to properly segment the point
cloud of a biological body into sub-clouds so that each sub-cloud
represents just a unique feature that is suitable for biomimetics
study.
(3) For image-based 3-D modelling, especially for medical and ani-
mation applications, the techniques for edge detection and seg-
mentation are more promising that point cloud based. This is
due to the nature of data information and the methods for image
data acquisition. But the accuracy could be a problem for many
biomimetics studies.
10.6.1.4. 3-D model creation

Errors can be introduced in 3-D model creation. This is especially
true for 3-D modelling of complex biological bodies with non-smooth
features.8,41
10.6.2. Limitations with data interpretation

and sharing
(1) Most software systems output the 3-D models in the formats
that are very difficult for biomimetics researchers to understand.
So even if some models are built, their potential cannot be fully
exploited for biomimetics.
(2) Most software systems are Windows based, so 3-D models built by
different people cannot be shared by others in an efficient way.
10.6.3. Research interests in the areas of 3-D

modelling of biological objects
10.6.3.1. Analysis and synthesis of functions of biological
systems
It is critical to understand the underlying mechanics of biological sys-
tems for biomimetics research. Functional analysis and synthesis should
be the first step for biomimetics research and applications. Simply
copying geometrical features is not really what biomimetics research
means. There are functional, structural, biologically evolutional reasons
why biological systems exist in their existing shapes. The relationships
between shapes and functions have to be established. Considerable
work has been done. In particular, biologists have been making valuable
efforts in areas such as functional, engineering and constructional
morphology.112,113 However, the interests of biologists are more in
the explanation of why biological systems have those functions rather
than from an engineering perspective. Besides, in nature most parts of
organisms do two or more jobs. Usually our technology is interested
in only a single function. So it is not scientifically meaningful and
economically effective to simply learn the research results of biologists’
work and copy biological systems for innovative product introduction.
It is important to analyze and synthesize functions of biological systems
from the biomimetics point of view based upon the biologists’ work.
This work should address issues such as:
(1) What functions of biological systems are economically meaningful

for biomimetics study?
(2) Among all body shapes, what geometrical features are critical for
those biological functions?
(3) The relationships between functions and shape geometry.
(4) Besides geometrical factors, are there any other important factors to
support and dictate those functions? These factors can be chemical,
physical and material properties, etc.
(5) All findings should be stored in data formats that are accessible
to most computer systems (platform and computer software). The
information should be easily interpreted by both biologists and
engineers in order to do further analysis, simulation and even
animation.
This work demands close collaboration between biologists, engineers,

mathematicians, their colleagues in the fields of computing, chemistry
and physics, as well as other related professions, coordinated by people
who have expert knowledge and experience in biomimetics. Some
valuable efforts have been made.114
The work in this area should lead to the requirements of 3-D
modelling of biological systems. Documents should be organized to
answer questions like what kinds of 3-D model are required and in
which way the geometrical information is recorded and stored.
10.6.3.2. Methodology for 3-D modelling of biological

objects
Based on the requirements of 3-D biological objects in Section 10.3.1,
the available 3-D scanning technologies/scanners, raw data processing
algorithms, 3-D geometry computation theory and computer graphical
visualization methods, work should be carried out to investigate
all computer-based enabling 3-D modelling technologies to propose
methodologies that can be used as a guideline (not a bible) for
biomimetics researchers. However due to the fact that most existing
3-D modelling technologies have been developed for industrial, med-
ical and animation purposes, one is very likely to reach the conclusion
that it is impractical, if not impossible, to implement the proposed
methodologies using just the existing technologies available on the
market, which would demand new research and development work
leading to other research interests areas that will be addressed in the

following sections.
The results of methodology proposition work should lead not only
to some diagrams to depict the procedures for each methodology and
documents to explain what should be done and how to do them, but
also to the requirement documents for each technology used in various
steps, their availability on the market, and study and development of
new technologies.
10.6.3.3. New 3-D scanning system (scanners and

associated software)
From biologists’ functional analysis of biological systems, we know
some functions can be used as a clue/hint to or a “natural prototype”
of an innovative product introduction. Then we begin to analyze this
special function again from an engineering point of view. We often
look around for a method of capturing geometrical information of the
biological system. 3-D scanners are very likely our choice.
At the moment, it is technologically possible and economically
viable to configure and build a 3-D scanner or purchase it from the
commercial market. As seen in the discussion in Section 10.3, there are
many 3-D scanners and associated software on the market. But they
have been mainly built for industrial, medical and animation purposes.
Therefore, they are suitable only for some biomimetics research. The
functional limitations of existing 3-D scanners often lead to difficult
situations for biomimetics research and applications.
As discussed in Section 10.6.1.1, the problems are typically either:
(1) the workable area is big enough for a single setting/scanning but
accuracy and resolution is low (most scanners industrial and animation
and some scanners for medical applications) or (2) accuracy and res-
olution are high, but workable areas are too small (a few industrial
scanners, some medical ones and most nano-technology ones). As far
as the authors are aware, there are very few 3-D scanners available on the
market that have been designed and developed specially for biomimetics
research purposes. So we believe that it is necessary to design and
develop 3-D scanning systems suitable for biomimetics research. The
functional performance should be:
(1) The scanner’s accuracy is probably as high as 0.5 µm to 1 µm.

(2) Ideally, scanners can capture both geometrical and color informa-
tion with options choices for users.
(3) The scanning resolution is probably as high as 0.2 µm to 1 µm
with adjustable ranges.
(4) The effective measurement ranges in x-, y- and z-directions should
be large enough to capture most typical features of biological
systems. It is arguable that the effective measurable areas in the
x-y plane should not be less than 50 mm × 50 mm.
(5) The scanning force contact scanners can be pre-adjusted to low
level to avoid deformation of body during scanning.
(6) The data can be exported to the other software systems either in
their native formats or neutral formats such as IGES and STEP, etc.
(7) The technology used in the new systems should be reliable; special
consideration should be given to various light source units (bulbs,
for example) and mechanical adjustment mechanisms.
10.6.3.4. 3-D computation theory and 3-D modelling

software
In the same way, most existing computation theory and mathematical
models for 3-D modelling are suitable for industrial, medical and
animation applications. They often require only point clouds and their
polygonization for medical and animation applications. Bezier patches
and NURBS mathematical models are suitable for a considerable
portion of 3-D modelling of biological systems. But, as far as the
authors are aware, there is little report of suitability studies to answer the
question of which spatial geometry computation methods are suitable
for 3-D modelling of which kinds of biological surfaces. Study in this
area should cover:
(1) Geometrical feature analysis of typical biological surfaces to clas-

sify them into various primitive geometry features; the primitive
features can be either as simple as those used in industry or a
combination of them, or defined by using mathematical models
such Bezier curves and NURBS.
(2) Performance analysis of typical mathematical models such as
Bezier and NURBS for 3-D modelling of biological systems with
particularly features. For example, sharp tip corner of tarsus of

insects.
(3) New methods for raw data pre-processing including noise filtering,
point cloud sampling, and segmentation. These new methods
should address the problems discussed in Section 10.5.1.
(4) Web-based mathematical modelling and presentation techniques.
(5) Theoretical study of a feature-based methodology for 3-D mod-
elling of biological systems.
(6) Software should be designed and developed for 3-D modelling
of biological systems. Besides the normal functions of most typical
3-D modelling software, some special modules should be included:
(a) a data base that is accessible to the biomimetics community to
store the pre-built class library (multi-media database) of modules
of 3-D models of those primitive features defined in 1), (b) an
animation module, (c) a module for outputting the explicit form
of single curve and mathematical model based upon the user’s
instruction, and (d) a module for publication of geometrical
information generated onto an information sharing portal over the
World Wide Web.
(7) Software implementation should be with characteristics of platform
independence, neutral file formats and ideally industrial standard
(like XML) for information sharing.
10.7. Conclusion
To learn from nature, one of the fundamental issues is to understand

the natural systems such as animals, insects, plants and human beings,
etc. The geometrical characterization and representation of natural
systems is an important fundamental work for biomimetics research.
3-D modelling plays a key role in geometrical characterization and
representation, especially in computer graphical visualization. This
chapter has presented the typical procedure of 3-D modelling methods
and then reviewed the previous work of 3-D geometrical modelling
techniques and systems developed for industrial, medical and animation
applications. In particular, the chapter has discussed the problems
associated with existing techniques and systems when they are applied
to 3-D modelling of biological systems. Three case studies have been

presented to illustrate the applications of 3-D modelling to biomimetics
research. Based upon the discussions, the chapter has proposed some
areas of future research interest in 3-D modelling of biological systems
and for biomimetics.
References
[1] Vincent JFV. (2001) Stealing Ideas from Nature. In: Pellegrino S (ed.),
Deployable Structures. Springer, Vienna, pp. 51–58.
[2] Vincent JFV, Bogatyreva OA, Bogatyrev NR, et al. (2006) Biomimetics: Its
practice and theory. Interface 3: 471–482.
[3] Zhang S, Hapeshi K, Ashok K, Bhattacharya AK. (2004) 3D Modelling of
Biological Systems for Biomimetics. Journal of Bionics Engineering 1: 20–40.
[4] Tong J, Guo ZJ, Ren LQ, Chen BC. (2003) Curvature features of three
soil-burrowing animal claws and their potential applications in soil-engaging
components. International Agricultural Engineering Journal 12: 119–130.
[5] Ren LQ, Zhang L, Tong J, Shi YW. (2001) GA-based analysis of the
fractalness of soil-burrowing animal body surfaces. International Agricultural
Engineering Journal 10: 13–27.
[6] Kobayashi H, Kresling B, Vincent JFV. (1998) The geometry of unfolding tree
leaves. Proc R Soc Land 5: 147–154.
[7] Tong J, Sun JY, Chen DH, Zhang SJ. (2005) Geometrical features and
wettability of dung beetles and potential biomimetic engineering applications
in tillage implements. Soil and Tillage Research 80: 1–12.
[8] Li SW, Tong J, Zhang SJ, Chen BC. (2004) Geometrical modelling
of the exterior configuration of a cattle hoof by reverse engineering
technique. Transactions of the Chinese Society of Agricultural Engineering 20:
156–160.
[9] Haas F. (1994) Geometry and mechanics of hind-wing folding in Dermaptera
and Colroptera.M Phil thesis, University of Exeter, UK.
[10] Herbig A, Kull U. (1991) Leaves and ramification. Proc 2nd Int Symp. SFB
230 Part 2, Stuttgart, Germany, pp. l09–117.
[11] Kresling B. (1991) Folded structures in natural lesson in design. Proc 2nd Int
Symp. SFB 230 Part 2, Stuttgart, Germany, pp. 155–161.
[12] Kesel AB. (1994) The insect wing — a multifunctional mechanical system. Proc
3rd Int Symp. SFB 230, Stuttgart, Germany, pp. 181–184.
[13] Wootton RJ. (1981) Support and deformability in insect wings. J Zoot Land
193: 447–468.
[14] Okabe M. (2003) 3D modelling of trees from freehand sketch. BSc Thesis,
University of Tokyo, Japan.
[15] Deussen O, Lintermann B. (1999) Interactive modelling of plants. IEEE

Computer Graphics and Applications 19: 56–65.
[16] Mech R, Prusinkiewicz P. (1996) Visual models of plants interacting with their
environment. Proceedings of SIGGRAPH 96: 397–410.
[17] Ghanem TA, Rabbitt RD, Tresco PA. (1998) Three-dimensional reconstruc-
tion of the membranous vestibular labyrinth in the toadfish. Opsanus tau.
Hearing Research 124: 27–43.
[18] Angelo J, Beraldin JA, Blais F, et al. (2000) Real world modelling through
high resolution digital 3D imaging of objects and structures. ISPRS Journal of
Photogrammetry and Remote Sensing 55: 230–250.
[19] Gledhill D, Tian GY, Taylor D, Clarke D. (2003) Panoramic imaging — a
review. Computers and Graphics 27: 435–445.
[20] Huang D, Yan H. (2003) NURBS curve controlled modelling for facial
animation. Computers and Graphics 27: 373–385.
[21] Craig G, Cramer KE, Cody DD, et al. (1999) Premature partial closure and
other deformities of the growth plate: MR Imaging and Three-dimensional
Modeling. Radiology 210: 835–843.
[22] Kim JS, Park TS, Park SB, et al. (2000) Measurement of femoral neck
anteversion in 3D, Part 1: 3D imaging method. Medical and Biological
Engineering and Computing 38: 603–609.
[23] Kim JS, Park TS, Park SB, et al. (2000) Measurement of femoral neck
anteversion in 3D, Part 2: 3D modelling method. Medical and Biological
Engineering and Computing 38: 610–616.
[24] Gruen A, Murai SJ. (2002) High-resolution 3D modelling and visualization
of mount Everest. ISPRS Journal of Photogrammetry and Remote Sensing 57:
102–113.
[25] Swingen CM, Seethamraju RT, Jerosch-Herold M. (2003) Feed-back-assisted
three-dimensional reconstruction of the left ventricle with MRI. Journal of
Magnetic Resonance Imaging 17: 528–537.
[26] Pollefeys M, Koch R, Vergauwen M, Van Gool L. (2000) Automated
reconstruction of 3D scenes from sequences of image. ISPRS Journal of
Photogrammetry and Remote Sensing 55: 251–267.
[27] Sarti A, Gori R, Lamberti C. (1999) A physically based model to simulate
maxillo-facial surgery from 3D CT images. Future Generation Computer Systems
15: 217–221.
[28] Han M, Kanade T. (2002) A perspective factorization method for Euclidean
reconstruction with uncalibrated cameras. The Journal of Visualization and
Computer Animation 13: 211–223.
[29] Streicher J, Miiller GB. (2001) 3D modelling of gene expression patterns.
TRENDS in Biotechnology 19: 145–148.
[30] Behr J, Choi SM, Grokopf S, et al. (2000) 3D modelling for diagnosis and
treatment planning in cardiology. Radiologe 40: 256–261.
[31] Kehl HG, Jager J, Papazis N, et al. (2000) 3D heart modelling from biplane,
rotational angiocardiographic X-ray sequences. Computers and Graphic 24:
731–739.
[32] Courrioux, G, Nullans, S, Guillen, A, et al. (2001) 3D volumetric modelling
of Cadomian terrains (Nonhem Brittany, France): an automatic method using
Voronoi diagrams. Tectonophysics 333: 181–196.
[33] Zhang X, Zhang S, Hapeshi K. (2014) A New Parameterised Feature-based
Generic 3D Human Face Model for Emotional Bio-robots. Appl Mech Mater
461: 838–847.
[34] Sun J, Tong J, Chen D, Zhang S. (2014) Quasi-static and dynamic nanoin-
dentation of some selected biomaterials. Journal of Bionics Engineering 11:
114–150.
[35] Li M, Chen D, Zhang S, Tong J. (2014) Biomimetic Design of a Stubble-
Cutting Disc Using Finite Element Analysis. Journal of Bionics Engineering
10: 118–127.
[36] Zhang X, Zhang S, Hapeshi K. (2010) A New Method for Face Detection in
Colour Images for Emotional Bio-robots. Science China, Technological Science
53: 2983–2988
[37] Zang H, Zhang S, Hapeshi K. (2010) A Review of Nature-Inspired Algorithms.
Journal of Bionics Engineering 7S: s232–s237.
[38] Wen H, Zhang S, Hapeshi K, Wang X. (2008) A Methodology for Innovative
Product Design: Inspiration from Nature. Journal of Bionic Engineering 5:
75–84.
[39] Zhang S, Tong J, Hapeshi K, et al. (2007) Optimising Body Layout Design of
Limbed Machines — Observation from Nature and Theoretical Investigation,
Journal of Bionic Engineering 4: 117–122.
[40] Tamas V, Ralph RM, Jordan C. (1997) Reverse engineering of geometric
models an introduction. Computer Aided Design 29: 225–268.
[41] Zhang S, Raja V, Lim W, et al. (2002) Error analysis of data acquisition
of reverse engineering process using design of experiment. In: Cheng K,
Webb O (eds.), Advanced Manufacturing Technology XVI. Professional Engi-
neering Publishing Limited, pp. 145–150.
[42] Renishaw. http://www.renishaw.com/client/product/UKEnglish/PGP-890.
shtml (Accessed on 11-06-2015.)
[43] Wicks and Wilson. http://www.wwl.co.uk/ (Accessed on 11-06-2015.)
[44] 3D Cafe. http://www.3dcafestore.com/3dmodels1.html (Accessed on 11-06-
2015.)
[45] 3Dlenta. http://3dlenta.com/en/plants.html (Accessed on 11-06-2015.)
[46] Anderson RC. (1982) Photograrnmetry: the pros and cons for archaeology.
World Archaeology 14: 200.
[47] Jeyapalan K. (1980) Photogrammetry in recording the historic ship iantic.
Photogrammetric Engineering and Remote Sensing 46: 15–31.
[48] Schenk I. (2005) Introduction to Photogrammetry. http://www.mat.uc.

pt/∼gil/downloads/IntroPhoto.pdf (Accessed 12-06-2015.)
[49] Urquhart CW, Sieben JP. (1993) Towards real-time dynamic close range
photogrammetry. Proc SPIE Videometrics II 2067: 240–251.
[50] Sieben JP, Marshall SJ. (2000) Human body 30 imaging by speckle texture
projection photogrammetry. Sensor Review 20: 218–226.
[51] William J, Wing N. Geometric Analysis, Visualization, and Conceptualization
of 3D Image Data. Imaging and Distributed Computing Group, Information
and Computing Sciences Division, Lawrence Berkeley Laboratory, Publication
number: LBL-35329. http://froggy.lbl.gov/papers/Reports/LBL.35329.
html (Accessed 05-06-2015.)
[52] Zhang S, Hapeshi H. (2014) 3D Geometry Modelling of Pores Copper
Materials. Industrial Report, The University of Gloucestershire.
[53] Zhang S, Raja V, Ryall C, Wimpenny D. (2003) Rapid prototyping models
and their quality evaluation using reverse engineering. Journal of Mechanical
Engineering Science 217(C): 81–96 .
[54] Jun YT, Raja VH. (2002) Extracting geometric attributes directly from scanned
point set for feature recognition. Int J of Computer Integrated Manufacturing
15: 50–61.
[55] Jun YT, Raja VH, Park S. (2001) Geometric feature recognition for reverse
engineering using neural networks. Int J of Advanced Manufacturing Technol-
ogy 17: 462–470.
[56] Park S, Jun YT. (2001) Automated segmentation of scanned point data in
feature-based reverse engineering system. 2001 I nternational CIRP Design
Seminar. Stockholm, Sweden, pp. 353–358.
[57] Lindeberg T. (1998) Edge detection and ridge detection with automatic scale
selection, International Journal of Computer Vision 30: 117–154.
[58] Ziou D, Tabbone S. (1998) Edge detection techniques: An overview. Interna-
tional Journal of Pattern Recognition and Image Analysis 8: 537–559.
[59] Canny J. (1986) A computational approach to edge detection. IEEE Trans
Pattern Analysis and Machine Intelligence 8: 679–714.
[60] Imageware Surfacer. http://www.dipro.co.jp/english/products/cad_cam/
imageware/productfamily/index.html (Accessed 11-06-2015.)
[61] Raindrop Geomagic Studio. http://www.geomagic.com (Accessed 11-06-
2015.)
[62] Faro 3D X330. http://www.faro.com/products/3d-surveying (Accessed 11-
06-2015.)
[63] Mullins J. (1997) Secrets of a Perfect Skin: Sharks do it, dolphins do it —
now even aircraft engineers are managing to banish turbulence and drag. New
Scientist 2065: 28.
[64] Ridgway SH, Carder DA. (1993) Features of dolphin skin with potential
hydrodynamic importance. IEEE Engineering in Medicine and Biology 12:
83–88.
[65] Zhang S, Howard D, Sanger D, Miao S. (1998) Multi-legged walking machine

body designs. Robotica 16: 593–598.
[66] Howard D, Zhang S, Sanger DJ. (1998) Multilegged walker design — the
joint torque versus workspace compromise. Journal of Mechanical Engineering
Science 212 (C): 477–488.
[67] Chung A. (2002) Bionics information exchange portal system. MSc Thesis,
University of Warwick, UK.
[68] Mitutoyo. http://www.mitutoyo.co.uk (Accessed 11-06-2015.)
[69] http://what-when-how.com/digital-imaging-for-cultural-heritage-preservat-
ion/processing-sampled-3d-data-reconstruction-and-visualization-technolog-
ies-digital-imaging-part-1 (Accessed 11-06-2015.)
[70] Reid GT, Rixon RC, Stewart H. (1998) Stripe scanning for engineering. Sensor
Review 8(2): 67–71.
[71] Reid GT, Marshall SJ, Rixon RC, Stewart H. (1988) A laser scanning camera
for range data acquisition. Phys D: Appl Phys 21: S1–S3.
[72] Reid GT, Rixon RC, Stewart H. (1988) Moire topography with large contour
intervals. Proc SPJE 814: 307–313.
[73] Isdale J. (1998) 3D scanner technology review. http://www.isdale.com/jerry/
VR/3DScanners/3DScannerReview.html (Accessed 11-06-2015.)
[74] Stanford School of Medicine, 3D Reconstruction. http://biocomp.stanford.
edu/3dreconstruction (Accessed 11-06-2015.)
[75] Stevens JK, Mills LR, Trogadis JE. (1994) Three-dimensional Confocal
Microscopy: Volume Investigation of Biological Systems. Academic Press, London.
[76] MicroScribe. http://www.3d-microscribe.com/G2%20Page.htm (Accessed
11-06-2015.)
[77] Photogrammetry. http://www.realitymeasurements.com/crdp-low-altitude.
htm (Accessed 11-06-2015.)
[78] Gerald EF. (1988) Curves and Surfaces for Computer Aided Geometric Design:
A Practical Guide. Academic Press Inc, London.
[79] Gerald F. (1993) Curves and Surfaces for CAGD (a Practical Guide), 3rd ed.
Academic Press Inc, London.
[80] Josef H, Dieter L. (1993) Fundamentals of Computer Aided Geometric Design.
A. K. Peters, Ltd, USA.
[81] Huang D, Yan H. (2003) NURBS curve controlled modelling for facial
animation. Computers and Graphics 27: 373–385.
[82] Piegl L, Tiller W. (1997) The NURBS Book, 2nd ed. Springer, Berlin.
[83] Wen H, Zhang S, Hapeshi H, Wang X. (2008) A Methodology for Innovative
Product Design: Inspiration from Nature. Journal of Bionic Engineering 5:
75–84.
[84] Cross N. (1984) Developments in Design Methodology. John Wiley, New York,
USA.
[85] Edmonds E, Soufi B. (1992) The computational modelling of emergent shapes

in design. In: Gero JS, Sudweeks F (eds.), Computational Models of Creative
Design. University of Sydney, Australia.
[86] Raja V, Zhang SJ, Garside J, et al. (2006) Rapid and cost-effective design and
manufacturing of high-integrity aerospace components. International Journal
of Advanced Manufacturing Technology 27: 759–773.
[87] Zhang X, Zhang S, Hapeshi K. (2014) A New Parameterised Feature-based
Generic 3D Human Face Model for Emotional Bio-robots. Applied Mechanics
and Materials 461: 838–847.
[88] Hashimoto T, Hiramatsu S, Kobayashi H. (2006) Development of face
robot foremotional communication between human and robot. Proceedings of
IEEE International Conferenceon Mechatronics and Automation (ICMA 2006),
Luoyang, China, pp. 25–30.
[89] Takahashi Y, Hatakeyama M, Kanno M. (2007) Experiments on human facial
expressions for improvement of simplified robot face. Annual Conference SICE,
Takamatsu, Japan, pp. 480–483.
[90] Takahashi Y, Sato H. (2010) Compact robot face with simple mechanical
components. Proceedings of International Conference on Control Automation
and Systems (ICCAS 2010), Gyeonggi-do, South Korea, pp. 2300–2303.
[91] Waters K, Terzopoulos D. (1991) Modelling and animating faces using
scanned data. The Journal of Visualization and Computer Animation 2:
123–128.
[92] Yongli H, Mingquan Z, Zhongke W. (2009) An automatic non-rigid point
matching method for dense 3D face scans. Proceedings of International
Conference on Computational Science and Its Applications (ICCSA 2009),
Suwon, Korea, pp. 215–221.
[93] Tognola G, Parazzini M, Svelto C, et al. (2003) A fast and reliable system for
3D surface acquisition and reconstruction. Image and Vision Computing 21:
pp. 295–305.
[94] Guenter B, Grimm C, Malvar H, Wood D. (1998) Making faces. Proceedings of
the 25thAnnual Conference on Computer Graphics and Interactive Techniques,
Orlando, FL, USA, pp. 55–66.
[95] Rydfalk M. (1978) CANDIDE, a parameterized face. PhD thesis, Linköping
University, Linköping, Sweden.
[96] Ahlberg J. (2001) CANDIDE-3 — an updated parameterised face. Report
No. LiTH-ISY-R-2326, Dept. of Electrical Engineering, Linköping University,
Sweden.
[97] Welsh B. (1991) Model-based coding of videophone images. Electronics &
Communication Engineering Journal 3: 29–36.
[98] Mayer C, Wimmer M, Stulp F, et al. (2008) A real time system for model-based
interpretation of the dynamics of facial expressions. Proceedings of 8th IEEE
International Conference on Automatic Face & Gesture Recognition (FG’08),

Amsterdam, Netherlands, pp. 1–2.
[99] Blanz V, Vetter T. (1999) A morphable model for the synthesis of 3D
faces. Proceedings of the 26th Annual Conference on Computer Graphics and
Interactive Techniques, Los Angeles, USA, pp. 187–194.
[100] Davies A, Ek CH, Dalton C, Campbell N. (2012) Generating 3D morphable
model parameters for facial tracking: factorising identity and expression,
Proceedings of the International Conference on Computer Graphics Theory and
Applications and International Conference on Information Visualization Theory
and Applications, Rome, Italy, pp. 309–318.
[101] Au BL, Ly QN, Minh HP. (2012) An individualized system for face ageing
and facial expressions modeling based on 3D practical faces data. Proceedings
of International Conference on Control, Automation and Information Sciences
(ICCAIS), Saigon, Vietnam, pp. 398–403.
[102] Ahyoung S, Seong-Whan L, Bulthoff H, Wallraven C. (2012) A morphable
3D-model of Korean faces. Proceedings of IEEE International Conference on
Systems, Man, and Cybernetics (SMC), Seoul, Korea, pp. 2283–2288.
[103] He Y, Qin H. (2004) Surface reconstruction with triangular b-splines. Proceed-
ings of Geometric Modeling and Processing, Beijing, China, pp. 279–287.
[104] Larrabee WF, Makielski KH, Henderson JL. (2004) Surgical Anatomy of the
Face. Lippincott Williams & Wilkins, Philadelphia, USA.
[105] Rose EH. (1998) Aesthetic Facial Restoration. Lippincott-Raven, New York,
USA.
[106] Sun N, Ayabe T, Nishizaki T. (2007) Efficient spline interpolation curve
modeling. Proceedings of the 3rd International Conference on Intelligent
Information Hiding and Multimedia Signal Processing, Splendor Kaohsiung,
Taiwan, 2: 59–62.
[107] Hartley PJ, Judd CJ. (1978) Parametrization of Bézier-type b-spline curves
and surfaces. Computer Aided Design 10: 130–134.
[108] Boor CD. (1972) On Calculating with B-Splines. Journal of Approximation
Theory 6: 50–62.
[109] Zhang X, Zhang S, Hapeshi K. (2010) A new method for face detection in
colour images for emotional bio-robots. Science China Technological Sciences
53: 2983–2988.
[110] Meille S, Garboczi EJ. (2001) Linear elastic properties of 2D and 3D models
of porous materials made from elongated objects. Modelling Simul Mater Sci
Eng 9: 371.
[111] Karageorgiou V, Kaplan D. (2005) Porosity of 3D biomaterial scaffolds and
osteogenesis. Biomaterials 26: 5474–5491.
[112] Gudo M, Gutmann M, Scholz J. (2002). Concepts of functional, engineer-
ing and constructional morphology: introductory remarks. Senckenbergiana
lethaea 82: 7–10.
[113] Schmidt K, Vogel KPJ. (1991) Constructional Morphology and Evolution.

Springer, Berlin, Heidelberg, New York, Tokyo.
[114] Vincent JFV, Mann DL. (2002) Systematic technology transfer from biology
to engineering. Phil Trans R Soc A 360: 159–174.
Chapter 11
Superhydrophobic Surfaces
with Hierarchical Structures
Inspired by Nature Leaves
Yuying Yan∗ and Nan Gao
Faculty of Engineering
University of Nottingham, UK
∗ yuying.yan@nottingham.ac.uk
Many superhydrophobic surfaces (both manmade and natural) nor-

mally exhibit micro- or nanosized roughness as well as hierarchical
structure, which somehow can influence the surface’s water repellence.
In this chapter, the recent progress in preparing manmade superhy-
drophobic surfaces through a variety of methodologies, particularly
within the past several years, and the fundamental theories of wetting
phenomena related to superhydrophobic surfaces are reviewed, and the
perspective of natural superhydrophobic surfaces utilized as mimicking
models are discussed. The physics related to superhydrophobic surfaces
is also discussed with attention of its potential applications. A facile
method for preparing superhydrophobic surfaces based on micro and
nano scaled structures is reported. Composite thin films are formed
by using SiO2 nanoparticles and poly(dimethylsiloxane) (PDMS). It is
demonstrated that the hierarchical structure in micro and nano scale on
the surface, plays an important role in prompting the superhydrophobic
(water-repelling) properties. Wetting phenomena and related theories
are also discussed within the paper.
393
11.1. Introduction
The superhydrophobicity (hydrophobicity) of solid surfaces has been
investigated with considerable attention over the past few years and
remarkable progress has been achieved. It has been discovered that
water droplets on hydrophobic surfaces can exhibit a contact angle
higher than 90◦ , with some even approaching 180◦ .1−3 In particular,
the contact angles related to superhydrophobic (or ultrahydrophobic)
surfaces are greater than 150◦ . And those superhydrophobic surfaces
are very likely to have phenomenal roughness with micro- or nanosized
(or even smaller) protrusions coming out of the surface.4,5 Therefore,
the liquid might contact only a few bits of the superhydrophobic surface
without fully wetting it. Indeed, fluid interacting with superhydropho-
bic surfaces is one important discipline of research in the 21st century,
and can essentially influence a lot of cutting-edge topics in engineering
and biotech research which involve surface structures, fluid motivation,
and their physical and chemical properties. Basically, contact angle
related wetting phenomena are of great interest and importance to
current research progress. A considerable amount of work has been
carried out to study the involved mechanisms and principles, and
these have been reviewed in detail recently.6,7 Interestingly, many
methods that are used to create manmade superhydrophobic surfaces
are inspired by the “Lotus Effect.”8 Recent research has indicated that
there are even more superhydrophobic surfaces in nature9,10−12 ; and
this helps to promote the applications of biomimetic ideas into practical
fields.13−16
11.2. Theoretical Models
The physics related to superhydrophobic (or ultrahydrophobic) sur-

faces has been intensively investigated over the past decade, and
remarkable progress has been achieved. The starting point of a wetting
model is defined by the Young equation, derived for a sessile drop on
an ideal rigid, homogeneous, flat and inert surface. The drop contacts
its substrate on a disc (radius = R) where the three phases of the system
Superhydrophobic Surfaces with Hierarchical Structures Inspired by Nature Leaves 395
Fig. 11.1. A liquid drop balanced by three interfaces. The letter ‘A’ indicates the
interfaces as well as their contact areas. The contact angle θ is represented by its
supplementary angle.
coexist, and the so-called three phase contact line is formed, as shown
in Fig. 11.1.
The liquid joins the solid at an angle θ, which determines the size
of the contact. Each interface draws the contact line so as to minimize
the corresponding surface area, balancing the surface tensions on the
direction of potential motion so that there will yield a relation attributed
to Young, as shown in Eq. (11.1), although it does not explicitly show
up in Young’s publication17 :
γSV − γSL
cos θ = , (11.1)
γLV
where γ is the surface tension, which indicates the energy per unit
surface area of the interface. SV, SL and LV correspond to the inter-
faces between solid, liquid and vapor, respectively. Young’s equation
unambiguously relates the intrinsic contact angle θ (Young Contact
Angle) to interfacial tensions (energies). However, since real surfaces
usually vary in the surface conditions, most scenarios regarding contact
angles cannot be explained by the Young equation. Wenzel proposed an
equation,18,19 including surface roughness, liquid drop contact angle
and the corresponding surface energies, which can be written as:
∗
r(γSV − γSL ) = γLV cos θW , (11.2)
∗
where θW is the apparent Wenzel contact angle, which measures the
apparent contact angle influenced by the roughness of solid surfaces.
r corresponds to the “roughness factor,” also referred to as the
roughness area ratio of the actual surface with respect to the geometric
surface. This model is called the Wenzel equation, which is normally
expressed as:
∗
cos θW = r cos θ. (11.3)
Wenzel assumed the water would penetrate into the grooves caused
by the surface roughness, and therefore the Wenzel equation is related
to the homogeneous wetting regime.20 The “apparent” Wenzel angle
θ ∗ is in contrast with the “real” angle θ, because close inspection of
a rough surface would always reveal the “real” contact angle on any
element that could be regarded as smooth.21 The measurement of
contact angles is therefore somewhat arbitrary, for one might pursue
this argument to molecular dimensions when the surface would appear
rough and heterogeneous. Fortunately, for practical purposes, it is
possible to have a reasonable division of surfaces into “smooth” and
“rough.” The Wenzel equation states that the wettability is improved
by roughness for a superhydrophilic surface, but gets worse for a
superhydrophobic one.22
There are two types of stable wetting states, namely, homogeneous
wetting state and heterogeneous wetting state, and the metastable states
in-between. While the Wenzel equation applies to the homogeneous
wetting regime, its succeeding theory the Cassie–Baxter equation
corresponds to the heterogeneous wetting regime. If the surface
roughness is illustrated as pillars or protrusions, the two-dimentional
schematics of those wetting states can be shown in Fig. 11.2.
Thus, basically there are two types of wetting states, namely, the
homogeneous wetting state and the heterogeneous wetting state.
A homogeneous state corresponds to the fact that the liquid drop fills up
the roughness grooves (Fig. 11.2a), and the Wenzel equation is applied;
a heterogeneous state refers to the fact that air bubbles are entrapped
inside the grooves underneath the liquid (Fig. 11.2b), and the Cassie–
Baxter (CB) equation is then applied. For very rough hydrophobic
materials, the energy stored for following the solid surface could be
much larger than the energy associated with the air pockets. In this
state, the liquid only contacts the solid at the top of the protrusions,
on a fraction denoted as φS ,23,24 which is the ratio of the total area of the
solid–liquid interface with respect to the total area of solid–liquid and
Fig. 11.2. The homogeneous wetting state (a) and heterogenous wetting state
(b) The liquid drop in the homogeneous wetting state follows the solid surface
and penetrates into the grooves caused by the protrusions; the liquid drop in the
heterogeneous wetting state only contacts the top of the protrusions, leaving air below
into the grooves.
liquid–air interfaces in a plane geometrical area of unity parallel to the

rough surface. If only air were present between the solid and the liquid,
the “contact angle” would be 180◦ — the smaller φS is, the closer this
extreme situation is achieved, and thus the higher the hydrophobicity is.
More precisely, the contact angle θ ∗ of such a “fakir” drop is an average
between the angles on the solid and the air, respectively weighed by
the fractions φS and 1 − φS , which yields17,24 :
cos θ ∗ = −1 + φS (cos θ + 1) = φS · cos θ + φS − 1. (11.4)
θ ∗ monotonously increases as φS decreases, suggesting that φS should
be made as small as possible. However reducing φS also makes the
roughness decrease if one cannot manage to present only air between
the solid and the liquid. Besides, if we turn to the ratio of the actual
wetted area to the projected area, rf , that is to say the roughness
ratio of the fraction, this will give rise to the modified form of the
CB equation 20,25 :
∗
cos θCB = rf · φS · cos θ + φS − 1. (11.5)
When φS = 1, rf = r, the CB equation turns into the Wenzel equation.
Equation (11.5) interprets the multilayered roughness and is more
suitable for the hierarchical surface structure, which has been found
much closer to the natural model.10,26 Thus it is understood that both
surface roughness and its structure (morphology) can have an impact
on the value of contact angles and therefore the wetting states. If the
intrinsic contact angle of water-repelling surfaces is fixed for a particular
material, it will be the way for preparing the surface that determines
the wettability of surfaces.
To sum up briefly for this part, the Young equation builds the
basement for wetting on ideal smooth surfaces. The Wenzel equa-
tion is valid if the surface is homogeneously wetted, while the CB
equation applies to a heterogeneous surface. At this point, it seems
that people have already got sufficient theoretical support to study
those involved scenes. However, the real situation is not that simple.
As recently realized,20,27,28 both Wenzel equation and CB equation
could be correct only if the drop is sufficiently large compared with
the typical roughness scale. It is still under debate when and how a
stable wetting state can be achieved. Also, there might be a transition
between homogeneous state and heterogeneous state, and it is even not
fully understood how this transition happens. Thus, we can say only
that those mentioned theories are to some extent necessary, but not
sufficient for describing wetting phenomena entirely on solid surfaces.
To solve these questions, further investigation related to and crucial to
understand the wetting process is needed. In the following sections,
we will discuss the issues of contact angles and wetting modelling to
interpret the involved mechanisms.
Tsujii and his co-workers conducted a series of experiments testing
the water repellence on fractal surfaces.3,29−31 One typical process was
to prepare fractal surfaces made of alkylketene dimer (AKD) and to
measure the contact angles. The maximum value of contact angles
reported in their work was 174◦ . They argued that the maximum
contact angle could approached 180◦ if there was no adsorption.3
This was a sparking point in a long run of studying contact angles
on superhydrophobic surfaces, for it opened the door of possibility
for succeeding researchers to hypothesize approaching the extremely
high contact angles in practical fields.22,32−45 Actually, afterwards there
were also other cases reported of high contact angles.33,36−43 Some
researchers even reported a contact angle of 180◦ .44,45 Figure 11.3
shows a few reported cases of extremely high contact angles of liquid
droplets on superhydrophobic surfaces.
Fig. 11.3. High contact angles of liquid droplets on artificial superhydrophobic

surfaces. (i) 171.2 ± 1.6◦ , achieved by poly(vinyl alcohol) nanofibers38 ; (ii) 175 ±
0.7◦ , achieved by the coating of SiO2 nanoparticles43 ; (iii) 180◦ , achieved by
submicrometered particles of tetrafluoroethylene oligomers44 ; (iv) 180◦ , achieved by
fluorinated nanoparticles45 ; (v) 175 ∼ 180◦ , achieved by using DRIE patterning of
controlled parameters42 ; (vi) 173(−3, +6)◦ , achieved by using silicon etching and gas
phase isotropic etching.33
However, there was no clear interpretation for the contact angle

hysteresis in Tsujii’s experiment and theory. Also, the lack of completely
consistent measurement standards might make other high contact
angles uncertain to decide. Because of the contact angle hysteresis
(CAH), the maximum value of a measured contact angle is not the
only criterion for defining a stable superhydrophobic state. To describe
a superhydrophobic state, the static angle as well as the contact angle
hysteresis should be measured. For an extremely stable superhydropho-
bic state, its static contact angle should be as high as possible, and its
CAH should be as small as possible, otherwise the achieved wetting
state might transit to a different one.
In addition, the fitting modes that people adopted to measure the
contact angles can affect their values46 : It was reported that ellipse
fitting, circle fitting, tangent searching, and Laplace–Young fitting
caused various values of the contact angle from similarly shaped droplets
as shown in Fig. 11.4. That is the reason why the contact angles of those
liquid drops in Fig. 11.3 just look similar but have different values.
Fig. 11.4. Ellipse fitting (a), circle fitting (b), tangent searching (c), and Laplace–
Young fitting (d) make static contact angles of the same water droplet different. The
figures include the simulation lines of the shape of the water droplets and the horizontal
baselines.42
Therefore, the fitting mode should be clearly mentioned to reflect the

real situation of wetting.
The contact angle hysteresis can be explained schematically in two
manners. As shown in Fig. 11.5, the liquid drop will advance at the
down side and recede at the up side when the substrate is inclined at
θD , which is the sliding (tilt43 or slip44 ) angle that a substrate must
be tilted at/above, in order for a water drop to slide off. In this case,
the liquid wets the substrate at the advancing point and de-wets it at the
receding point. Thus, θA is the advancing contact angle and θR is the
receding contact angle. They stay constant during the sliding process
as long as the surrounding conditions do not change.
Fig. 11.5. (a) A liquid drop theoretically sliding on a declination of θD without

acceleration. θA is the advancing angle, θR is the receding angle, and θD is the sliding
(rolling off or tilt) angle. (b) The schematic of static contact angle and dynamic contact
angles.
Fig. 11.6. (a) The droplet is pinned at the three-phase contact line until θR is reached
at stage 2 and θR remains constant during subsequent volume decreasing (b) The
droplet is pinned at the three-phase contact line until θA reaches stage 6 (c) A growing
and shrinking drop sequence exemplarily.
If one withdraws water from a droplet on the solid surface, the

volume of the droplet as well as the contact angle will decrease, but
the contact area of the droplet on the surface will not change until it
begins to recede, as shown in Fig. 11.6a. Similarly, if one adds water to
the droplet, the volume as well as the contact angle will increase, but
the contact area will not change until the droplet begins to advance,
as shown in Fig. 11.6b. Thus, the contact angles will, because of
the characteristic of the surface chemistry and structure, stay constant
during receding and advancing as well, and they are denoted as receding
contact angle θA and advancing contact angle θR , respectively. The
difference between advancing and receding contact angles is termed
as the Contact Angle Hysteresis.47,48 Corresponding to the description
along with Figs. 11.6a and 11.6b, a typical experimental process of
contact angle hysteresis is shown in Fig. 11.6c. The contact angle of a
metastable droplet formed on a solid surface can be at any value between
receding and advancing contact angles. Some contact angle hysteresis is
quite low, within 5◦ ,39 which does not affect the superhydrophobicity
dramatically. However, some can be as much as 40◦ ,49 which might
even change the wetting state. From this point of view, it is not only the
static and metastable contact angles but also the receding and advancing
contact angles that can characterize a surface.
11.3. Surface Tension, Surface Free Energy

and Wetting Transitions
Although the surface tension is referred to as the free energy per
unit area, it can be equally thought of as the force per unit length,
which is measured in N/m.50 The surface free energy is an important
characteristic of every surface or interface. In the bulk of the body,
chemical bonds exist between the molecules. To break the chemical
bonds, a certain amount of energy has to be applied. Molecules that
do not form bonds at the side of the surface have higher (potential)
energy than those that form the bonds. This additional energy caused
by the higher energy is called surface or interface free energy and is
measured in the energy per area units, that is, in the SI system, J/m2 .51
Thermodynamically, the concepts of surface tension, surface tension
force, surface energy density, and surface free energy are different.52,53
However, these concepts, in most cases studying superhydrophobic or
superhydrophilic surfaces, are now numerically equivalent to each other
when temperature and pressure are assumed constant.54 In addition,
the restriction should also include that the scale of the substance is small
enough to neglect the affect caused by those factors. It is necessary to
assume that there is no adsorption at the interfaces as well. That is the
ideal condition where we say the surface tension is equivalent to the
surface free energy.
Now consider a water drop wetting a solid surface. For a super-

hydrophobic surface, the dry area should have lower surface energy
than the wet, and the shape of the drop then spontaneously becomes
more spherical to minimize its energy. As the area under the drop
becomes de-wetted, a release of energy will occur as the solid–liquid
area decreases; the released energy will promote the liquid–air surface
increasing. It is also understood from this point that the solid surface
will try to minimize its energy if the dry area has lower energy than the
wet, and therefore repel the liquid drop. Surface roughness can adjust
the effect of surface energy as an amplifier. Hence a drop forms a high
contact angle on rough low-energy surfaces and a low contact angle on
rough high-energy surfaces.55
There is no doubt that those obtained high contact angles men-
tioned above were due to the superhydrophobicity that the surfaces
exhibited. People can achieve superhydrophobicity in quite a few ways
(see subsequent sections). However, for most cases, it was not presented
clearly yet that what the wetting state was when the high contact angle
was shaped up. As stated previously, Wenzel and Cassie–Baxter states
might both be suitable for superhydrophobic surfaces; but whether
or not the air can remain trapped below the drop is still under
investigation. The contact angle hysteresis in the CB state is normally
lower than in the Wenzel state, and therefore the CB state can make
a more stable contact angle and wetting state.17 There is an apparent
difference between those two states. If the contact angle hysteresis is
relatively big, a transition might happen, which makes the wetting less
stable or metastable.56,57
Bico et al. suggested a critical contact angle θc between Wenzel
state and CB state by equating Eq. (11.3) with Eq. (11.4).22 This
value is determined by the surface design. That the air is entrapped
into the surface grooves should be favored only if θ is larger than θc .35,58
However, more importantly, this critical contact angle should include
the effect caused by rf , the roughness ratio of projected protrusions or
grooves. Then after equating Eq. (11.3) with Eq. (11.5), one can get:
1 − φS
cos θc = , (11.6)
rf · φ S − r
where θc is the critical contact angle with the effect of rf .60
11.4. Superhydrophobic Surfaces in Nature

