Discrete Modeling of Ultra High Performance Concrete With Application To Projectile Penetration - 2014 - International Journal of Impact Engineering PDF

International Journal of Impact Engineering 65 (2014) 13e32
Contents lists available at ScienceDirect
International Journal of Impact Engineering

journal homepage: www.elsevier.com/locate/ijimpeng
Discrete modeling of ultra-high-performance concrete

with application to projectile penetration
Jovanca Smith a, Gianluca Cusatis b, *, Daniele Pelessone c, Eric Landis d, James ODaniel e,
James Baylot e
a
Department of Civil and Environmental Engineering, Northwestern University, Tech Building Room A323, Evanston, IL 60208, USA
Department of Civil and Environmental Engineering, Northwestern University, Tech Building Room A125, Evanston, IL 60208, USA
ES3, San Diego, CA 92101, USA
d
Department of Civil and Environmental Engineering, University of Maine, 303 Boardman Hall, Orono, ME 04469, USA
e
Geotechnical and Structures Laboratory, U.S. Army Engineer Research and Development Center, Vicksburg, MS 39180, USA
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 21 April 2013
Received in revised form
15 October 2013
Accepted 22 October 2013
Available online 1 November 2013
In this paper, the Lattice Discrete Particle Model for ber reinforced concrete (LDPM-F) is calibrated and
validated with reference to a new high-strength, ultra-high-performance concrete (UHPC) named
CORTUF and applied to the simulation of projectile penetration. LDPM-F is a three-dimensional model
that simulates concrete at the length scale of coarse aggregate pieces (meso-scale) through the adoption
of a discrete modeling framework for both ber reinforcement and embedding matrix heterogeneity. In
this study, CORTUF parameter identication is performed using basic laboratory ber pull-out experiments and experiments relevant to a CORTUF mix without ber reinforcement. Extensive comparisons of
the numerical predictions against experimental data that were not used during the calibration phase
(relevant to both plain CORTUF and CORTUF with ber reinforcement) are used to validate the calibrated
model and to provide solid evidence of its predictive capabilities. Simulations are then carried out to
investigate the behavior of protective CORTUF panels subjected to projectile penetration, and the
numerical results are discussed with reference to available experimental data obtained at the Engineering Research and Development Center (ERDC).
2013 Elsevier Ltd. All rights reserved.
Keywords:
Ultra-high-performance concrete
Lattice Discrete Particle Model
Fracture
Fragmentation
Penetration
1. Introduction
The design exibility and natural durability of concrete help to
make it the most used man-made material in the world. As opposed
to other building materials, such as brick and stone, concrete has
revolutionized previous construction methods with its intrinsic
versatility and high compressive strength. Within the last century, a
tremendous effort has been made to increase concrete strength.
Over time, each new generation of the material became stronger
and more durable, birthing the title we know today as ultra-highperformance concrete (UHPC). The material is sought for many
* Corresponding author. McCormick School of Engineering and Applied Science,

Northwestern University, Tech Building Room A125, 2145 N Sheridan Rd, Evanston,
IL 60208-3109, USA. Tel.: 1 847 491 4027.
E-mail addresses: jsmith2014@u.northwestern.edu (J. Smith), g-cusatis@
northwestern.edu (G. Cusatis), peless@es3inc.com (D. Pelessone), landis@maine.
edu (E. Landis), James.L.ODaniel@usace.army.mil (J. ODaniel), James.T.Baylot@
usace.army.mil (J. Baylot).
0734-743X/$ e see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijimpeng.2013.10.008
uses including its adoption in high-rise structures, long-span