The lotus flower is considered a symbol of purity in a few religions.
Its leaves can be kept from contamination or pollution without being
folded even when the lotus is surrounded by muddy water. This
phenomenon illustrates that nature can protect itself from omnipresent
dirt and pathogenic organisms. And ideally, if this property is applied
to functional surfaces, a self-cleaning effect can be prompted in almost
any materials in the open air by rain water. As is mostly accepted,
Barthlott and Neinhuis’ early work started the recent research of
superhydrophobic surfaces inspired by nature.8,60 Surface roughness
has been recalled to explain the surface’s extreme repellence against
liquid droplets, as shown in Fig. 11.7 (a). Subsequent and further
research has indicated that the plant cuticle is technically a composite
material mainly built up by a network of cutin and hydrophobic waxes,
where surface structuring arises at different hierarchical levels.61 The
composite or hierarchical surface structure, formed by a combination
of two (or even more) layers in different sizes, is built by convex
cells and a much smaller superimposed layer of hydrophobic three-
dimensional wax tubules,10 as shown in Fig. 11.7 (b and c) with
different magnifications. It has been argued that wetting of such
surfaces is minimized because air is enclosed in the cavities of convex
cell sculptures.
Along with the lotus leaves, there are other natural superhy-
drophobic surfaces in the plant kingdom. For example, taro (Colocasia
Fig. 11.7. (a) A glycerol drop on Euphorbia myrsinites, which is a robust specimen
and well suited to show the surface’s repellence against the liquid droplet. Scale
bar = 80 µm (b) The upper side surface of the lotus leaf without the shrinkage artifact.
Scale bar = 8 µm (c) The wax tubules from the upper side of the lotus leaf. Scale bar =
1 µm.
esculenta) leaves have been used to demonstrate the self-cleaning effect

in the original paper.8 It has been observed that the elliptic protrusions
with an average diameter of about 10 µm form the microstructure on
the taro leaf, and the nanoscaled pins form the hierarchical structure
with the microstructure, which is however similar to the lotus leaf.63
The surface structure of a colocasia leaf has similar bumps to that of
the lotus, but the bumps are separated by the surrounding ridges. The
bump and ridge both contribute to the hydrophobic nature of colocasia
since they might create air pockets between the water droplet and the
surface. At this point, it is necessary to emphasize that the surface cuticle
and its waxes are important to surface wettability by folding the cuticle
or by forming three-dimensional wax crystals on the plant surface.10
It has been brought into consideration that some natural surfaces
with hierarchical structure and roughness can produce significant
superhydrophobicity, not only from plants but also animals. It has
been revealed that the leg of a water strider has numerous oriented
needle-shaped setae with their diameters ranging from 3 µm down to
several hundred nanometers.63,64 Many elaborate nanoscaled grooves
are noticeable on each microseta, forming a hierarchical structure,
which is the source the superhydrophobicity of the water strider’s legs,
along with the assistance of the hydrophobic secreted wax.
It is becoming clear from the discussions mentioned previously
that the mechanisms of individual superhydrophobic surfaces are not
necessarily identical with each other in every aspect. Unlike the lotus
effect, the so-called petal effect describes the phenomenon that a
water droplet on the petal surface of a red rose (rosea Rehd) forms
a spherical shape, but does not roll off even when the petal is turned
upside down,9 as shown in Fig. 11.8. The diverse design in surface
microstructure on rose petal results in a different kind of dynamic
wetting from lotus leaves. When a small water drop wets the petal
surface, the liquid film impregnates the textured regime. However, the
liquid wets only the grooves between the projected pillars, leaving the
“plateaus” dry, which form the Cassie impregnating state, as shown
in Fig. 11.9. For the petal surfaces, the dimensions of hierarchical
micro- and nanostructures both are found to be larger than those
related to the lotus leaf. In the Cassie impregnating wetting regime,
Fig. 11.8. SEM images (a, b) of the surface of a red rose petal, showing a periodic
array of micropapillae and nanofolds on each papillae top (c) A water droplet on
the petal’s surface, indicating its superhydrophobicity, with a contact angle of 152.4◦
(d) Profile of a water droplet on the petal surface when turned upside down.
water droplets enter into the “large” grooves of the petal but not
into the “small” ones. This indicates why small water drops sealed
in micropapillae would be clinched to the petal’s surface, showing
a high CAH even if the surface was turned upside down. Thus, the
mechanisms for different natural superhydrophobic surfaces and their
wetting states might seem vague to follow. However, the common
characteristics of superhydrophobic surfaces rest on the congenially
periodical structures that are hierarchically organized in micro- and
nanoscale. In particular, the natural models that are related or similar
to the lotus effect have been mostly used due to their prominent
features. Following this, we continue to review recently published work
that successfully manufactured superhydrophobic surfaces with typical
features.
Fig. 11.9. Schematic illustrations of a drop of water in contact with the petal of a
red rose (Cassie impregnating wetting state) and a lotus leaf (the Cassie state).9
11.5. Artificial Superhydrophobic Surfaces

and Test
It is noticeable that many of the natural superhydrophobic surfaces
discussed above have intense roughness and complex structure in
micro- and nanoscales. Some symbolic surfaces have been utilized as
the primary models for researchers to mimic superhydrophobicity. For
example, a combination of poly(vinyl chloride) (PVC) and ethanol has
been used to form superhydrophobic films.65 The formed PVC surfaces
become rougher with the increase of ethanol content (under 50%) in the
PVC solution, and more pores and nanocomposites are formed. These
superhydrophobic surfaces form lotus-leaf-like structures consisting of
many nanoparticles with the sizes ranging from 100 nm to 300 nm.
Also, as pointed out previously, typical superhydrophobic surfaces
discovered in the nature have shown multi-scaled roughness consisting
of nanometer sized flakes on top of micrometer sized protrusions. This
type of morphology for superhydrophobic surfaces can be achieved
using lithographic methods, template-based techniques, plasma treat-

ment, self-assembly and self-organization, chemical deposition, layer-
by-layer (LBL) deposition, colloidal assembly, phase separation, and
electrospinning. Most of these methods are in accordance with the
theoretical and natural models discussed in previous sections, and
one strategy commonly used to invoke superhydrophobicity for many
processes is producing rough surfaces with hierarchical structure and
low surface energy. The following section will discuss those methods as
well as manmade superhydrophobic surfaces. As some of the previous
research has already been discussed in Refs. 32 and 66, we will be
concentrating on the most up-to-date articles that have been published
over the past few years.
Theoretically, to provide superhydrophobicity on solid surfaces, it
is necessary to either lower the surface energy or to prompt a proper
surface roughness. In some cases, it is necessary to work on the combi-
nation of surface energy and surface structure, as this can provide a more
remarkable effect on forming superhydrophobic surfaces. Practically,
a variety of techniques including plasma treatment,67 lithography,49
sol-gel process,68 chemical vapor deposition,69 electrospinning,70 and
colloidal assemblies71 have been utilized to produce superhydrophobic
surfaces with extreme water-repelling properties.
Moreover, it is noteworthy that the hierarchical structure, which
provides multi-layers of surface roughness at micro- and nanoscale, can
play an important role in prompting superhydrophobicity of surfaces.
The authors have recently developed a method to form superhy-
drophobic surfaces based on hierarchical micro- and nanostructures.21
It is shown that such hierarchical structures can be formed by using
SiO2 nanoparticles and polydimethylsiloxane (PDMS) solutions, which
will form a solid thin film exhibiting large contact angles and small
contact angle hysteresis, and consequently superhydrophobic surfaces
are formed.
In the authors’ approach, PDMS (average Mn 50,000 and average
Mw 95,000) was purchased from Sigma-Aldrich and used without
further purification. It was dissolved in white spirit to form stock
solutions of 6 wt %. Silica (SiO2 ) particles (fumed powder, Aldrich) with
two different average sizes (7 nm and 14 nm) were added, separately,
into the PDMS solutions. Thus, the PDMS-SiO2 nano-suspensions had

been prepared, and the nanoparticles within each individual sample had
only one single average size. The concentrations of SiO2 nanoparticles
in the nano-suspensions were 0.5, 1.0, and 2.0 w/v %, respectively. The
nanoparticle-added PDMS solutions were stirred at room temperature
for at least 30 minutes before use. Micro glass slides were washed in
acetone and distilled water repeatedly for approximately 20 minutes and
dried. The glass slides were immersed into the nano-suspensions for 10
minutes to form superhydrophobic composite films, and then rinsed in
distilled water gently to remove the unstably absorbed nanoparticles,
if any. The coated substrates were dried at room temperature overnight
and kept in a sealed chamber for three days. The combination of 7 nm
and 14 nm SiO2 particles was dispersed into the PDMS solutions,
of which the concentrations of 7 nm and 14 nm SiO2 nanoparticles
were both 0.5 w/v % eventually. Nano-suspensions prepared from the
combination of nanoparticles were also coated on the glass slides
following the same procedures mentioned above. The morphologies
of the coated substrates were investigated using a SEM (Phillips XL30
ESEM-FEG). Contact angle measurements were performed using a
contact angle meter (CAM200, KSV Instruments) on three different
spots for each substrate.
11.6. Results and Discussions

The static contact angles and contact angle hysteresis were measured
to characterize the wettability of the coated substrates. Particularly, the
substrates coated with PDMS solutions which had no nanoparticles
were also measured for reference. Overall, the measurement showed
that the contact angle of the water droplet increased and the contact
angle hysteresis decreased with the concentration of SiO2 nanoparticles
dispersed in the nano-suspensions. This applied to both 7 nm and
14 nm SiO2 particles. The average contact angle of water droplets
on the substrates simply coated with PDMS solutions was approx-
imately 85◦ . Typically, in the presence of 7 nm SiO2 nanoparticles
only, the average contact angle that was exhibited on the substrates
coated with the PDMS-SiO2 nano-suspensions was 93◦ , with the
concentration of nanoparticles being 0.5 w/v %; the average contact
angle was increased to 99◦ , with the concentration of nanoparticles

being 1.0 w/v %; and the average contact angle was increased up
to 130◦ , with the concentration of nanoparticles increased 2.0 w/v
%. On the other hand, in the presence of 14 nm SiO2 nanoparticles
only, the average contact angle was 95◦ , with the concentration of
nanoparticles being 0.5 w/v %; the average contact angle was increased
to 103◦ , with the concentration of nanoparticles being 1.0 w/v %;
and the average contact angle was increased even higher, up to
147◦ , with the concentration of nanoparticles being 2.0 w/v %.
Phenomenally, when only 7 nm SiO2 nanoparticles were involved, the
contact angle hysteresis was quite large (almost 40◦ ), even when the
concentration of nanoparticles was increased to 2.0 w/v % (around
30◦ ). However, in the presence of 14 nm SiO2 nanoparticles only,
the contact angle hysteresis could drop to 20◦ , and in some cases
even smaller (around 10◦ ). The combination of 7 nm and 14 nm SiO2
nanoparticles led to moderate results, showing average contact angle
of 116◦ and contact angle hysteresis of around 30◦ . Figure 11.10
shows the water droplets with typical contact angles on PDMS-SiO2
surfaces.
As shown in Fig. 11.11, the nanoparticle protrusions (grooves) of
14 nm and 7 nm (both at 2.0% w/v) were respectively interspersed on
the substrates in roughly regular orders. As the magnification of the
image increases, the surface structures are revealed with more detail. It
can be seen that porous structures were formed among the aggregations
on both types of the surfaces (from 14 nm and 7 nm nanoparticles).
Those nanoparticle aggregations had relatively large compositions at
the bottom, which were in tens of micro-meters, and the compositions
at the top were even smaller.
This combination of nanoparticle compositions formed the hier-
archical structure. However, the difference between those two types
of protrusions was also noticeable. The distribution of 14 nm SiO2
particles on the substrate was more serried, covering most of the area
of the substrate, whilst the distribution of 7 nm SiO2 particles had
relatively more open gaps on the substrate. This made the distance
between each protrusion of 7 nm particles larger than that between the
14 nm particles. Also, the order of the 14 nm nanoparticle protrusions
Fig. 11.10. Contact angles of water droplets on the SiO2 -PDMS surfaces. The
contact angles are 93◦ in (a), 113◦ in (b), 142◦ in (c), 97◦ in (d), 117◦ in (e), and
153◦ in (f). Images (a), (b), and (c) represent the 7 nm silica particles; images (d), (e),
and (f) represent 14 nm silica particles.
was more regular than that of the 7 nm ones. And the surface properties
based on the 14 nm nanoparticle protrusions would be even more
uniform, which should be the reason for the lower contact angle
hysteresis. The difference in terms of protrusion distribution was
attributed to the size of the nanoparticles. The PDMS had a high
viscosity, and nanoparticles were prone to form aggregations inside the
solution. This led to the formation of the protrusions of nanoparticles.
Besides, as both the 14 nm and 7 nm nanoparticles were small and
light enough, their motions would be overwhelmed by the solution.
However, it was still easier for small scaled particles to have random
motion with the solution, therefore forming a less regular order.
The heavier 14 nm particles were more prone to get connected for a
continuous coating than the same number of 7 nm particles. Thus, the
14 nm particles formed larger and denser aggregations than the 7 nm
particles on the surface.
Further to the protrusions formed on the substrate, their pattern
would influence the contact mode of the water droplet on the solid
Fig. 11.11. SEM images of SiO2 -PDMS composite films. (a)–(c) are based on 14 nm
silica particles, and (d)–(f) are based on 7 nm silica particles. Scale bars represent
100 µm in (a) and (d), 10 µm in (b) and (e), and 5 µm in (c) and (f).
surface with hierarchical structure at micro- and nanoscale. A well-

ordered hierarchical structure includes the substrate, first layer of
protrusions (microscale), second layer of protrusions (nanoscale), and
possibly more tiny layers of protrusions (even smaller than nanoscale).
Figure 11.12 illustrates the contact modes of liquid droplets on
hierarchically structured surfaces with two layers of protrusions. In the
stable wetting regime, the liquid droplet can fully contact the surface
with hierarchical structure, as shown in Fig. 11.12(1). Since both
layers beneath the liquid are fully wet in the Wenzel wetting state, this
regime corresponds to the Wenzel–Wenzel (W–W) state. Alternatively
the liquid contacts only the very top of the surface, as shown in
Fig. 11.12(4). Both layers have air inside, exhibiting the CB–CB
wetting state. To explain the wetting phenomena related to these two
states, the overall surface roughness or the fraction of the protrusions
wet by liquid on one single layer with respect to the total plain area
is taken into account. However, in a metastable wetting regime, the
liquid droplet would contact only the bottom layers (CB–W state),
shown in Fig. 11.12(2), or the top layers (W-CB state), shown in
Fig. 11.12(3); the roughness of individual layers and fraction of
protrusions on individual layers has to be considered for the wetting
Fig. 11.12. Illustration of contact modes between water droplets and solid surfaces
with hierarchical structure. Connection among the four states.
phenomena involved. The transition between those four wetting states

can be triggered by external force or energy. Thus, the contact mode
would condition the connection between the liquid and the solid. Apart
from the surface roughness and surface energy, this is one element that
influences the contact angle of liquid droplets on solid surfaces. And
that’s why the contact angle could be different even on the same surface.
11.7. Conclusion
Although superhydrophobicity is only a recently developed concept,

it has already become important to research and will be potentially
important to our lives. Much effort has been made in research to
understand the mechanisms that are related to superhydrophobicity
on solid surfaces. This nature-inspired theory is an interdisciplinary
subject which involves physics, chemistry, material science, and even
biology. In cutting edge research, nanotechnologies are effectively used
to explore the intrinsic essence of the subject, the “real” structures of
the superhydrophobic models that are found in nature, and fabricate
manmade superhydrophobic surfaces which have significant water-

repellence and durability. The achievement of superhydrophobicity-
related study is phenomenal and this tendency will be maintained and
continued. The Young equation, Wenzel equation, and Cassie–Baxter
equation are fundamental theories for describing wetting phenomena
on superhydrophobic surfaces; they are very necessary and useful,
but not sufficient to fit all situations. When a liquid drop wets a
superhydrophobic surface, its state depends on the particular surface
property and the wetting process. A transition between the wetting
states might take place if the energy balance of the liquid drop is
upset. The question for theoretical research in the future lies in how to
work out governing equations that are adequate to illustrate the stable
wetting regimes as well as the transitions on superhydrophobic surfaces.
With further research into the theories, tremendous work has been
made to understand the wetting phenomena on superhydrophobic
surfaces. We thus review the superhydrophobic surfaces in nature and
the most recent progress in fabricating manmade superhydrophobic
surfaces from a biomimetic point of view. Creating a superhydrophobic
surface may involve several procedures, during which a number of
techniques are used together. But in some cases, it is also possible to
create superhydrophobic surfaces using relatively simple and prompt
methods. A typical superhydrophobic (ultrahydrophobic) surface can
repel water droplets from wetting itself, and the contact angle of a
water droplet resting on a superhydrophobic surface is greater than
150◦ , which means extremely low wettability is achievable on super-
hydrophobic surfaces. Many superhydrophobic surfaces (both artificial
and natural) normally exhibit micro- or nanosized roughness as well
as hierarchical structure, which can somehow influence the surface’s
water repellence.
This chapter discusses the authors’ recent work of investigating
superhydrophobic surfaces made from SiO2 -PDMS films. Silica nano
particles of 7 nm and 14 nm were dispersed in PDMS solutions with
variable concentrations respectively to prepare nanoparticle suspensions
for application. Composite films were formed on the test surfaces
coated with the nanoparticle suspensions. The contact angles of water
droplets on the composite films were measured, and it has been
found that the contact angle increases with the concentration of

nanoparticles in solution. Also, the contact angles of water droplets
on the coated surfaces regarding to 14 nm silica particles are larger
than those with the 7 nm ones. In particular, the composite surface
coated with the SiO2 -PDMS solutions, which has 2.0 wt/v % 14 nm
silica particles inside, can prompt a contact angle up to more than
150◦ . Water repelling properties are rendered by the composite films
and superhydrophobic surfaces are thus formed. The SEM observation
reveals that the hierarchical structure in micro- and nanoscale can
be created among the composite films on superhydrophobic surfaces.
Based on the measurement and observation, we have demonstrated that
superhydrophobic surfaces can be created by using the SiO2 -PDMS
solutions. Indeed, the hierarchical structure in micro- and nanoscale
plays an important role in prompting the superhydrophobic (water-
repelling) properties and wetting states.
References
[1] Burke JE. (1964) Scattering of surface waves on an infinitely deep fluid. J Math
Phys 6: 805–819.
[2] Poynor A, Hong L, Robinson IK, et al. (2006) How water meets a hydrophobic
surface. Phys Rev Lett 97: 4.
[3] Onda T, Shibuichi S, Satoh N, Tsujii K. (1996) Super-water-repellent fractal
[4] Yoshimitsu Z, Nakajima A, Watanabe T, Hashimoto K. (2002) Effects of
surface structure on the hydrophobicity and sliding behavior of water droplets.
Langmuir 18: 5818–22.
[5] Kijlstra J, Reihs K, Klamt A. (2002) Roughness and topology of ultra-
hydrophobic surfaces. Colloid Surf A-Physicochem Eng Asp 206: 521–9.
[6] Yan YY, Gao N, Berthlott W. (2011) Mimicking natural superhydrophobic
surfaces and grasping the wetting process: A review on recent progress in
preparing superhydrophobic surfaces. Adv Colloid Interf 169(2): 59–79.
[7] Gao N, Yan N. (2012) Characterisation of surface wettability based on nanopar-
ticles. Nanoscale 4(7): 2202–2218.
[8] Barthlott W, Neinhuis C. (1997) Purity of the sacred lotus, or escape from
contamination in biological surfaces. Planta 202(1): 1–8.
[9] Feng L, Zhang YA, Xi JM, et al. (2008) Petal effect: a superhydrophobic state
with high adhesive force. Langmuir 24: 4114–9.
[10] Koch K, Bhushan B, Barthlott W. (2008) Diversity of structure, morphology

and wetting of plant surfaces. Soft Matter 4: 1943–63.
[11] Fang Y, Sun G, Cong Q, et al. (2008) Effects of methanol on wettability of the
non-smooth surface on butterfly wig. J Bionic 5: 127–33.
[12] Byun D, Hong J, Ko JH, et al. (2009) Wetting characteristics of insect wing
surfaces. J Bionic Eng 6: 63–70.
[13] Barnes WJP. (2007) Biomimetic solutions to sticky problems. Science 318:
203–4.
[14] Nosonovsky M, Bhushan B. (2007) Multiscale friction mechanisms and hierar-
chical surfaces in nano- and bio-tribology. Mater Sci Eng R-Rep 58: 162–93.
[15] Tuteja A, Choi W, Ma ML, et al. (2007) Designing superoleophobic surfaces.
Science 318: 1618–22.
[16] Gao N, Yan YY, Chen XY, Mee DJ. (2011) Superhydrophobic surfaces with
hierarchical structure. Mater Lett 65: 2902–5.
[17] Callies M, Quere D. (2005) On water repellency. Soft Matter 1: 55–61.
[18] Wenzel RN. (1936) Resistance of solid surfaces to wetting by water. Ind Eng
Chem 28: 988–94.
[19] Wenzel RN. (1949) Surface roughness and contact angle. J Phys Colloid Chem
53: 1466–7.
[20] Marmur A. (2003) Wetting on hydrophobic rough surfaces: to be heterogeneous
or not to be? Langmuir 19: 8343–8.
[21] Cassie ABD. (1948) Contact angles. Discuss Faraday Soc 3: 11–16.
[22] Blossey R. (2003) Self-cleaning surfaces — virtual realities. Nat Mater 2: 301–6.
[23] Quere D. (2005) Non-sticking drops. Rep Prog Phys 68: 2495–532.
[24] Cassie ABD, Baxter S. (1944) Wettability of porous surfaces. Trans Faraday Soc
40: 0546–50.
[25] Patankar NA. (2004) Transition between superhydrophobic states on rough
[26] Herminghaus S. (2000) Roughness-induced non-wetting. Europhys Lett 52:
165–70.
[27] Wolansky G, Marmur A. (1999) Apparent contact angles on rough surfaces: the
Wenzel equation revisited. Colloid Surf A 156: 381–8.
[28] Brandon S, Haimovich N, Yeger E, Marmur A. (2003) Partial wetting of
chemically patterned surfaces: the effect of drop size. J Colloid Interf Sci 263:
237–43.
[29] Tsujii K, Yamamoto T, Onda T, Shibuichi S. (1997) Super oil-repellent surfaces.
Angew Chem Int Ed Engl 36: 1011–2.
[30] Shibuichi S, Yamamoto T, Onda T, Tsujii K. (1998) Super water- and oil-
repellent surfaces resulting from fractal structure. J Colloid Interf Sci 208:
287–94.
[31] Shibuichi S, Onda T, Satoh N, Tsujii K. (1996) Super water-repellent surfaces
resulting from fractal structure. J Phys Chem 100: 19512–7.
[32] Roach P, Shirtcliffe NJ, Newton MI. (2008) Progess in superhydrophobic

surface development. Soft Matter 4: 224–40.
[33] Kwon Y, Patankar N, Choi J, Lee J. (2009) Design of surface hierarchy for
extreme hydrophobicity. Langmuir 25: 6129–36.
[34] Patankar NA. (2003) On the modeling of hydrophobic contact angles on rough
[35] Quere D, Lafuma A, Bico J. (2003) Slippy and sticky microtextured solids.
Nanotechnology 14: 1109–12.
[36] Choi CH, Kim CJ. (2006) Large slip of aqueous liquid flow over a nanoengi-
neered superhydrophobic surface. Phys Rev Lett 96: 4.
[37] Jung YC, Bhushan B. (2009) Dynamic effects induced transition of droplets on
biomimetic superhydrophobic surfaces. Langmuir 25: 9208–18.
[38] Feng L, Song YL, Zhai J, et al. (2003) Creation of a superhydrophobic surface
from an amphiphilic polymer. Angew Chem Int Ed 42: 800–2.
[39] Kim H, Kim MH, Kim J. (2009) Wettability of dual-scaled surfaces fabricated
by the combination of a conventional silicon wet-etching and a ZnO solution
method. J Micromech Microeng 19: 7.
[40] Chen MH, Hsu TH, Chuang YJ, Tseng FG. (2009) Dual hierarchical biomimic
superhydrophobic surface with three energy states. Appl Phys Lett 95: 3.
[41] Journet C, Moulinet S, Ybert C, et al. (2005) Contact angle measurements on
superhydrophobic carbon nanotube forests: effect of fluid pressure. Europhys
Lett 71: 104–9.
[42] Kim JW, Kim CJ. (2002) Nanostructured surfaces for dramatic reduction of
flow resistance in droplet-based microfluidics. Fifteenth IEEE International
Conference on Micro Electro Mechanical Systems. Tech Digest : 479–82.
[43] Jindasuwan S, Nimittrakoolchai O, Sujaridworakun P, et al. (2009) Surface
characteristics of water-repellent polyelectrolyte multilayer films containing
various silica contents. Thin Solid Films 517: 5001–5.
[44] Gao LC, McCarthy TJ. (2007) A commercially available perfectly hydropho-
bic material (theta(A)/theta(R)=180 degrees/180 degrees). Langmuir 23:
9125–7.
[45] Sawada H, Suzuki T, Takashima H, Takishita K. (2008) Preparation and proper-
ties of fluoroalkyl end-capped vinyltrimethoxysilane oligomeric nanoparticles —
a new approach to facile creation of a completely superhydrophobic coating
surface with these nanoparticles. Colloid Polym Sci 286: 1569–74.
[46] Zhang X, Shi F, Niu J, et al. (2008) Superhydrophobic surfaces: from structural
control to functional application. J Mater Chem 18: 621–33.
[47] Gao LC, McCarthy TJ. (2002) Contact angle hysteresis explained. Langmuir
22: 6234–7.
[48] Extrand CW. (2002) Model for contact angles and hysteresis on rough and
ultraphobic surfaces. Langmuir 18: 7991–9.
[49] Jeong HE, Kwak MK, Park CI, Suh KY. (2009) Wettability of nanoengineered
dual-roughness surfaces fabricated by UV-assisted capillary force lithography.
J Colloid Interface Sci 339: 202–7.
[50] Adamson AW, Gast AP. (1997) Physical Chemistry of Surfaces, 6th ed. John
Wiley & Sons, New York.
[51] Nosonovsky M, Bhushan B. (2008) Multiscale Dissipative Mechanisms
and Hierarchical Surfaces: Friction, Superhydrophobicity, and Biomimetics.
Springer, Berlin.
[52] Orowan E. (1970) Surface energy and surface tension in solids and liquids. Proc
R Soc LondA Math Phys Sci 316: 473–491.
[53] Kumikov VK, Khokonov KB. (1983) On the measurement of surface free-energy
and surface-tension of solid metals. J Appl Phys 54: 1346–50.
[54] Ip SW, Toguri JM. (1994) The equivalency of surface-tension, surface-energy
and surface free-energy. J Mater Sci 29: 688–92.
[55] Blossey R. (2003) Self-cleaning surfaces — virtual realities. Nat Mater 2: 301–6.
[56] Chen W, Fadeev AY, Hsieh MC, et al. (1999) Ultrahydrophobic and ultra-
lyophobic surfaces: some comments and examples. Langmuir 15: 3395–9.
[57] Wang JD, Chen HS, Sui T, et al. (2009) Investigation on hydrophobicity of
lotus leaf: experiment and theory. Plant Sci 176: 687–95.
[58] Lafuma A, Quere D. (2003) Superhydrophobic states. Nat Mater 2: 457–60.
[59] Gao N, Yan YY. (2009) Modeling superhydrophobic contact angles and wetting
transition. J Bionic Eng 6: 335–40.
[60] Neinhuis C, Barthlott W. (1997) Characterization and distribution of water-
repellent, self-cleaning plant surfaces. Ann Bot 79: 667–77.
[61] Koch K, Bhushan B, Barthlott W. (2009) Multifunctional surface structures of
plants: an inspiration for biomimetics. Prog Mater Sci 54: 137–78.
[62] Guo ZG, Liu WM. (2007) Biomimic from the superhydrophobic plant leaves
in nature: binary structure and unitary structure. Plant Sci 172: 1103–12.
[63] Feng XJ, Jiang L. (2006) Design and creation of superwetting/antiwetting
surfaces. Adv Mater 18: 3063–78.
[64] Gao XF, Jiang L. (2004) Water-repellent legs of water striders. Nature 432: 36.
[65] Chen H, Yuan ZQ, Zhang JD, et al. (2009) Preparation, characterization and
wettability of porous superhydrophobic poly(vinyl chloride) surface. J Porous
Mat 16: 447–51.
[66] Li XM, Reinhoudt D, Crego-Calama M. (2007) What do we need for a
superhydrophobic surface? A review on the recent progress in the preparation
of superhydrophobic surfaces. Chem Soc Rev 36: 1350–68.
[67] Kim TI, et al. (2008) Optical lithography with printed metal mask and a simple
superhydrophobic surface. Small 4(2): 182–185.
[68] Mahltig B, Bottcher H. (2003) Modified silica sol coatings for water-repellent
textiles. J Solgel Sci Technol 27(1): 43–52.
[69] Jung YC, Bhushan B. (2009) Mechanically durable carbon nanotube-composite

hierarchical structures with superhydrophobicity, self-cleaning, and low-drag.
ACS Nano 3: 4155–63.
[70] Wang N, Zhao Y, Jiang L. (2008) Low-cost, thermoresponsive wettability of
surfaces: Poly(N-isopropylacrylamide)/polystyrene composite films prepared by
electrospinning. Macromol Rapid Commun 29(6): 485–489.
[71] Motornov M, et al. (2008) Superhydrophobic surfaces generated from water-
borne dispersions of hierarchically assembled nanoparticles coated with a
reversibly switchable shell. Adv Mater 20(1): 200–205.
Chapter 12
Bio-Inspired Macro-Morphologic
Surface Modifications to Reduce
Soil–Tool Adhesion
Peeyush Soni∗ and Vilas M. Salokhe

∗ soni@ait.asia; soni.ait@gmail.com
The force of attraction between particles is a natural phenomenon, and

can be termed as cohesion — if it occurs among similar molecules, and
as adhesion — if found among unlike molecules. With a focused interest
on agricultural soils, soil cohesion is of greater importance with regard
to soil erosion and its binding strength, while its adhesive nature results
in undesired increase in draft forces on soil-interacting tools. May it be
farm implements, construction machinery or any soil-engaging tool,
their larger force requirement increases the input energy and hence less
work is done per unit effort supplied. Working performance is badly
impaired with a large mass of adhered soil carried along with the tool
under operation. This is especially significant in sticky soils.
Keeping in view the enormous associated losses, scientists and
engineers have been trying to solve the problem to reduce burden of
such excessive energy consumption in sticky soils. This chapter presents
a broader approach to describe soil–tool adhesion in cohesive soils and
reviews various methods, both under practice and under investigation,
used to reduce energy losses due to adhesion. Principles involved under
theories of soil adhesion and its remedial measures are elaborated.
421
Further, this chapter theoretically analyzes microscopic forces at

the soil–tool interface and hence opens a window to further evaluate
these concepts for various soil types with different tool shapes.
NOTATIONS USED
Subscripts
a : Adsorption
A : Adhesion
c : Capillary
d : Dry
e : Equilibrium
f : Friction
G : Gravitational
i : Individual
m : Meniscus
p : Soil pores (assumed to be filled with water with negligible air)
s : Soil particles
t : Tangent
v : Viscosity
w : Wet
γ : Surface tension
Symbols
(mw /mm ) : Ratio of mass of water to mass of soil

hω : Lifishitz–Van der Waals constant, J
A : Change in surface area, m2
A : Surface area of film, m2
b : Vertical distance between soil particle surface to its
lowest tip, m
d : Separation between parallel planes of soil particle and
tool surface, m
Bio-Inspired Macro-Morphologic Surface Modifications 423
Symbols (Continued )
d0 : Separation of the tip of the summit from the solid surface

d0 : Separation between the summit tip and tool surface, m
de : Effective intermolecular separation
de : Equilibrium intermolecular distance (shortest possible
distance), m
E : Surface energy of film, J
E(d) : Energy of adhesion per unit area of separation, J
E(d) : Interaction energy per unit area of separation, J/m2
Ea : Energy of adhesion at equilibrium intermolecular
distance de
EA : Energy of adhesion, J/m2
F : Tensile force, N
F : Total microscopic force at the interface, N
Fa : Normal adhesion force, N
FAm : Adhesion force of water meniscus, in the direction
of surface tension, N
Fcohesion : Force of cohesion, N
Fgravity : Gravity force, N
Ft : Tangential adhesion force, N
Fviscous : Viscous force, N
g : Acceleration due to gravity, m/s2
h : Height of capillary rise, m
h1 , h2 : Separation between discs, m
HES : High energy surfaces
J : Some property of interlayer water (including partial
specific heat, partial specific volume, and viscosity)
J0 : Same property of the water in bulk
k : Universal constant for liquid, unitless
L : Latent heat of the liquid, J
l : Length of cylindrical liquid layer, m
LES : Low energy surfaces
lt : Total accumulative length of surface tension, m
N : Total normal force at interface, N
NA : Normal adhesive force, N
NAm : Normal adhesion force caused by meniscus, N

NAs : Normal adhesion force caused by bare soil particles,
N
NAv : Normal adhesion force caused by water film due to
viscosity, N
Nca : Normal adhesion force caused by capillary negative
adsorption, N
Nf : Normal friction, N
NG : Total gravitational force normal to interface, N
Np : Weight of water-filled soil pores at interface, N
Ns : Weight of soil particles at interface, N
Nwfm : Normal wet friction of interfacial water film
meniscus, per unit length of the line of contact, N
P : Laplace pressure, N/m2
p : Perimeter of the smallest cross-section parallel to the
solid surface,
p : Percent of pore space at the soil-tool interface, % of
total cross sectional area
P : Pressure inside liquid column near the meniscus, Pa
P (d) : Attraction pressure per unit area of separation, Pa
P0 : Atmospheric pressure, Pa
P1 , P2 : Pressure inside liquid column near the upper and
lower meniscus, respectively, Pa
PAC : Negative pressure existing at the liquid wedge at the
level AC, Pa
Patm : Atmospheric pressure, Pa
PE : Polyethylene
PES-PTFE : Polyethersulphone-Polytetrafluoroethylene
PP : Polypropylene
PS : Polystyrene
PSD : Particle size distribution
Pstatic : Static pressure of water column at corresponding
height, Pa
PTFE : Polytetrafluoroethylene
r : Curvature radius of meniscus, m

R : Radius of soil particle asperity, m
r1 , r2 : Radii of curvature, m
RAC : Radius of curvature of the interface at point A, m
Rc : Radius of capillary tube, m
S : Initial spreading coefficient
S : Total contacting area of soil–tool interface, m2
s : Distance between the lowest point of upper
meniscus and the shearing plane, m
Si : Section area of individual soil pore, m2
SMT : Soil moisture tension, Pa
Sp : Total area of soil pores, m2
Ss : Total contact area of soil particles, m2
T : Temperature of liquid ◦ C
t : Time required to pull two parallel plates, s
Tc : Critical temperature, ◦ C
UHMWPE : Ultra high molecular weight polyethylene
v : Velocity of flow, m/s
WA : Work of adhesion, J
Wadhesion : Work of adhesion
Wcohesion : Work of cohesion
x : Distance of liquid layer from the tool surface, m
β : A constant, which depends upon the solid surface
characteristics
γ, γLV : Surface tension at liquid–vapor interface, N/m
γb, γf : Backside and frontside surface tension, respectively,
N/m
γSL : Surface tension at solid–liquid interface, N/m
γSV : Surface tension at solid–vapor interface, N/m
η : Coefficient of dynamic viscosity, dyne-sec/cm2 or
poise
θ : Contact angle between meniscus and tool surface,
degree
θ1 , θ2 : Contact angles are upper and lower meniscus,

respectively, degree
θA : Advancing contact angle, degree
θb : Contact angle at the back of meniscus, degree
θf : Contact angle at the front of meniscus, degree
θR : Receding contact angle, degree
µdfs : Coefficient of dry friction of soil
ρ : Density of liquid, kg/m3
ρ, ρ2 : Density of liquid, kg/m3
ρ1 : Density of solid soil particles, kg/m3
ρ3 : Density of second fluid surrounding the whole system,
kg/m3
σp : Mean gravitational stress of soil pores, Pa
σs : Mean gravitational stress of soil particles, Pa
τ : Total tangential/sliding (lateral) resistance at
interface, N
τA : Adhesive component of tangent resistance, N
τdfs : Tangent resistance from dry friction of soil particles, N
τdrag : Drag component of tangent resistance (τ), N
τf : Frictional component of tangent resistance (τ), N
τtAm : Tangent resistance due to tangent adhesion of water
meniscus, N
τtAγ : Tangent resistance due to tangent adhesion of surface
tension, N
τtv : Tangent resistance due to viscous forces, N
ϕ : Angle, with vertical, made by the radius with the point at
which the liquid surface meets sphere, degree
12.1. Introduction
With a focused interest on agricultural soils, its adhesive nature results

in undesired increase in draft forces on soil-interacting tools — their
larger force requirements increase energy consumption per unit soil
operation. Furthermore, soil adhesion decreases quality of work and

reduces efficiency of soil-tools. For components of sowing machines
such as furrow openers and covering devices, it reduces the emergence
rate of seeds, and for fertilizer applicators and transplanters, it lowers
application quality. Studies reveal a reduction of as high as 30–50% in
work efficiency due to excessive adhesion of soil-engaging components
of farm implements, earth-moving and construction machinery, includ-
ing dumpers, bulldozers and excavators. Keeping in view the enormous
associated losses, scientists and engineers have been trying to solve the
problem to reduce the burden of excessive energy consumption and
undesired wear and tear of tools in sticky soils.
Working performance of a soil-engaging tool is largely dependent
on soil properties — soil type and prevailing soil-moisture content,
and tool properties — tool geometry and tool surface properties. For
agricultural purposes, soil properties have more or less strict limitations
to be altered for a given crop or tillage operation; thereby leaving
tool properties that can be maneuvered by operator. Furthermore,
tool geometry is greatly a function of forces associated with a specified
operation, which has a complicated response to any alteration. Hence in
order to improve working performance of a tool, its surface properties
seem to be an attractive option for possible modifications.
Tool surfaces can either be modified by altering intrinsic proper-
ties of the base material — micro-morphologic modifications, which
include heat treatment and fusing metallic/non-metallic elements
into the base material; or by modifying surface structure — macro-
morphologic modifications, using coating or plating and mounting
protrusions. The latter offers greater practical ease due to minimal
know-how required at the operator’s level. Moreover, toughness and
hardness — considered important material properties that should be
present in sufficient amount in soil-engaging materials — can’t be
compromised by replacing the whole tool component. This makes the
possibility of completely replacing metal with polymers obsolete. This
favors only coating or mounting of such hydrophobic material onto the
metallic base tool material — to retain its strength.
Biomimetics or bionic-engineering recently opened up a new win-
dow to address the problem of soil adherence onto tool surfaces —
especially in sticky soils. This chapter reviews the underlying concepts

and applications of bio-inspired macro-morphologic surface modifica-
tion techniques to reduce soil–tool adhesion. Protuberance patterns
available on the dung-beetle and other similar soil-burrowing animal
enable them to keep their body surface soil-free. Such patterns have
been identified to be copied on various tools as macro-morphological
surface modificaiton. In addition, it has been shown by various
researchers that coating can also contribute to improved performance
of tillage tools in sticky soil.
Attempts have also been made to lower rolling resistance of soft-
ground off-road vehicles on paddy fields, swamp and beaches, reducing
plowing resistance of tillage implements including moldboard and disc
plows, and decreasing power requirement of construction machinery
including dozers, levelers and excavators. Farmers, especially paddy
growers, rural transporters using off-road vehicles, and construction
workers at remote sites who lack modern machinery would be among
the beneficiaries of this research.
12.2. Soil–Tool Adhesion
From an agricultural point of view, soil–tool adhesion is of great

importance, especially when the prevailing soil is rich in clay content
resulting in it being sticky by nature. Various investigations have been
carried out to study the adhesion mechanism, to quantify soil-adhesion
properties, to determine their contributing factors, and for developing
means of reducing soil-adhesion.1,2,3 As per the definition of adhesion,
the soil-engaging tool and soil particles are the two rigid bodies with
unlike material properties, where an additional force, acting on their
mutual contact surface, is often required to pull them apart. This
additional force of attraction, depending on various factors, often
attains noticeable magnitude and has been inviting special attention
for decades.
Soil adhesion to tillage implements lowers quality of work and
increases energy consumption. For components of sowing machines
such as furrow openers and covering devices, it reduces the emergence
rate of seeds, and for fertilizer applicators and transplanters, it decreases
application quality.3
There are various mechanisms through which two particles in

nature exert force on each other; the force of attraction between them
is a usual natural phenomenon. Depending on the particle type, it
can be cohesion — if the two particles are from the same parent
material, or adhesion — if the two come from different parent material.
Cohesion and adhesion are the molecular phenomenon occurring at
the interface of a liquid and gas, a result of electrical interactions
of the microscopic particles comprising the materials. The maximum
distance up to which the force of cohesion between molecules can
act is known as their molecular-range (≈10−7 cm).4 An imaginary
sphere drawn around a molecule with its radius equal to the molecular
range and center coinciding with the center of the molecule is called
the sphere of influence of the molecule. For agricultural soils, their
cohesive properties are mainly studied for various agronomic and soil-
conservation interests, whereas their adhesive characteristics are greatly
dealt by agricultural engineers and soil physicists. In farm implements,
construction machinery or any soil-engaging tool, their larger force
requirement increases the input energy. This results in smaller output
per unit effort supplied.
12.2.1. Background
Soil–tool adhesion and abrasion are considered tribological processes
in a soil–tool system that adversely affect the working quality, energy
consumption, and service life of the system. For obtaining large
control over such impacts, inclusion of related concepts has been
felt mandatory for numerous fields including agricultural engineering,
terramechanics, tribology, soil chemistry, soil physics and material
science and engineering. Adhesion has been envisaged as a dynamic
soil property, though expressed theoretically in terms of soil moisture
and other basic soil physical properties. This approach limited the
quantitative evaluation of soil adhesion and hence resulted in relatively
less understanding of soil–tool interaction.5,1
There have been identified two important modes of appearance of
soil adhesion, which should be included for truly expressing a soil–tool
mechanism: sliding friction and stickiness.1
Soil adhesion has invariably two components, normal to the surface

and tangential to the tool surface over which soil is to be moved.
The amount of adhered soil differs with system properties including
tool material, soil condition and system dynamics. The magnitudes of
adhesion forces are different in the two components, and scientists
suggest inclusion of the both aspects for determining cumulative
effect.1
12.2.2. Soil–tool system

When two solid surfaces are pressed closely, the adhesion force between
them greatly depends on their properties, no matter whether the
interfacial water film exists or not. If the two solids are hard, presence
of interfacial water film increases adhesion, but for hard and soft
surfaces together, the adhesion force is a function of applied pressure
and is reduced upon introducing interfacial water film. For practical
purposes, the soil–tool system can be considered as the interaction of
two hard surfaces, whose adhesion increases with water film between
them. A better visualization can be achieved upon assuming the soil–
tool system as comprised of three constituents — the soil, the water
and the tool surface (Fig. 12.1). A careful observation leads to further
classification of these three constituents into five system-layers, namely
the soil, the soil–water interface, the water, the water–tool interface,
and the tool surface. As mentioned earlier, the presence of a continuous
water film — satisfying Laplace pressure — is important for soil–tool
adhesion description. For the sake of computational ease, soil particles
can be assumed to be spherical.6−9
Soil-water interface
Interfacial Soil particles
water film
Tool Tool-water interface
Fig. 12.1. Three constituents and five layers of soil–tool system (figure not to scale).
The contacts at the soil–tool interface can be both solid and

liquid, but solid–solid attraction (soil particles and tool surface) can
be neglected as compared to that due to the water film. Solid–solid
interaction is mostly of abrasion type rather than adhesion.10 Sliding
resistance of soil–tool is composed of two components, frictional and
adhesive. Shearing, cutting and scrapping forces of abrasive particles,
soil disturbance and soil deformation are included under the frictional
component, whereas the adhesive component comprises tangential
viscous resistance of continuous thin water film.11
Based on the direction of adhesion forces, the soil–tool system can
be sub-divided into the following two categories:
(a) Normal adhesion system: Pull is exerted normal to the surface of

soil–tool interaction. Use of polymers has been gaining popularity
for reducing adhesion in such systems.
(b) Tangential sliding system: Pull is applied tangentially at the soil–
tool interaction. Both polymers and enamel coatings have been
found satisfactory for minimizing tangential adhesion.
Plowing resistance could be assumed to comprise the following three

components:
(a) Resistance due to deforming and fragmenting soil during plowing

(b) Momentum of soil mass traveling over plow, and
(c) Sliding resistance
Reference 12 estimated that sliding resistance between soil and plow

surface during plowing operation constitutes 25–28% of the total
plowing resistance. Sliding resistance includes tangential adhesion force
(Ft ) and the frictional force.
Sliding resistance, at the interface, largely depends on the adhesion,
the stress normal to the interface, and the angle of friction between soil
and interface. This resistance can be lowered by bringing adhesion value
lower than soil cohesion and by bringing interface friction angle lower
than the soil–soil friction angle.11
12.2.3. Mechanisms involved in soil adhesion

Soil adhesion, produced by continuous thin water film and loops
between soil and tool surface, is caused by surface tension, meniscus
tension, and viscous resistance.1,3,5,8,9 The thickness of the water film
directly governs soil–tool adhesion by altering the above three basic
parameters — the thinner the film, the greater the Laplace pressure
and viscous resistance, while the nature and properties of tool surface
material determines wettability.10
The following individual basic phenomena constitute soil adhesion
in a totality:
i. Molecular attraction
ii. Surface tension
iii. Wettability and contact angle
iv. Viscous resistance
i. Molecular attraction
Between the tool surface and non-contacting asperities or water point
contacting asperities of soil, there exists a force of molecular attraction.
This molecular attraction is significant only when the separation
between them attains equilibrium intermolecular distance.13 Contri-
bution of this attraction force can be neglected when the separation
is larger than equilibrium intermolecular separation. Situations arise
while working with a soil–rubber system, when vacuum pockets develop
within closed structural units, which contributes to soil adhesion.
ii. Surface tension

Strong intermolecular attractive forces are present in molecules in
a liquid state. They are referred to as cohesive forces for like
molecules — as the case of molecules of liquid. Strong cohesive forces,
at the surface, are responsible for surface tension. The liquid molecules
on the surface experience inward pull and the liquid surface behaves
as a stretched membrane, tending to reduce its surface area — which
is known as surface tension. The molecules at the surface, unlike other
molecules inside the liquid, are not surrounded by other molecules
on all the sides, and are more strongly cohered. Surface tension is
numerically equal to the work done in unit increments of its surface
area.4 The cohesive forces among the liquid molecules are mainly
responsible for surface tension. Surface tension makes dragging of an
object through the surface of a liquid more difficult than moving it
under submerged conditions.
Continuity of water film at the interface is a key factor deter-
mining adhesion forces; a discontinuous film may result in significant
reduction — as appears in sandy loam and sandy soils. Surface tension
and moisture tension were assumed to be primary means for force
transmission through moisture, and are included for enlisting factors
affecting adhesion. It was believed that those which directly alter
moisture tension would influence adhesion. Wettability, which changes
contact angle of a material, was thus recognized to address adhesion.1, 8
iii. Wettability and contact angle

Wettability describes the degree to which water adheres to the surface
of a given material, and is measured in terms of contact angle (θ). Water
on a highly wettable surface forms a thin layer, whereas it will remain
merely a large drop on a non-wettable material surface. At the liquid–
solid interface, the surface of a liquid is generally curved. Depending
on the resultant sum of cohesion and adhesion, the liquid rises up or
is depressed down near the wall of the solid vessel. The angle drawn
inside the liquid, between the planes tangent to liquid and the solid
surface, is known as the contact angle (θ), and is measured in degrees.
The greater the angle, the less spreading will occur, and the material
will become less wet with prevailing liquid.
The contact angle of a liquid is a result of mechanical equilibrium
among three surface tensions, γLV (at the liquid–vapor interface), γSL
(at the solid–liquid interface), and γSV (at the solid–vapor interface) for
a drop resting on a solid (tool) surface.14 Figure 12.2 shows different
phases interacting each other on three possible values of contact angle,
i.e. acute, right angle and obtuse.
From the figure, under equilibrium condition, Young’s equation
can be stated as,
γSV − γSL = γLV Cos θ. (12.1)
LV
Vapor
Liquid
SL
SV
Tool
(a)
LV LV
(b) (c)
Fig. 12.2. Wettability of a film, variety of contact angles; (a) acute angle, (b) right
angle, (c) obtuse angle (figure not to scale).
Reference 14 revealed that the work of adhesion (Wadhesion ) of liquid

and solid is reversible in nature. It can be expressed in terms of surface
tensions as
Wadhesion = γSV + γLV − γSL . (12.2)
The above equation is a thermodynamic expression stating that the
reversible work of separating liquid–solid phases must be equal to
the change in free-energy of the system. Thus it can be viewed that
the right-hand-side term (γSV + γLV − γSL ) is equivalent to the free
energy of the system. More precisely, these terms individually are
the free energy per unit surface area of solid–vapor, liquid–vapor and
solid–liquid interfaces, respectively.
Reference 15, as cited in Ref. 14, introduced the well-known
condition for spreading of liquid b on a solid or liquid a,
For spreading, S > 0;
For nonspreading, S ≤ 0,
where,
S = γa − (γab + γb ). (12.3)
Reference 16, as cited by Ref. 14, named S as the Initial Spreading

Coefficient, and proposed its following relation with the reversible work
of adhesion (Wadhesion ) and the reversible work of cohesion of liquid
(Wcohesion ),
S = [Wadhesion − Wcohesion ]. (12.4)
The force of attraction among dissimilar molecules is called adhesive

force. An upward turning meniscus results when the adhesive force
between liquid molecules and the walls of a vessel exceeds the cohesive
force within the liquid molecules, and is known as capillary rise. Contact
angle is a function of both the soil and surface material; the smaller the
angle, the greater the wetting of the surface. The water layer, in the
condition of adhesion, is attracted by both the soil particles and the soil-
engaging tool surface.1 Experiments conducted in Ref. 17 indicated
that wetting angle (wettability) of various materials by different soil
solutions was constant for a given soil–tool system. Table 12.1 presents
typical values of surface tension and wetting angle of various soils with
different material.
Contact angle is affected by the surface finish of the tool surface;
the roughness might act as capillaries that allow water to rise in and
thereby increase the film contact area. Contact angles (both advancing
and receding) of distilled water with polymers (PTFE, UHMWPE,
PES-PTFE) were found higher than those with plasma sprayed Al2 O3
coating, Ni-base alloy coating, enamel coating, and the hardened
and tempered steel-45. It was demonstrated that polymers, being
Table 12.1. Surface Tension and Wetting Angle of Various Soils
Wetting angle (degrees)

Surface
Soil solution tension* (N/m) Cast iron Stainless steel Plow steel
Cecil clay 0.722 76.7 80.7 78.5

Greenville sandy loam 0.732 73.7 81.8 77.6
Sumpter clay 0.727 65.5 81.5 76.5
Lufkin clay 0.705 66.5 80.9 75.6
*Surface tension at room temperature 0.728 N/m

(Compiled from Refs. 1, 17)
low-energy surfaces (LES), having surface tension lesser than water,

couldn’t absorb water.5
Metals and inorganic non-metals are categorized as high-energy
surfaces (HES). High energy surfaces have higher binding strength
between water-HES than that of water itself and therefore possess
higher tendency to absorb moisture, even at atmospheric relative
humidity of less than 0.6%.5 Table 12.2 gives an overview of surface
properties of some materials.
The moisture layer between soil and soil-engaging tool surface
is predominately important to governing the adhesion force. These
forces could satisfactorily be explained using moisture tension and
surface tension information of the soil-solution at the interfacial layer.
Table 12.2. Contact Angles and Normal Soil Adhesion of Some Materials
of Agricultural Interest
Contact angle∗1 (degree)

Normal soil
Advancing Receding adhesion∗2
Material (θA ) (θR ) (kPa)
Hardened and low-tempered Steel-45 71 0 1.47

Hardened and low-tempered 73 0 1.46
medium carbon steel
Iron base alloy-epoxy 83 0 0.82
Al2 O3 coating*3 82 0 1.05
PES-PTFE coating*3 105 93 0.25
Polycarbonate 84 68 —
Polyethylene (PE) 96 62 —
Polypropylene (PP) 108 — —
Polystyrene (PS) 91 84 —
PTFE 113 103 0.26
Teflon 109 106 —
UHMWPE sheet*4 94 75 0.32
∗1 contact angle of distilled water at 20◦ C by sensible drop method

∗2 disk area = 1,963 mm2 ; M.C. = 26.8% (d.b.); Soil [11% clay, 64% silt, 25%
sand]; Soil was cut with 0.2 mm thick steel wire; Disks were pressed under 26.5 kPa
pressure and held for 15 s before pulled up
∗3 coating of 0.5 mm thickness
∗4 sheet of 3 mm thickness
(Compiled from Refs. 5, 8, 13)

Researchers demonstrated how moisture tension affects adhesion. The

tension in the water would attain similar value throughout the soil mass
at equilibrium condition. This condition of equilibrium is obtained by
automatic movement of water–air meniscus inside pore-channels when
a continuous film of water exists in a saturated soil. In order to equalize
tension in the continuous water film, such auto-adjustment of menisci
occurs in the channels whose dimensions correspond to the tension
governed by the pressure–surface tension relationship. Not all menisci
may reach the channel of appropriate size; such menisci have to assume
forced curvatures. The adhesion force per unit area of continuous water
film is thus generated by the moisture tension. Experimental results
established almost proportional relationship between moisture tension
(P) and adhesive force per unit area. Upon vertical pulling at the joint,
water attempts to travel through soil into the water film and failure
results when the exerted force exceeds the moisture tension. The high
energy layer existing in the water film will become the rupture layer.1,8
No surface tension exists in the interfacial liquid film when the
clearance between two solid surfaces is completely immersed in a
liquid.3 In soils partially saturated with water there is moisture tension,
which is equal in magnitude but opposite in sign to the soil water
pressure. The force that holds moisture in the soil per unit area is called
soil moisture tension (SMT). SMT is generally measured in units of
pressure (or suction) called bars. One bar equals one atmosphere of
pressure (14.7 pounds per square inch) or 1,030 centimeters of water
column height. Moisture tension is equal to the pressure that must be
applied to the soil water to bring it to a hydraulic equilibrium, through
a porous permeable wall or membrane, with a pool of water of the same
composition.19,20
Air squeezes in the interfacial zone of continuous water film
between the soil–tool system and forms capillary tubes between
neighboring asperities between soil layer and tool surface. The value
of the receding contact angle (θR ) of water on tool surface determines
the magnitude of capillary pressure; two surfaces having similar θR ,
would be having identical capillary attraction.5
Reference 9 depicted soil particles as spheres of identical dimen-
sions — a geometrically simplified shape. They described soil adhesion
for two states: for a continuous water film state and for water menisci
contact state. In their findings, they reported water film capillary

pressure as a source of adhesion in the former case, while both capillary
pressure and surface tension were the sources in the latter case. Moisture
tension and the capillary pressure are due to the surface tension.3
iv. Viscous resistance

Viscosity is the property which opposes relative motion between two
layers of liquid.4 Viscosity and loading rate (travel speed) affect the
adhesion.1 Reference 6 confirmed that the normal tensile force required
to separate two solid surfaces depends on the speed of separation.
Viscous resistance plays an important role in the normal adhesion (Fa )
between two solid plates completely immersed in liquid.
Viscous resistance also plays an important role in soil–tool adhesion,
especially with a continuous water film.5 Viscous resistance is developed
due to the continuous water film at the interface. Since the water film
is not uniformly distributed, viscous resistance too is uneven and varies
from place to place. This uneven distribution results in the pulling off
of some of the micro-sized soil particles and water molecules. Material
properties of the tool surface govern the amount of such soil particles
pulled off, which is different for different material.13 Some materials,
including polyethylene, provides a soil-free surface, while some show
soil-affinity. Viscous resistance largely depends on surface activity of
the soil and tool surface and thickness of the interfacial water film.
A system having higher activity of soil-tool with thinner water film
would have greater viscous resistance than a system with smaller activity
and thicker film. The system with high capillary attraction and high
viscous resistance will obviously exhibit greater soil–tool adhesion.5
12.2.4. Soil–tool surface morphology and contact

models
Reference 3 described surface morphologies of soil at contact interfaces.
The roughness of soil surfaces formed at the contact interface were
found in following three sizes:
(a) The micro-aggregate
(b) The particle
(c) The asperity on particles
Status of water film at the interface differs with soil moisture content.
With high moisture content or at higher normal loads, the interface
is filled with water and a continuous film links soil to tool surface.
In contrast, with low moisture contents or at lower normal loads, a non-
continuous water film exists. Based on the surface morphology at the
contact interface, the following soil–tool contact models at the soil–tool
interface could be defined for non-continuous water film interface.3,13
(i) Completely non-contacting asperity
(ii) Water-point contacting asperity
(iii) Water-loop (water-meniscus) contacting asperity
(iv) Water-film contacting particles and micro-aggregates
(v) Water-film contacting clods
For a non-contacting semi-spherical asperity (radius R), attraction
pressure per unit area of separation P (d) can be expressed as function of
its separation (d) between asperity summit and the solid surface using
the Lennard-Jones function, as cited by Ref. 10.
The energy of adhesion will be Ea , at equilibrium intermolecular
distance de .
Thus, E(d) = Ea at d = de ,
3 9
8Ea de de
P (d) = − , (12.5)
3de d d
where,
de = equilibrium intermolecular distance.
For a single asperity without any contact (Fig. 12.3), the adhesion force
(Fa ) can be expressed as,

8 de 2 1 de 8
Fa = πREa − , (12.6)
3 d0 4 d0
where,
d0 = separation of the tip of the summit from the solid surface.
If the separation is much larger than the equilibrium intermolecular
distance, it will be a completely non-contacting asperity and the
adhesion force would be negligible.
Thus, for do de ; Fa ≈ 0.
R
Asperity
Tool surface
d0
Tool
Fig. 12.3. Contact model schematic at soil–solid interface for completely non-
contacting asperity (figure not to scale).
Asperity
Tool surface
Tool
Fig. 12.4. Contact model schematic at soil–solid interface for water point contacting
asperity (figure not to scale).
For single asperity contacting solid at one point (Fig. 12.4),

do = de ,
Fa = 2πREa (12.7)
For a water-loop contacting asperity (Fig. 12.5), contact angle of water
on soil is zero.
For water film contacting particles and micro-aggregates
(Fig. 12.6), the film links the soil particle to a solid surface in an
irregular loop. It takes a very short time for Laplace pressure to attain
equilibrium.
For water film contacting clods, it takes a longer time for Laplace
pressure to attain equilibrium in a continuous film between the solid
surface and clods. To describe the adhesion mechanism in such models
Laplace pressure, meniscus tension and viscous resistance should be
included. Air enters the interface in a continuous water film, which
forms a capillary tube between adjacent asperities (radius R) and the
R
Asperity
Tool surface
Tool
Fig. 12.5. Contact model schematic at soil–solid interface for water loop contacting
asperity (figure not to scale).
Fig. 12.6. Schematic of contact model at soil-solid interface for water film contacting
soil particles and micro-aggregates (figure not to scale).
solid surface while pulling solid from soil. In a special condition of such
a model, adjacent asperities are tangents and their tips touch the solid
surface (Fig. 12.7).
12.3. Microscopic Forces at Soil–Tool Interface
12.3.1. Background
Soil–tool interaction has been a prime concern of research about soil-
engaging tools used both in agricultural and construction machinery.
Forces acting on a soil-engaging tool can be broadly understood by
classifying them into two groups: macroscopic — the forces extensively
depend on the system parameters including tool type, tool speed,
soil properties, and operating depth; and microscopic — the forces
largely rely on intrinsic properties of tool surface, soil composition,
physical and chemical properties of soil and molecular phenomena.
Practical difficulty and complications experienced by researchers in
demarcating adhesive and frictional forces at the soil-tool interface
Asperity 1 Asperity 2
R R
Tool surface
Tool
Fig. 12.7. Formation of capillary tube between two tangent adjacent asperities and
tool surface (figure not to scale).
led to a newer approach of grouping such forces into mutually

perpendicular components: normal (N) and tangent (τ) at the interface
for determining total microscopic force (F). Normal force acting at the
interface comprises three primary components: normal gravity, normal
adhesion and normal friction, whereas tangent resistance constitutes
three components: tangential friction, drag and tangential adhesion.
Earlier research on tangent resistance has clearly indicated that it
is composed of two components, frictional and adhesive.21 In Ref. 22,
the distinct element method (DEM) was used to analyze soil forces;
it was successful from both the soil behavior and reaction points
of view.
The practical difficulty and natural complication of clearly demar-
cating the adhesive and frictional components21 of the total microscopic
forces are behind the need for a simultaneous approach of analysis.
Based on various theories previously suggested, the present description
attempts to classify such microscopic forces at the soil–tool interface
into normal forces and tangential resistance and consequently elabo-
rates their individual nature and constituents.
Scientists7−9,21,23 proposed mechanical models to analyze soil
resistance at the soil–tool interface. Such models assume cutting of soil
by a planer tool surface, which could reasonably represent the actual
mechanism.21
Greater maneuverability can be achieved upon knowing micro-

scopic forces at the soil–tool interface to reduce the draft requirement
by lowering various components contributing to the total interfacial
force. Microscopic forces are both important and necessary to address24
while designing a soil-engaging tool, especially in soils with high clay
content. Many researchers have proposed different models to analyze
such forces with particular emphasis on normal adhesion,25 tangential
adhesion,1,9 chemical adsorption,26 friction,27 capillary attraction28−30
and liquid bridge force.31−34 The present study combines intermolecu-
lar, gravitational, adhesive, capillary, frictional and viscous forces occur-
ring at the planer soil–tool interface. A theoretical model is proposed
after reviewing existing principles and grouped under two mutually
perpendicular components viz. normal and tangent to the interface.
In this analysis, following assumptions are used:
(a) To avoid complexity of analysis, the interface is assumed to be a
planer surface.
(b) The total gravitational stress at the interface is mainly due to soil
particles and water-filled soil pores. The maximum gravitational
force can be calculated when all soil pores are filled with water,
since voids have no contribution to it.
(c) Soil particles are spherical, smooth and identical. Deformation
at the interface is neglected. For simplicity in computation, soil
particles are assumed to be un-clustered.
(d) Tool surface is smooth and free from micro-irregularity.
(e) Contact angle of water on soil is zero.
(f) Surface tension of water remains constant, irrespective of temper-
ature change.
12.3.2. Normal forces acting at soil–tool interface

Normal forces (N) acting at the interface have three components, viz.
normal gravity, normal adhesion and normal wet friction.
12.3.2.1. Normal gravitational force (NG )

Contact area at the soil–tool interface is a key parameter affecting
normal forces at the tool surface. A representative soil sample can
be cut with a thin steel wire13 to examine spatial constitution of soil

particles and pores. A unit cross section at the soil–tool interface can
be considered, where soil pores randomly occupy p% of the total cross
sectional area at the interface. Thus only (1−p) fraction of the area falls
under direct soil–tool contact.
If Si is the area of an individual soil pore, the total porous area at the
interface will be the summation of all such pores. It should be noted
that these randomly distributed pores may have different individual
dimensions. Thus the total porous area at the interface Sp ,

Sp = Si . (12.8)
The total porous area Sp and the total area of soil–tool contact at the
interface (Ss ) can be related to the interfacial cross sectional area (S) as
Sp = p.S, (12.9)
Ss = (1 − p).S, (12.10)
where,
S = Sp + Ss . (12.11)
The maximum gravitational force acting normally downward at the
interface can be expected when all the soil pores are filled with water.
If σp is the mean gravitational stress of soil pores, and σs is the mean
gravitational stress of soil particles, then the total gravitational force at
interface NG , can be given as,
NG = Np + Ns , (12.12)
where,
Np = Weight of water-filled soil pores
Ns = Weight of soil particles
Or,
NG = σp .Sp + σs .Ss . (12.13)
Gravitational force of soil particles also contributes to the frictional
component of normal forces, whereas water-filled soil pores play a role
in the adhesive component of normal forces. Substituting values of Sp
and Ss from Eqs. (12.9) and (12.10), Eq. (12.13) can be reduced for
total interface area (S) as
NG = σp .p.S + σs .(1 − p).S, (12.14)
NG = σs .S − (σs − σp ).p.S, (12.15)
12.3.2.2. Normal Adhesive Force (NA )

Normal adhesive force at the soil–tool interface depends on various
factors including soil properties, tool material, tool surface, soil water
content and interfacial conditions. Reference 35 claimed that soil
adhesion to the soil-engaging tool increased with normal stress on it.
The normal adhesion can be treated as having action similar to that of
normal load at the interface.1,8
Soil water content variation, at the micro level, causes significant
alteration in status of soil particles and its respective interface with
the tool surface. Reference 25 revealed a direct relationship of soil
adhesive force with the interface situation between soil and tool surface.
References 21 and 35 described such interfaces at four water content
levels, which are summarized in Table 12.3.
Various possible contact models of soil particles on tool surface
at different water content stages are shown in Fig. 12.8. Dry friction
dominates between soil particles and tool surface for the oven dried
bare soil particles and particles enveloped by closely confined water.
With increase in water content up to a certain limit, viscous forces and
wet friction also become significant.
For some physically isolated soil particles, an oven dried bare condi-
tion can be assumed, where theoretically zero water content attributes
to the maximum water adsorptive tendency (Figure 12.8a). Depending
on the system parameters, such particles rapidly get converted into
other water stages. Friction generated from the rubbing and scratching
action of dry soil particles on the tool surface27 is the major contributor
to interfacial forces, while the viscous, adsorptive and adhesive forces
might be neglected for this water stage.21
In the process of adsorption, at the water content between maxi-
mum water adsorption to maximum molecular holding (Figure 12.8b),
a bare particle is bounded with only a few water molecule thick
water layer (≈10−7 cm),4,35,36 supported by hydrogen bonds. Strong
intermolecular force between soil and closely confined water molecules
neither facilitates viscidity nor dissolving. No water film is formed
at the soil–tool interface. Behavior of this stage can be characterized
intermediately between its former and latter water stages.
Table 12.3. Status of Soil Particle and Soil–Tool Interface with Soil Water
Variation
Contribution
Soil water Status of Soil–tool Water status to interfacial
content range soil particle interface at interface forces
(a) Zero — Max Oven dried Bare soil No water • Dry friction
water bare particle– • Small
adsorption particle tool adhesion
content interface
(b) Max water Enveloped by Closely Film-like • Physico-
adsorption closely confined water chemical
content — confined water adsorption
Max film-like enveloped of soil
molecular water soil particles
holding particle–
water tool
content interface
(c) Max Enveloped by Water film Water film • Physico-
molecular water film enveloped produced chemical
holding soil by adsorption
water particle– film-like of soil
content — tool water particles
Field interface • Capillary
holding negative
water adsorption
content • Meniscus
adhesion
(d) Capillary Enveloped by Gravitational • Meniscus

water — gravita- water tension
Saturated tional enveloped • Viscidity
water water soil • Wet friction
content particle– • Imperfect
tool capillary
interface adsorption
A further increase in water content brings soil particles at the

interface under an envelope of water film of 10–20 water molecules
in thickness (≈10−6 cm).21 Physico-chemical adsorption is the main
contributing factor at this water content stage (Fig. 12.8c). The
Fig. 12.8. Various possible contact models of soil particles on tool surface at different
water stages: (a) adsorption water; (b) molecular water; (c) field holding water and
(d) gravitational water; 1 — soil particle; 2 — tool surface; 3 — soil-tool interface;
4 — closely confined water; 5 — water meniscus and 6 — gravitational water (figure
not to scale).
intensity of physico-chemical adsorption greatly depends on soil com-

position. For soils with high clay content, about 30% of total soil water
is in the form of water film, whereas it is 15% in loamy soils and only
1.5% in sandy soil.21
Increasing water content introduces gravitational water, allowing
capillary rise and meniscus formation. Meniscus tension, viscous force
and wet friction are summing up to total interfacial forces at this water
content stage (Fig. 12.8d).
Reference 3 described surface morphologies of soil at contact
interfaces. The authors reported that roughness of soil surfaces formed
at the contact interface could be found in three sizes, micro-aggregate,
particle, and asperity on particles. The soil particles, to avoid complex-
ity, could be assumed to be smooth and spherical.25 Soil water content
directly governs interfacial contact condition. As a broad classification
there can be soil–tool contact with or without water ring.
The normal adhesion force at the interface comes from forces

mainly caused by intermolecular attraction of bare soil particles (NAs ),
attraction of water meniscus (NAm ), attraction of water film due to
viscosity (NAv ) and capillary negative adsorption (Nca ).
Normal Adhesion Caused by Intermolecular Attraction of Bare Soil
Particles (NAs )
Considering the contact models where situation states absence of water
between soil particle and tool surface (Fig. 12.9 (a) and (b)), inter-
molecular attraction is caused by molecular interaction between them.
Fig. 12.9. Contact models of soil–tool interface: (a) completely non-contacting

asperity without water ring; (b) water-point contacting asperity without water ring; (c)
water-ring contacting asperity; (d) water-loop contacting asperity and (e) continuous
water film (figure not to scale) (modified and redrawn after Ref. 25).
For a non-contacting (Fig. 12.9 (a)) spherical asperity (radius R),

attraction pressure per unit area of separation (unit area of separation
is defined as the separation of the tip of the summit to the tool
surface) P (d), can be expressed as a function of its separation (d)
between asperity summit and the solid surface using the Lennard-Jones
function13
∂E(d)
P (d) = , (12.16)
∂d
where E(d) is interaction energy per unit area of separation. The inter-
action energy will be the energy of adhesion (EA ) at the equilibrium
intermolecular distance de ,13
3 9
8EA de de
P (d) = − ; E(de ) = EA at d = de ,
3de d d
(12.17)
The interaction energy per unit area of separation between two parallel
planes, as a function of d, was defined by Lifishitz’s theory18
hω
E(d) = , (12.18)
16π2 d 2
where
hω = Lifishitz–Van der Waals constant.
At the separation equal to intermolecular distance, E(de ) gives the work
of adhesion,25
hω
WA = E(de ) = EA = , (12.19)
16π2 de2
where EA was expressed as25

hω R d0 + R
EA = − ln . (12.20)
8π d0 d0
The normal adhesion caused by intermolecular attraction force of
bare soil particles (NAs ) is expressed as25

−∂EA −hω R R
NAs = = − , (12.21)
∂d0 8π d02 d0 (d0 + R)
where,
d0 = Separation between the summit tip and tool surface.
The negative sign in Eq. (12.21) shows that NAs is of attraction
characteristics, i.e. if tool is assumed stationary then soil is attracted
towards the tool surface.
From Eqs. (12.19) and (12.21), magnitude of NAs ,25

1 1
NAs = 2π.WA de R. 2 −
2
. (12.22)
d0 d0 (d0 + R)
For a water-point contacting asperity without water ring (d0 = de ), the
magnitude of normal adhesion due to intermolecular interaction can
be expressed by simplifying Eq. (12.22) as25

de
NAs = 2π.WA R. 1 − at (d0 = de ). (12.23)
(de + R)
Alternatively, if the Lifishitz-Van der Waals constant is unknown or
difficult to calculate, Eq. (12.22) can be expressed in terms of EA ,13

8 de 2 1 de 8
NAs = π.R.EA − . (12.24)
3 d0 4 d0
In Eq. (12.24), for a reasonable approximation, if no water molecule
is absorbed at the surface of the soil particle, then EA can be expressed
as the work of adhesion (WA ),13
EA = WA = γSV + γLV − γSL , (12.25)
where,
γLV = Surface tension at liquid–vapor interface
γSL = Surface tension at solid–liquid interface
γSV = Surface tension at solid–vapor interface.
But the surface of the soil particle normally absorbs at least a layer of
water molecules. EA will then be expressed as13
EA = 2γLV . (12.26)
Normal Adhesion Caused by Water Meniscus (NAm )
When the water film becomes thicker, gravity water also joins the
interfacial film, and normal adhesion is produced by the meniscus
adjacent to the soil particle.21 Water ring formed between the soil
particle and planer tool surface (Fig. 12.9 (c) and (d)) results in
meniscus tensile force, acting in the direction of surface tension, i.e.
tangent at the meniscus–tool contact. Assuming the contact angle of
water on soil particle as zero, θ is the contact angle between meniscus
and tool surface, Reference 25 suggested an expression for magnitude

of adhesion force (FAm ), in the direction of surface tension,
2π.R.γ
FAm = (1 + Cos θ), (12.27)
1 + db0
where,
FAm = Adhesion force of water meniscus, in the direction
of surface tension
b = Vertical distance between soil particle surface to its lowest tip
The vertical component of Eq. (12.27) will give magnitude of the
normal adhesion caused by water meniscus (NAm ),
2π.R.γ
NAm = FAm. Sin θ = (1 + Cos θ).Sinθ. (12.28)
1 + db0
Consequently, Ref. 37 explained that the ring-shaped liquid wedge

(density ρ2 ) between a solid spherical particle (radius R, density ρ1 )
and a horizontal tool surface cause the downward force to increase by
the amount, which is the normal adhesive force (NAm ),
π
NAm = 2πRγSin2 ϕ − (ρ2 − ρ3 ).gR 3 (1 − Cos ϕ)2 (2 + Cos ϕ)
3
− (πR 2 Sin2 ϕ.PAC ), (12.29)
Where
NAm = Normal adhesion caused by water meniscus
ρ3 = Density of second fluid surrounding the whole system
ϕ = Angle, with vertical, made by the radius with the point
at which the liquid surface meets sphere
PAC = Negative pressure existing at the liquid wedge at the level AC
The first (positive) term in Eq. (12.29) represents the downward com-
ponent due to surface tension; the second (negative) term represents
the upward force experienced by the volume ABC, which is submerged
in liquid; and the third (negative) term represents the upward force
resulted from the negative pressure PAC existing in the liquid wedge at
level AC, relative to the pressure in the third fluid (density ρ3 ) at the
same level,37
−γ
PAC = , (12.30)
RAC
where
RAC = Radius of curvature of the interface at point A.
For small values of R and ϕ, and zero contact angles, circular profiles
are obtained. Thus, in such cases, the radius of this circular meniscus
(r) is given as37
(1 − Cos ϕ)
r= .R, (12.31)
(1 + Cos ϕ)
and the negative pressure in the wedge, relative to the outside, can be
given as37
γ
P − P0 = PAC = − , (12.32)
r

γ (1 + Cos ϕ)
PAC = − . (12.33)
R (1 − Cos ϕ)
Substituting the values of r and PAC in Eq. (12.29) and neglecting the
second term for small values of ϕ, we get
NAm = πRγ(1 + Cos ϕ)(3 − Cos ϕ). (12.34)
The normal adhesive force (NAm ) between the soil particle and plate
can be computed as per the concept of capillary rise, which states that
the force of adhesion will be equal to the contact perimeter (2 * 2πR),
surface tension (γ) and the cosine of the contact angle (θ).1
Thus,
NAm = 4πRγCos θ. (12.35)
The maximum normal adhesive force obtained24,25,29,37 would be
(NAm )max = 4πRγ. (12.36)
Normal Adhesion Caused by Attraction of Water Film Due
to Viscosity (NAv )
Viscosity is the property that opposes relative motion between two
layers of liquid,4 and it can be understood as the internal friction of
the liquid that brings resistance to flow.38 Reference 1 revealed that

viscosity and loading rate affect adhesion. For the contact model with
water film between two parallel solid planes (Fig. 12.9e), an analogy
of two circular discs of radius R immersed in a liquid of viscosity η
could be applied.25 If they are subjected to a tensile force (NAv ), their
separation increases from h1 to h2 , just before the liquid film does not
break during pull. If the time required to pull is t , the NAv can be
expressed3,6,28 as

3π.η.R 4 1 1
NAv = − . (12.37)
4t h12 h22
Normal Adhesion Due to Capillary Negative Adsorption (Nca )

Soil porosity contributes an important role to adhesive forces. These
pores are either filled with air or water. Soil pores can be classified into
three categories,36 namely inactive pores (equivalent diameter <2 µm),
capillary pores (equivalent diameter 2–20 µm) and air pores (equivalent
diameter >2 µm). The soil particles interact with soil water through
either physico-chemical or physico-mechanical processes. The friction
and adhesion of air can be neglected whereas inactive pores contribute
little either. The adhesion between capillary pores and tool surface is
noticeable and is mainly due to the capillary adsorption.
Upon immersing the tube (radius Rc ) in a liquid (density ρ), the
capillary rise of liquid will take place due to the surface tension of
the liquid (γ), and a meniscus will form corresponding to the contact
angle (θ) of the liquid on the tube material. The height of capillary
rise (h) can be determined at its equilibrium condition, where upward
force caused by surface tension would balance weight of the lifted liquid
column,
2πRc γCos θ = πRc2 ρgh, (12.38)

2γCos θ
h= . (12.39)
ρgRc
A pressure difference (P )exists at the two surfaces of the meniscus,

which causes capillary ascent, and is expressed by the Young–Laplace
Fig. 12.10. Force analysis for snipped capillary. (a) Capillary tube in liquid reservoir
with a snipping plane; (b) hypothetical snipped capillary section in air; (c) snipped
capillary section on tool surface and its equivalent soil particle (figure not to scale).
(Modified and redrawn after Ref. 21.)
equation, at static condition (Fig. 12.10a),
2γ.Cos θ
P = P − P0 = , (12.40)
Rc
where
P = Pressure inside liquid column near the meniscus

P0 = Atmospheric pressure
When a tool acts on soil mass, the process can be regarded as shearing of
soil capillaries at the interface.21 Reference 39 described the mechanism
of capillary snipping in their snipping-capillary test. To understand the
process, a section of ascending sheared capillary (Fig. 12.10b) can be
considered as having two menisci — upper and lower. Applying the
Young–Laplace equation to calculate pressure difference at the upper
meniscus (P1 ) and at the lower meniscus (P2 ), we get,

−2γ.Cos θ1
P1 = P1 − P0 = , (12.41)
Rc
−2γ.Cos θ2
P2 = P2 − P0 = , (12.42)
Rc
where P1 and P2 are pressures inside the liquid near the upper and lower
meniscus, respectively, and θ1 and θ2 are contact angles of liquid with
tube at the upper and lower meniscus, respectively. For an ascending
capillary, P1 and P2 should obviously be negative.
At any instant, the net ascending pressure difference for the liquid
segment should balance the static pressure of lifted liquid column
(Pstatic ) that is the weight of liquid column segment. Thus, for the
instantaneous partial equilibrium,
P1 − P2 = Pstatic . (12.43)
If s is the distance between the lowest point of the upper meniscus and
the shearing plane,

−2γ.Cos θ1 −2γ.Cos θ2
− = ρgs. (12.44)
Rc Rc
Rearranging Eq. (12.44) for s , we get,
2γ.(Cos θ2 − Cos θ1 )
s= . (12.45)
ρgRc
Equation (12.38) suggests a positive value on the right-hand-side
of the equality sign for capillary rise (s > 0) and negative value for
capillary depression (s < 0). Soil capillary pores undergo a capillary
rise with water, thus the contact angle at the lower meniscus (θ2 ) will
always be smaller than the contact angle at the upper meniscus (θ1 ). It
reveals that the lower meniscus has a smaller curvature than that of the
upper.40
Figure 12.3c describes a schematic of a sheared capillary, which is
cut by a tool and closely touches its surface. The tool surface holds
capillary water tightly under the action of normal adhesion of water
meniscus (NAm ) on to it. The three agents act on the capillary water
viz. the upward pulling pressure at the upper meniscus (P1 ), the
downward normal adhesion of the meniscus by tool surface (NAm )
and the downward static pressure (Pstatic ). At the partial equilibrium

condition,
NAm
P1 − = ρgs. (12.46)
πRc2
But situations occur with shearing tools where interfacial water forms
a film and completely seals capillary pores at their contact with the
tool surface. Under such circumstances the net ascending pressure
difference (P1 − NAm /πRc2 ) will exceed the static pressure (ρgs),21
and a vacuum is created at the tool surface. Capillary adsorption of
negative air pressure can be envisaged on the entire area (πR2c ) encircled
by the meniscus. In absence of the atmospheric pressure at the lower
meniscus, P1 will be the Laplace pressure, thus,
NAm Nca
P − = ρgs + , (12.47)
πRc
2 πRc2
where,
Nca = Normal adhesion force caused by capillary negative adsorption.
Reference 29 confirmed the importance of understanding surface

tension to evaluate adhesion. Under saturated conditions water film
contacts tool surface in circular vicinity whose diameter was empirically
verified as 4R, where R is the radius of spherical water asperity
(Fig. 12.3c).
A careful observation of Fig. 12.10c reveals the relationship
between the capillary tube radius (Rc ) and the radius of corresponding
soil particle (R). The magnitude of NAm can be expressed as
NAm = 2πRc γCosθ. (12.48)
From Eqs. (12.40) and (12.48), Eq. (12.47) can be modified as

2γ.Cos θ 2γ.Cos θ2 Nca
− = ρgs + . (12.49)
Rc Rc πRc2
Substituting the value of static pressure from Eq. (12.44),
2γ.Cos θ 2γ.Cos θ2 2γ.Cos θ2 2γ.Cos θ1 Nca
− = − + . (12.50)
Rc Rc Rc Rc πRc2
Substituting θ1 → θ and θ2 → 00 for a snipped capillary with

completely sealed lower meniscus, we get
2γ.Cos θ 2γ 2γ 2γ.Cos θ Nca
− = − + , (12.51)
Rc Rc Rc Rc πRc2
Nca 4γ.Cos θ 4γ
= − , (12.52)
πRc
2 Rc Rc
which gives
Nca = −4πγRc (1 − Cos θ). (12.53)
A negative value of Eq. (12.53) reveals that it is a drawing vacuum,
which is caused by capillary adsorption of negative pressure at the tool
surface. The theoretical maximum value of Nca can be given as
(Nca )max = −8πγRc at [θ → 180◦ ]. (12.54)
Furthermore, using Eq. (12.39), Eq. (12.49) can be modified for
θ2 → 00 ,
2γ Nca
ρgh − = ρgs + , (12.55)
Rc πRc2
which results in
Nca = −[2πγRc − ρg (h − s ). πRc2 ]. (12.56)
12.3.2.3. Normal friction force (Nf )

The friction of an ascending capillary contributes to the wet friction
of interfacial water film meniscus and can be expressed as the normal
wet friction of meniscus at the soil tool interface (Nwfm ). Neglecting
contribution of the meniscus at the individual soil particle, the total
normal friction force (Nf ) reasonably assumed to come from Nwfm .
Reference 6 confirmed that the normal tensile force required to
separate two solid surfaces depends on the speed of separation. Viscous
resistance also plays an important role in soil–tool adhesion, especially
with continuous water film.5 Viscous resistance is developed due to
the continuous water film at the interface. Since the water film is not
uniformly distributed, viscous resistance too is uneven and varies from
place to place. This uneven distribution results in pulling off some of
the micro-sized soil particles and water molecules. Material properties
of the tool surface govern the amount of such soil particles pulled off,
which is different for different materials.10 Some materials, including
polyethylene, provide a soil-free surface, while some shows soil-affinity.
Viscous resistance largely depends on surface activity of the soil, tool
surface and thickness of the interfacial water film. A system having
higher soil–tool activity with thinner water film would have higher
viscous resistance than a system with lower activity and thicker film.
The system with high capillary attraction and high viscous resistance
will obviously exhibit greater soil–tool adhesion (Tong et al., 1999).5
When a liquid moves over a solid surface, a frictional force, which
acts at a right angle to the contact plane, opposes the motion.41 If Nwfm
is the wet frictional force per unit length of the line of contact, acting
normally, and θA & θR are advancing and receding contact angles,
γ
Nf = Nwfm = [Cos θR − Cos θA ]. (12.57)
2
The value of Nwfm is independent of direction of travel, i.e. Nwfm doesn’t
change when the direction of motion is reversed, which suggests that
the solid–vapor and solid–liquid interfaces rapidly attain equilibrium.
The above analysis reveals that the total normal force (N) acting at
the soil–tool interface comprises three major components: the normal
gravitational force (NG ), the normal adhesion (NA ) and the normal
friction (Nf ). Normal gravitational force consists gravity of soil particles
and the pores; normal adhesion comprises the normal adhesion from
bare soil particles, from water meniscus, from viscidity, and capillary
negative adsorption; whereas normal friction constitutes the normal
wet friction of water meniscus. Using the magnitudes of the afore-
mentioned relations, N can be expressed as
N = NG + N A + N f , (12.58)
N = [Np + Ns ] + [NAs + NAm + NAv + Nca ] + Nwfm , (12.59)


1 1
N = [σs .S − (σs − σp ).p.S] +  2π.WA de2 R. 2 −
d0 d0 (d0 + R)
 
 2π.R.γ 
3π.η.R 4 1 1
+ (1 + Cos θ). Sin θ + − 2
 1 + d0  4t 2
h1 h2
b

γ
+{4πγRc (1 − Cos θ)} + (Cos θR − Cos θA ) . (12.60)
2
Alternatively,

8 de 2 1 de 8
N = [σs .S − (σs − σp ).p.S] + π.R.EA −
3 d0 4 d0

3π.η.R 4 1 1
+ {πRγ(1 + Cos ϕ)(3 − Cos ϕ)} + −
4t h12 h22

γ
+ {2πγRc − ρg (h − s ). πRc } + (Cos θR − Cos θA ) .
2
2
(12.61)
12.3.2.4. Tangential forces acting at soil–tool interface

Tangential forces (τ) at the interface can be grouped under three
components, viz. tangential friction, drag resistance and tangential
adhesion.
12.3.2.5. Tangent frictional resistance (τf )

Neglecting friction generated by organic matter, impurities and the
soil pores at the interface, the tangent frictional resistance (τf ) can be
approximately equal to the tangent resistance due to the dry friction of
soil particles.
Tangent Resistance due to Dry Friction of Soil Particles (τdfs )
At very low water contents (water range — a in Table 12.3) with
negligible adhesion between soil particles and tool surface, bare soil
particles produce rubbing action against the surface. The dry friction,
caused by the bare soil particles (τdfs ), contributes to tangent frictional
resistance and acts in the lateral tangential direction, opposite to the
direction of soil travel over tool surface. If µdfs is the coefficient of dry
friction of soil particles on tool surface,27
τf = τdfs = µdfs Ns . (12.62)

12.3.2.6. Drag resistance (τdrag )

With increase in water content at the interface, free water is available
to undergo a relative motion with the moving tool. The backward
dragging force acts tangentially on any liquid layer. The drag resistance
(τdrag ) is mainly dominated by the tangential viscidity at the interface.
Drag Resistance Due to Tangential Viscidity (τtv )

Newton showed that viscous force is directly proportional to the
surface area (S) of the layer and velocity (v), and inversely proportional
to its distance (x) from the stationary layer. If the coefficient of
proportionality is η, which is the coefficient of viscosity, then,
dv
τdrag = τtv = −ηS . (12.63)
dx
12.3.2.7. Tangential adhesive resistance (τA )

Tangential adhesive resistance comes from surface tension and water
meniscus adhesion.
Tangential Adhesive Resistance Due to Surface Tension (τtAγ )

Relative motion of soil particles with respect to the tool surface distorts
the associated meniscus. In the situation of pulling a water film over a
tool surface, the meniscus obtains a distorted shape. Viewing from the
front of the interface, the contact angle there (θf ) will be larger than the
static contact angle (θ), while the contact angle at the back (θb ) will be
smaller (Fig. 12.11). Depending on the velocity of pull, the value of the
distortion changes and so with the tangential adhesion force (Ft ). With
higher speeds and higher normal adhesion force, it is possible to witness
θf as an obtuse angle and θb be zero.42 The horizontal component of
θb , along the interface plane, resists the motion and as θb approaches
zero, this component claims a higher contribution.21 If lt is the total
accumulative length of surface tension (γ) relative to θf and θb , then
τtAγ can be expressed as
τtAγ = [γb Cos θb − γf Cos θf ]lt , (12.64)

R
Before pull
R
During pull
b f
b
f
Fig. 12.11. Meniscus deformation producing tangent adhesion during pull (figure
not to scale).
where γb and γf are the back side and front side surface tension of the
liquid, respectively.
For simplicity, it can be assumed that γb = γf = γ,35 thus
τtAγ = γ[Cos θb − Cos θf ]lt . (12.65)
Tangential Adhesive Resistance Due to Water Meniscus (τtAm )

For the film-like water at the interface, water ring results in meniscus
tensile force, whose magnitude is expressed by Eq. (12.27). Com-
ponent of FAm in the direction parallel to the interface would give
tangential resistance due to water meniscus (τtAm ),
2π.R.γ
τtAm = FAm. Cos θ = (1 + Cos θ). Cos θ. (12.66)
1 + db0
Thus from Eqs. (12.65) and (12.66), the total tangential adhesive
resistance (τA ) could be expressed as
τA = [τtAγ + τtAm ] = γ[Cos θb − Cos θf ]lt
2π.R.γ
+ (1 + Cos θ).Cos θ. (12.67)
1 + db0
The above analysis reveals that the total tangent force (τ) acting at
the soil–tool interface comprises three major components: the tangent
frictional resistance (τf ), the drag resistance (τdrag ) and the tangent
adhesive resistance (τA ). Tangential frictional resistance consists dry
friction of soil particles; drag resistance contains the tangential viscidity;
whereas, tangent adhesive resistance constitutes the surface tension and
water meniscus. Using the aforementioned relations, magnitude of τ
can be expressed as
τ = τf + τdrag + τA , (12.68)
dv
τ = µdfs Ns + ηS + γ[Cos θb − Cos θf ]lt
dx
2π.R.γ
+ (1 + Cos θ).Cos θ. (12.69)
1 + db0
12.3.2.8. Total microscopic forces at interface (F )

Practical difficulty and complications experienced by researchers21 in
demarcating adhesive and frictional forces at the soil-tool interface led
a newer approach of grouping such forces into mutually perpendicular
components — normal (N) and tangent (τ) to the interface. The total
microscopic force at the interface (F), using equations Eqs. (12.58 &
12.68), can be expressed,

F = (N 2 + τ 2 ) (12.70)

F = (NG2 + NA2 + Nf2 ) + (τf2 + τdrag
2
+ τA2 ) (12.71)
12.4. Reducing Soil–Tool Adhesion
As realized from the above discussion, soil–tool adhesion is a highly

undesirable phenomenon, especially with sticky soils. Whether it is
farm implements in an agricultural field or earth-moving machinery

at construction sites, researchers have shown great interest in dealing
with this problem.
Toughness and hardness were considered as important material
properties that should be present in sufficient amount in soil-engaging
materials.43 In addition, those soil-engaging tools that are to be used
in soils rich in clay and colloids must have anti-adhesion ability.
Reference 44 demonstrated that reduction in soil adhesion could
increase tractive performance of soft ground terrain machinery. By
reducing soil adhesion, plowing resistance could be lowered.45−47
Decreasing soil adhesion enhanced quality of work with furrow
openers48 and covering shovels.34
12.4.1. Factors affecting soil–tool adhesion

Surfaces are never truly flat, i.e. there are micro undulations even on a
very smooth surface, and such geometrical roughness alters wettability
and thus affects soil adhesion. The effects of such roughness of tool
surfaces on soil–tool adhesion are a function of the chemical properties
of the surface, wettability and the moisture content of interacting soil.7
Soil–tool adhesion, at large, depends on the nature and properties of
the soil, properties of the tool material, and working conditions.
12.4.1.1. Effect of soil properties on soil–tool adhesion

Soil properties, for characterizing static and dynamic behavior of soil
reaction, can be visualized as: static state properties, including position,
structure and density; behavior capacity properties, comprising strength
and transmission; and material properties, consisting of texture, min-
erals and chemical composition.1
Among various other factors, the soil properties dominantly gov-
erning soil–tool adhesion were identified as: mineral composition, soil
texture, soil moisture content, organic matter content, soil porosity and
nature of soil solution. To our routine experience, one can verify the fact
that soil consisting of Illite-Montmorillonite, Sanidine, Tridymite and
Cristoballite (with rougher grain surfaces and higher fractal dimension
of PSD) shows greater adhesion than that with Kaolin, Feldspar and
Quartz. Soil adhesion is higher with clay rich soils than sandy soils.
Clay content increases the fractal dimension of soil PSD, which in
turn results in higher adhesion and frictional losses. Higher values of
molecular fractal dimension provide larger specific surface area, which
favors adhesion. Fine-textured soil exhibits a higher degree of adhesion
than coarse-textured soil. Similarly, soil moisture, in a certain range
(probably between plastic limit and liquid limit moisture contents),
favors adhesion.2,3,5
Table 12.4 exhibits some relative characteristics: particle size, pore
size, and capillary rise, for different soil solutions. It is evident from the
table that as the pore size reduces, capillary rises to higher heights.
Reference 4 described soil consistency as the manifestation of
the physical forces of cohesion and adhesion within soil at different
moisture content. Phenomena of soil consistency are friability, plasticity,
stickiness and resistance to compression and shear. A schematic plot
shown in Fig. 12.12 clearly reveals that cohesion (molecular attraction)
is higher in dry soils and decreases with moisture content. Adhesion first
increases and then decreases the consistency with increase in moisture
content. However, as consistency is the resultant effect of both soil
cohesion and adhesion forces, in the wet range, decrease in adhesion is
such that increase in cohesion due to increasing area of contact fails to
compensate for it. Thus after this certain moisture content consistency
decreases. The plot of soil consistency has two maxima and two minima.
Table 12.4. Capillary Rise in Various Soil Solutions
Soil solution Particle diameter (µm) Pore radius (µm) Capillary rise (cm)
Fine silt 8 2 750

Coarse silt 25 5 300
Very fine sand 75 15 100
Fine sand 150 30 50
Medium sand 300 60 25
Coarse sand 500 100 15
Very coarse sand 2,000 400 4
Fine gravel 5,000 1,000 1.5
(Source: Ref. 49)

Fig. 12.12. Soil consistency variation with moisture content — contribution of

cohesion and adhesion.
12.4.1.2. Effect of liquid properties on soil–tool adhesion

The soil solution in the soil–tool system can also alter adhesion values.
When interfacial water films are very thin, the molecular structure of
water varies due to the presence of Montmorillonite molecules. This
results in increased difficulty in distortion and breakdown of hydrogen
bonds. A general equation was proposed, to which thermodynamic,
dynamic and spectral properties of liquid follow in a similar trend,3,50

β
J = J0 exp , (12.72)
mw /mm
where,
J = Some property of interlayer water (including partial
specific heat, partial specific volume, and viscosity)
J0 = Same property of the water in bulk
β = A constant, which depends upon the solid surface
characteristics
mw
= Ratio of mass of water to mass of soil
mm
It can be easily visualized from Eq. (12.72) that the ratio (mw /mm ) has
an important contribution to alter the corresponding property. This
ratio can be expressed as the function of the thickness of interlayer
water; the thinner the water layer, the smaller the ratio, and hence the
larger (more severe) the property of the water film.
12.4.1.3. Effect of tool properties on soil–tool adhesion

Both tool geometry and tool material affect the amount of soil adhered
to it. A small tool will certainly have less adhesion force as compared to
a giant tool that provides larger area for soil to travel along. Similarly,
surface geometry of the tool may alter adhesive forces.
Since most tool designs are based on the work they are intended for,
including soil adhesion into design criteria to bring a major change in
tool geometry is neither practical nor wise. Despite research on minor
design modifications in surface geometry, significant attention has been
paid to altering the surface material instead.
When solids are immersed in a soil solution, their physical and
chemical properties are changed because of physical adsorption, chem-
ical adsorption, chemical reaction and soil adhesion. Based on the
surface free energy, surfaces could be LES (organic compounds and
organic polymers) or HES (metals, metallic compounds, inorganic
compounds including oxides, nitrides, silica and diamond). HES have
surface tensions ranging from 0.2 to 0.5 N/m, whereas LES have
surface tensions less than 0.1 N/m. Water is a low surface energy
substance with surface tension of 0.0728 N/m at 20◦ C.3
Affinity towards water is known as hydrophilicity and is expressed
by the contact angle of water on the surface. A larger contact angle
makes the surface hydrophobic. Hydrophobic materials show less soil
adhesion, which exhibits the inverse relationship between contact
angle and adhesion. HES are hydrophilic surfaces while LES are
hydrophobic.3
12.4.1.4. Effect of working conditions on soil–tool

adhesion
Working conditions affecting soil–tool adhesion may include normal
load, loading rate, loading time, temperature, electro-magnetic field,
and vibration. Sliding resistance, for most of the materials, monotoni-
cally increases with normal load. But soil adhesion increases with normal
load and then decreases, probably because of the interfacial lubrication
by water squeezed out of the soil solution.3,51
Temperature elevation results in decreased surface tension and

viscosity of water, which eventually shows less adhesion.
12.4.2. Means of reducing soil–tool adhesion

Peasants and construction workers had started addressing the problem
of soil–tool adhesion since earlier centuries, however modern attempts
are more scientific with an additional advantage of technological
advancement. Better understanding of the mechanisms involved in the
soil–tool system has also contributed positively. Table 12.5 presents
an overview of historical developments towards reducing soil adhesion
through various means. This section briefly describes major avenues
examined for reducing soil–tool adhesion.
12.4.2.1. Altering surface design

Altering surface design of soil-engaging tools based on kinematic and
dynamic analysis of soil flow over tool surface was the conventional
method of reducing soil–tool adhesion. Additional mountings on
surfaces could decrease the real contact area and break the continuity
of the water film. Scales (Fig. 12.13), convex corrugations and comet-
type oblique holes (Fig. 12.14) drilled through the plowshare and the
breast of the moldboard were such attempts.23,84,69
12.4.2.2. Heating
In early 19th century, Ref. 54 used thermal heating of plows for
reducing soil adhesion and improving scouring. This method was
based on the concept of lowering the surface tension by increasing
temperature and thereby reducing adhesion. However the experiments
couldn’t lead to any significant decrease which, according to Ref. 1,
was probably due to lack of uncontrolled experimental conditions.
12.4.2.3. Lubrication
Lubricants, as with those used with mechanical parts, attracted suf-
ficient attention of researchers to test their effectiveness with sticky
soils; however, possible contamination of topsoil and cumbersome
handling of continuity in lubricant supply are probably the reasons
Table 12.5. A History of Steps Taken for Soil–Adhesion and Contemporary

Milestones∗
Year∗∗ Event∗∗ Reference
1907 Leo Bakeland fabricated the oldest recorded Lewis, 2005 [52 ]
synthetic plastic, known as Bakelite
1918 Investigation of various moldboard materials Bacon, 1918 [53 ]
ranging from glass to plaster of Paris to hog
hides in waxy soils of Texas
1920 • Heating of plow Bacon, 1920 [54 ]
• Staudinger published his classic paper Uber Lewis, 2005 [52 ]
Polymerization
1927 Mass production of PVC started Lewis, 2005 [52 ]
1930 Polystyrene (PS) invented Lewis, 2005 [52 ]
1938 Nylon introduced Lewis, 2005 [52 ]
1941 Poyethylene (PE) developed Lewis, 2005 [52 ]
1960 Airblast proposed to reduce soil–metal sliding Bertelsen, 1960 [55 ]
resistance of a moldboard bottom → Aeroplow
1961 PTFE, Teflon (tetrafluoroethylene) and PE plastics Cooper and
used on moldboard plow in sticky clay soils in McCreery,
Georgia could reduce draft by 25% 1961 [56 ]
1965 Teflon-covered tillage tools Fox and Bockhup,
1965 [57 ]
1968 • Teflon-covered plow bottom could reduce draft Wismer et al.,
by 23% 1968 [58 ]
• Water addition through small holes increased Gill and Vanden
scouring quality of clay soils Berg, 1968 [1 ]
1973 Liquid polymer used as lubricant on bulldozer Pittsburg and
blade in mines Midway Coal
Mining Company,
1973 [59 ]
1975 Sliding coefficients of liquid polymer lubricated Schafer et al.,
steel-on-soil evaluated and effectiveness of 1975 [60 ]
polymer as lubricant investigated — could cause
10–20% reduction in draft of moldboard with
adhesive soil
1977 • Polymer lubrication with motorized injecting Schafer et al.,
system, mounted on tractor, could reduce 1977 [61 ]
plowing draft of moldboards 15–32%
• High frequency Ultrasonic vibrations Sharma et al.,
1977 [62 ]
1982 Electro-osmosis with electric field application Shen, 1982 [63 ]
1984 Air-jet and water injection at soil–tool interface Araya and Kawanishi,
1984 [64 ]
(Continued )
1988 Ceramic tile, Teflon tape, silicon lubricant oil, lead Salokhe and
oxide paint, gloss paint, varnish, chromium plating Gee-Clough,
and enamel coating of cage wheel lugs 1988 [65 ]
1990 • Lubrication with liquid substances secreted on Chen et al., 1990;
body surfaces of soil animals [Biomimetics] Li et al.,
199066,67
• Enriching cast iron by phosphorus and silicon — Tong, 199068
white iron
• Surface design modifications (convex corrugations Cong et al., 1990;
and comet-type drilled hole on moldboards) could Zhu et al.,
reduce plowing resistance by 2.5–3.5% in moist 199223,69
fields and by 8–12% in paddy fields
• Flexible floor bucket could reduce soil accretion by Ren et al., 199070
59.4% and loading resistance by 15.8%
• Flexible tongue scraper mechanism for loader Yin et al., 199071
bucket could increase loader productivity by 20%
1991 Magnetic field on plows could reduce plowing Han and Zhang,
resistance by 30–55% and tractor fuel consumption 1991; Zhang
by 11–30% and Han,
199245,34
1992 • Biomimetics Ren et al., 1992;
Tong et al.,
1994; Li et al.,
199572,13,73
• Enamel coating on moldboard plow could reduce Salokhe et al.,
specific draft 8–23%; on disc plow could reduce 1992; Salokhe
4–21% and Shirin,
199274,75
• Modified and unsmoothed moldboard plow; Qaisrani et al.,
ceramic coating (2–12%), enamel coating (2–16%), 199246
steel convex domes (1–30%), UHMWPE domes
(1–34%) reduction in plowing resistance
1993 • Heat treatment of metals (tempering) Li et al., 199351
• Embossed bulldozing blade with UHMWPE Qaisrani, 199376
convex domes could reduce bulldozing resistance
by 34%
1995 • Electric field on non-smooth surface Cong et al., 1995;
electro-osmosis Chen et al.,
1995; Ren et al.,
199577,78,79,12
(Continued )
• Magnetized plowshare Guo and Liu,

199512
• Embossed bulldozing blade with convex domes Ren et al., 199579
could reduce bulldozing resistance by 18%
1996 Mechanical vibrations Wang, 199680
1997 Flexible body surfaces [Biomimetics] Ren et al., 1996;
Wang et al.,
199781,82
1999 Polymers (PTFE, UHMWPE, PES-PTFE), Tong et al.,
plasma-sprayed coatings, enamel coatings, 1999 [5 ]
2001 Bionic electro-osmosis on loading shovel could Ren et al., 200183
reduce adhesion at low energy consumption
2006 Theoretical analysis of microscopic interfacial forces Soni and Salokhe,
at soil–tool interfaces — considering normal 200687
gravity, normal adhesion and normal friction for
Normal force; and tangential friction, drag and
tangential adhesion (Tangential force)
• Normal force
• Tangential force
2006 Height-to-Diameter ratio was introduced to Soni and Salokhe,

UHMWPE protuberances 200688
• With Bangkok clay soil, protuberances with HDR
<0.5 lowered sliding resistance by 10%–30% and
reduced normal adhesion by 10%–60%.
2007 Modification of moldboard plow surface using Soni et al., 200789
arrays of UHMWPE protuberances
• Percent reduction in plow resistance of bionic
moldboard plow with HDR = 0 was 1–6% in dry
soil, 16–22% in sticky soil, 14–20% in wet soil
and 8–12% in flooded soil.
• With HDR = 0.25 plowing resistance reduced
by 2–7% in dry soil, 18–36% in sticky soil,
17–33% in wet soil and 15–28% in flooded soil.
• With HDR = 0.5, reduced by 10–16% in sticky
soil, 6–17% in wet soil and 12–26% in flooded
soil.
*As compiled from numerous sources

**As suggested by respective reference
15mm
40mm
30mm
Fig. 12.13. Scale mounting on bulldozing plate for reducing adhesion (Ref. 84).
Sliding direction
Fig. 12.14. Comet-type oblique hole drilled through moldboard plow (Ref. 69).
behind discouraging further research of this option. Reference 59,

as cited by Ref. 60, reportedly used liquid polymer for lubricating a
bulldozer blade to minimize soil sticking on the blade during strip mine
reclamation. In line with liquid polymer lubrication, Ref. 60 compared
sliding coefficients of the polymer-lubricated steel-on-soil with that of
plain steel and of Teflon-on-soil. They also evaluated the effectiveness
of polymer as a lubricant. They found that polymer lubrication could
reduce the draft of a moldboard plow 10 to 20% when plowed in
adhesive soil.
Air-jet and water injection were also tried to lubricate the inter-
face between soil and tool surface.64 When air under pressure was
introduced into soil having a high moisture content and a low air
permeability, it was observed that large horizontal cracks were produced
around the nozzle port.
12.4.2.4. Material and coating

Development of excellent anti-adhesion and sufficiently wear-resistant
materials/coatings has been considered as one of the options for
reducing adhesion.5 Various materials, ranging from glass to plaster
of Paris to hog hides, were examined with moldboards in waxy soils of
Texas by Ref. 53. Tetrafluoroethylene (commercially known as Teflon)
and PE (polyethylene) plastics were effectively used by Ref. 56 on
moldboard plows with sticky clay soil in Georgia. Reference 61, on use
of plastics on tillage tools, accepted that plastic coating helps control
sticking, but questioned its durability. Reference 85 described the effect
of surface properties on exchange of electrons at the soil–metal interface
(Fig. 12.15).
Selection of suitable material for the soil-engaging tool requires a
clever compromise between its abrasion and adhesion properties. For
Free-flow of electrons No flow of electrons

at the soil-metal interface at the soil-metal interface
Fig. 12.15. Effect of surface property on electron behavior at soil–metal interface

(Ref. 85).
example, PES-PTFE coating offers very good anti-adhesion charac-

teristics, yet its abrasion resistance is very poor. Enamel, as another
example, having medium abrasion characteristics, can be implemented
in soils with high adhesion but not in soil containing sand and
gravel. Reference 5 concluded that UHMWPE, having higher abrasion
resistance than PTFE and PES-PTFE, was highly suitable for soil-
engaging components.
Reference 65 examined the soil-adhesion characteristics of enamel
coating. It could reduce rolling resistance of cage wheels and the
plow boat to give better tractive performance. Enriching cast iron,
a commonly used material for tillage implements, with phosphorus
and silicon could improve anti-adhesion properties.68 A comparative
photograph is presented in Fig. 12.16, showing the smoother surface
with enamel coating.85
Uncoated mild steel surface
25KV x1.000 10 m 000001
Enamel coated surface
25KV x1.000 10 m 000001
Fig. 12.16. Uncoated and enamel-coated steel surfaces (Ref. 85).

12.4.2.5. Vibrations
Reference 62 tried reducing soil–metal adhesion by using high fre-
quency (10 kHz) ultrasonic vibrations. Reference 82 studied the effect
of mechanical vibrations on soil adhesion, and found that such vibra-
tions (60–100 Hz) could successfully reduce soil adherence on tool
surfaces.
12.4.2.6. Electro-magnetic field

An electric field was applied to generate an electro-osmotic non-
smooth surface over lugs and other implements77−79 to reduce soil
adhesion on agricultural and construction machinery. Reduction in
soil–tool adhesion was also achieved using a magnetic field on plows.45
Figure 12.17 depicts a schematic of embossed bulldozing blade using
electro-osmosis method of reducing soil adhesion.
Water from the soil moves from the positive pole to the nega-
tive pole when an electric field is applied, which is termed electro-
osmosis of soil. Non-scouring areas of soil are negatively charged to
increase the thickness of the interfacial water film, reducing adhesion.
Electro-osmosis was found better for clayey soils than sandy soil. This
method is limited to only application with longer contact time with soil.
High energy consumption and high voltage requirement are some of
the other limitations of this method.63,77,3 A strong magnetic field can
increase the variable negative electric charge and permeability, which
helps reduce soil adhesion.
Fig. 12.17. Electro-osmotic embossed bulldozing blade (Ref. 77).

12.4.2.7. Biomimetics
Recently, a number of investigations into application of polymers for
macro-morphological modification of tool surface have been carried
out. An adaptive science popularly known as biomimetics, which
is learning from nature, has opened up a window to reduce soil
sliding resistance onto tool surface by altering surface designs. These
studies, with extensive stress on convex or domed protuberations as
one of the widely used construction units, have tried to harness the
benefits of using polymers in agriculture. Enamel coating and ultra-high
molecular weight polyethylene (UHMW-PE) have proved themselves
an emerging modification with attractive compatibility with adhesive
soils. Soil-shedding ability of various non-smooth shapes, surface
conditions or a combination of both has highlighted an interesting
field of research — popularly called bionics or biomimetics. This aims
at application of such bionic surface patterns, which are “copied” from
soil-burrowing animals, onto soil-engaging tools with the object of
reducing undesirable adhesion with soil and thereby lowering draft
force and energy required in the operation, eventually. Such research
might help farmers of the region better manage their tillage tools by
allowing them to select smaller power-units for a given implement size.
Research outcomes may also be used by construction professionals to
lower fuel consumption by their earth-moving machinery.
Biomimetics or biomimicry or bionics engineering are the syn-
onymous terms used for lessons learnt from nature. Soil-burrowing
animals, including earthworms and dung beetles, have evolved signif-
icant ability to manage the sticky environment under adhesive soils.
Scientists13,72,73 have shown that the non-smooth body surfaces of such
creatures provide anti-adhesion and hydrophobic abilities.
Reference 46 designed a biomimetic, embossed moldboard plow.
It was prepared by gluing small convex domes on the surface of a
conventional moldboard (steel-35). The small convex domes were
made from UHMWPE (ultra-high molecular weight polyethylene).
The arrangement pattern of convex domes on the moldboard is shown
in Fig. 12.18. The plowing resistance of this biomimetic non-smooth
plow was reduced by 26% and 34% at the forward speeds of 3.6 and
5.0 km/h, respectively, as compared with a corresponding traditional
Fig. 12.18. Biomimetic embossed moldboard plow (Ref. 46).
Fig. 12.19. Flexible lining on bucket (Ref. 86).
smooth moldboard plow. The soil adhering to the biomimetic plow

surface was less than that to the smooth plow.
Biomimetics primarily comprises three approaches to address the
soil–tool adhesion problem, viz. non-smooth body surface (analogous
to domed, wavy and scale-shaped construction units), liquid substance
secreted (analogous to lubricating action), and flexible body surface
(Fig. 12.19) (analogous to mechanically flexible or retractable tool
surfaces).
Scientists79 were inspired by the dung beetle to design embossed
bulldozing blades based on the pseudo-variable approximation design
optimizing method. Reference 76 discusses the designing of embossed
bulldozing blades with UHMWPE convex domes, which could reduce
bulldozing resistance.
12.5. Epilogue
Microscopic interfacial forces at the soil–tool interface were reviewed,
and major forces were clubbed into two mutually perpendicular groups:
normal and tangent to the interface. Water content plays an important
role to alter the relative strength of each component. Other factors
influencing microscopic forces include intrinsic properties of tool,
surface conditions, tool material, physical and chemical composition
of soil. Capillary suction or negative pressure is induced from the soil
pores, and is the main cause of normal adhesive resistance, while the
contact area at the soil–tool interface largely contributes to the tangent
resistance.
Normal force acting at the interface comprises three primary
components: normal gravity, normal adhesion and normal friction.
Gravitational force is contributed by soil particles and soil water; normal
adhesion comes from intermolecular attraction of bare soil particles,
water meniscus, the water film’s viscidity and capillary negative adsorp-
tion; and normal friction is composed of wet friction of the water
meniscus.
Tangent resistance at the interface constitutes three components:
tangential friction, drag and tangential adhesion. Tangent friction is
mainly caused by dry friction of soil particles; drag resistance comes
from tangential viscidity; and tangential adhesion is due to surface
tension and water meniscus.
It can be predicted that resistance reduction at the interface could
be achieved by eliminating capillary negative pressure and lowering the
physico-chemical adsorption at the soil–tool interface. Having closer
insight to microscopic descriptors of forces at soil–tool interface would
certainly allow efficient and energy-saving design of soil-engaging tools.
Moreover, it may lead to appropriate design modifications of interacting
surfaces. Experimental validation can be done by precise investigations
to further explore microscopic phenomena at the soil–tool interface.
Soil–tool adhesion, especially in sticky soils, creates an unfavorable
liability on the power source to develop additional effort for overcom-
ing adhesive forces. This reduces quality of work and increases energy
consumption and working resistance in both farm implements and
construction machinery. The soil–tool interaction can be understood

as a system comprising soil, water and tool surface. Laplace pressure,
meniscus tension and viscous resistance are produced by a thin interfa-
cial film of water at the soil–tool interface. These forces, among other
factors, predominantly govern the system and are sensitive to the nature
and properties of tool surfaces.
Factors affecting soil–tool adhesion are soil properties, liquid
properties, tool characteristics, tool material and working conditions.
Various methods of reducing soil–tool adhesion have been researched:
Surface design modification, tool heating, heat treatments, lubrication,
material, coating, vibration, electro-magnetic field application, and
biomimetics are such often-reported attempts.
While various means of adhesion reduction account for their
individual benefits, macro-morphologic surface modifications using
bionic principles might offer greater flexibility, manoeuvrability and
applicability. Considering the cost and complexity involved in existing
methods of reducing soil–tool reduction, biomimicry — learning from
the soil-burrowing creatures of nature — appears to be promising.
Surface design modifications based on typical natural morphologies
have exhibited exciting results. The future trend is likely to incorporate
anti-adhesion properties of such techniques or their combinations to
improve working performance of soil-engaging tools.
References
[1] Gill WR, Vanden Berg GE. (1968) Soil Dynamics in Tillage and Traction. Agri-
cultural Handbook, No. 316. Agricultural Research Service. US Department of
Agriculture.
[2] Chancellor WJ. (1994) Friction between soil and equipment materials — review.
ASAE Paper No. 94-1034.
[3] Ren LQ, Tong J, Li JQ, Chen BC. (2001) Soil adhesion and biomimetics of soil-
engaging components: A review. Journal of Agricultural Engineering Research
79(3): 239–263.
[4] Oswal MC. (1994) Soil Physics. Oxford & IBH Publishing Co., New Delhi.
[5] Tong J, Ren LQ, Yan JL, et al. (1999) Adhesion and abrasion of several materials
against soil. Agricultural Engineering Journal 8(1): 1–22.
[6] Bowden FP, Tabor D. (1964) Friction and Lubrication of Solids. Oxford
Clarendon Press, London.
[7] Qian D. (1965) Study on the adhesion characteristics of the conventional
mouldboard material — white iron against wet clay soil. Transactions of the
Chinese Society of Agricultural Machinery 8(2): 145–150. (In Chinese.)
[8] Fountaine ER. (1954) Investigations into the mechanism of soil adhesion.
Journal of Soil Science 5(2): 251–263.
[9] Akiyama Y, Yokoi O. (1972) Study on soil adhesion, Part II. Theoretical analysis
of adhesion mechanism at the saturating stage. Journal of Japanese Soil Fertilizer
Science 43(8): 271–277. (In Japanese.)
[10] Tong J, Ren LQ, Chen B, Qaisrani AR. (1994b) Characteristics of adhesion
between soil and solid surfaces. Journal of Terramechanics 31(2): 93–105.
[11] Hendrick JG, Bailey AC. (1982) Determining components of soil-metal sliding
resistance. Transactions of the ASAE 25: 845–849.
[12] Guo PG, Liu PS. (1995) The effect of magnetized plowshare upon tillage
performance. Agricultural Research in the Arid Areas 13(2): 103–109.
[13] Tong J, Ren LQ, Chen BC. (1994a) Geometrical morphology, chemical
constitution and wettability of body surfaces of soil animals. Agricultural
Engineering Journal 3(1&2): 59–68.
[14] Zisman WA. (1964) Relation of Equilibrium Contact Angle to Liquid and Solid
Constitution. In: Gould RF (ed.), Contact Angle, Wettability, and Adhesion.
Advances in Chemistry Series. American Chemical Society, Washington.
[15] Cooper WA, Nuttall WA. (1915) Journal of Agricultural Science 7: 219.
[16] Harkins, WD. (1941) in Chemical Reviews. As cited by Zisman WA. (1964).
Relation of Equilibrium Contact Angle to Liquid and Solid Constitution.
In: Gould RF (ed.), Contact Angle, Wettability, and Adhesion. Advances in
Chemistry Series. American Chemical Society, Washington.
[17] Kummer FA, Nichols ML. (1938) The dynamic properties of soil — VII. A study
of the nature of physical forces governing the adhesion between soil and metal
surfaces. Agricultural Engineering 19: 73–78.
[18] Wu S. (1982) Polymer Interface and Adhesion. Marcel Dekker Inc., New York.
pp. 54–62.
[19] Schut PH. (1976) Glossary of Terms in Soil Science. Research Branch, Canada
Department of Agriculture, Ottawa. Publication No. 1459, 44 pp.
[20] Curtis LM, Tyson TW. (1998) Scheduling irrigation using soil moisture tension.
Publication No. ANR-467. The Alabama Cooperative Extension Service. Avail-
able online at URL: http://www.aces.edu/department/irrig/anr-467.htm
[21] Shi-qiao D. (2004) The mechanism of reducing soil adhesion and the design of
bionic plow. Ph.D. thesis, Jilin University. (Unpublished) (In Chinese.)
[22] Oida A, Momozu M. (2002) Simulation of soil behavior and reaction by machine
part by means of DEM. Agricultural Engineering International: the CIGR
Journal of Scientific Research and Development. Manuscript PM 01 004. Vol. IV.
October, 2002.
[23] Cong Q, Ren L, Chen B, Su F. (1990) A study on adhesion reducing methods of

terrain-machine. Transactions of the Chinese Society of Agricultural Engineering
6(1): 8–14. (In Chinese.)
[24] Israelachvili JN, Fisher LR, Horn RG, Christenson HK. (1982) Tribology Series
7: Microscopic Aspects of Adhesion and Lubrication, New York, pp. 55–66.
[25] Jia X. (2004) Theoretical analysis of the adhesion force of soil to solid materials.
Biosystems Engineering 87(4): 489–493.
[26] Chen B, Cong Q, Ren L. (1996) Soil adhesion mechanism of chemical
absorption. Proceedings of the 12th International Conference of the ISTVS,
pp. 546–551.
[27] Liu J. (1993) The Theory of Material Abrasion and its Anti-abrasion Character-
istics. Tsinghua University Press, Beijing, China. (In Chinese.)
[28] Bikerman JJ. (1974) Capillary attraction and hysteresis of wetting. Journal of
Adhesion 6: 331–336.
[29] McFarlane JS, Tabor D. (1950) Adhesion of Solids and the Effect of Surface
Films. Proc R Soc Lond A Math Phys Sci 202: 224–243.
[30] Orr FM, Scriven LE. (1975) Pendular rings between solids: Meniscus properties
and capillary force. J Fluid Mech 67(4): 123–142.
[31] Zhang R, Li J, Xu S. (2004a) Simulation of force on nonsmooth bulldozing
plate by Distinct Element Method. Proceedings of 2004 CIGR International
Conference, Beijing, China, October 11–14, 2004, 30-160D: III-166.
[32] Zhang R, Li J. (2006) Simulation on mechanical behavior of cohesive soil by
Distinct Element Method. Journal of Terramechanics 43: 303–316.
[33] Zhang R, Li J. (2004) Simulation on mechanical behavior of cohesive soil by
Distinct Element Method. Proceedings of the 7th Asia-Pacific Conference on the
ISTVS, Changchun, China, September 14–16, 2004, pp. 154–161.
[34] Zhang Y, Han GZ. (1992) Measurement and analysis of work resistance on
the magnetic coverer. Journal of Shenyang Agricultural University 23(Supple-
ment): 25–28.
[35] Zeng D. (1995) The Dynamics of Mechanical-Edaphic System. Agricultural
Handbook. Beijing Science and Technology Press, Beijing, China. (In Chinese.)
[36] Xiong S. (2001) Basic Agrology. China Agricultural University Press, Beijing,
pp. 96–131. (In Chinese.)
[37] Princen HM. (1969) The equilibrium shape of interfaces, drops, and bubbles:
Rigid and deformable particles at interfaces. In: Matijevic E (ed.), Surface and
Colloid Science, Vol. 2. Willey-Interscience, New York, pp. 1–84.
[38] Michael AM. (1998) Irrigation: Theory and Practice. Vikas Publishing House,
New Delhi, India.
[39] Shi-qiao D, Lu-quan R, Yan L, Zhi-wu H. (2005) Tangent resistance of soil
on moldboard and the mechanism of resistance reduction of bionic moldboard.
Journal of Bionics Engineering 2(1): 33–46.
[40] Zheng S. (2001) Capillary phenomena and contact angles. Journal of Gansu
Advance Normal College 6(2): 29–31.
[41] Schwartz AM, Rader CA, Huey E. (1964) Resistance to Flow in Capillary
Systems of Positive Contact Angle. In: Gould RF (ed.), Contact Angle,
Wettability, and Adhesion. Advances in Chemistry Series. American Chemical
Society, Washington, D. C., pp. 250–267.
[42] Hu Z, Chen G, Du Y. (2001) The Surface Interface of Materials. East China
University of Science and Technology Press, Shanghai, China. (In Chinese.)
[43] Yan EL. (1991) Low-carbon martensite steel mouldboard. Proceedings of the
International Conference of Agricultural Mechanization, 16–20 October 1991,
Beijing, China, pp. 245–250.
[44] Salokhe VM, Gee-Clough D, Manzoor S, Singh KK. (1990) Improvement
of the tractive performance of cage wheel lugs by enamel coating. Journal of
Agricultural Engineering Research 45(3): 209–224.
[45] Han GZ, Zhang Y. (1991) Measurement and analysis of ploughing resistance
on the magnetic plough. Transactions of the Chinese Society of Agricultural
Machinery 22(3): 25–28.
[46] Qaisrani AR, Bingcong C, Luquan R. (1992) Modified and unsmoothed plow
surfaces — A means to reduce plowing resistance. Agricultural Engineering
Journal 1(3): 115–124.
[47] Li JQ, Ren LQ, Liu CZ, Chen BC. (1996) A study on the bionic mouldboard
plow for reducing soil adhesion and plowing resistance. Transactions of the
Chinese Society of Agricultural Machinery 27(2): 1–4.
[48] Zuo CC, Zhang SQ, Ma CL, He HP. (1997) The study on the adhesion
decreasing and resistance-reducing of disc opener. Transactions of the Chinese
Society of Agricultural Machinery 28(Supplement): 37–40.
[49] Doran P. (2005) Earth and Environmental Sciences. University of Illinois at
Chicago.
[50] Low PF. (1979) Nature and properties of water in montmorillonite-water
systems. Soil Science Society of America Journal 43(8): 651–658.
[51] Li J, Ren L, Chen B, Gu W. (1993) The effect of microstructure of steel on
soil adhesion characteristics. Transactions of the Chinese Society of Agricultural
Engineering 9(3): 14–21.
[52] Lewis JA. (2005) History of polymers. Department of Materials Science and
Engineering, University of Urbana-Champaign. Site visited on June 22, 2005
at URL: http://matse1.mse.uiuc.edu/∼tw/polymers/time.html
[53] Bacon CA. (1918) Plow bottom design. Transactions of the ASAE 12: 26–30
(old series).
[54] Bacon CA. (1920) A Treatise on Plows and Plowing, 2nd ed. South Bend, 200 pp.
[55] Bertelsen WR. (1960) Something brand new in plows — the aeroplow. Des
Moines Register, August 21, Section H, pp. 17-H, as cited in Salokhe VM,
Suharno, Gee-Clough D. (1992). Effect of enamel coating on the field
performance of a mouldboard plough. Soil & Tillage Research 24(3): 285–297.
[56] Cooper WA, McCreery WF. (1961) Plastic surfaces for tillage tools. ASAE Paper
No. 61–649.
[57] Fox WR, Bockhup CW. (1965) Characteristics of a Teflon-covered simple tillage
tool. Transactions of the ASAE 8(2): 227–229.
[58] Wismer RD, Wegscheid EL, Luth HL, Raning BL. (1968) Energy application
in tillage and earth moving. Transactions of the Society of Automotive Engineers
77: 2486–2496.
[59] Pittsburg and Midway Coal Mining Company (1973) As cited in Schafer RL,
Gill WR, Reaves CA. (1975) Lubrication of soil-metal interfaces. Transactions
of the ASAE 18(5): 848–851.
[60] Schafer RL, Gill WR, Reaves CA. (1975) Lubrication of soil-metal interfaces.
ASAE Paper No. 74–1587.
[61] Schafer RL, Gill WR, Reaves CA. (1977) Lubricated plows vs. sticky soils.
Agricultural Engineering 58(10): 34–38.
[62] Sharma VK, Drew LQ, Nelson GL (1977) High frequency vibrational effects
on soil-metal friction. Transactions of the ASAE 20(1): 46–51.
[63] Shen Z. (1982) Mechanism of soil adhesion and its applications. Proceedings of
the First Conference of Chinese Society for Terrain-Machine Systems, Wuxi, China,
pp. 121–123.
[64] Araya K, Kawanishi K. (1984) Soil failure by introduction of air under pressure.
Transactions of the ASAE 27(5): 1292–1297.
[65] Salokhe VM, Gee-Clough D. (1988) Coating of cage wheel lugs to reduce soil
adhesion. Journal of Agricultural Engineering Research 41(3): 201–210.
[66] Chen BC, Ren LQ, Xu XB, et al. (1990) Research on reducing soil adhesion
of body surfaces of typical soil animals. Transactions of the Chinese Society of
[67] Li AQ, Ren LQ, Chen BC, Ciu XX. (1990) Constitution and mechanism analysis
of reducing soil adhesion for the body surface liquid of earthworms. Transactions
of the Chinese Society of Agricultural Engineering 6(3): 8–13.
[68] Tong J, Ren L, Chen B. (1990) Reducing adhesion of soil by phosphorus white
iron on the basis of bionics principles. Proceedings of the 10th International
Conference of ISTVS, 20–24 August, Kobe, Japan, pp. 57–63.
[69] Zhu H, Tan L, Wu S. (1992) An experimental study on the ‘comet-type passage-
holes’ plough for reducing resistance. Transactions of the Chinese Society of
Agricultural Machinery 23(4): 20–24.
[70] Ren L, Zhang J, Chen B, Qi B. (1990) Bionics research on reducing adhesion
and resistance of soil for flexible shovel. Transactions of the Chinese Society of
Agricultural Machinery 21(4): 33–39.
[71] Yin Y, Ren L, Cheng Y, Chen B. (1990) A study on cleaning buildup soil in
loader shovel on the principle of bionics. Transactions of the Chinese Society of
[72] Ren LQ, Cong Q, Chen BC, Wu LK. (1992) A study on the adhesion reducing
character of geometric non-smooth surface of typical animal. Transactions of
Chinese Society of Agricultural Machinery 23(2): 29–35.
[73] Li JQ, Ren LQ, Chen BC. (1995) The statistical analysis and mathematical
analogy of unsmoothed geometrical units on the surface of soil animals.
Transactions of the Chinese Society of Agricultural Engineering 11(2): 1–5.
[74] Salokhe VM, Suharno, Gee-Clough D. (1992) Effect of enamel coating on
the field performance of a mouldboard plough. Soil & Tillage Research 24(3):
285–297.
[75] Salokhe VM, Shirin AKM. (1992) Effect of enamel coating on the performance
of a disc plough. Journal of Agricultural Engineering Research 53: 71–80.
[76] Qaisrani AR. (1993) The effects of modified and unsmoothed surfaces on the
drafts of bulldozing plates and moldboards plows. PhD Thesis, Jilin University
of Technology, Changchun, China. (Unpublished.)
[77] Cong Q, Wu LK, Ren LQ, Chen BC. (1995) The principle experiment of
reducing soil adhesion and scouring soil by nonsmooth surface electro-osmosis.
Transactions of the Chinese Society of Agricultural Engineering 11(3): 19–23.
[78] Chen BC, Liu DW, Ning SJ, Cong Q. (1995) Research on the reducing
adhesion and scouring of soil of lugs by using unsmoothed surface electro-
osmosis method. Transactions of the Chinese Society of Agricultural Engineering
11(3): 29–33.
[79] Ren LQ, Cong Q, Wu LK, Chen DX. (1995) Study on the reduction of
soil adhesion and resistance by the application of non-smooth surface electro-
osmosis. Transactions of the Chinese Society of Agricultural Engineering 11(3):
24–28.
[80] Wang XL. (1996) Study on reducing soil adhesion to machines by vibrations.
In: Yu Q, Qiu LJ (eds.), Proceedings of the 12th International Conference of the
ISTVS, 7–11 October, Beijing, pp. 539–545.
[81] Ren LQ, Wang Y, Li J, Sun S. (1996) Bionic research on flexible non-
smooth surface of typical animals. Transactions of Chinese Society of Agricultural
Engineering 12(4): 31–36.
[82] Wang XL, Ren LQ, Li J, et al. (1997) A model test study on flexible bionics
technology of anti-adhesion to dump truck. Transactions of Chinese Society of
[83] Ren LQ, Cong Q, Tong J, Chen B. (2001a) Reducing adhesion of soil against
loading shovel using bionic electro-osmosis method. Journal of Terramechanics
38: 211–219.
[84] Qian C, Luquan R, Bingcong C, Beizhan Y. (1999) Using characteristics of
burrowing animals to reduce soil-tool adhesion. Transactions of the ASAE 42(6):
1549–1556.
[85] Salokhe VM, Hoki M, Sato K. (1993) Why does soil not stick to enamel coating?
Journal of Terramechanics 30(4): 275–283.
[86] Sun S, Ren LQ, Wang Y, Chen D. (1996) Design of anti-adhesion flexible lining
steel rings. Journal of Jilin University of Technology 26(4): 20–24.
[87] Soni P, Salokhe VM. (2006a) Theoretical analysis of microscopic forces at soil-
tool interfaces: A review. Agricultural Engineering International: CIGR Journal
of Scientific Research and Development, Manuscript PM 06 010. Vol. VIII. June,
2006.
[88] Soni P, Salokhe VM. (2006b) Influence of Dimensions of UHMW-PE Protu-
berances on Sliding Resistance and Normal Adhesion of Bangkok Clay Soil to
Biomimetic Plates. J Bionic Eng 3(2): 63–71.
[89] Soni P, Salokhe VM, Nakashima H. (2007) Modification of a Mouldboard
Plough Surface using Arrays of Polyethylene Protuberances. Journal of Ter-
ramechanics 44(6): 411–422.
Chapter 13
Application of Bio-Inspired
Surfaces in Reducing Adhesion
to the Surfaces of Soil-Engaging
Components of Agricultural
and Earth-Moving Machinery
Rashid Qaisrani∗ and Li Jianqiao†

∗ Department of Agriculture and Water Resources
Canberra, Australia
† Jilin University, Changchun, China
Soil adheres to the surfaces of soil-engaging components of various

machinery and equipment. The adhesion of soil to the surfaces of
soil-engaging components increases draft and energy consumption
and affects the quality of work adversely. The adhesion of soil is
primarily from the capillary pressure and the viscous resistance of the
water film at the interface due to molecular attraction and negative
air pressure. The factors affecting soil adhesion include the nature
and properties of the soil, surface characteristics of soil-engaging
components, the working conditions and the environment. Conven-
tional methods for reducing soil adhesion include working soil when
it is comparatively dry, improving surface shapes of soil-engaging
components, producing soil-engaging components from materials with
better scouring properties, and application of electro-osmosis, magnetic
fields, vibration and lubrication, etc. An ideal technique should be safe
and simple, economical to manufacture, easy to use, synchronize with
485
other components of the machine and tools, have no requirement for

extra controls and power, consume lesser energy and be efficient with
scouring abilities 90% or higher. For practical field conditions, many
attempts have been made to reduce friction and soil adhesion on the
surfaces of soil-engaging components of various machines. So far, many
of such attempts have not been very successful. Some techniques, such
as air injection at the soil–tool interface are useful but add weight to the
existing set up and in many cases make the system more complicated
to operate. Enamel coating is cheap and simple to apply but has poor
wear resistance and cannot be used in abrasive soil conditions.
Soil animals such as ground/dung beetles stay in moist sticky
soils for extended periods without soil adhering to their bodies. Since
1960, researchers have been exploring the soil adhesion preventing
mechanism of soil animals’ cuticles. They studied the structural prop-
erties of the biological system, energy transformation and information
process used to improve or modify the machines. A number of soil-
burrowing animals developed significant anti-adhesive characteristics
through evolution over time. The body surfaces of such animals
have better scouring properties due to a range of factors such as
geometrically non-smooth surface morphologies, chemical composi-
tion, hydrophobicity, microscopic electro-osmotic systems, flexibility
of their body surfaces and their bodies releasing liquid which acts
as lubricant. Various biomimetic methods have been developed on
the basis of soil-burrowing animals’ cuticles’ scouring mechanism and
properties. Some of the biomimetic methods to improve the soil-
scouring properties of soil-engaging components of agricultural and
earth-moving machinery include development of non-smooth surfaces
of soil-engaging components, electro-osmotic systems, and flexible
parts for removing soil, lubrication, coating and painting, etc. This
process revealed great vitality and brought fruitful results over time.
For example, manufacturing of off-road vehicles for desert and muddy
fields was made possible by simulating the special motions of kangaroo
and penguin, respectively. This chapter discuss the soil adhesion
preventing mechanism of soil animals’ cuticles and their application
in improving soil-scouring properties of soil-engaging components of
various machinery and equipment in detail.
Application of Bio-Inspired Surfaces in Reducing Adhesion to the Surfaces 487
13.1. Introduction
Background
Soil-engaging components of agricultural and earth-moving machinery
are used to apply forces to soil to cause some desired effects, such as
pulverization, cutting, inversion or movement of soil. Some of these
components produce multiple effects simultaneously. The ultimate aim
is to manipulate soil from its current condition into a different desired
one by mechanical means. The vehicles that take soil as an actuating
medium or implements, earth-moving machinery and hand tools that
treat soil as working objects, all suffer from soil adhesion. It affects the
efficiency and work quality adversely and increases energy consumption
of various machinery and equipment that operate under moist sticky
soil conditions. In extreme cases it does not allow the machine to
move and work. The energy consumption in overcoming adhesion and
friction between soil and tillage tools is estimated to be 50% of the
gross energy required in carrying out these operations. Therefore, it is
important to investigate the mechanism of soil adhesion to the surfaces
of soil-engaging components of various machinery and equipment and
develop techniques for scouring soil.1
Adhesion of soil to the surfaces of soil-engaging components of
various machinery and equipment is affected by a range of factors. Some
of the factors such as soil type, moisture content, time of operation,
mechanical and chemical composition of soil particles, number and
colloidal contents of soil particles and the presence of organic matter in
soil contributing towards the increase in adhesion are not fully under
our control. However, there are other factors that help to reduce
adhesion between working parts of machinery and equipment and
soil. The adhesion of soil to the surfaces of soil-engaging components
may be reduced by techniques such as lubrication, electro-osmosis,
vibration and modification of surfaces.2 The hydrophobic treatment
of steel components can reduce adhesion of soil to the surfaces of
treated components significantly.3 Various coating materials such as
silicon lubricant oil, lead oxide paint, gloss paint and varnish, chromium
painting, teflon tape, ceramic and enamel were found helpful in
reducing adhesion of soil to the surfaces of cage wheels.4,5,6 However,
practical problems such as higher cost and low durability limit the
use of lubricant oil, teflon tape and ceramic coating in practical field
conditions. Modification of the surfaces of soil-engaging components
of tillage and earth-moving machinery based on bionics principles are
useful in improving the soil scouring properties of these components.7
These findings are supported by other research studies.8,9,10 The ability
of bio-inspired surfaces to reduce adhesion of soil to the surfaces of
soil-engaging components is related to several factors such as chemical
composition, surface morphology, shape and structure, presence of
some elements in their body surface and secretion of special materials
while moving through moist sticky soils.11
Adhesion and friction of soil — causes, effects

and remedies
The adhesion force of soil to solid materials primarily consists of an
intermolecular force between soil and soil materials, and the attraction
force of the water film depending on the interface state between
soil and solid materials. The mechanism of soil adhesion has been
investigated from different perspectives by researchers depending on
their importance in various areas of research. A great deal of progress has
been made since the pioneer work of Amoutons in 1699 and Coulomb
in 1875. As a result of these investigations a number of theories have
been proposed on the phenomena of soil adhesion, including the well-
recognized Water Tension Theory and Capillary Theory.12 Three basic
elements including area of contact, strength of the body at the soil–tool
contact and the shearing mechanism of soil in and around the contact
areas are involved in the friction of unlubricated solids.12
Based on the research, soil adhesion can be reduced by a number
of techniques such as lubrication,13 electro-osmosis,14,15 mechanical
vibration,16 ultrasonic vibration17 and coating.4,5 The chemical com-
position of the contact surfaces has an important role in reducing
adhesion of soil to their surfaces. For example, phosphorus has the
ability to reduce adhesion between two surfaces.2 Soil adhesion can
be reduced by hydrophobic treatment of steel components. Another
approach for reducing adhesion by coating lug surfaces with different
materials such as silicon lubricant oil, lead oxide paint, gloss paint
and varnish, chromium painting, teflon tape, teflon sheet, ceramic

tile and enamel coating have been effective.4,5 However, there were
practical problems (expensive and low durability, etc.) in using silicon
lubricant oil, teflon tape, ceramic tile and enamel coating. Moreover,
soil animal cuticles respond to the changes occurring due to their ability
to exchange materials and energy with soil.9 The results of changes and
exchanges of the soil animal cuticles are helpful in reducing adhesion
of soil to their bodies and reduce sliding resistance. This is helpful in
investigating the soil adhesion preventing mechanism of soil animal
cuticles. The application of such techniques on the surfaces of soil-
engaging components of agricultural and earth-moving machinery is
highly desirable in improving their performance under wet sticky soil
conditions.
Theoretical analysis of soil adhesion forces

The soil adhesion forces have a direct relationship with the interface
condition between soil and soil materials. While assuming the surface of
soil particles as spherical, several kinds of interface contact situation may
occur between soil particles and solid material16 as shown in Fig. 13.1.
The adhesion force of soil to solid materials consists mainly of
intermolecular force between soil and the solid materials, the water ring
attraction, and the attractive force of the water film. The soil adhesion
forces may be dominated by the attraction force of a water ring or
water film depending on the interface state domination by water ring
or water film, respectively. In the absence of water ring and water film at
the interface, the intermolecular force between soil and solid materials
is the dominant component of soil adhesion force. Moreover, the soil
adhesion forces are inversely proportional to the hydrophobicity of solid
materials. The increase in the contact angle of water on solid material
between 0 and 180◦ decreases the soil adhesion force on solid material.
The soil adhesion force has a direct relationship with the contact
interface between soil and solid materials. For analysis, it is assumed
that the surfaces of soil particles are spherical and smooth with the
presence of several types of interface contact situation between the soil
particles and solid material plane, as shown in Fig. 13.1. For example
the water ring has not fully formed in Fig. 13.1(a) and (b); the water
(a) (b)
(c) (d)
(e)
Fig. 13.1. Contact models between soil and solid plane (a) and (b) no water ring
formed; (c) and (d) water rings formed; (e) continuous water film formed [16].
ring has formed on each side of the solid contact in Fig. 13.1(c) and (d);
and a continuous water film has formed between the solid interfaces
in Fig. 13.1(e). Therefore, the soil adhesion force is the result of
intermolecular attraction, attraction produced by the water ring, and
the attraction force resulting from the water film due to the viscosity
of water.
R = the radius of soil particle

D0 = shortest distance between soil particle and solid material plane
D = distance from soil particle surface to solid material plane
x = distance from soil particle surface to the vertical symmetrical axis
of soil particle
b = distance from soil particle surface to the surface to the tangent
plane at the lowest point of soil particle surface
φ = angle between R and the symmetrical axis
θ1 and θ2 = contact angles of water on soil and the contact material,

respectively
(a) Intermolecular attraction

Based on Lifishitz’s theory, the interaction energy “E” of unit area
(J/m2 ) between two parallel planes was described as [18]:
E(Z ) = hw/16π2 Z 2 , (13.1)
hω = Lifishitz–van der Waals constant, J
Z = distance between two parallel planes, m.
Therefore, when soil particles and solid materials are in contact
with each other as shown in Fig. 13.1(a) and (b), the interaction
energy between the soil particle and solid material plane “Ea ” can be
expressed as:
R
hw R R
Ea = 2πE12 (D) dx = − ln 1 + . (13.2)
0 8π D0 D0
R = radius of soil particle;
D0 = shortest distance between the soil particle and solid material plane
x = the distance from the soil particle surface to the axis perpendicular
to the solid plane.
The interaction force between the soil particle and solid material
plane “Fa ” is:

dEa hw R R R
Fa = =− − − . (13.3)
dD0 8π D0 2 D0 2 D0 (D0 + R )
When the distance between two solid objects is equal to the equilibrium
distance between two molecules Z0 , the interaction energy of unit
area between two parallel planes E12 (Z0 ) is the adhesion work “Wa .”
Therefore, Eq. (13.1) can be simplified as:
hw
Wa = E12 (Z0 ) = . (13.4)
16πxZ0 2
Equation (13.4) can be further simplified as:
hw
2
= 2πZ02 Wa . (13.5)
8πxZ0
Substituting the value of hw into Eq. (13.3) and simplifying results in

R 1
Fa = −2πZ02 Wa R − , (13.6)
D20 D0 (D0 + R )
when Z0 = D0 , Equation (13.6) can be re-written as

Z0
Fa = −2πWa R 1 − . (13.7)
(Z0 + R )
(b) Water Ring attraction

It is assumed that with the formation of water ring between the soil
particle and solid material plane as shown in Fig. 13.1(c) and (d), the
contact angles of water with soil and solid material plane are θ1 and θ2 ,
respectively. The surface area and surface tension are AS1V and γS1V
for soil and AS2V and γS2V for solid material plane, respectively. After
formation of the water ring, the new surface areas for soil and solid

material plane are “AS1V ” and “AS2V ,” respectively. The interface area
and interface tension between soil and water ring are AS1L and γS1L ,
respectively and for solid material plane and the water ring are AS2L
and γS2L , respectively. The surface area and surface tension of the side
face of the water ring are ALV and γLV respectively. The volume free
energies of soil particle, air phase, solid material plane and water ring
are GS1 , GV , GS2 and GL , respectively. Therefore, the energy of the soil
adhesion system E can be expressed as follows:
E = ASLV ϒSLV + AS1L ϒS1L + AS2L ϒS2L + A SLV ϒSLV + A S2L ϒS2L
+ GS1 + GV + GS2 + GL = ALV ϒLV + AS1L (ϒS1L − ϒS1V )
+ AS2L (ϒS2L − ϒS2V )+K, (13.8)
K = ASLV ϒSLV AS2L ϒS2L + GS1 + GV + GS2 + GL . (13.9)
As AS1L ≈ AS2L ≈ πR sin φ ≈ πR φ, R is the radius of the soil
2 2 2
particle and φ is the angle between R and the symmetrical axis as shown
in Fig. 13.1(d), Eq. (13.11) can be simplified as follows:
E = AS1L (ϒS1L − ϒS1V ) + AS2L (ϒS2L − ϒS2V ) + K (13.10)
From Eq. (13.10), the soil adhesion force can be obtained as follows:
dE d
Fa = = [(ϒS1L − ϒSLV ) + AS2L (ϒS2L − ϒS2V )]. (13.11)
dD0 dD0
As AS1L ≈ AS2L ≈ πR 2 sin2 φ ≈ πR 2 φ, R is the radius of the soil

particle and φ is the angle between R and the symmetrical axis as shown
in Figure 13.1(d), Eq. (13.11) can be simplified as follows:
dφ
Fa = −2πWa R φ [(ϒS1L − ϒSLV ) + (ϒS2L − ϒS2V )]. (13.12)
dD0
Similarly, the volume of the water ring V can be expressed as:
V = πR 2 Sin2 φ(D0 = b) − πR 3 (1 − Cos φ)2 (2 + cos φ)
R 3 φ4
≈ πR 2 φ2 D0 + , (13.13)
4
where b is the distance from the soil particle surface to the tangent plane
at the lowest point of the soil particle surface as shown in Fig. 13.1(d).
When differentiating V with respect to Do , Eq. (13.13) can be
written as:
dV dφ dφ
= πR 2 φ2 + 2πR 2 D0 φ + πR 3 φ3 . (13.14)
dD0 dD0 dD0
As the volume of the water ring V is a constant, dV/dD0 = 0, and Eq.
(13.14) can be further simplified as:
dφ φ
= (13.15)
dD0 Rφ2 − 2D0
Combining equations (13.12) and (13.15):
2πRϒLV
Fa = − (cos θ1 + cos θ2 ). (13.16)
1 + D0 /b
We know that the contact angle of water on soil “θ1 ” is zero, so
Eq. (13.16) can be simplified as:
2πRϒLV
Fa = − (1 + cos θ2 ). (13.17)
1 + D0 /b
The negative sign in Eq. (13.17) indicates that “Fa ” is an attraction
force.
Equation (13.17) shows that when “θ2 ” is between 0 to 180◦ , “Fa ”
is an attraction force. Therefore, when soil comes in contact with solid
materials, it always adheres to their surfaces. The greater the value of
“θ2 ,” the smaller the value of the soil adhesion force “Fa ” is. Similarly,
the worse the wettability of solid materials, the better it is for reducing
adhesion of soil to the surfaces of solid materials.
When “θ2 ” is 0, “Fa ” reaches a maximum value of 4πRγl V .
This result is compatible with earlier results obtained by Israelachvili
et al. (1982) and Mcfarlane and Tabor (1950) while analyzing forces
between soil and solid materials.
(c) Attraction force of the water film

A liquid film forms between two discs, when they are separated from
position h1 to h2 , the separation force required may be expressed
as [19]:

3ηπR 4 1 1
Fa = − , (13.18)
4t2a h12 h21
2
where:
η = viscosity of liquid, Pa s;
R = radius of the discs in m;
h1 and h2 = distances between the two discs before and after separa-
tion, m;
t = time required to separate the two discs from h1 to h2 , s,
Fa = the separation force in N, respectively.
When a continuous water film forms as shown in Fig. 13.1(e), the soil
adhesion force is generated by the attraction force of the water film. The
value of the soil adhesion force may be determined by Eq. (13.18). The
influence of solid materials on the soil adhesion force depends mainly on
the influence produced by the surface force field of the solid materials on
the viscosity of the water film. The higher the surface tension of the solid
materials and the thinner the water film, the higher are the viscosity of
water and the soil adhesion force. When the thickness of the water film is
enough to render the water in the water film unaffected by the surface
force field of solid materials, the water becomes free water. In such
situations the soil adhesion force depends on the viscosity of free water.
Soil–tool Interface
In the absence of water at the soil–tool interface, as shown Fig. 13.1(a)
and (b), the intermolecular attraction is primarily due to the molecular
interaction between them. For a non-contacting (Figure 13.1(a))
spherical asperity with radius R, attraction pressure per unit area of
separation, P (d), can be expressed as a function of its separation
distance (d) between asperity summit and the solid surface using the
Lennard-Jones function [20] as follows:

∂E(d)
P(d) = . (13.19)
∂d
The unit area of separation is normally defined as the separation of the
tip of the summit to the tool surface.
Where: E(d) = interaction energy/unit area of separation.
The interaction energy is the energy of adhesion (Ea ) at the
equilibrium intermolecular distance de was investigated by scientists.20
Therefore, Eq. (13.19) can be further simplified as:
9
8EA de 3 de
P(d) = − ; E(de ) = EA at d = de .
3de d3 d
(13.20)
The interaction energy per unit area of separation between two parallel
planes, as a function of d, defined by Lifishitz’s theory were determined
as follows18 :
hw
E(d) = , (13.21)
d16π2 de2
where hω = Lifishitz–Van der Waals constant.
As the separation equal to intermolecular distance, E(de) provides
the work of adhesion [16] as:
hw
WA = W(de) = Ea = . (13.22)
d16π2 de2
Mechanism of soil adhesion

There are various mechanisms through which two particles in nature
exert forces on each other and the force of attraction between them
is a natural phenomenon. It may be cohesion, when two particles are
of the same parent material, or adhesion, when they are of different
parent materials. Cohesion and adhesion are the molecular phenomena
occurring at the interface of liquid and gas, resulting from the electrical
interactions of solid microscopic particles. The maximum distance up
to which the force of cohesion between molecules can act is known as
their molecular range, which is ≈10−7 cm.22
The mechanism of soil adhesion has been investigated by various
scientists and researchers from different angles and perspectives. Based
on the research outcome, various theories are presented. Fountain

presented the theory of capillary in 1954. It was considered either the
normal or tangential force produced by the formation of water film
of constrained water. Five boundary layers for the adhesion of soil are
produced.23 Finally it was proposed that the soil adhesive force is the
algebraic sum of the six forces as follows [24]:
P = pm + pe + pc + pv + pw + pg . (13.23)
P = soil adhesive force
pm = sum of the molecular forces of soil molecules contact with other
bodies.
pe = sum of the forces of electro-static attraction between contact
surfaces.
pc = sum of the capillary forces produced by surface of meniscus formed
by soil liquid with contact surface.
pv = viscous resistance of soil liquid.
pw = wedged pressure produced by potential chemical non-equilibrium
between liquid film of area of contact and void liquid.
pg = negative air pressure produced by closing of soil pores of the area
of contact of soil and other surfaces.
The area of contact has significant effect on various components of soil
adhesive force. Capillary forces are one of the major components of soil
adhesive force followed by magnetic properties of soil such as cation
exchange capacity.25 The adhesive force of soil decreases according to
the order: K+, Mg++, Ca++, H+, Fe+++, Al+++.11 The presence
or absence of organic matter and its contents in soil has significant
effect on adhesive forces. For example, acidic soils are less adhesive than
soil containing fresh rotten materials. Adhesive force increases with the
increase in soil moisture content, and reaches its maximum value and
decreases with any further increase in moisture content as shown in
Fig. 13.2. Generally, adhesive force is maximum when the moisture
content is between plastic and liquid limits. The soil bulk density and
moisture contents have significant effects on adhesive properties of
soils.24 The results of this study showed that disturbed soils are two
to three times less adhesive than undisturbed soils.
The normal force has significant influence on adhesive pressure, as
shown in Fig. 13.2.
Plastic Liquid
limit limit
Fig. 13.2. Relationship between moisture content and adhesive pressure.
Law of soil adhesion

Soil adhesion is a complicated multiple-phase process influenced by
many factors. To formulate the law of soil adhesion, scientists have
conducted a series of experiments leading to the following facts:
(a) Soil type: Clayey soils are more adhesive than loamy and sandy
soils. In other words, the adhesive force is inversely proportional to the
diameter of soil particles. For example, adhesive forces of clay soils vary
from soil to soil depending on the type of clay particles, their size and
parent material, etc. These forces also depend on the cation exchange
capacity of soil, presence or absence of organic matter, and soil moisture
content.
(b) Material in contact with soil: The adhesive forces vary from
material to material. The force increases with increase in the free surface
energy of the contact surfaces. The geometry of contact surfaces also
influences the forces of adhesion.24 For example, spherical convexes
made from ultra-high molecular weight polyethylene (UHMW-PE)
reduced the adhesive forces and improved the scouring properties of
bionic bulldozing plates and moldboard plows.7,26
(c) External forces and environment: The adhesive pressure increases

linearly with the normal pressure, as shown in Fig. 13.3. This increase is
linear until normal force achieves a value of 30 kN. The adhesive force
then increases sharply with any further increase in the normal force.24
Fig. 13.3. Relationship between normal force and adhesive resistance [11].
Techniques used to reduce soil adhesion

Soil scouring could be improved in a number of ways. Some of the
common techniques used in scouring soil from the surfaces of various
tools include liquid injection, pneumatic injection, combination of
pneumatic and liquid injection, vibration, mechanical, heating, electro-
osmosis, modifying shapes and geometries of the tool surfaces, coating,
and bionics, etc.
(a) Surface shape design

Improvement of the surface design of soil-engaging components is one
of the conventional methods for reducing soil adhesion and interface
friction. This method is mostly based on the kinematic and dynamic
analysis of the interactive relation between soil and soil-engaging
components. Some convex corrugations made on the soil-contacting
surfaces can decrease the real contact area and break the continuity
of the interfacial water film between soil and solid surfaces, which
results in a reduction of the soil adhesion and sliding resistance against
solid surfaces.23 A plow with some “comet type” oblique holes drilled
through the plowshare and the main body of a moldboard reduced the
draft by 2.5–3.5% and 8–12% in a moist field soil and under paddy field
conditions, respectively.23
(b) Soil-engaging component materials

Modification of soil-engaging component materials is an important
method for reducing soil adhesion and interface friction. Many
researchers have examined the soil adhesion characteristics of polymeric
materials, owing to their lower surface energy. The tested polymeric
materials included Teflon (polytetra fluoroethylene PTFE), ultra-high
molecular weight polyethylene (UHMWPE), polyethersulphone poly-
tetra floroethylene (PES-PTFE) coating and others.6 The applications
of polymeric materials for soil-engaging components were limited due
to their poor abrasion resistance against soil.29
The soil adhesion characteristics of enamel coating on cage wheels
was assessed by various researchers5,6,30 . The enamel coating reduced
soil adhesion to the surfaces of cage wheels significantly. Rolling
resistance of enamel-coated cage wheels and the plow boat reduced,
and their tractive performances improved as a result of soil scouring.
However, the enamel coating surface did not perform well under
abrasive soil conditions.31 The abrasion resistance of enamel coating
may be increased to some extent by modifying the materials for the
coating and by improving the preparation techniques. For example,
cast iron materials are most commonly used for manufacturing the
majority of soil tillage implements. White iron, one of the conventional
plow moldboard materials, containing more phosphorus and silicon,
improved the anti-adhesive properties.32,33 However, it was noted
that excessive phosphorus content tended to increase brittleness,
which limits the application of phosphoric white iron to soil-engaging
components.
(c) Air and liquid injection to scour soil

Air or a liquid is injected at the soil–tool interface to improve soil
scouring. The liquid acts as a lubricant and prevents adhesion of soil
by aiding free flow of soil over the surface. The formation of an air
cushion or liquid film at the soil–tool interface is helpful in reducing
the contact of soil with the tools. The air and liquid injection at the soil–
tool interface improves the performance of the tools and requires some
additional components for the application of fluid. This technique has
been useful in minimizing the adhesion from the surfaces of excavators,

earth-moving machinery, plows, etc. However, the quantity of water
required as lubricant makes the technique impractical in a number of
situations. For example, more than 16 litres/ha water is required to
achieve a desired level of scouring from the plow surfaces. Moreover,
it is better to use air as lubricant for soils which absorb large quantities
of water.11
(d) Lubrication
Araya and Kawanishi (1984) and Schafer et al. (1977) studied The
effects of the flow of air, water and polymer–water solution on the
friction between soil and solid surfaces was investigated.34 The fluids
injected between soil and soil-engaging components had a lubrication
effect and reduced draft of equipment. Air and water lubrication can
be used to reduce the soil adhesion to excavator buckets.35 One
of the major disadvantages of the fluid lubrication technique is the
requirement of additional equipment for reducing soil adhesion on the
surface of soil engaging components and interfacial friction between
these components and soil. Air lubrication may generate dust under
low soil moisture content situations.
(e) Heat to reduce adhesion and score soil from the tools
The application of heat in reducing adhesion of soil to the surface
of tools has been explored in a number of experiments. The heat
energy requirements for achieving a desired scouring rate were not
studied thoroughly in these experiments and no practical methods have
been developed. Some quantitative data on heat and soil adhesion
was produced.36 The coefficient of the hot slider was considerably
less than that of the wet slider. No information is available on the
temperature values used and the quantity of heat and heat loss from the
slider. These experiments were conducted on dry sand with negligible
adhesion. However, they may provide some useful information on the
effects of heat on friction as shown in Table 13.1. The efficacy of this
technique cannot be verified as it has never been tried under practical
field conditions. The temperature has significant effect on the adhesion
Table 13.1. Effect of Heat on Soil Adhesion

Measured as Coefficient of Friction (µ) of Steel
Slider on Dry Sand [36]
Slider conditions
Slider weight, g Dry Wet Hot
1500 0.266 0.333 0.250

3000 0.266 0.333 0.250
4500 0.266 0.333
Fig. 13.4. Experimental and predicted adhesion for saturated clay soil as a function
of temperature.
of clay soil to the surface of steel. With an increase in temperature from

5◦ C to 30◦ C, the adhesion of soil decreased by 82% (Fig. 13.4).37
Although this technique reduces soil adhesion by minimizing
surface tension of the water film at the soil–tool interface, adhesion
may increase if a suboptimum level of heat is applied. Exhaust air could
be used as a source of heat in practical field conditions. This technique
may be useful when excavating frozen soils. However, this technology
is complicated because of its construction and may not be safe to use
for the operators.
(f ) Electro-osmosis to reduce adhesion

When soil animals come in contact with soil, a microscopic electro-
osmotic system forms in between the stimulated body parts and other
Fig. 13.5. Schematic diagram of the existence of a microscopic electro-osmostic

system on the body surface of soil animals coming in contact with soil [10].
parts nearby (Fig. 13.5). As a result, water in the adjacent soil moves
to the contact zones owing to the action of the potential difference,
the water film at the contact interfaces becomes thicker, so that the soil
adhesion to the body surfaces would be reduced through lubrication.
Although the amplitude of the action potential of soil animals is small, a
microscopic electro-osmotic system can be formed because the distance
between the positive pole and the negative pole is very short. The zone
of negative polarity produced by stimulation from the contacting soil
is on the same surface as the resting body part near to the stimulating
zone. For example, the positive pole and the negative pole were on the
segments near to each other. The action potential of the body surfaces of
soil animals has a dynamic distribution feature. Therefore, this electro-
osmosis was called non-smooth surface electro-osmosis.14 In addition,
the action potential would stimulate such soil animals as earthworms
to produce more secretions.
The increase in the thickness of water film at the soil–tool interface
reduces the adhesive forces significantly. Soil electro-osmosis can be
used to increase the thickness of water film. There is an electrochemical
double layer on the surface of clay particles, and metallic tools are
good conductors of electricity. Soil adhesion can be reduced by forcing
out some of the soil water to act as lubricant at the soil–tool interface
by electro-osmosis. The electro-osmosis time is shorter because of the
continuous operation of tools. Therefore, its potential applications in
field conditions may be very limited. Moreover, when the tools are
operating at higher speeds, the time for electric current to pass through
the soil is further shortened. This will reduce the thickness of water film
acting as lubricant at the soil–tool interface. Therefore, the application
of this technique in practical field conditions is limited. The effects of
electro-osmosis were investigated by using a range of voltages (50–
100V) at operational speeds from 1.25 m/s to 150 m/s. The draft of
the steel slider was expressed mathematically with the inclusion of all
the potential parameters as follows [35]:
F = 0.636 + 0.000988V − 0.00159E − 0.00922W − P2 . (13.24)
F = draft, N
V = speed, m/s
E = potential, volts
W = soil moisture contents, %
P = normal force, N.
This technique may be successfully employed for reducing adhesion
and scouring soil from the surfaces of some tools such as excavators
where soil and tools remain in contact for a comparatively longer
time. For example, the time of static touch between soil and shovel
is approximately 13 seconds, and electro-osmosis may be applied
effectively during this time interval. This technique was successfully
evaluated in the laboratory. The assessment of adhesive forces for
electro-osmosis and non-electro-osmosis techniques at a range of soil
moisture content [11] is presented in Fig. 13.6.
Figure 13.7 shows the relationship between adhesive force and the
duration of electro-osmosis. The longer the electro-osmosis time, the
more water moves out and creates a thick water film at the soil–tool
interface. This water film ultimately reduces the adhesive force. The
higher the soil moisture content (MC), the better the result that can
be achieved via electro-osmosis techniques.
Figure 13.7 shows the relationship between adhesive force and the
duration of electro-osmosis. The longer the electro-osmosis time, the
more water moves out and creates a thick water film at the soil–tool
interface. This water film ultimately reduces the adhesive force.
The voltage, electrode distance, and the area of the electrode plate
all have significant effects on adhesive forces. The application of this
technique requires energy. To make it practical in field situations, the
benefits of reducing the adhesive forces have to be greater than the
25
non-electroosmosis
20 under electro-
osmosis
Adhesive Force, g/m2
15
10
0
0 10 20 30 40 50
Moisture Conetnts, %
Fig. 13.6. Relationship between adhesive force and electro-osmosis at two soil
moisture contents [11].
14
12
Adhesive Force, g/m2
10
30%
8 MC
0
0 10 20 30 40
Time, Second
Fig. 13.7. Relationship between adhesive force and electro-osmosis duration [11].
energy spent to achieve them. The experimental model suggested that

it is fully feasible for excavators.11
The electro-osmotic effect is influenced by: the soil moisture
content, soil texture, electric voltage, contact time for electro-osmosis,
the separation between the zones of positive and negative polarity and
their area ratio, and the contact pressure. A high voltage for electro-
osmosis and a long contact time are required in order to increase the
thickness of the interfacial water film between the soil and the zone of
negative polarity.
Soil adhesion to the zone of negative polarity could be reduced
by shortening the separation between the two terminals. The electro-
osmotic effect is better for clay soil than for sandy soil.27,35 The
application of the electro-osmotic method to a sliding component in
contact with soil was limited because of the long contact time required
for electro-osmosis. The method of electro-osmosis may be applied
successfully only to reducing adhesion of soil to the soil-engaging
components with a long contact time between them, for example,
excavator buckets.35 Much more energy is consumed when high voltage
is used for electro-osmosis and further research is required into the
effective use of lower voltages. It was reported that the voltage can be
decreased by changing the position of the zones of polarity and their
arrangement.35 .
(g) Soil scouring by vibration

Vibration can be used in reducing adhesion of soil to the surface of soil-
engaging components of various machines and tools. Soil contact with
tools is reduced when vibration is applied perpendicular to the soil–
tool interface. This reduces the soil–tool area of contact and lets the
air and water flow smoothly. This technique is helpful in improving
both lubrication and soil scouring. Although, it improves scouring
characteristics of the tools and machines, it has not been practically
applied because of the potential damage it may inflict on machine itself.
The soil adhesion can be reduced by vibration of parts of the machines
that come in contact with soil.38
The effect of ultrasonic vibration on the friction of metal against
soil was investigated.17 The vibration frequency used was 10 kHz.
A mechanism used for testing the effect of mechanical vibration on
soil adhesion was developed. The vibration was produced by blowing
compressed air. The soil used for the tests was a silty loam with a
moisture content of 38.7% d.b. Vibration at frequencies of between
Fig. 13.8. Soil separation versus frequencies of vibration showing the effects of
vibration on scouring of soil-engaging components [40]; soil spraying speed during
preconditioning: 5.0–5.9 m/s; 6.0–7.0 m/s.
60 and 100 Hz was remarkably effective in removing soil from the

plate (Fig. 13.8).39,40
(h) Soil scouring by mechanical means

Soil scouring can be improved by making the contact surfaces elastic
and flexible. The productivity of the shovel improved by 15% when
this technique was used to excavate frozen soils [11]; chains have
been used to improve soil scouring from excavators. Sometimes, the
structure becomes too complicated and requires regular service to keep
the system functioning properly. Scraping the soil is commonly used to
scour soil from the surfaces of a number of soil-engaging components
of various machines and tools. This technique is simple to construct,
easy to use and requires no extra power. Scarpers are commonly used on
tillage and sowing equipment. Wear and tear is a major issue associated
with this technique. However, it is more complicated to use scrapers
on the surfaces of shovels.
An additional mechanism such as scrapers can be attached to
remove the soil adhering to the soil-engaging components.35 This
method is primarily applied to excavator buckets and blades. Scrapers
are also used to remove soil from discs and press-wheels used on
Scrapers
Fig. 13.9. Scrapers used to remove soil from presswheels while working under moist
sticky soil conditions.
tillage and sowing machinery (Fig. 13.9). This technique adds extra
cost, increases the weight of the equipment and requires frequent
changing of scrapers under abrasive soil conditions. Improved soil
scouring properties of a bucket with a flexible floor which reduced soil
adhesion on the buckets by 59.4%. This technique reduced the loading
resistance by 15.85% at the same time.41 A tongue scraper mechanism
for a loader bucket to remove the soil accretion in the bucket. The
shovel with this mechanism proved to increase loader productivity by
20% while handling soil and powdered coal.
(i) Soil scouring by surface modifications

Soil scouring can be improved by modifying the surface characteristics
of soil-engaging components of machines and tools. For example,
silicon lubricant oil was used in China to improve the scouring
properties of a wooden plow.11 Some of the materials such as ultra-high
molecular weight polyethylene (UHMW-PE), polytetra fluoroethylene
(PTFE), polyethylene and porcelain are used for coating the surfaces of
moldboard plows.9 The coefficient of friction of PTFE can be up to 50%
less than steel 45. PTFE improves soil scouring and reduces the draft by
up to 25% as shown in Table 13.2.11 Enamel coating was employed on
the surface of moldboard plows to improve soil scouring properties.
The draft of enamel-coated plow surfaces was reduced by 14% and
Table 13.2. Effects of PTFE Plow Covering on Draft (kN) of Moldboard Plows
Operating at Two Different Speeds [11]
Draft, kN
Decatur clay Deviation clay
Plow
surface 1.60 km/h 5.6 km/h 1.60 km/h 5.6 km/h
Steel (conventional) 2.00 2.18 2.14 2.56

PTFE covered moldboard 1.73 2.16 1.89 2.36
with steel share
PTFE covered moldboard 1.62 1.91 1.38 1.96
and share
16% at 3.6 km/h and 4 km/h working speeds, respectively (Salokhe

et al., 1990). In another laboratory experiment, enamel coating on the
moldboard plow surface reduced the draft by up to 26%, depending
on soil moisture content and working speed.6 Some of these materials
have poor wear resistance and practically cannot be used in abrasive soil
conditions. For example, PTFE coating of 0.50–0.80 cm thickness lasts
for plowing of 20 ha and high-density polyethylene covering of 0.50–
0.80 cm lasts for about 8 ha in abrasive soils (Salokhe et al., 1990). The
higher costs and lower life are the main limiting factors in using these
materials practically in the field.
( j) Soil scouring by bionic

Soil animals have the ability to scour soil from their bodies through
adapting to their surrounding due their evolution in biological system
over time. There are a range of features assisting these animals to scour
soil from their bodies. A number of these animals spend the majority
of their life underground; it is hard to get a good grasp of their daily
activities and their life cycles. For example, stronger digging legs are the
result of active adaptation for burrowing. The forelegs of mole crickets
(Gryllotalpidae) and the claws of black ground beetles (Carabidae),
dung beetles (Scarabaeidae) and burrowing bugs (Cydnidae) are some
important features of evolution.43 Moreover, the adaptation of animals
to the soil habitat resulted in shorter or vestigial additional legs, reduced
body size, thinner or flatter with eyes and wings diminished, as shown
Fig. 13.10. Macroscopic features of soil animal evolution (after Xin, 1986).
in Fig. 13.10. With the continuous interaction between soil animals and
the soil in which they live and move, their existence would threatened
if soil stuck to their bodies for a longer period. The soil scouring
characteristics of their body surfaces which evolved over time are
important for their survival and functioning. These features can be used
to improve the soil scouring properties of soil-engaging components
of agricultural and earth-moving machinery. Figure 13.11 shows the
surface morphology of some soil animal cuticles.
The surface morphology of the ground beetle was employed for
the modification of moldboard plows and bulldozing plates. The plow
and bulldozing plates modified by bionics improved soil scouring
and reduced the draft of these implements significantly. The bionic
Fig. 13.11. Non-smooth surface morphology of soil animal cuticles. (a, b) Head
of dung beetle (c) abdomen of ground beetle, (d) head of black ant, pangolin scale,
(e) pangolin scale [34,44].
modification of the surfaces of various machines and tools based on

the surface morphology of a number soil animal cuticles are simple and
can be used in practical field conditions. However, the manufacturing
techniques are comparatively complicated unless produced on mass
scale.7
Soil animals possess a significant capability of preventing soil
adhesion to their bodies and diminishing sliding resistance. It is because
of several factors. Special chemical composition and unsmooth surface

morphology of soil animal cuticles are two of the important factors in
scouring soil.
The shape and size of the soil-engaging components of machinery
and equipment have significant effect on soil adhesion. For example,
reducing the soil–tool area of contact and unsmoothed surface mor-
phology to discontinue the moisture film at the soil–tool interface are
helpful in scouring soil. The unsmoothed moldboard plow surfaces
with ultra-high molecular weight polyethylene (UHME-PE) convexes
improved soil scouring and reduced the draft by 25% and 30% at 3.6
km/h and 4 km/h plowing speeds respectively.
The application of the concept of bionic non-smooth surface on the
disc plows significantly reduced the draft of the modified plows through
reducing adhesion of soil to their surfaces. Moreover, the higher or
deeper bionic units always resulted in less soil resistance. Convex bionic
units resulted in reducing the soil resistance by up to 19% as compared
to concave bionic units. The surfaces with a bionic unit density of 30%
gave the highest resistance reduction compared to 10% density. The
concept of bionic non-smooth units can be applied to the modification
of disc plows for improving soil scouring characteristics and reducing
soil resistance.45
(k) Magnetic fields

The effect of a magnetic field on plowing resistance were investigated by
researchers.46,47 Several permanent magnets were attached to the back
of the plowshare. Both the plowing resistance and the fuel consumption
of the tractors used for pulling the plows were reduced by 10.55% and
11.30%, respectively, for the magnetized plow as compared to the non-
magnetized plow under identical field conditions (in a medium loam,
with a moisture content of 15.98% d.b.).46 In another study the fuel
consumption of the tractors was reduced by 7–15% for the magnetized
plow in field soil with a moisture content of 12.26–14.28% d.b. and for
a plowing depth of 25–30 mm.47 The effect of the magnetic field on
the forward resistance of the covering blade in a medium loam soil at a
moisture content of 12% d.b were investigated in the laboratory. The
forward resistance of the magnetized covering blade can be reduced by
15.4–23.3% as compared to conventional blades of same specifications.

It was also found that the largest reduction in the forward resistance
occurred at a magnetic induction of 23 tonne.48
(l) Soil scouring by vibration

Vibration can be used in reducing adhesion of soil to the surface of soil-
engaging components of various machines and tools. Soil contact with
tools is reduced when vibration is applied perpendicular to the soil–
tool interface. This reduces the soil–tool area of contact and lets the air
and water flow smoothly. This technique is helpful in improving both
the lubrication and the soil scouring. Although it improves scouring
characteristics of the tools and machines, it has not practically applied
because of the potential damage it may inflict on machine itself. The
soil adhesion can be reduced by vibration of parts of the machines that
come in contact with soil.38
(m) Soil scouring by hydrophobicity

It was found that the water film on the body surfaces of soil animals
would rapidly contract and fall from their bodies as they came out of wet
soil. This phenomenon indicated that their body surfaces had a strong
hydrophobic nature. The organic substances, similar to resin and paint,
including hydrophobic membranous protein and non-polar radicals of
phospholipid on the body surfaces contributed to their hydrophobic
nature. The hydrophobic ability of the body surfaces suggested that the
attraction between the body surface substances and water molecules was
very weak. It can be considered that the body surface substances were
low surface energy materials. The inherent hydrophobic ability of the
body surfaces of soil animals would be enhanced by their geometrically
non-smooth or rough structure. The combination of the geometrically
non-smooth structure and its hydrophobic nature is even better in
preventing soil from sticking to the body surfaces of soil animals.
Mechanics of reducing adhesion and draft of bionic

bulldozing blades
Most theories of soil adhesion identical that the water membrane of
soil is the main cause of adhesion of soil to the surfaces of other bodies.
For example, Fountain’s theory describes soil adhesion mathematically

as follows:
P = a.S (13.25)
P = Normal adhesive force at the interface of soil and surfaces of other
bodies, N
S = Area of water membrane connecting the soil and the surface of
other bodies, cm2
a = Water tension in soil, Pa
The adhesive force of a particular soil may be reduced by reducing
the area of the water membrane connecting the soil and the surfaces
of material and reducing the water tension. Figure 13.3 shows the
relationship between normal and adhesive forces.
Factors affection soil adhesion to the contact surfaces

Soil adhesion is a complicated multiple phase process influenced by
many factors. Scientists conducted a series of experiments on the
relationship between soil types, properties of contact materials, external
forces and soil adhesive forces to formulate the laws of soil adhesion.
(a) Soil type

Clayey soils are more adhesive than loamy and sandy soils, and the
adhesive force is inversely proportional to the diameter of soil particles.
The adhesive force is directly proportional to the gum body contents of
the soil.25 Moreover, adhesive forces of clay soils vary from soil to soil
depending on the type of particles, particle size parent material, cation
exchange capacity of soil, presence of organic matter contents and
moisture contents, etc. The adhesive force of soil decreases according to
the order: K+, Mg++, Ca++, H+, Fe+++, Al+++.11 The presence
or absence of organic matter and its contents in soil has significant
effect on adhesive forces. For example, acidic soils are less adhesive
than soil containing fresh rotten materials. Adhesive force increases
with the increase in soil moisture contents, reaches its maximum value
and decreases with any further increase in moisture contents, as shown
in Fig. 13.3. Generally, adhesive force is higher at moisture contents
between plastic and liquid limits. The effects of soil bulk density and
moisture contents on their adhesive properties. Loose dry soils are two
to three times less adhesive than the moist dense soils.24
(b) Material in contact with soil

The adhesive forces vary from material to material. It increases with the
increase in the free surface energy of the contact surfaces. The geometry
of contact surfaces also influences the forces of adhesion.24 The
spherical convexes made from ultra-high molecular weight polyethylene
(UHMW-PE) reduced the adhesive forces and improved the scouring
properties of bionic bulldozing plates and moldboard plows.7,26
(c) External forces and environment

The adhesive force increases linearly with the normal pressure as shown
in Fig. 13.17. This increase is linear until normal force achieves a value
of 30 kN. The adhesive force then increases sharply with any further
increase in the normal force.24
Factors affecting soil adhesion

On the basis of the above explanation of soil adhesion, it is understood
that the adhesion between soil and solid surfaces is dependent upon
the nature and properties of soil, the properties of the soil-engaging
components’ material and the experimental conditions or working
surroundings.
(a) Soil properties

Soil factors affecting soil adhesion include soil texture, moisture content
and water tension, porosity, organic matter content and others.49
Soil adhesion varies with the types of minerals present in the soil
composition. It increases with the increase in the proportion of clay
particles in the soil. Adhesion of soil is highest when the soil moisture
content is between the plastic limit and the liquid limit (Fig. 13.3).
Moreover, an increase in soil water tension elevates soil adhesion.
Application of fractal geometry was investigated to evaluate the impacts
of irregular morphologies and soil structures on soil adhesion.50 The
soil physical properties such as bulk density, pore size distribution,
aggregate-size distribution, ped shape and soil micro-topography can

be described by fractals. Moreover, soil processes such as adsorption,
diffusion, transport of water and solutes, brittle fracture and fragmen-
tation can also be modelled by fractals. At the same time the soil spatial
variability can be quantified by fractals.51 As an indicator of soil texture,
the soil particle-size distribution displays fractal structure in many
cases. Ultimately, Tyler and Wheatcraft (1992) derived a relationship
between the cumulative mass of soil particles and the particle size as
follows:
M(r < T)/MT = (R /RL)3−D . (13.26)
M(r(R)) = cumulative mass of soil particles of size requal to or more
than a characteristic size R,
MT = mass of the total soil particles,
RL = size of the largest particle
D = the fractal dimension of the soil particle-size distribution.
If the particle-size distribution of a soil is a fractal, then the log–
log plot of M(r(R)) and R/RL is a straight line. The fractal dimension
D can be calculated from the slope of the straight line. Generally, the
fractal dimension of particle size distributions of soils increases with the
increase in the clay particle contents. Therefore, the soil adhesion and
frictional forces are increased with the increase in fractal dimension of
soil particle size.13,20,52 The molecular fractal dimension of soil particle
surfaces is also an important factor affecting soil adhesion. The larger
molecular fractal dimension increases the special surface area and the
soil adhesion.
(b) Effects of properties of soil-engaging component

surfaces
The atomic or molecular structures of the surfaces of solids or liquids
are different from that of their bulk materials due to the lost structural
symmetry of atoms or molecules. The physical and chemical properties
of surfaces are changed as well. This results in surface phenomena
such as physical adsorption, chemical adsorption, chemical reaction
and soil adhesion. The surfaces may be classified into high-energy and
low-energy surfaces on the basis of surface free energy. High surface
energy materials such as metals, metallic compounds and inorganic

compounds (oxides, nitrides, silica and diamond) have surface tension
ranging from 0.2 to 0.5 N/m. Low surface energy materials such as
organic compounds and organic polymers has surface tension generally
less than 0.1 N/m. Water is a low surface energy material with a
surface tension of 0.0728 N/m at 20◦ C. High energy surfaces are
hydrophilic and low-energy surfaces are hydrophobic. The hydrophilic-
ity or hydrophobicity of a surface can be expressed by the contact angle
of water on the surface. The higher the contact angle of water on a
surface, the higher is the hydrophobicity of that material and it has
the lower hydrophilicity. Table 13.3 lists advancing contact angles θA
and receding contact angles θR of water on the surfaces of some solid
materials. The adhesion between soils on the surfaces of hydrophobic
materials is poor. Therefore, the larger the contact angles, the lower will
be the soil adhesion. Therefore, metallic materials, metallic compounds
and inorganic compounds display higher soil adhesion, and polymeric
materials have a lower soil adhesion.
The effects of microstructure on the adhesion of soil to the surface
of conventional steel (steel-35) were investigated by researches.53,54,55
Table 13.3. Contact Angles of Water on Surfaces of Various Solid

Materials
Advancing Receding
contact contact
Materials angles, θA angles, θR References
Polytetra fluoroethylene 109 106 Wu (1982)

(Teflon) — PTFE
Polystyrene 91 84 Wu (1982)
Polyethylene 96 62 Wu (1982)
Polypropylene 108 Adamson (1976)
Paraffin wax 110 Adamson (1976)
Polycarbonate 84 68 Wu (1982)
Human skin 90 Adamson (1976)
Gold 66 Adamson (1976)
Platinum 40 Adamson (1976)
Plain Carbon Steel 73 0 Tong et al. (1994c)
Enamel 29 0 Tong et al. (1994c)
1.8
1.7
Adhesion Force, kPa
1.6
1.5
1.4
1.3
1.2
1.1
1
100 150 200 250 300 350 400 450 500 550 600
Temperature, C
Fig. 13.12. Soil adhesion of steel-35 after tempering at various temperatures [55].
They found that the soil adhesion to the surfaces of steel-35 tempered
martensite microstructure and sorbite microstructure was lower than
that of the microstructure obtained through quenching and low-
temperature tempering treatment, quenching and average-temperature
tempering treatment, and the tempered sorbite through quenching
and high-temperature tempering treatment. Figure 13.12 illustrates
the relationship between soil adhesion and the tempering temperature
for steel-35. Soil adhesion to the surface of steel-35 is higher within
the average-tempering temperature range soil. The surface tension
of water is also a function of temperature. The surface tension and
viscosity of water decrease with an increase in temperature. Therefore,
any increase in temperature reduces the surface tension of the interfacial
water film which in turn minimizes the soil adhesion. When an electric
field generated by a negative pole and positive pole is applied to soil, the
soil water can move from the positive pole to the negative pole. This
is called the electro-osmotic phenomenon of soil. If the non-scouring
areas of soil contact are given negative polarity, the thickness of the
water film increases owing to electro-osmosis, which can reduce the
soil adhesion.27,56 The nature and properties of the soil can be changed
under a strong magnetic field, for example: the variable negative electric
charge and the permeability are increased, and the positive electric
charge and the expandability are decreased.57 These changes, in general,
reduce the adhesion of soil to solid materials.47,48 Ultrasonic vibration
and mechanical vibration can reduce the soil adhesion and interface
friction or separate the adhered soil from the solid surfaces.17,39,40
Biomimetics of soil-engaging components

of agricultural and earth-moving machinery
(a) Components with biomimetic non-smooth surfaces —
bulldozing blades with biomimetic, embossed surfaces
The surface morphology of dung beetle was imitated on the surfaces
of to bulldozing blades using pseudovariable approximation design
optimizing method. The biomimetic bulldozing blades were made
from plain carbon steel and were 400 mm long by 200 mm wide.
There were two arrangement patterns, a rectangle and a parallelogram
(Fig. 13.13). The small convex domes embossed on the surface were
2–8 mm in height, 16–32 mm in base diameter and 20–50 in number.52
Bulldozing tests were conducted in an indoor soil bin using a clay
soil with an average moisture content of 27.8% d.b. It was found
that the bulldozing resistance of the embossed blades was reduced,
on an average, by 13.02% in comparison to a conventional (smooth)
blade.
The structural parameters of the optimum embossed surface were
convex domes, 7 mm in height, 25 mm in base diameter and 45 in
number, with a parallelogram arrangement. In the optimum case,
bulldozing resistance was reduced by 18.09% as compared to that of
the conventional smooth blade.
Fig. 13.13. Schematic diagrams showing two regular distributions of convex domes
on the biomimetic embossed non-smooth bulldozing blades: (a) rectangle distribution;
(b) parallelogram [52].
Fig. 13.14. Schematic diagram of a biomimetic moldboard plow [59].
Other researches designed embossed bulldozing blades with six

arrangement patterns (Fig. 13.14).7,59 The small convex domes were
made from steel-45 and UHMWPE. The small convex domes, prepared
in advance, were glued on a conventional smooth steel-45 plate
250 mm in length and 130 mm in width. The tests were conducted
on a soil cutting test table. The depth and the angle of cut were 15 mm
and 353◦ , respectively.
Research on the soil adhesion preventing mechanism of non-
smooth bionic plates at three operating speeds of 0.01, 0.02 and
0.06 m/s found that the bulldozing resistance of the six biomimetic
non-smooth bulldozing blades whose convex domes were made from
UHMWPE was reduced to different extents and the fifth blade
(Fig. 13.15) had the lowest resistance, which was reduced by 27.50,
28.59 and 34.0% at the cutting speeds of 0.01, 0.02 and 0.06 m/s,
respectively. Of the six bulldozing blades whose convex domes were
made from steel-45, the fifth blade was the only blade with a reduction
in the bulldozing resistance, which was reduced by 1.3, 1.5 and
15.55% at cutting speeds of 0.01, 0.02 and 0.06 m/s, respectively. The
resistance of the other blades was higher than that of the conventional
blade.7
The resistance of bionic bulldozing plates were compared them
with the conventional plates. The drafts of the bionic plates were less
than those of the conventional ones as a result of their better scouring
properties. These results are represented in Fig. 13.16.60
Fig. 13.15. Six non-smooth bionic bulldozing plates tested in the laboratory [7].
Fig. 13.16. Forward resistance versus forward velocity, showing a reduced resistance
for the biomimetic embossed non-smooth bulldozing blade with a random distribution
of convex domes compared with a conventional blade (derived from Ren et al., 1995b);
biomimetic blade; conventional blade.
A statistical analysis of the characteristic parameters of the surface

morphology of the unsmooth bulldozing blades, including the charac-
teristic dimensions, and the number of the small convex domes, using
a scanning electron microscopy photograph was undertaken. It was
Fig. 13.17. Schematic diagram of a biomimetic embossed non-smooth bulldozing

blade with a random distribution of small convex domes.60,61
found that the position distribution of the small convex domes on the
bulldozing blade’s surface followed a statistically uniform distribution
because the number of the convex domes NL through proportion to
the length, of this segment(Fig. 13.17) [61]:
NL = a + bL, (13.27)
where a and b were parameters to be estimated, which were −0.2430
and 0.1692, respectively, for the dung beetle tested. The base diameter
of the convex domes ranged from 0.033 to 0.749 mm. The size of the
base diameter of the convex domes followed a Gaussian distribution
on the basis of the s2 test, and the mean and the root-mean square
deviation were 0.052 and 0.008 mm, respectively.
Based on the above statistical analysis, a biomimetic embossed
bulldozing blade was designed by researchers.60 The blade and the
convex domes were made from plain low carbon steel (Q235A). The
tests were run on a soil cutting test table using a clay soil with a moisture
constant of 27% d.b. the angle of cut of a straight line segment along any
direction was in direct 31-463 (the angle of the blade with respect to
the direction of travel in the horizontal plane) and the cutting speeds
were 13.33–58.82 mm/s. It was demonstrated that the biomimetic
blade had a lower resistance to forward travel (Fig. 13.16). This has
been confirmed by earlier research.7
(b) Bulldozing blades with biomimetic, corrugated surfaces

The wavy structure of body surfaces of some soil animals was imitated
by reaserchers in China while designing a biomimetic, corrugated
Fig. 13.18. Schematic diagram of biomimetic corrugated non-smooth bulldozing

blade [62].
bulldozing blade (Fig. 13.18). The dimensions of the biomimetic

blades prepared were 300 mm long by 150 mm wide. Six biomimetic
blades of different wave parameters were prepared (Fig. 13.25) and
cutting tests demonstrated that the soil adhesion to the conventional
blade did not occur on the corrugated blade. The bulldozing resistance
of the biomimetic, corrugated blade was reduced considerably in
comparison with the conventional blade. A finite element analysis of
the interaction between the soil and a corrugated bulldozing blade
predicted the horizontal and vertical components of the bulldozing
resistance during operation.62
(c) Bulldozing blades with biomimetic dimpled and

scaly surfaces
Biomimetic, non-smooth bulldozing blades imitating the dimpled form
and the scaly form of surfaces of soil animals was designed (Fig. 13.19).
A comparative test of the bulldozing resistance of the two biomimetic
bulldozing blades in clay soil, wet sand and dried sand was conducted.
It was found that the resistance against wet sand and dried sand was
increased markedly, while the resistance in disturbed yellow clay was
reduced considerably.63
(d) Moldboard plows with biomimetic embossed surfaces

A biomimetic, embossed plow moldboard was developed. It was
prepared by gluing small UHMWPE convex domes on the surface of
a conventional moldboard (steel-35). The small convex domes had a
base diameter of 20 mm and a height of 2 mm and were made from
Fig. 13.19. Schematic diagram of the biomimetic bulldozing blades: (a) dimpled,
and (b) scaly [63].
Fig. 13.20. Soil adhesion to steel-35 versus its surface relative potential [63].
UHMWPE.59 The angles of the sliding direction of the soil on the

moldboard surface with the horizontal surface were between 55 and
65◦ .64 Based on this phenomenon, an arrangement pattern of convex
domes on the moldboard was designed (Fig. 13.20). The biomimetic
plow was tested in an indoor soil bin using black clay soil. The plowing
resistance of this biomimetic non-smooth plow was reduced by 26
and 34% at the forward speeds of 3.6 and 5.0 km/h, respectively,
as compared with a corresponding smooth moldboard plow. The soil
adhering to the biomimetic plow surface was less than that to the
smooth plow.
A biomimetic, embossed moldboard plow with the same non-
smooth structure as the biomimetic bulldozing blade surface shown in
Fig. 13.17 was designed. The base diameter of small convex domes on
the moldboard surface was a Gaussian distribution and the position of

the convex domes was a statistical uniform distribution. The convex
domes were made from white iron alloy. The white iron alloy was
manufactured into a metal pencil. The white iron alloy convex domes
were welded on the surface of the conventional moldboard made from
steel-35. The resistance of the biomimetic plow was reduced by 6.6–
12.7% as compared to the conventional moldboard plow operating
under identical conditions.58 Figure 13.20 shows adhesion of soil to
the surface of steel-35 with respect to its relative potential.
Modification of soil-engaging materials

(a) Iron and steel
The effects of the chemical composition and microstructures on the
water wettability, corrosion resistance and soil-scouring properties of
three commercial materials used for cast-iron moldboards in China
were examined. The moldboards made in Yangcheng in the Shanxi
province of China showed better soil scouring properties of the three
traditional moldboards tested in these trials. Its composition featured
higher carbon (4.27%) and phosphorus (0.58%,) and lower manganese
(0.07%,), silicon (0.05) and sulphur (0.032%) content. The Yangcheng
moldboard possessed the best corrosion resistance.33
The effect of the corrosion resistance of steel-35 on its anti-adhesive
characteristic to soil were studied in the laboratory. The microstructure
of steel-35, hardened and tempered at a low temperature or hardened
and tempered at a high temperature, had higher corrosion resistance
than that for the microstructure hardened and tempered at an average
temperature, as did their anti-adhesive characteristics to soil.66 The soil
adhesion to iron and steel was related to their individual relative surface
potential. A calomel electrode was taken as a standard electrode for
comparison to determine the relative surface potentials of steel-35 of
different microstructures using 0.1% NaCl solution as electrolyte.65
The soil adhesion to steel-35 with a microstructure of a lower relative
surface potential was lower. This phenomenon seems to have a cor-
responding relation with the characteristic phenomenon of the action
potential of soil animals.
Li (1993) developed a white iron alloy and applied it for preparing

convex domes for the embossed surfaces. The white iron alloy had
a lower relative surface potential. In addition, the iron had a better
abrasion resistance. The biomimetic embossed plow whose convex
domes on the moldboard were made from the iron displayed a better
anti-adhesion characteristic and lower plowing resistance.66
(b) Polymer composites and coatings

Although they had a better anti-adhesive characteristics, applications of
polymeric materials to soil-engaging components were limited because
of their lower abrasive wear resistance against soil.13,21,68 The composite
material technique is a better method for increasing abrasive wear
resistance of polymeric materials. The second phase particles or fibers
existing in composite materials can make their surfaces possess a feature
similar to the geometrically non-smooth topography of the body
surfaces of soil animals.
An alumina particle reinforced Teflon composite material was
prepared for assessing its wear resistance against soil.69,70 When the
alumina particle content was less than 20% by weight, its abrasive wear
resistance was increased considerably and the soil adhesion was affected
a little (Fig. 13.22). The effects of quartz particles and glass beads as the
reinforced materials were investigated upon soil adhesion and abrasive
wear of UHMWPE. The anti-adhesive characteristic of the UHMWPE
matrix composite was not affected significantly as long as the content
of the reinforced materials in the composite was not more than 9% by
volume.71,72
The water wettability of composite materials reinforced with par-
ticles was analyses. Based on the assumption that (a) the particles in
composite materials were spherical and had the same diameter which
was smaller than the water droplet during tests of contact angles of
water on the surfaces of composites, (b) the particles were uniformly
distributed in the matrix of the composite materials, and (c) the outer
surfaces of particles were circular with variable diameters, they derived
a theoretical equation for the particle reinforced composite materials as
follows73 :
√
Cos θc = Cos θ1 + 3 2/4V(Cos θ2 − Cos θ1 ) (13.28)
Fig. 13.21. Schematic diagrams showing six regular distribution patterns of convex
domes on biomimetic embossed non-smooth bulldozing blades [7].
θc = contact angle of water on the surfaces of a composite material;

θ1 = inherent contact angle of water on the matrix material;
θ2 = inherent contact angle of water on the reinforcing material in
bulk; and
V = volume content of reinforcing particles in a composite material.
The contact angles of water on nylon 1010 (PA1010) composite

materials reinforced with alumina particles followed Eq. (13.28).
The alumina particle reinforced PA1010, epoxy (EP), polyurethane
(PU) matrix composite coatings and the steel-T8 coating impregnated
with polysiloxane was prepared.74,75,76,77 These coatings were applied
to a smooth bulldozing blade and/or the fifth embossed blade shown
in Fig. 13.21 and tested for their bulldozing resistance in a soil cutting
test table using clay soil with moisture contents of 29.4 and 33.0%
d.b. The cutting speed was 0.02 and 0.04 m/s, the depth of cut
was 20 mm and the angle of cut was 45◦ . In the comparison with
an uncoated smooth blade, the reductions in cutting resistance were
Fig. 13.22. Effects of size and content of Al2 O3 particles on the polytetrayuoroethy-
lene (PTFE) matrix composite reinforced with Al2 O3 particles: (a) soil adhesion;
(b) abrasive wear; 120 mesh; • 280 mesh [70].
26.38, 23.23 and 22.27% for the smooth bulldozing blades coated
with PA1010, EP and steel-T8 coatings, respectively, and 23.58 and
35.20% for the non-smooth bulldozing blades coated with PA1010 and
PU composite coatings, respectively. After EP and PA1010 composite
coatings were applied to the plow moldboards, their plowing resistance
was reduced by 30.31 and 35.88%, respectively. Moreover, no soil stuck
to the moldboard as compared to the conventional one (steel-35).
The soil used for the test was black clay with a moisture content of
20.5% d.b. Both the depth and the width of the plow were 200 mm.
The forward speed was 0.95–1.88 m/s (3.4–6.8 km/h). The above
results showed that the combination of hydrophobic materials and
geometrically non-smooth surfaces was a better method for reducing
adhesion and interfacial friction of soil-engaging components.
Biomimetic non-smooth electro-osmosis

(a) Basic structure of biomimetic electro-osmotic surfaces
Based on the characteristics of the surface action potential of soil
animals, a combined measuring dome was designed14 with a positive
pole in the center and a negative pole at the periphery. This combined
disc was used for testing, in principle, the effect of the electro-osmosis
upon soil adhesion. The soil used had a moisture content of 23) 5%
d.b. The voltage for electro-osmotic tests was 12 V. The applied normal
stress was 26) 78 kPa. It was found that there existed an area ratio for
the zones of positive and negative polarities that resulted in minimum

soil adhesion. In other words, the ratio of the area of positive pole
to negative pole was an important factor affecting soil adhesion. This
electro-osmotic system in which the positive pole and negative pole
were on the same surface and the positive pole was on a convex dome
was called non-smooth surface electro-osmosis or biomimetic electro-
osmosis.
(b) Bulldozing blades with biomimetic electro-osmosis
According to the principal test results,14,15 designed two biomimetic
bulldozing blades with non-smooth surface electro-osmotic structures.
One was the convex dome positive pole (Fig. 13.23(a)) and another was
the corrugated form of the positive pole (Fig. 13.23(b)). Bulldozing
tests were conducted on a soil cutting test table. The forward speed was
from 43 to 172 mm/s, the depth cut was 50 mm, the angle of cut was
60◦ , the voltage for electro-osmosis was 12 V and the ratio of area of
the positive pole to the negative pole was 1:7.3. It was found that while
soil stuck on the surface of the conventional blade and the non-smooth
surface without electro-osmosis, only a little or no soil stuck to the
surface of biomimetic, electro-osmotic plate under identical conditions.
Biomimetic electro-osmosis can further reduce bulldozing resistance
by 9–12% and 15–32% for the embossed and corrugated non-smooth
surfaces, respectively, as compared with the corresponding non-smooth
surface without electro-osmosis.
Fig. 13.23. Photographs of two biomimetic bulldozing blades with electro-osmosis:

(a) embossed; and (b) corrugated [15].
Fig. 13.24. Schematic diagrams of biomimetic electro-osmotic surface structures

used for the lugs of paddy-field wheels [78].
(c) Wheel lugs with biomimetic electro-osmosis

The electro-osmotic principle for non-smooth surfaces was used78 for
the lugs of paddy-field wheels. They designed varied convex forms of
biomimetic lugs (Fig. 13.24). The tests were conducted in an indoor
soil bin using supersaturated black clay soil similar to that found in
paddy fields. The working depth of the lugs was 150 mm. Soil blockage
due to adhesion occurred on the conventional lugs without electro-
osmosis and the blocked soil was not self-cleaning, while only a little
soil adhered to the lugs with biomimetic electro-osmosis, and the lugs
were self-cleaning. When a voltage of 24 V was used for electro-osmosis
and convex domes with a base diameter of 24 mm were taken as positive
poles, the soil shedding property of lugs was the best at an area ratio
for the zones of positive and negative polarity of 8–12.5%. The traction
force and the driving efficiency of paddy-field wheels were increased
owing to biomimetic electro-osmosis (Fig. 13.25).
Flexible components for removing soil

(a) Loading bucket with a flexible lining
A flexible steel mesh based on the flexibility features of soil-burrowing
animals was designed.79,80,81 The steel mesh was made from linked steel
rings and slotted steel rings (Fig. 13.26). The steel mesh was made into
a lining suitable for such soil-engaging components as digger buckets
and wheel buckets. The position of the lining could be changed with
the position of the bucket as shown in Fig. 13.27. When the bucket
with a steel mesh lining was filled with soil, the lining was pressed close
Fig. 13.25. Effects of biomimetic electro-osmotic lugs on a paddy-field wheel:

(a) tractive force, (b) tractive efficiency [78] • with electro-osmosis; without electro-
osmosis.
Fig. 13.26. Schematic diagram of structure of a flexible steel mesh lining (after Sun
et al., 1992).
to the inside of the bucket. At the beginning of unloading, both soil

and the lining slide out from the bucket simultaneously, and then the
soil was separated from the lining gradually by gravity and, finally, the
lining was hung from the bucket. Only the smaller size of soil particles
could enter between the lining and the inside of the bucket through
the gap in the lining, since the size of the gaps in the lining was very
small. In order to protect soil from blocking between the lining and
the inside of the bucket, several holes were drilled through the bottom
of the bucket shell to release the trapped soil.
Fig. 13.27. The position of the flexible lining with respect to the bucket [81].
(b) Self-unloading box of dump truck with flexible chains

The effect of three flexible linings, steel mesh, steel chain and nylon-
knitted net, on a model of a self-unloading box were analyzed.82 The
tests were conducted using yellow clay with a moisture content of 24.5%
on dry basis. After 20 successively loading-unloading operations, the
soil adhering to the box without flexible lining was 14% by volume of
the total holding capacity of the box. Heavy soil adhesion took place
after 10 loading-unloading operations for the nylon-knitted net and,
afterwards, it lost the flexible function and soil-shedding action. The
nylon-knitted net was found to be broken easily. No soil was built up
on the steel mesh lining and the steel chain lining after 20 loading-
unloading operations. The steel mesh lining and the steel chain lining
would increase the deadweight of the self-unloading box, which is their
disadvantage.
Wear characteristics of some of the potential
materials used for bionics modification
of soil-engaging components with of agricultural
and earth moving machinery
(a) Mechanics of damage as result of wear
The failure of various components of soil-engaging components or
structures can occur due to damage by plastic deformation, formation
and propagation of cracks, corrosion or wear. Elastic and particularly
plastic deformation of a component may result in it being unable to
support the forces, perform the intended function or avoid interference
Damage
Corrosion Wear
Fig. 13.28. Mechanisms resulting in damage of components or structure.
with other operating components. Figure 13.28 explains the mech-

anism of damage. Wear mechanism of soil-engaging components of
various machines will be described briefly in this chapter.
In abrasive wear, “material is removed or displaced from a surface
by hard particles, or sometimes by hard protuberances on a counter
face, forced against and sliding along the surface.” Examples of two-
body and three-body abrasion are shown in Fig. 13.29. In erosion, wear
is caused by hard particles striking the surface, either carried by a gas
stream or entrained in a flowing liquid.
In two-body abrasion, wear is caused by hard protuberances on
the counter face (e.g., as in the wear of drill bits cutting rock), while in
three-body abrasion the hard particles are free to roll and slide between
the two surfaces.
Figure 13.30 is relevant for when abrasive particles are hard
compared to the material being abraded, i.e. the material being abraded
is less hard than the abradent. However, materials with the same
hardness can have widely different values of abrasive wear resistance.
Thus, in the diagram below, for the same value of hardness, ceramics
have a lower resistance to abrasion than martensitic steels. Austenitic
steels have higher resistance to abrasion than martensitic steels.
Models of abrasion recognize that abrasive wear can arise either
from (i) plastic deformation forming a groove in a material or (ii)
abrasive wear by brittle fracture. In abrasive wear by brittle fracture,
lateral cracks formed beneath a plastic groove produce chips which are
subsequently removed from the surface by the abrasive process.
The relative potential has some effect on soil adhesion on the
surfaces of materials. For example, the soil adhesion to steel-35 with a
microscopic structure of a lower relative surface potential was lower.10
Fig. 13.29. Schematic diagram of wear mechanism.
(b) Physical properties affecting wear

Flow pressure or hardness, fracture toughness, work hardening, duc-
tility, strain distribution, mechanical instability and crystal anisotrophy.
Figure 13.31 shows the relationship between fracture toughness and
wear resistance.
Under severe abrasion conditions, metals which are tougher but
less hard, such as tool steels, tend to have good abrasive wear resistance
and suffer abrasive wear by plastic deformation in contrast to cast iron,
which is less tough, but harder, and suffers abrasive wear by brittle
fracture.
Fig. 13.30. Abrasive wear resistance of different materials in the pin abrasion test as
a function of bulk hardness [83]. b.c.c. = body centered cubic; f.c.c. = face centered
cubic.
Fig. 13.31. Relationship between material toughness and wear resistance of various
materials under abrasive conditions [84].
Soil is extremely weak in tension, very strong in compression

and in practice mainly fails in shear. When soil is strained the shear
stress builds up to a peak value which in certain soils (loose dry soils)
remains constant with increasing strain and in other soils, falls off before
levelling out to a small constant value, residual stress. The wear and tear
of soil-engaging components of tillage and earth-moving machinery
during field operations leads to increased power requirements, frequent
downtime for tool replacement, loss of production and increase in the
equipment and operational costs. The quality of the work and the over-
all performance of the machinery and equipment are adversely affected
as a result of excessive wear and tear of soil-engaging components.
The wear and tear of soil-engaging components are affected by a wide
range of parameters such as the soil moisture content, soil particle
sizes, operation speed, depth of operation, tool material properties and
their geometries. Therefore, it is important to understand the primary
mechanisms behind soil-engaging component wear and develop to
effective methods to mitigate wear rates. Figure 13.32 shows abrasive
soil conditions for seeders, and Fig. 13.33 shows the worn-out soil-
engaging component of a tillage tool requiring replacement.
The wear of soil-engaging components of tillage and earth-moving
machinery is treated as a two-body abrasion where a rigid particle moves
against the surface of the soil-engaging components at a prescribed
speed. A finite element model is developed to model the scratching
process. The deformations predicted by the finite element models are
used in conjunction with classical plowing theory and the material
removal factor to identify the mechanisms underlying the wear of the
soil-engaging components. The predicted material removal factors are
Fig. 13.32. Tillage equipment operating under abrasive conditions.

Fig. 13.33. Worn-out soil-engaging components of sowing equipment require

expensive and time-consuming replacement.
compared with experimental results and a parametric study is carried

out to study the effect of various parameters on tool wear.
A major portion of the energy and wear losses in soil-engaging
components of various machinery and equipment can be attributed to
the movement of these components in soil. The primarily process by
which these components experience wear and tear evidently include
impact, abrasion, fretting, other physical and chemical actions. The
wear is one of the main parameters that determines the useful life of
soil-engaging components. Moreover, the efficiency and work capacity
of these components and the machinery are also determined by the soil
response to the tools. The focus of research has primarily been on the
industrial materials with little emphasis on soil-engaging components of
agricultural machinery. An NRCC85 report estimated that total annual
losses in the agricultural sector due to friction are $320 million and
those due to wear are approximately $940 million. Although these are
very old figures, a significant amount can be saved annually through
improving wear resistance and friction of soil-engaging components
of agricultural machinery. Therefore, it is important to produce these
components from material with better scouring properties and wear
resistance.
A range of materials have been used to modify the surfaces of
soil-engaging components of various machinery and equipment to
Table 13.4. Wear Resistance of Some Commonly Used Materials Applied to

Production and Modification of Soil-Engaging Components with Agricultural
and Earth-moving Machinery [7]
Relative
wear resistance
Materials by weight, Ew Av Ew Av Ev
E: Polytetra Fluoro Polyethylene (PTFE) 42 49 53 48.0 12.9

F: Ultra-High Molecular Weight 476 398 339 404.3 48.5
Polyethylene (PTFE)
D: PES-PTFE 60 83 71.5
K: Enamel 97 100 90 97.7 30.7
Z: Al2 O3 123 127 121 123.7
J: Ion Base Alloy 53 43 68 54.7
L1 : Nickel Alloy Coating 55 69 93 72.3
L2 :FeWCrNi Coating 70 65 62 65.7
A: Quenching and Tempering of Steel-35 81 93 96 90.0
at Low Temperature
H: Silconized Steel-45 122 136 113 123.0
I: Steel-45 Hardened by Laser 110 127 120 119.0
S: Normalized Steel-45 63 43 79 61.7
Av Ev = Av Ew x Specific weight of material used/Specific weight of steel (7.8)

Ew = Relative wear resistance by weight = Loss of weight of the standard specimen/
loss of weight of test specimen
Ev = Loss of volume of the standard specimen/loss of volume of test specimen
improve soil scouring and reduce friction. Some of these materials have
poor wear resistance and practically cannot be used in abrasive soil
conditions. For example, Poly Tetra Fluoro Ethylene (PTFE) coating
of 0.50–0.80 cm thickness lasts for plowing of 20 ha and high-density
polyethylene covering of 0.50–0.80 cm lasts for about 8 ha in abrasive
soils.30 The higher costs and lower life are the main limiting factors in
using these materials practically in the field. The abrasive resistance of
various materials was investigated in the laboratory.7 The results of his
research studies are presented in detail in Table 13.4. Figure 13.34
shows the test apparatus used to measure the wear characteristics
of various materials. Figure 13.35 shows the abrasive sand particles
with sharp angles and edges used in testing various materials for wear
resistance. Figure 13.30 explains the working principle of the apparatus
to a large extent. The apparatus consists of a rotating wheel with four
4,5,6,7 = Test specimens 8 = Sub soiler
Fig. 13.34. Schematic diagram of the apparatus used for wear test (JMM Abrasion
Tester).
Fig. 13.35. Abrasive sand particles showing sharp edges and angles used to test the
wear resistance of various materials for modifications of the surfaces of soil-engaging
components to improve their soil scouring properties [7].
mounted test specimens of the materials with one always as a standard

specimen, three compaction wheels and a chisel.
(c) Wear resistance of PTFE and UHMW-PE

The wear mechanism of untreated material is mainly adhesion wear,
abrasive wear, and oxidation wear. The bio-inspired wear is due mainly
to adhesion wear and abrasive wear. Worn surfaces of PTFE and
UHMW specimen are shown in Fig. 13.36.
Fig. 13.36. Morphology of worn surfaces of PTFE and UHMW-PE specimens

(PTFE = a to c; UHMW-PE = d to f) (after Qaisrani, 1993).
(d) Wear resistance of polymers

Polymers coating possesses poor abrasive characteristics. The Poly
ether sulfone-Polytera fluoro ethylene (PES-PTFE) coating of 50 µ
was completely worn out within 13 minutes.7 It has to be improved
before it can be recommended for coating the surfaces of soil-engaging
components of earth-moving and tillage machinery. Some polymers can
be used under sliding conditions. However, some researchers some of
the disagree with earlier findings on the basis of these studies primarily
focusing on the pin-on-disc abrasion with the abrasive material used
being either abrasive paper or cloth. The hardness of abrasive material
has significant effects on the wear resistance of material tested. This does
not represent the abrasive field conditions soil-engaging components
are subjected to. For example, the wear characteristics of various
polymers such as Polyethylene (PE), Polytera fluoro ethylene (PTFE),
Polystyrene (PS), Polypropylene (PP) and Polyamide (PA). Their
wear rates are inversely proportional to their fracture stress time total
strains.26
(e) Wear resistance of enamel coating

The soil adhesion characteristics of enamel coating was thoroughly
examined and found that enamel coating has the ability to reduce
soil adhesion. Rolling resistance of cage wheels and the plow boat
can be reduced owing to the enamel coating on the wheel lugs and
on the boat surfaces, and their tractive performance was improved.5,6
However, spalling rupture of the enamel coating easily took place under
the abrasive wear conditions.31 The wear resistance of enamel coating
was evaluated.86 Vitreous enamel coatings are primarily applied to
protect metallic substrates from corrosion and improve the scouring
characteristics of coated surfaces. The durability of the coating is
related to the resistance of the surface to mechanical damage such as
wear which can impair the characteristics of the enamel layer. Low-
carbon steel panels were coated with three distinct enamels applied
by two different methods: wet dip coating and electrostatic powder
coating. The abrasion resistance of enamel coating may be increased
by modifying the rare materials for the coating and by improving the
preparation techniques. For details, refer to Fig. 13.37.
(f ) Plasma spray coating

The ceramic coating possessed excellent abrasive resistance against
soil.87 Worn surface morphology of ceramic coating investigated by
others7 is shown in Fig. 13.38. The wear on the surface is primarily
from the separation of Al2 O3 particles from each other because of their
poor interfacial strength.
Fig. 13.37. Worn surface morphology of enamel coating (after Qaisrani, 1993).
The worn surface morphology of ion-base coating has many

grooves on the surface. It has very poor wear resistance under highly
abrasive conditions. Its use is not recommended for coating the
surfaces of ground-engaging components of earth-moving and tillage
machinery.7
(g) Heat treatment by laser

Modification of steel by laser could improve its strength and durability
by improving its wear resistance. Worn surfaces of heat-treated steel
samples tested in the laboratory showed the width and depth of cuts
were smaller than in the control sample. Worn surfaces of heat-treated
steel contained some micro concaves similar to size and shape to the
abrasive soil particles used in the experiments.26 Moreover, the concaves
formed on the worn surface of conventional steel samples were deeper
and larger than those on the laser-treated specimens (Fig. 13.39).
(h) Hard facing

When a metallic component is pulled through the soil, the metal will
eventually wear out over time. How fast that occurs depends on the pull
speed, the angle of movement, soil texture and the presence of rocks
or sand or other abrasive materials. As operational speeds increase, one
Fig. 13.38. Morphology of worn surfaces of (a) ceramic coating, (b) ion-base alloy
coating, (c, d) FeCrNi coating, and (f) nickel alloy coating [7].
can expect more wear on soil-engaging components even if all the other
soil and material parameters are unchanged.
To date, ground-engaging tools provided by most manufacturers
share a fairly similar metallurgy, one that gives the best long-term
performance at a price the user will pay. In cases where excessive wear
may be the norm, hard facing has provided a solution for soil-engaging
components. That trend may be changing, however, as a number of the
tillage designers we spoke with say “custom alloys” (based on regional
needs) for points and shovels may begin to gain favor as growers
demand less downtime and more productivity. Chrome-boron alloys
Fig. 13.39. Morphology of worn surfaces of (a, b) conventional steel and (c, d)
silconized by laser, (e, f) remelt solidified by lazer [7].
are already popular as standard equipment on many tools, such as coul-

ters and subsoiler points. In addition, users may become more aware of
the actual wear on their tillage equipment as some engineers and prod-
uct developers foresee systems that will provide real-time wear readouts
from the implement to the tractor cab. Tool wear is becoming more of a
problem, particularly as operational field speeds increase. Figure 13.40
shows improvement in wear characteristics of hard-faced chisel points.
The hard-faced and conventional (un-treated) samples are tested
in the actual field environment. All the samples are degraded under
the effect of abrasive wear in the actual field environment. The field
Fig. 13.40. These subsoiler points after plowing the same number of hectares show
the wear advantages of hard-facing treatments (right) vs untreated point (left).
Table 13.5. Results of Hard-Facing Treatment of Various Excavators

in the Field Environment [88]
Initial Final Loss of Weight

Sample Weight, g Weight, g Weight, g loss, %
OK 84.78 (Green) 2.650 2.306 0.344 12.98

OK 83.50 (Red) 2.600 1.910 0.690 26.54
OK 83.65 (Yellow) 2.650 1.732 0.918 34.64
OK 84.80 (White) 2.500 2.113 0.387 15.48
Conventional (Untreated) 2.350 1.435 0.195 38.94
results showing the percentage of weight loss of hard-faced and non-

hard-faced samples are as shown in Table 13.5.
Hard-faced samples with all the four types of hard-facing alloys (OK
84.78, OK 83.50, OK 83.65 and OK 84.80) demonstrated better wear
resistance on the Low Manganese steel substrate. The wear resistance
of the Low Manganese steel substrate can be increased by up to three
times by hard facing with OK 84.78. It is evident from the results
(Table 13.6) that as the percentage composition of Chromium (Cr)
increases, the abrasive wear resistance improves.89 SEM techniques
may be useful in analyzing the wear mechanism and characteristics
of various hard-facing techniques. Figure 13.40 shows the wear
resistance of QLT Steel, siliconized by laser and remelt solidified by
laser.
Table 13.6. Composition of Hard-Facing Alloys [88]
Hard Welding
facing technique
alloys Percentage Composition used
OK 83.50 C — 0.4, Si — 0.6, Mn — 1.0, Cr — 6.0, Mo — 1.3, SMAW

Nb — 0.6
OK 84.78 C — 4.5, Si — 0.8, Mn — 1.6, Cr — 33 SMAW
OK 84.80 C — 5, Si — 2, Cr — 23, Mn — 0.7 SMAW
OK 83.65 C — 0.75, Si — 0.4, Mn — 0.6, Cr — 2, P — 0.03, SMAW
S — 0.03
Fig. 13.41. Worn surface morphologies of (a, b) QLT Steel 45; (c, d) silconized by
laser and (e, f) remelt solidified by laser [7].
Fig. 13.42. Graphical representation of wear resistance of various engineering mate-

rials [7]. E: PTFE; F: UHMW-PE; D: Polyethersulphone/polytetrafluoroethylene
(PES-PTFE); K: Enamel; Z: AL2 O3 ; J: Ion base alloy; L1: Nickle alloy coating; L2:
FeWCrNi coating; A: Quenching + Tempering of steel-35 at low temperature; H:
Siconized Steel-45; I: Steel-45 hardened by laser; S: Normalized Steel-45.
Figure 13.42 shows the comparative wear resistance of various

materials under identical conditions.
Table 13.6 shows the composition of hard-facing alloys used in
modification of components conventional buckets.
Conclusion
This chapter reviewed some of the methods employed to improve soil

scouring properties of soil-engaging components of different machines
and tools. Each of these techniques has its advantages and disadvan-
tages. However, an ideal technique employed to reduce adhesion of
soil to the surface of soil-engaging components of various machinery
should be simple, ease to manufacture and use, low cost, synchronized
with other components of the machines with no extra controls and
power requirements, and effective and efficient with scouring rate 90%
or higher. No one technique has universal application. Some techniques
are better for one system and others are useful for another. The soil
adhesion preventing mechanism of soil animal cuticles can be employed
in modifying the surfaces of soil-engaging components of a number

of machines and equipment. The benefits of research work could be
achieved by applying these techniques practically on the surfaces of
machines and equipment. The material used for modifications should
have better wear resistance for making this technique practical. Some
techniques such as enamel coating are simple and cheap, but their use
cannot be recommended under abrasive soil conditions because of poor
wear characteristics.
The research results of a number of studies showed that the drafts
of moldboard plows and bulldozing plates modified by bionics were
reduced by improving the soil scouring properties of these implements.
However, there have been some inconsistencies in the results mainly
because these trials were conducted under varying conditions. It is
very difficult to repeat these experiments under similar conditions.
Moreover, the operation of a number of soil-engaging components
of agricultural and industrial machinery is different from that of soil
animals. Therefore, for improving the efficiency of these tools, each
implement has to be designed differently based on its operating
environment. However, the soil adhesion preventing mechanism of
soil animals can successfully be used as a guideline in reducing
adhesion and scouring soil from the surfaces of various soil-engaging
components of agricultural and industrial machinery. In some situation
a combination of these techniques may be used to achieve the desired
outcome.
References
[1] Qian DH., Zhang JX. (1984) Research and Adhesion between Soil and Solid
Materials. Arbeiten Landwirtschafiliche Hochschule Iloheim 18: 39–125 (NIAE
Transactions No. 165).
[2] Tong J., Ren LQ., Chen BC. (1990a) Reducing adhesion of soil by phosphorus
white iron on the basis of bionics principles. Proceedings of the 10th International
Conference of the ISTVS, Kobe, Japan, pp. 57–63.
[3] Zhang J., Sang Z., Gao L. (1986) Adhesion and friction between soils and solids.
Transactions of the Chinese Society of Agricultural Machinery 17: 32–40.
[4] Salokhe VK., Gee-Clough D. (1987) Studies of effects of lug surfaces coating on
soil adhesion of cage wheel lugs. Proceedings of the 9th International Conference
of the ISTVS, Barcelona, Spain, pp. 389–396.
[5] Salokhe VK., Gee-Clough D. (1988) Cage wheel blocking in wet clay soil.
Journal of Agricultural Engineering Research 39: 277–286.
[6] Salokhe VM., Gee-Clough D. (1989) Applications of enamel coating in
agriculture. Journal of Terramechanics 26: 275–286.
[7] Qaisrani AR. (1993) The effects of modified and unsmoothed surfaces on the
draft of bulldozing plates and mouldboard Plow. A PhD Dissertation No. 0067,
Jilin University of Technology, Changchun, People’s Republic of China.
[8] Qaisrani R., Li JQ., Khan MA., Rashid I. (2010) Soil Adhesion Preventing
Mechanism of Bionic Bulldozing Plates and Mouldboard Ploughs. Advances in
Natural Science 3: 100–107.
[9] Ren LQ,. Chen DX., Hu J. (1990a) D-Optimum experiment research on the
bionic bulldozing plates. Proceedings of the 10th International Conference of the
IST VS, Kobe, Japan, pp. 1015–1023.
[10] Ren LQ., Tong J., Li JQ., Chen DX., Hu J. (2001) Soil Adhesion and
Biomimetics of Soil-engaging Components: a Review. Journal of Agricultural
Engineering Research 79: 239–263.
[11] Chen BC., Ren LQ., Chen DX., Tong J., Cong Q. (1990) Theory and
Techniques of Reducing adhesion and scouring soil for terrain machines by
bionics. A technical report. Jilin University of Technology, Changchun, People’s
Republic of China.
[12] Tabor D. (1981) Friction-the present state of our understanding. Journal of
Lubrication Technology 103: 169–179.
[13] Tong J., Ren LQ., Yan J., Ma Y., Chen BC. (1999) Adhesion and abrasion of
several materials against soil. International Agricultural Engineering Journal 8:
1–22.
[14] Cong Q., Wu L., Ren LQ., Chen BC. (1995a) The principled experiment of
reducing soil adhesion and scouring soil by non-smooth surface electro-osmosis.
Transactions of the Chinese Society of Agricultural Engineering 11: 19–23.
[15] Ren L., Cong Q., Wu L., Chen D. (1995a) Study on the reduction of soil
adhesion and resistance by the application of nonsmooth surface electro-osmosis.
Transactions of the Chinese Society of Agricultural Engineering 11: 24–28.
[16] Jia X. (2004) Theoretical analysis of the adhesion force of soil to solid materials.
Biosystem Engineering 87: 489–493.
[17] Sharma V.K, Drew LO., Nelson GL. (1977) High frequency vibrational effects
on soil-metal friction. Transactions of the ASAE 20: 46–51.
[18] Wu S (1982) Polymer Interface and Adhesion. Marcel Dekker, Inc., New York.
[19] McFarlane JS., Tabor D. (1950) Relationship between Friction and Adhesion.
Proceedings of the Royal Society London 202: 224–243.
[20] Tong J., Ren L., Chen B., Yan B. (1994a) Fractal dimensions of soil particle-size
24 distributions and their effects on soil adhesion behavior. Transactions of the
Chinese Society of Agricultural Engineering 10: 27–33.
[21] Tong J., Ren L., Chen B., Qaisrani AR. (1994b) Characteristics of adhesion
between soil and solid surfaces. Journal of Terramechanics 32: 93–105.
[22] Oswal MC. (1994) Soil physics. Oxford & IBH Publishing Co. Pvt. Ltd. New
Delhi, India.
[23] Hua QD. (1965) Study on adhesive forces of conventional steel in heavy clay
soil (in Chinese). Journal of Agricultural Soil Mechanics 5.
[24] Zhang JG. (1985) Studies on the adhesion and friction of soil to solid materials.
PhD Thesis, Jiangsu Technical College, China.
[25] Nichols ML. (1931) The dynamic properties of soil II, Soil and metal friction.
Journal of Agricultural Engineering 12: 321–324.
[26] Tong J. (1992) Study on reducing adhesion and resistance of soil to soil engaging
components of machinery for land locomotion by bionics. A Ph.D. Dissertation,
Jilin University of Technology, Changchun, China (in Chinese).
[27] Cong Q., Ren L., Chen B., Su F. (1990a) A study on adhesion reducing
methods of terrain-machine. Transactions of the Chinese Society of Agricultural
[28] Zhu H., Tan L; Wu S (1992). An experimental study on the ‘comet-type
passage-holes’ plough of reducing resistance. Transactions of the Chinese Society
of Agricultural Machinery 23: 20–24.
[29] Tong J, Ren L., Chen Y., Qaisrani AR., Chen B. (1994c). Abrasive properties
and mechanism of polytetra fluoroethylene and ultra high molecular weight
polyethylene. Tribology 14: 65–72.
[30] Salokhe VM., Cheunpakaranant W., Niyampa T. (1990) Effects of enamel
coating on the performance of tractor drawn rotavator. Journal of Terramechanic
36: 127–138.
[31] Tong J., Zhang M., Jiang M., Ren L .(1998). Free-abrasive wear of an enamel
coating. Journal of Materials Science Letters 17: 523–525.
[32] Tong J., Ren L., Chen B. (1990) Reducing adhesion of soil by phosphorus
white iron on the basis of bionics principles. Proceedings of the 10th International
Conference of ISTVS, 20–24 August, pp 57–63, Kobe, Japan.
[33] Chen B., Li J., Ren L. (1995a) Research on scouring soil for materials of the
traditional plow moldboard. Transactions of the Chinese Society Agricultural
Machinery 26: 46–49.
[34] Araya K., Kawanishi K. (1984) Soil failure by introduction air under pressure.
Transactions of the ASAE 27: 1292–1297.
[35] Cong Q., Ren L., Chen B. (1990b) Research on reducing adhesion results of
soil electro-osmosis and its affecting factors. Proceedings of the 10th International
Conference of the ISTVS, 20–24 August, pp 45–55, Kobe, Japan.
[36] Nichols ML. (1931) The dynamic properties of soil II, Soil and metal friction.
Journal of Agricultural Engineering 12: 321–324.
[37] Azadegan B., Massah J. (2012) Effect of temperature on adhesion of clay soil
to steel. Cercetări Agronomice în Moldova XLV: 21–27.
[38] 16. Wang XL., Ito N., Kito K., Garcia PP. (1998) Study on use of vibration to
reduce adhesion. Journal of Terramechanics 35: 87–101.
[39] Wang XL, Ito N, Wang X L, Ichikawa M, Tajiri I, Noro A (1993) Study on
prevention of soil adhesion to rotary cover by vibration (Part I). Journal of the
Japanese Society of Agricultural Machinery 55: 41–46.
[40] Wang X L, Ito N, Kito K (1996) Study on reducing soil adhesion to machines
by vibration. Proceedings of the 12th International Conference of ISTVS, 7–10
October, 1996, pp 539–545. China Machine Press, Beijing, China.
[41] Ren L, Zhang J, Chen B, Qi B (1990b) Bionics research on reducing adhesion
and resistance of soil for flexible shovel. Transactions of the Chinese Society of
Agricultural Machinery 21: 33–39.
[42] Yin Y., Ren L., Cheng Y., Chen B. (1990) A study on cleaning buildup soil in
loader shovel on the principle of bionics. Transactions of the Chinese Society of
Agricultural Engineering 6: 1–6.
[43] Xin J (1986) Introduction to Soil Animals. Science Press, Beijing, China.
[44] Tong J., Ren L., Chen B. (1995) Chemical constitution and abrasive wear
behavior of pangolin scales. Journal of Materials Science Letters 14: 1468–1470.
[45] Chirende B., Li JQ., Wen LG., Simalenga TE. (2010) Effects of bionic non-
smooth surface on reducing soil resistance to disc ploughing. Science China
Technological Sciences 53: 2960–2965.
[46] Han G, Zhang Y (1991) Measurement and analysis of ploughing resistance
on the magnetic plough. Transactions of the Chinese Society Agricultural
Machinery 22: 25–28.
[47] Guo P, Liu J (1995) The effect of magnetized plowshare upon tillage perfor-
mance. Agricultural Research in the Arid Areas (China) 13: 103–109.
[48] Zhang Y., Han G. (1992) Measurement and analysis of work resistance on the
magnetic coverer. Journal of Shenyang Agricultural University 23: 25–28.
[49] Chancellor WJ. (1994) Friction between soil and equipment materials review.
American Society of Agricultural Engineer Paper No. 94–1034.
[50] Mandelbrot BB. (1982) The Fractal Geometry of Nature. W. H. Freeman, New
York.
[51] Perfect E., Kay BD. (1995) Applications of fractals in soil and tillage research: a
review. Soil & Tillage Research 36: 1–20.
[52] Yan B., Ren L., Tong J. (1996) Effects of soil particle size distribution on
friction of soil-solid surfaces. In: Proceedings of the 12th International Conference
of ISTVS, 7–10 October (Yu Q, Qiu L, eds), pp 498–505. China Machine Press,
Beijing, China.
[53] Li J., Ren L., Chen B., Gu W. (1993a) A study on the soil adhesion characteristics
of 35 steel. Transactions of the Chinese Society of Agricultural Engineering 9:
19–25.
[54] Li J., Ren L., Chen B., Gu W. (1993b) The effect of microstructure of steel on
soil adhesion characteristics. Transactions of the Chinese Society of Agricultural
[55] Li J., Ren L., Chen B., Jiang M. (1996a) Study on surface characteristics of
materials of plough moldboard and their effects on soil adhesion. Transactions
of the Chinese Society of Agricultural Engineering 12: 45–48.
[56] Shen Z. (1982) Mechanism of soil adhesion and its applications. Proceedings of
the First Conference of Chinese Society for Terrain Machine Systems, pp 121–123,
Wuxi, China.
[57] Zhuang J., Yi Y., Wu L., Liu X. (1992) Effect of magnetic energy on soil
electrochemical properties. Journal of Shenyang Agricultural University 23:
27–31.
[58] Li J, Ren L, Liu C, Chen B (1996b) A study on bionic plow moldboard of
reducing soil adhesion and plowing resistance. Transactions of the Chinese
Society of Agricultural Machinery 27: 1–4.
[59] Qaisrani AR., Chen B., Ren L. (1992) Modified and unsmoothed plow surfaces
- a means to reduce plowing resistance. International Agricultural Engineering
Journal 1: 115–124.
[60] Ren L., Tong J., Zhang S., Chen B. (1995b) Reducing sliding resistance
of soil against bulldozing plates by unsmoothed bionics surfaces. Journal of
Terramechanics 32: 303–309.
[61] Li J., Ren L., Chen B. (1995) The statistics analysis and mathematical analogy
of the unsmoothed geometrical units on the surface of soil animal. Transactions
[62] Cong Q., Wang L., Ren L., Chen B., Zhou S., Li A. (1996) Biomimetic bull-
dozing plates with wave-form non-smooth surfaces for reducing soil adhesion
and resistance. Construction Machinery (China) 3: 28–30.
[63] Cong Q., Ren L., Wang G., Chen B. (1995b) Study on the reduction of soil
adhesion and resistance about bionic nonsmooth surface by test. Proceedings
of the International Conference of Agricultural Mechanization, 10–13 April, pp
33–39, Beijing, People’s Republic of China.
[64] Suminitrado D., Koike M., Konoba T., Yuzawa S., Kuroishi I. (1988) A
mathematical model to predict the trajectory of soil motion on a moldboard plow
surface. Proceedings of the Second Asia-Pacific Conference of ISTVS, Bangkok,
Thailand, pp 195–204.
[65] Li J., Jiang M., Li Y. (1997) The relationship between soil adhesion and surface
potential of steel. Transaction of the Chinese Society of Agricultural Engineering
13: 42–45.
[66] Li J., Ren L., Chen B., Jiang M. (1996b) Study on surface characteristics of
materials of plough moldboard and their effects on soil adhesion. Transactions
[67] Li J. (1993) A study on the bionic plow moldboard of reducing soil adhesion
and plowing resistance. PhD Thesis, Jilin University of Technology, Changchun,
China, pp 77–93.
[68] Chen Y., Tong J., Ren L., Chen B. (1994) Abrasive wear behaviour of several soil
adhesion-reducing materials. Transactions of the Chinese Society of Agricultural
[69] Tong J., Lu X., Chen Y., Ren L., Chen B. (1990b) Soil adhesion and
abrasive wear of PTFE-matrix composites. Transactions of the Chinese Society
of Agricultural Engineering 6: 13–19.
[70] Lu X., Wen S., Tong J., Chen Y., Ren L. (1996) Wettability, soil adhesion,
abrasion and friction wear of PTFE (+PPS) + Al2O3 composites. Wear 193:
48–55.
[71] Liu C., Ren L., Tong J. (1998a) Wettability and soil adhesion of glass bead
reinforced UHMWPE matrix composites. Acta Materiae Compositae Sinica 15:
29–33.
[72] Liu C., Ren L., Tong . (1998b) Behavior of soil adhesion on soil-engaging
components. Transactions of the Chinese Society of Agricultural Engineering 14:
37–42.
[73] Jia X., Ren L., Tong J., Chen B. (1993) Wettability of chemical unsmoothed
surfaces and rough surfaces. Transactions of the Chinese Society of Agricultural
[74] Jia X., Ren L., Chen B. (1995) Characteristics of reducing adhesion and
resistance of bionic coatings prepared for touching-soil components of terrain
machinery. Transactions of the Chinese Society of Agricultural Engineering 11:
10–13.
[75] Jia X., Ren L., Chen B. (1996a) Wettability of unsmoothed surfaces of soil
animals’ cuticles and bionic composite coatings. Chinese Journal of Materials
Research 10: 556–560.
[76] Jia X., Ren L., Chen B. (1996b) Characteristics of bionic composite coating
reducing adhesion and resistance. Chinese Journal of Materials Research 10:
210–214.
[77] Jia X., Ren L., Chen B. (1996c) The functions of reducing adhesion and
resistance of PA1010, EP composite coatings and surface improved moldboards.
Transactions of the Chinese Society of Agricultural Machinery 27: 7–11.
[78] Chen B., Liu D., Ning S., Cong Q .(1995b) Research on the reducing adhesion
and scouring of soil of lugs by using unsmoothed surface electro-osmosis
method. Transactions of the Chinese Society of Agricultural Engineering 11:
29–33.
[79] Sun S, Ren L, Tong J, Chen P (1992) A shovel bucket with steel-cloth pocket by
bionics. In: New Technology and Application of Terrain-Vehicles and Machinery,
Proceedings of the Third Asia-Pacific Conference of ISTVS, 10–13 August,
Changchun, China; pp 231–235.
[80] Sun S, Ren L, Tong J (1996a) Design and applications of flexible lining for anti-
adhesion. Transactions of the Chinese Society of Agricultural Engineering 12:
65–70.
[81] Sun S, Ren L, Wang Y, Chen D (1996b) Design of anti-adhesion flexible lining
steel rings. Journal of Jilin University of Technology 26: 20–24.
[82] Wang Y., Ren L., Li J., Sun S., Zhou R. (1997) A model test study on flexible
bionics technology of anti-adhesion to dump truck. Transactions of the Chinese
Society of Agricultural Engineering 13: 139–143.
[83] Zum Gahr KH. (1987) Microstructure and Wear Materials. Elsevier Science
Publishers B.V. Amsterdam, the Netherlands.
[84] Diesburg DE., Borik F. (1974), Optimizing abrasion resistance and toughness
in steels and irons for the mining industry. Proceedings of the Symposium of the
Materials for the Mining Industry, Climax Molybdenum Co., 15–41.
[85] NRC (1986) A strategy for Tribology in Canada. National Research Council,
Ottawa, Canada, pp. 30–39.
[86] Rossi S., Scrinzi E. (2013) Evaluation of the resistance of enamel coatings.
Chemical Engineering and Processing: Process Intensification 68: 74–80.
[87] Foley AG., Chisholm CJ., McLees VA. (1988) Wear of Ceramic-Protected
Agricultural Subsoilers. Tribology International 21: 97–103.
[88] Singhla S., Shibe V., Grewal JS. (2011) Performance evaluation of hard
faced excavator bucket teeth against abrasive wear using MMAW process.
International Journal of Mechanical Engineering Application Research 2: 73–78.
[89] Tyler SW, Wheatcraft SW (1992) Fractal scaling of soil particle-size distributions:
analysis and limitations. Soil Science Society of America Journal 56: 362–369.
Chapter 14
Application of Bionic Technologies

for Soil-Engaging Tillage Components
in Northeast China
Ji-yu Sun,∗,‡ Zhi-jun Zhang,† Jin Tong,∗

Hong-lei Jia∗
∗ Key Laboratory of Bionic Engineering (Ministry of Education),
Jilin University, Changchun, 130022, China
† Institute of Mechanical Science and Engineering, Jilin
University, Changchun 130022, China

‡ sjy@jlu.edu.cn
The reduction of water resources and soil fertility in Northeast China

requires increasing the working efficiency of agricultural soil-engaging
components. Adhesion and resistance are the main problems for
soil-engaging tillage components. However, soil-burrowing or soil-
digging animals give inspiration for resolving those problems. Their
fair, claws, toes, textured surfaces and scales have anti-adhesion or
resistance reduction functions. Those results provide a way to realize
the sustainable development of modern agriculture by developing
novel bionic agricultural machinery systems with independent intel-
lectual property rights to meet conservation tillage requirements in
the Northern China region. This chapter reviews biological structures
of some soil-burrowing or soil-digging animals, such as the beetle,
mole cricket, earthworm, mole, vole, pangolin and snake, then presents
‡ Corresponding author.
555
their mechanism of anti-adhesion or resistance reduction. Bioinspired

applications which have been used in Northeast China are also
presented in the chapter, including moldboard, subsoiler components,
furrow opener, roller, and bionic rototilling-stubble-breaking blade.
Subsequently, the existed problems in agricultural engineering and the
future development trend are discussed.
14.1. Introduction
In China, cultivated land resources and agricultural water resources are
scarce, with cultivated land area decreasing at a mean annual rate of
0.3%. Moreover, soil fertility is also declining. Recently, the organic
matter content in the black soil of Northeast China has decreased
from 3.5–5% at the start of modern agriculture use to about 2%,
and this rate is continuing to accelerate. This decline in soil fertility
has become a bottleneck constraining the sustainable development of
agriculture in China.1 Many existing agricultural machines in Northeast
China are single-task machines, with high-energy consumption and low
efficiency that have not been adapted to the requirements of modern
agriculture. Thus, there is a need to develop as well as widely popularize
through application demonstration intelligent bionic systems with high
efficiency, low energy consumption, good soil moisture storage and
preservation capability, and good soil fertility improvement. This has
become an important topic for ensuring high and stable yields of grain
and realizing the sustainable development of modern agriculture.2−4
High energy consumption, poor economic efficiency, serious soil
moisture loss, and other negative equipment factors severely constrain
the popularization of soil moisture storage and preservation tillage
systems.5−7 The results of field experiments in Northeast China show
that seedling emergence rate had decreased by 5%–12% due to soil
adhesion and serious soil moisture loss during various operation
processes such as furrowing, covering and rolling, which thus greatly
reduces grain yield. Therefore, it is important to develop high-
efficiency and energy-saving bionic tillage implements that meet the
requirements of modern agriculture in order to reduce tillage resistance
and soil adhesion. Soil adhesion causes increased working resistance
during operation of soil-engaging components, which increases energy
Application of Bionic Technologies for Soil-Engaging Tillage Components 557
consumption, decreases operational quality, causes significant soil

moisture loss, and can even cause mechanical failure. For example,
during the operation of a traditional agricultural tillage component
such as a furrow opener and tillage blade, the energy consumption
for overcoming the slide resistance between the soil and the tillage
surface accounts for 30%–50% of total energy consumption due to
soil adhesion and friction. If there is serious soil adhesion occurring
on the surfaces of the furrow opener, roller and other soil-engaging
implements, traction resistance increases, soil moisture preservation
becomes more difficult, and mechanical failures occur more frequently,
which all affects seedling emergence, growth, and grain yield. It was
pointed out that the crop yield per unit area has decreased by an
average of 3% per year in Northeast China owing to poorly timed post-
sowing rolling or poor rolling quality. In addition, there are higher
equipment costs associated with larger (and thus more effective) soil-
engaging implements. A deep-loosening implement has the largest
energy consumption per unit disking width among field tillage and
seeding machines, and can operate only if towed by a high-power
tractor. The development of such implements is a significant factor in
raising efficiency, reducing cost, raising natural water utilization rate,
improving grain yield, increasing farmers’ income, and overall realizing
more efficient agriculture.1 It is therefore of necessity to assess the role
played by tillage tool design in controlling the energy consumption that
is accounted for by soil-engaging implements.8
Nichols9 presented a comprehensive discussion of soil-on-material
sliding that included a discussion of friction and adhesion. The concepts
of adhesion are well developed.10 Different from cohesion, soil adhesion
is the attraction of water molecules to solid surfaces.11 Many studies were
done to investigate the causes of soil adhesion, such as water film,12−14
soil type and soil condition,10 soil normal pressure,13 temperature15
and interfacial friction.16 Interfacial tension must also be considered as
it leads to water films in interaction with solid walls, this in addition
to the capillary meniscus, which is linked to surface properties and
the topology of porosity, respectively.17 Therefore, various approaches
have been developed to reduce soil adhesion,18 such as modifying the
surface material,19−22 lubrication,23 electro-osmosis method,24,25 and
vibration method.26 However, those technologies are difficult to apply
in mechanical equipment due to poor abrasion resistance, high cost,

complex operation, unstable technique or difficult repair.
However, in nature the phenomena of soil adhesion is almost absent
in soil-burrowing or soil-digging animals when they move through
soil. With the continued development of bionics, researchers have
noticed that soil-burrowing or soil-digging animals, such as the dung
beetle, ground beetle, mole cricket, house mouse, yellow mouse and
pangolin, have the functions of anti-adhesion or reducing resistance
through long-term evolution (Fig. 14.1, from macro-scale to meso- or
micro-scale.27−38 These excellent anti-adhesion or resistance reduction
abilities of soil animals are partly the result of their non-smooth surface
morphologies.39,40 For example, dung beetles can break up dung pads
and compact dung into balls, and their cuticle surfaces do not stick
dung or soil due to non-smoothness or roughness in micro scales on the
morphological body surface.30 In terms of bionic non-smooth surface
self-cleaning technologies,41 there have been more than 200 patents
applied for or authorized all over the world between 1998, when the
first patent was obtained in Europe, and 2004. Researchers have found
that the convex-concave shape of the flippers of Megaptera novaeangliae
(humpback whale) have a resistance reduction function — the results
of wind tunnel experiments show that a convex-concave shape flipper
can increase the motive power by 8% and decrease the resistance by 40%
compared with a smooth flipper of the same size.42 Biomimetic studies
reveal that creatures achieve excellent functional capabilities by adapting
to their living environment through long-term evolution.29,35 Some
animals continually living in a soil environment have evolved various
activity modes to adapt to different soil environmental conditions.
Meanwhile, these animals have gradually evolved appendages with
optimized geometric structures and excellent mechanical properties to
reduce cutting resistance during digging. It is that which provides the
basis for biomimetic studies on optimization of the geometric structure
and mechanical properties of soil cutting tools.27,29,30,43
But biomimetics is not simple copying from nature. From Fig. 14.1,
it can be seen that the geometric structures include ones from macro-
scale to meso- or micro-scale. For example, the convex hulls on the
surface of the dung beetle cuticle are in micro-dimensions, which if used
May 24, 2016
13:33
Application of Bionic Technologies for Soil-Engaging Tillage Components

9in x 6in
Fig. 14.1. Some soil-burrowing or soil-digging animals that have functions of anti-adhesion or resistance reduction by body length
(Tong et al., 1994; Ren et al., 2001; Ren, 2009; Tong et al., 2005; Dai et al., 2006; Liu et al., 2008; Ma et al., 2008; Chirende and
b2378-ch14
Li, 2009; Koch et al., 2009; Ji et al., 2010b; Zhang et al., 2006, 2007).
559
page 559
in the design of agricultural engineering, will have a larger error effect

than the biomimetic geometric structure effect. The result showed that
power consumption in the direction of drafting would increase when
there was salience on a cutting blade.44 So the researchers found other
ways to resolve the questions. In this paper, biological structures of
some soil-burrowing or soil-digging animals, such as the beetle, mole
cricket, earthworm, mole, vole, pangolin and snake, are reviewed. In
addition, their mechanisms of anti-adhesion or resistance reduction are
presented. Bioinspired applications which have been used in Northeast
China are also presented. Finally, existing problems in agricultural
engineering and the future development trend are discussed.
14.2. Biological Structures of Soil-Burrowing

or Soil-Digging Animals and their
Mechanisms of Anti-Adhesion or Resistance
Reduction
Bionic studies reveal that creatures achieve excellent functional capa-
bilities by adapting to their living environment through long-term
evolution.29,35 Some animals continually living in a soil environ-
ment have evolved various activity modes to adapt to different soil
environmental conditions. Meanwhile, these animals have gradually
evolved appendages with optimized geometric structures and excellent
mechanical properties to reduce cutting resistance during digging. It
is that which provides the basis for bionic studies on optimization
of the geometric structure and mechanical properties of soil cutting
tools.27,29,30,43 In the following section, we present biological structures
of soil-burrowing or soil-digging animals which have functions of anti-
adhesion or resistance reduction, from macro-scale to meso- or micro-
scale.
The shape of the clypei of Copris ochus Motschulsky decreases resis-
tance during earth cutting and burrowing.45 In addition, Copris ochus
Motschulsky can automatically reduce resistance by adjusting the wedge
angle through rotation of its head to execute efficiently different func-
tions such as bulldozing and earth cutting. By applying a bionic curved
surface as the stress-bearing surface, adjusting the angle of the compo-

nents, and switching between various soil-engaging components based
on slide cutting principles, the soil-engaging components can be func-
tionally optimized and structurally innovated to minimize resistance.45
The resistance reduction characteristics of the surface morphology
and surface wettability of the earthworm were studied based on biolog-
ical coupling theory.46 The lubrication mechanism of the earthworm
surface was analyzed according to the parameters of surface dorsal pore
and corrugation. The distribution of the pores and surface morphology
were designed, and the bionic coupling samples were prepared. The
positive pressure, lubricant flow rate and advancing velocity were
chosen as the experiment factors while the soil friction resistance was
chosen as observed object. According to the obtained data of bionic
coupling samples from the testing system, the optimal samples from
the bionic coupling reducing resistance tests were selected through the
range analysis. Compared to the normal ones, the soil resistance of
the bionic coupling samples was reduced by 76.8%. This is of great
significance and offers bright prospects for reducing energy loss in
terrain mechanics.46
The forefeet of Gryllotalpa (mole cricket) have excavating feet with
ultra-strong cave excavating capabilities. For example, the forefeet of
Gryllotalpa orientalis Burmeister have a toothed structure, with toe tips
bending towards the outside of body, so that the toes of the forefeet
can exert force towards the outside of its body during soil excavation.
This geometry is different from the excavating feet of most soil animals,
which bend inwards and exert force towards the inside or underside of
the body.47 A comparison of the driving forces of the two geometries
shows that the morphological structure and the force-exerting mode
of the forefeet of Gryllotalpa orientalis Burmeister may require a larger
driving force. However, the fact that the morphological structure of
the forefeet of Gryllotalpa orientalis Burmeister has not significantly
evolved for some time indicates that such a structure has a very high
efficiency and is worth further study.48
The mole is soil-burrowing animal that can excavate a 91
m underground passage in one night. The mole exhibits an
extremely high earth-excavating efficiency, despite its body length
of only 100–180 mm. By analyzing the geometric structure of the
Fig. 14.2. (a) The arrangement of the middle three toes of the fore claw of mole
rat (Scaptochirus moschatus); (b) The geometrical structures of the bionic specimen
(Ji et al., 2010b).36
soil-engaging surface of the toes of its forefeet, particularly the curva-

ture and second derivative of the fitted contour curve of the toe surfaces,
developed a change law for the characteristic contour lines. The first
aspect of this model is that the minimal change in the longitudinal
contour curvature of the toe means that the longitudinal surface of the
toe will be smooth to ensure that the inner wall of the excavated passage
is flat (Fig. 14.2).36,49 Second, a structure with convex longitudinal
and transverse contour curves, where the transverse contour line bends
more at the middle and the end than the tip, makes the tip of the toe
more easily able to enter the soil and allows the middle and end of
the toe to enlarge the passage once dug in. The geometric structural
features of the toe, and the characteristic contour curve changing law
of the excavating organ of a mole, can provide a technical reference for
the design of high-efficiency and energy-saving bionic soil cutting and
excavating tools.50
Spermophilus dauricus (a type of vole) exhibits real-time flexibility
and adjustability of its toes and the structures that exist between its
toes during digging. In other words, it has excellent biomechanical
self-feedback capability, which is extremely favorable for anti-adhesion
or reducing resistance during digging. For this kind of vole, the
model describing the curvature trend line of the inside and outside
contour lines of one of its toes can be roughly divided into three types
(Fig. 14.3).51,52
Fig. 14.3. Clawed toe geometric features of a vole (Tong et al., 2003; Guo et al.,
2009). (a) Photograph of the middle toe of a vole’s right fore claw (×10); (b) The
curvature profile of the inside contour line of the same clawed toe.
Pangolin scales are extremely resistant, which not only plays an

important protective role, but also serves as an effective digging tool.
The surface of the pangolin’s scale showed non- equal-lattice geometric
network morphology with transverse grooves and longitudinal ridges
crossing each other, and which would be conducive to anti-adhesion.33
Meanwhile, the corrugations on the pangolin scales can reduce wear
under free abrasive wear conditions.53
The ventral scale of Trimeresurus stejnegeri was found to be formed
with a periodic arrangement of microfibrils, micropits and pits by
atomic force microscopy (AFM).37 The microfibrils are helpful for
reducing the contact area with soil, thus reducing the adhesion; the
excreted and stored polar lipids by micropits, pits, and the hydropho-
bicity of the ventral scales can reduce the adhesion of water. The similar
microstructures were found in the ventral scale surface of the Burmese
python, which has a periodically arranged micro-convex and plank
structure.38 The frictional force of the ventral scale surface was caused
by mechanical tooth function of micro-convex, acting intermolecular
force and the material’s elastic hysteresis.
14.3. Bionic Design of Key Tillage Components
The black soil area of Northeast China is facing a series of serious prob-
lems, such as soil erosion and decreasing soil fertility. Implementation
of conservation tillage in the Northeast black soil area is necessary for
the stable development of agriculture and for the long-term effective

supply of agriculture products. Designing the soil tillage has important
implications for conservation tillage effects, such as the functions of
anti-adhesion or resistance reduction. In the following section, the
bionic design of components used in Northeast China is presented,
including components such as the moldboard, subsoiler, roller, furrow
opener, and the rototilling-stubble-breaking blade.
14.3.1. Bionic moldboard

A pimpled surface for the moldboard plow, as first cast in Yangcheng
China in the 1960s, significantly reduces soil adhesion during animal
power tillage. However, the design does not provide the same anti-
adhesion qualities for a tractor-towed moldboard plowing at high
working speeds and with significant impact force. In fact, a number
of different moldboard designs have been studied, such as plastic-
covered moldboards, through-hole moldboards, grated moldboards,
hydrophobe silicone oil treated moldboards, magnetic moldboards,
and vibrating moldboards. However, none of those designs has moved
beyond research in the laboratory or limited trials. In the 1940s
and 1950s, mole plows were developed abroad and were used to
excavate drainage ditches underground. This kind of plow imitates
a mouse’s underground caving. Such a caving mode can reduce the
resistance and earthwork quantity, compared to open ditch excavation.
In the 1980s, a type of vibrating mole plow was studied in China
to further reduce the resistance in drainage ditch excavations. Ren
et al.40 successfully developed a novel bionic non-smooth adhesion-
reduction and resistance-reduction moldboard, which possessed a sim-
ple structure and offered abrasion resistance, corrosion resistance, low
cost, low energy consumption, and high efficiency. The design worked
reliably, provided good soil turn up, and significant soil removal. This
moldboard design reduced a tractor’s tillage resistance by 8%–12%, thus
decreasing the tractor’s fuel consumption by 5%–12% compared with
traditional smooth moldboards.54
14.3.2. Bionic subsoiler components

One way to reduce the working resistance of tillage components is to
use a vibrating deep-loosening component. However, vibrating systems
increase the complexity of the structure of the implement, which then
increases manufacturing costs.26 The handle of a traditional subsoiler
implement generally has a transition section that is a straight line,
broken line, or arc shape to connect the tip of the subsoiler with the
frame, so the soil tillage resistance of the implement remains high.
Currently, this problem of large traction resistance during tillage by a
subsoiler implement is primarily solved by using a high-powered tractor.
As an alternative to this brute power approach, Tong et al.51 designed a
bionic bend-type subsoiler component based on analysis of the contour
shape of the toes of a field mouse (Fig. 14.4), and determined its tillage
resistance through a large-scale indoor soil bin test. The experiment
showed that the designed bionic deep-loosening component had an
obvious resistance-reduction effect of 6%–10%.
14.3.3. Bionic flexible compression roller

The bionic flexible roller shown in Fig. 14.5a was designed based
on the study of the typical movements of soil animals including
Pheretima and Spirobolus bungii.55 The design of the modified roller
exhibited a number of improved features over regular rollers: It
virtually eliminated soil adhesion, improved operational quality and
Fig. 14.4. The experimental cultivator parts (Tong et al., 2003; Guo et al., 2009):
(a) Components 1–8; (b) Force measurement device.
Fig. 14.5. (a) Bionic flexible compression roller (Ren et al., 2002); (b) The flexible
compression rollers have been used in the field.
efficiency, reduced energy consumption, created a clear and neat

appearance and uniform compactness of the rolled soil (Fig. 14.5b),
and was beneficial in maintaining soil humidity and temperature, thus
increasing the germination rate and seedling emergence rate after
sowing. In the early stages of land reclamation in Northeast China
in the 20th century, the soil was extremely adhesive when it was wet.
In fact, people could walk in the fields only by wearing shoes lined
with Carex meyeriana, from which the soil is easy to remove due
to the deformation of Carex meyeriana during walking. This same
concept inspired the design of the modified roller, where wrapping
the roller with rubber generates a large deformation that improves soil
removal.
14.3.4. Bionic furrow opener

The bionic furrow opener solves the problems that widely existed
in furrowing components, such as soil adhesion, grass winding, and
increased resistance. Geometrical structures (domes, cylindrical section
ridges, and tubular section ridges) on the cuticle surface of the dung
beetle Copris ochus Motschulsky were imitated and applied to the
furrow opener surface (Fig. 14.6).56 A corrugated bionic structural
morphology was formed using a series of ultra-high molecular weight
polyethylene forms attached to the surface of a general double share
Fig. 14.6. (a) Electric carriage used for soil resistance tests; (b) conventional furrow
opener and (c) bionic furrow opener with force sensor; (d) furrow opener surfaces,
from left to right: conventional smooth, bionic convex domes, bionic cylindrical section
ridges, and bionic surface structure with tubular section ridges (Tong et al, 2009b).
furrow opener. The goal of the altered geometric structure was to

reduce soil adhesion and working resistance of the furrow blade.
Sometimes, furrow openers that cut the soil for fertilization and
seeding can cause blockage of the seeder, which then cannot work
normally. Therefore, the design of a fertilizer opener with good cut
stubble capacity is the focus of the development of the no-tillage
planter.57 Learning from the toe curve of mole, the gear-tooth stubble
cutting mechanism of a no-tillage seeder was designed, as shown in
Fig. 14.7a. Figure 14.7b shows that the sliding-knife notched disc
opener worked through the straw serious coverage area. Through the
performance test of the improved no-till narrow-row planter, the pass
rate of the seed spacing was 91% (≥75%), the replay rate was 10.5%
(≥20%), and the leak seeding rate was 1.79% (≥10%). The displacement
instability coefficient and the inconsistency coefficient of each line were
11.9% and 0.82%, respectively.
14.3.5. Bionic rototilling-stubble-breaking blade

When a mole cuts and excavates soil, its front paws move in a rotating
manner on the sides of its body, with the front ends of its excavating
limbs moving elliptically.36 The motion is similar to the rotating
implements used in rotary tillage and stubble-breaking operations. Ji
et al.36 designed the front cutting surface of a universal rotary tillage and
Fig. 14.7. (a) Schematic diagram of sliding-knife notched disc opener (Jia et al.,
2013). (b) It passes through the straw serious coverage area.
stubble-breaking blade based on the geometric structure of the toes of

the mole. The redesign optimized the bionic structure of the blade by
evenly distributing three inwards arcs with central angles of 60◦ along
the front cutting surface to form a notch (Fig. 14.8a). Figure 14.8b
shows the tillage machine assembled with those bionic rototilling-
stubble-breaking blades.58 A comparison of power consumption results
for a single blade of a universal rototilling-stubble- breaking blade
versus a single blade of the bionic rototilling-stubble-breaking blade
shows that the torque of the bionic blade was 3.91% lower than
that of the universal blade in rotary tillage, and 1.62% lower than
that of the universal blade in stubble-breaking. The results of field
experiments indicate that the quality of both the rototilling-stubble-
breaking operations with the bionic and the universal rototilling-
stubble-breaking blade meet China’s national standards for rotary
tillage blades (Fig. 14.8c).58 Furthermore, the notch created by the
bionic structure offers an average dimension loss less than that of the
China national standard specified for rotary tillage blades.49,50
14.3.6. Ridging shovel

Learning from the microstructure of the convex (concave) and ridging
(triangle and arc-shaped) shapes of fresh lotus leaves and shark skin,
Fig. 14.8. (a) Diagram of bionic rotary stubble blade (Ji et al., 2012). 1 is the shaft
of the blade, 2 is the side-edge, 3 is the transitional surface, 4 is the transition edge,
5 is the scoop surface, 6 is the sidelong edge, 7 is the meander-line, L is the working
width, θ is the sliding cutting angle, and r is the sliding cutting radius; (b) The tillage
machine was assembled with bionic rototilling-stubble-breaking blades; (c) The soil
surface condition after the stubble-breaking operation (Ji, 2010).
bionic ridging shovels were designed with the characteristics of adhe-

sion and resistance reduction.59 The ridging shovel model consists of
the ridging shovel prototype, the bionic convex (concave) triangular
trough ridging shovel, and the bionic convex (concave) arc-shaped
trough ridging shovel (Fig. 14.9).
Finite element analysis software was used to simulate and analyze
the interaction process of the 10 ridging shovel models and the soil
model. The model was also used to explore and compare the stress
analysis results to achieve the shovel wall type structure of the bionic
ridging shovel, something that is more obvious in the visbreaking and
resistance reduction effect. Finite element analysis shows that the bionic
convex triangular trough ridging shovel and the bionic convex arc-
shaped trough ridging shovel have more obvious resistance-reduction
effects than other types of shovel. The indoor soil bin experiment
with the bionic ridging shovel and the prototype ridging shovel were
completed (Fig. 14.10). The ridging resistance of the three types of
Fig. 14.9. Models used for ANSYS analysis, where RS is the ordinary ridging shovel.
CVT is the bionic convex triangle trough cross-section of the ridging shovel; CCT is the
bionic concave triangle trough cross-section of the ridging shovel; CVA is the bionic
convex arc trough cross-section of the ridging shovel; CCA is the bionic concave arc
trough cross-section of the ridging shovel; CH is the convex hull.
Fig. 14.10. Three types of ridging shovel. (a) is the RS, (b) is the CVT, (c) is the
CVA.
ridging shovel was tested, under the condition of two soil moistures
(18.61% and 20.9%) and three different ridging speeds (0.68 m/s, 0.87
m/s, 1.11 m/s). The average resistance reduction rate of the bionic
convex triangular trough ridging shovel is 3.65% and that of the bionic
convex arc-shaped trough ridging shovel is 1.85%, so it is concluded
that the bionic convex triangular trough ridging shovel has a more
obvious resistance-reduction effect than the bionic convex arc-shaped
trough ridging shovel. These conclusions are consistent with the finite
element simulation results of the ridging shovel and soil model.
14.4. Conclusion
Bionic anti-adhesion or resistance-reduction technologies have been
used on farming machines, earthwork construction machines, coal
production machines and pulverized coal transport machines in six
provinces and autonomous regions in China now.29 Nine new products
were developed, and bionic technologies have been transferred and
applied to 11 companies, which increase profits by 42.6992 million
RMB, saving 48.07 million RMB.60 Bionic non-smooth plow mold-
board can achieve an anti-adhesion rate of 90%, reduce drag resistance
8–12%, and reach fuel savings of 7.2%. Bionic flexible lining for dump
trucks has an anti-adhesion rate of more than 95%. Bionic flexible rollers
have an anti-adhesion rate of up to 88% and reduce drag resistance by
7.2%. Bionic subsoilers can reduce the resistance by 6–8% compared
to Japanese-imported ones, which brings huge economic and social
benefits.29
Bionic anti-adhesion or resistance-reduction technologies were
also used in researching farming machines for soil water storage and
preservation of ridge tilling area in Northeast China. Twenty-two series
of machines have been developed which are suitable for minimum
tillage or non-tillage. Compared with convention tillage, the soil
moisture content can be increased by 30% compared, and production
and water production efficiency increased by 11,320 T and 25.6%,
respectively, through 11,333 hm2 application (Si, 2007).
Bionic agricultural machinery can be directly used in production
to generate economic and social benefits, help farmers, and advance
agricultural modernization. Recent innovations in bionic implements
are as following:
(1) New bionic soil-engaging implements can reduce the amount of

moved soil and disturbance to the soil, reduce tillage resistance, and
offer good anti-adhesion capabilities to achieve higher efficiency,
lower energy consumption, and better preserve soil moisture.
(2) Bionic technology has been combined with information tech-

nology, intelligent control, engineering technology, and modern
agriculture techniques of minimum- or non-tillage systems to
form integrated technologies with high-efficiency mechanization
in modern agriculture. These systems provide better soil moisture
storage and preservation, and better soil fertilization, to improve
production and yields. This higher efficiency means bionic imple-
ments have significant market potential.
(3) These core capabilities of multi-functional bionic implements can
help innovate existing agricultural equipment manufacturing and
increase the comprehensive production capability of agriculture
in Northeast China. These improvements will ensure efficient
and environmentally friendly utilization of agricultural resources,
promote agriculture restructuring, increase farmer income, and
provide advanced equipment to support modern agriculture
techniques.
Tillage will disrupt soil aggregates, compact soil, disturb plant

and animal communities that contribute to aggregation and lower soil
organic matter, cation exchange capacity, nutrients, microbial activity
and faunal activities that contribute to aggregation.12,60−64 However,
reasons for the increasing soil degradation due to soil compaction may
be found in the increase in weight of agricultural machinery, in the more
intense use of machinery even under unfavorable soil conditions, and
in bad crop rotations due to economic pressure and structural changes
in modern agriculture.65 Compaction implies an increase in soil bulk
density, and soil strength will increase associated with this, and decreases
in air permeability and hydraulic conductivity.66 The soil becomes
denser, and stronger due to compaction, and the increasing energy
requirement also negatively influences the farmer’s budget; the costs
for fuel are higher compared to the income from yield, and therefore it
is very important to minimize the costs for tillage in order to optimize
the profit.65 The intensity and timing of tillage determine the extent of
the effect of tillage on soil organic capacity.67 One impossible resolve
method is to combine tillage and seeding implements in one machine,
which is minimum tillage implement adapted to the requirements of
modern agriculture, and combines multiple working components in
one to complete all major seed bed preparation and crop planting
processes, including deep-loosening, sowing, fertilizing, covering, and
rolling.1,68 Soil compaction is also an environmental problem.69 It is
one of the causes of erosion and flooding.70−72 In addition, it directly or
indirectly increases nutrient and pesticide leaching to the groundwater
and nitrous oxide (a greenhouse gas) emissions to the atmosphere.73
Therefore, the design of lightweight machinery will be the development
trend of agricultural engineering in future.
Reducing the mass of agricultural equipment will only improve
the performance of agricultural machinery, but also can reduce fuel
consumption. If the overall weight of agricultural equipment were
reduced by 10%, fuel efficiency could increase by 6%–8%. Reducing
the equipment weight by 100 kg, fuel consumption can reduce from
0.3–0.6 liters per hundred kilometers. Along with a large number of
applications of lightweight and high-strength materials, combination of
high-quality materials and optimized design will lighten the combined
equipment by 30%–40% or even more.
Future research of lightweight design of agricultural engineering
will focus on how to diversify manufacturing processes, and how to
configure the associated policy orientation, which would be the only
way to ensure consistency between the high complexity product and
its prototype design concepts. There are four main ways to make
agricultural equipment lightweight. First, develop new manufacturing
processes for agricultural engineering components under the premise
of retaining the main parameters of size specifications to enhance
local reinforcement, and reduce the amount of materials. Second, use
lightweight materials, such as plastic, nano aluminum, magnesium,
ceramic, glass fiber or carbon fiber composite materials. Third, use
computer design, such as topology optimization method, to enhance
the structural strength of the whole machine. The main approach
currently is the use of lightweight materials, but this still does not fully
achieve the purpose of saving materials. It will be more meaningful
through structural optimization to guarantee the dynamic performance
to achieve lightweight machinery.
Acknowledgements
This work was supported by the National Natural Science Foundation
of China (No. 31172144), by the National Science & Technol-
ogy Pillar Program of China in the Twelfth Five-year Plan Period
(2014BAD06B03), by the Development Program of Science and
Technology of Jilin Province of China (No. 201303040NY), by the
Basic Operation Foundation of Jilin University (No. 201200007), and
by “Project 985” of Jilin University.
References
[1] Jia HL, Wang LC, Li CS, Tan HJ. (2010) Combined stalk-stubble breaking and
mulching machine. Soil and Tillage Research 107(1): 42–48.
[2] Jean RE, Christel A, Michel B, Guy R. (2010) Tillage and soil ecology: Partners
for sustainable agriculture. Soil and Tillage Research 111(1): 33–40.
[3] Dalal RC, Allen DE, Wanga WJ, et al. (2011) Organic carbon and total nitrogen
stocks in a Vertisol following 40 years of no-tillage, crop residue retention and
nitrogen fertilisation. Soil and Tillage Research 112: 133–139.
[4] Zhang SX, Li Q, Zhang XP, et al. (2012) Effects of conservation tillage on soil
aggregation and aggregate binding agents in black soil of Northeast China. Soil
and Tillage Research 124: 196–202.
[5] Schneider F, Steiger D, Ledermann T, et al. (2012) No-tillage farming- Co-
creation of innovation through network building. Land Degradation and
Development 23: 242–255.
[6] Chaplain V, De’fossez P, Richard G, et al. (2011) Contrasted effects of no-till
on bulk density of soil and mechanical resistance. Soil and Tillage Research 111:
105–114.
[7] López MV, Blanco-Moure N, Limón MA, Gracia R. (2012) No tillage in rainfed
Aragon (NE Spain): Effect on organic carbon in the soil surface horizon. Soil
and Tillage Research 118: 61–65.
[8] Maswaure J. (1995) Tillage forces and soil loosening as influenced by tool
geometry. Doctoral dissertation, McGill University.
[9] Nichols ML. (1931) The dynamic properties of soil. II. Soil and metal friction.
[10] Schafer RL, Johnson CE. (1982) Changing Soil Condition – The Soil Dynamics
of Tillage. In: Unger PW, Van Doren Jr. DM, Whisler FD, Skidmore EL (eds.),
Predicting Tillage Effects on Soil Physical Properties and Processes. American
Society of Agronomy Soil Science Society of America, Wisconsin, USA.
[11] Brady NC, Weil RR. (2002) The Nature and Properties of Soils, 14th ed. Prentice
Hall Inc., NJ.
[12] Fountaine ER. (1954) Investigations into the mechanism of soil adhesion.
Journal of Soil Science 5: 251–263.
[13] Gill WR, Vandenberg GE. (1968) Soil Dynamics in Tillage and Traction.
Agriculture Handbook No. 316, USDA, Washington.
[14] Sakharov SM, Voronin PP, Shipkov AB. (1973) Study of soil adhesion. Bulletin
of the Permanent International Association of Navigation Congresses 1: 49–51.
[15] Vilde A. (2001) Physical and mechanical properties of soil affecting energy
capacity of its tillage. 1st International Conference of BSB of ISTRO. Meeting of
Working Group 3 of the INCO-COPERNICUS Concerted Action on Subsoil
Compaction, Tartu (Estonia), pp. 21–24.
[16] Chancellor WJ. (1994) Friction between soil and equipment materials review.
In American Society of Agricultural Engineers Meeting, No. 94-1034.
[17] Bachmann J, van der Ploeg RR. (2002) A review on recent developments in soil
water retention theory: interfacial tension and temperature effects. Journal of
Plant Nutrition and Soil Science 165: 468–478.
[18] Sakane H, Maeyama T. (1993) The mudless rotary tiller. Kubota Technical
Report 26: 27–31.
[19] Salokhe VM, Gee-Clough D. (1987) Studies on effect of lug surface coating on
soil adhesion of cage wheel lugs. Proceedings of the 9th International Conference
of ISTVS, Barcelona, Spain, pp. 389–396.
[20] Salokhe VM, Gee-Clough D. (1988) Coating of cage wheel lugs to reduce soil
adhesion. Journal of Agricultural Engineering Research 41: 201–210.
[21] Salokhe VM, Gee-Clough D. (1989) Application of enamel coating in agricul-
ture. Journal of Terramechanics 26: 275–286.
[22] Jia X. (2006) Unsmooth cuticles of soil animals and theoretical analysis of
their hydrophobicity and anti-soil-adhesion mechanism. Journal of Colloid and
Interface Science 295: 490–494.
[23] Tong J, Ren LQ, Yan JL, et al. (1999) Adhesion and abrasion of several materials
against soil. International Agricultural Engineering Journal 8: 1–22.
[24] Ren LQ, Cong Q, Tong J, Chen BC. (2001) Reducing adhesion of soil against
loading shovel using bionic electro-osmosis method. Journal of Terramechanics
38: 211–219.
[25] Spagnoli G, Klitzsch N, Fernandez-Steeger T, et al. (2011) Application of
electro-osmosis to reduce the adhesion of clay during mechanical tunnel driving.
Environmental and Engineering Geoscience 17: 417–426.
[26] Wang XL, Ito N, Kito K, Garcia PP. (1998) Study on use of vibration to reduce
soil adhesion. Journal of Terramechanics 35: 87–101.
[27] Tong J, Ren LQ, Chen BC. (1994) Geometrical morphology, chemical constitu-
tion and wettability of body surfaces of soil animals. International Agricultural
Engineering Journal 3: 59–68.
[28] Ren LQ, Tong J, Li JQ, Chen BC. (2001) Soil adhesion and biomimetics of soil-
engaging components: a review. Journal of Agricultural Engineering Research
79: 239–263.
[29] Ren LQ. (2009) Progress in the bionic study on anti-adhesion and resistance
reduction of terrain machines. Science in China Series E: Technological Sciences
52: 273–284.
[30] Tong J, Sun JY, Chen DH, Zhang SJ. (2005) Geometrical features and
wettability of dung beetles and potential biomimetic engineering applications
in tillage implements. Soil and Tillage Research 80: 1–12.
[31] Dai ZD, Tong J, Ren LQ. (2006) Researches and developments of biomimetics
in tribology. Chinese Science Bulletin 51: 2681–2689.
[32] Liu GM, Li JQ, Zou M, et al. (2008) Feature and the characteristic of sliding
resistance reduction of the earthworm’s non-smooth surface. Transactions of the
CSAE 24: 62–65.
[33] Ma YH, Tong J, Zhou J, et al. (2008) Geometric shape and performance of the
scale of the pangolin. Journal of Chinese Electron Microscopy Society 27: 336–340.
[34] Chirende B, Li J. (2009) Review on application of biomimetics in the design of
agricultural implements. Biotechnology and Molecular Biology Reviews 4: 42–48.
[35] Koch K, Bhushan B, Barthlott W. (2009) Multifunctional surface structures of
plants: an inspiration for biomimetics. Progress in Materials Science 54: 137–178.
[36] Ji WF, Chen DH, Jia HL, Tong J. (2010) Experimental investigation into soil-
cutting performance of the claws of mole rat (Scaptochirus moschatus ). Wear 7:
S166–S171.
[37] Zhang ZL, Yang JC, Ding JN, et al. (2006) Ultrastructure and reducing-
resistance mechanism of ventral scales of Trimeresurus stejnegeri. Journal of
Jiangsu University (Natural Science Edition) 27: 471–474.
[38] Zhang ZL, Ding JN, Yang JC, et al. (2007) Frictional Mechanism and
Anisotropy of Burmese Python’s Ventra1 Scale Surface. Tribology 27: 362–366.
[39] Ren LQ, Wang YP, Li JQ, Tong J. (1998) The flexible unsmoothed cuticles of
soil animals and their characteristics of reducing adhesion and resistance. Chinese
Science Bulletin 43: 166–169.
[40] Ren LQ, Deng SQ, Wang JC, Han ZW. (2004) Design principles of the non-
smooth surface of bionic plow moldboard. Journal of Bionic Engineering 1:
9–19.
[41] Barthlott W, Neinhuis C. (1997) Purity of the sacred lotus, or escape from
contamination in biological surfaces. Planta 202: 1–8.
[42] Miklosovic DS, Murray MM, Howle LE, Fish FE. (2004) Leading-edge
tubercles delay stall on humpback whale (Megaptera novaeangliae) flippers.
Physics of Fluids 16: 39–42.
[43] Tong J, Moayad BZ, Ren LQ, Chen BC. (2004) Biomimetics in soft terrain
machines: a review. International Agricultural Engineering Journal 13: 71–86.
[44] Adachi N, Oida A, Nakashima H. (2005) Effect of blade surface properties

on soil cutting reaction using a cohesive soil. Proc International Agricultural
Engineering Conference, Bangkok, 2005.
[45] Tong J, Wu N. (2006) Mathematical models and analysis of profiles of the clypeus
surface of dung beetle Copris ochus Motschulsky. Transactions of the Chinese
Society for Agricultural Machinery 37: 113–116.
[46] Li JQ, Kou BX, Liu GM, et al. (2010) Resistance reduction by bionic coupling
of the earthworm lubrication function. Science China Technological Sciences 53:
2989–2995.
[47] Tong J, Gao H, Sun JY. (2009) Microstructure and nanoindentation properties
of foreleg’s tarsal claw of the oriental mole cricket (Gryllotalpa orientalis
Burmeister). Transactions of the Chinese Society of Agricultural Machinery 40:
190–193.
[48] Gao H, Li YZ, Tong J. (2011) Profile data acquisition and quantitative analysis
of the pronotum of oriental mole cricket. Transactions of the Chinese Society of
[49] Ji WF, Tong J, Jia HL, et al. (2010) Quantitative characteristic features of
geometric structures of claws of mole rat. Transactions of the Chinese Society
for Agricultural Machinery 41: 193–198.
[50] Ji WF, Jia HL, Tong J. (2012) Experiment on working performance of bionic
blade for soil-rototilling and stubble-breaking. Transactions of the Chinese Society
of Agricultural Engineering 28: 24–30.
[51] Tong J, Guo ZJ, Ren LQ, Chen BC. (2003) Curvature features of three
soil-burrowing animal claws and their potential applications in soil-engaging
components. International Agricultural Engineering Journal 12: 119–130.
[52] Guo ZJ, Zhou ZL, Zhang Y, Li ZL. (2009). Bionic optimization research of
soil cultivating component design. Science in China 52: 955-965.
[53] Tong J, Lü TB, Ma YH, et al. (2007) Two-body abrasive wear of the surfaces
of pangolin scales. Journal of Bionic Engineering 4(2): 77–84.
[54] Ren LQ, Deng SQ, Tian LM. (2004b) The forming mechanism and rules
of tangent resistance on the interface between soil and moldboard. Proc
Joint Meeting 7th Asia-Pacific Conf ISTVS & 25th Ann Meeting Japanese Soc
Terramechanics. Changchun, Jilin University Press, pp. 459–469.
[55] Ren LQ, Li JQ, Tong, J, Chen BC. (2002) Applications of the bionics flexibility
technology for anti-adhesion between soil and working conponents. Proc 14th
Int Conf ISTVS. Vicksburg, ISTVS.
[56] Tong J, Moayad BZ, Ma YH, et al. (2009b) Effects of biomimetic surface designs
on furrow opener performance. Journal of Bionic Engineering 6: 280–289.
[57] Jia HL, Zhao JL, Jiang XM, et al. (2013) Design and experiment of anti-blocking
mechanism for inter-row no-tillage seeder. Transactions of the Chinese Society of
[58] Ji WF. (2010) Biomimetic blades for soil-rototilling and stubble-breaking.

Doctoral dissertation, Jilin University.
[59] Zhang ZJ, Jia HL, Sun JY. (2015) Bioinspired design of a ridging shovel with
anti-adhesive and drag reducing. Advances in Mechanical Engineering 7: 1–11.
[60] Karlen DL, Mausbach MJ, Doran JW, et al. (1997) Soil quality: A concept,
definition, and framework for evaluation (a guest editorial). Soil Science Society
of America Journal 61: 4–10.
[61] Linn DM, Doran JW. (1984) Effect of water-filled pore space on carbon dioxide
and nitrous oxide production in tilled and nontilled soils. Soil Science Society of
America Journal 48: 1267–1272.
[62] Plante AF, McGill WB. (2002) Soil aggregate dynamics and the retention
of organic matter in laboratory-incubated soil with differing simulated tillage
frequencies. Soil and Tillage Research 66: 79–92.
[63] Filho CC, Lourenco A, Guimaraes MDF, Fonseca ICB. (2002) Aggregate
stability under different soil management systems in a red Latosol in the state of
Parana, Brazil. Soil and Tillage Research 65: 45–51.
[64] Bronick CJ, Lal R. (2005) Soil structure and management: a review. Geoderma
124: 3–22.
[65] Keller T. (2004) Soil Compaction and Soil Tillage – Studies in Agricultural Soil
Mechanics. Doctoral dissertation, Swedish University of Agricultural Sciences.
[66] Whalley WR, Dumitru E, Dexter AR. (1995) Biological effects of soil com-
paction. Soil and Tillage Research 35: 53–68.
[67] Studdert GA, Echeverria HE. (2000) Crop rotations and nitrogen fertilization
to manage soil organic carbon dynamics. Soil Science Society of America Journal
64: 1496–1503.
[68] Jia HL, Ma CL, Li GY, Huang DY. (2007) Combined rototilling-stubble-
breaking-planting machine. Soil and Tillage Research 96: 73–82.
[69] Pagliai M, Vignozzi N, Pellegrini S. (2004) Soil structure and the effect of
management practices. Soil and Tillage Research 79: 131–143.
[70] Horn R, Domzal H, Slowinska-Jurkiewicz A, van Ouwerkerk C. (1995) Soil
compaction process and their effects on the structure of arable soils and the
environment. Soil and Tillage Research 35: 23–36.
[71] Soane BD, van Ouwerkerk C. (1995) Implications of soil compaction in crop
production for the quality of the environment. Soil and Tillage Research 35:
5–22.
[72] Gieska M, van der Ploeg RR, Schweigert P, Pinter N. (2003) Physikalis-
che Bodendegradierung in der Hildesheimer Börde und das Bundes-
Bodenschutzgesetz. Berichte über Landwirtschaft 81: 485–511.
[73] Lipiec, J, Stepniewski W. (1995) Effects of soil compaction and tillage systems
on uptake and losses of nutrients. Soil and Tillage Research 35: 37–52.
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-index page 579
Index
3D modeling, 325, 326 condensation, 185–191, 193–198,

200–205, 207–217, 219–223
adhesion, 421, 422, 427–440, 442, conservation tillage, 555, 563, 564
443, 448–453, 455–460, 462–467, contact angle, 426, 432, 433, 436, 437,
472–475, 477, 478 440, 450, 452, 453, 455, 458, 460,
advanced wetting theory, 2 466
aerodynamics coefficients, 271
applications, 1, 2, 4, 7–9, 11, 18, 21, dimpled surface, 271, 284
22, 26, 32–37 draft forces, 421, 426
drag reduction, 229, 230, 238–240,
bio-forming, 230, 245 244, 258, 266, 271–274, 278, 279,
bio-replication, 229–240, 242, 244, 285, 288
245, 248, 249, 256–258, 260–263, drag reduction mechanism, 114
266 drag-reducing technology, 143, 144,
bioinspired/biomimetic surface, 143, 150–152, 180, 182
150
biological shark skin/shark, 45 fabricate vivid sharkskin, 69, 109
biological systems, 325–327, 335, 336, fabricating methods of
355, 378, 380–385 superhydrophobic surfaces, 2, 19
biomimetic, 69–71, 73, 75, 76, 79, 86, fluid dynamics, 299
87, 109, 135, 136, 187, 195, 205, friction, 429, 431, 442
222, 223, 295–297, 325–329, 335,
336, 338–340, 343–346, 348, gas pipelining, 143, 144, 148, 182
351–353, 377–382, 384, 385, 394,
414 heat transfer enhancement, 220
biomimetic surface, 33, 229, 230 hydraulic smooth pipe, 143, 147–149
bionics, 325, 326, 475, 555, 556, 558,
560–572 lotus effect, 1, 394, 405, 406
capillary adsorption, 453, 456, 457 microscopic interfacial forces, 477

capillary attraction, 437, 438, 443, 458 modeling, 55
CFD model, 271, 289 morphology, 113
579
May 24, 2016 13:33 Bio-Inspired Surfaces and Applications 9in x 6in b2378-index page 580
nano-chain, 114, 123, 135–137 superhydrophobic surface, 394, 398,

nanoparticles, 393, 407–411, 415 399, 403–408, 414, 415
nozzle, 296–301, 303–306, 308–310, surface tension, 423, 432–438,
313, 315–317 450–453, 456, 460–462, 466, 467,
477
passive flow control, 272 surface texture, 87
properties, 45, 48
template, 394, 404, 407
riblet surface, 297 turbulent boundary layer, 297, 298
self-cleaning, 1, 2, 8, 9, 17, 27, 28, 32, viscosity, 438, 448, 452, 453, 460, 467
33, 36, 37 viscous drag, 82
shark skin, 114, 133, 229–240, viscous sublayer, 143, 144, 147–149,
242–244, 246, 247, 252–258, 162
261–263, 265, 266
soil adhesion, 421, 427–430, 432, 437, wall friction, 69, 100
445, 463, 466, 467, 474 wettability, 432, 435, 463
soil–tool interface, 422, 431, 439, wetting, 393, 394, 396–400, 402–407,
441–445, 458, 459, 462, 477, 478 412, 414, 415
soil-engaging tools, 441, 463, 467, wind tunnel, 271, 274, 277, 278, 282,
475, 477, 478 287, 288

Bio-Inspired Surfaces and Appli - Eddie Y K NG

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Bio-Inspired Surfaces and Appli - Eddie Y K NG

Transféré par

Droits d'auteur :

Formats disponibles

May 2, 2013 14:6 BC: 8831 - Probability and Statistical Theory PST˙ws

This page intentionally left blank

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing-in-Publication Data

Copyright © 2016 by World Scientific Publishing Co. Pte. Ltd.

Typeset by Stallion Press

Devi - Bio-Inspired Surfaces and Applications.indd 1 19/5/2016 11:10:21 AM

The book addresses the latest technology developments in the area of

vi Bio-Inspired Surfaces and Applications

Chapter 1 examines different aspects of the self-cleaning function

viii Bio-Inspired Surfaces and Applications

on micro tubes may be of interest to the biomedical industry. When

power source to develop additional effort for overcoming adhesive

This page intentionally left blank

1. Self-Cleaning and Superhydrophobic Surfaces 1

2. Treatments and Constructing Digital Model

3. Different Approaches to Manufacture Low Viscous

4. Different Characteristic Analysis of Drag-Reducing

xii Bio-Inspired Surfaces and Applications

5. Application of Biomimetic Shark Skin Surface

6. Biomimetic Surfaces for Enhanced Dropwise

7. Large-Scale Fabrication of Biomimetic

8. Study of Flow over Dimpled Cylinder

9. Fluid Flow in Biomimetics Simulated Vessel Having

10. 3-D Modelling of Biological Systems for Biomimetics 325

11. Superhydrophobic Surfaces with Hierarchical

12. Bio-Inspired Macro-Morphologic Surface

13. Application of Bio-Inspired Surfaces in Reducing

14. Application of Bionic Technologies

This page intentionally left blank

xvi Bio-Inspired Surfaces and Applications

List of Contributors xvii

xviii Bio-Inspired Surfaces and Applications

G.G. Li,∗ Y.T. Zhao,∗ L. Zhang,∗ B.D. Liu,† Y. Luo,‡

University, Xi’an, 710072, China

Washington D.C., 20052, USA

Technological University, 639798, Singapore

Recently superhydrophobic surfaces have attracted considerable inter-

2 Bio-Inspired Surfaces and Applications

surfaces are systematically illustrated, including advanced wetting

A proverb originating in ancient China states that “The lotus grows

Fig. 1.1. Lotus leaf.

Self-Cleaning and Superhydrophobic Surfaces 3

Surface height map and 2D profile of lotus leaf

Surface height maps of micro bump on lotus leaf

Fig. 1.2. Superhydrophobicity and morphology of lotus leaf.

directional adhesion, photocatalysis, self-sterilizing, and anti-fogging

4 Bio-Inspired Surfaces and Applications

Fig. 1.3. Different superhydrophobic surfaces of insects.

taking away powder-like contaminants. This interesting phenomenon

Self-Cleaning and Superhydrophobic Surfaces 5

Fig. 1.4. Water droplets on plant leaf.

surface roughness can be more important in achieving dewetting

1.2. Advances in Wetting Theory

It is usually believed that wettability is the tendency of a liquid to

6 Bio-Inspired Surfaces and Applications

Fig. 1.5. Contact angle.

Fig. 1.6. Hydrophilic and hydrophobic surfaces.

Furthermore, it has been conﬁrmed by researchers all over the

Self-Cleaning and Superhydrophobic Surfaces 7

Fig. 1.7. Wenzel’s model and Cassie’s model.

cylindrical groove model, a model with paraboloids of revolution and

8 Bio-Inspired Surfaces and Applications