bridges, offshore structures, and structural rehabilitation [1,2].
There are various methods used today to develop higher
compressive strength in cementitious composites. One method is
strengthening the interfacial transition zone (ITZ), the weak area
surrounding the aggregate particle, which is often accomplished
through a reduction of the aggregate size; this enhances homogeneity and reduces stress concentrations between the aggregate and
mortar under loading. Another method to enhance the ITZ mechanical properties is reducing the water-to-binder ratio [3]. This
process can be counter productive if not balanced, since the high
rates of hydration often results in autogeneous shrinkage and can
lead to premature cracking during the setting phase [4]. Many highperformance concretes use silica fume, which acts like a microller
and can also react with calcium hydroxide to increase the nal
strength of the composite [5]. Other methods reported in the
literature that enhance the compressive strength of concrete are
particle-size gradation for optimum packing of aggregates [6] and
heat treatment [7e10].
14
J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32
Although there are many methods focused on increasing the

compressive strength of concrete, techniques to increase tensile
resistance and toughness are still in their infancy. One method
typically used for this purpose is the addition of bers in the matrix
[5,11e14]. Multi-scale ber reinforcing, in which bers of various
lengths and cross sectional sizes are used in the same mix, has also
been explored in recent times [15]. When ber reinforcing is used,
the type and amount of bers need to be optimized in order to
maximize tensile strength and toughness while preserving workability and efciency in terms of cost-to-performance ratio.
Enhancement of material toughness is particularly important in
the case of the UHPC materials that are being tested to determine
their efciency against blast impact and penetration and for their
use in particular areas susceptible to man-made and natural hazards [16,17]. As it pertains to projectile impact testing, many
researchers found the penetration depth to increase with
increasing impact velocities, and to decrease with increasing
compressive strength. The perforation ballistic limit (the minimum
velocity at which the bullet completely perforates the specimen) is
typically reported to increase with higher compressive strength
[18]. Aggregate size also plays an important role in material impact
response. Stronger and larger sized aggregate particles improve
impact resistance as long as enough workability is maintained
[3,12,19]. For UHPC without ber reinforcing, penetration events
cause dynamic propagation of many micro-cracks and often lead to
shattering of the target. On the contrary, inclusion of bers connes
cracks in the crater region in both front and rear faces and creates a
more localized damage around the crater. However, for increasing
ber volume fraction, a saturation limit seems to exist after which
no signicant change is observed in both crater area and penetration depth [20e22].
When dealing with the effect of dynamic events on concrete
structures it is also important to consider the strain-rate dependence of concrete response. Typical experimental observations on
regular strength concrete report an increase of macroscopic
mechanical properties such as, Youngs modulus, compressive
strength, tensile strength, and fracture energy, for increasing strain
rate (see, among many others, [23e25, 27e32]).
Evidences of strain-rate dependence of UHPC behavior are more
scarce and contradictory compared to regular strength concrete.
A few studies have been performed on the effect of strain rate on
UHPCs that investigated the effect of ber reinforcement and
increasing strain-rate under tensile and compressive loadings. Some
researchers found UHPCs with bers to be less sensitive to strainrate effects than similar strength unreinforced material [33,34].
Some other researchers [35] found no rate sensitivity for concrete
reinforced with hooked steel bers. Caverzan et al. [36] investigated
the effect of temperature on strain rates of high-performance berreinforced concrete (HPFRC) and found an increase in strength for
increased strain rate up to a rate of 150 s1 for temperatures up to
200 C, while material strength decreased for strain-rates above
150 s1 for higher temperatures due to ber failure.
In most cases, rate effect is quantied through the Dynamic
Increase Factor (DIF) dened as the ratio between the dynamic
property of interest and the associated quasi-static value. The
evolution of DIF with strain-rate is often approximated through a
bilinear curve in which the second linear branch, starting for strainrates around 1e10 s1, has a much steeper slope than the rst
branch. In many cases, the DIF curves are obtained by interpolating
between experimental data showing huge scatter, especially for
high strain-rates. Furthermore, it is common modeling procedure
to equip concrete constitutive equations with rate dependence by
using the DIF curve as a multiplicative factor for various model
parameters. As some authors have shown recently [37e39], this
approach is not correct in general because it does not distinguish
between intrinsic phenomena, which should be included in the

constitutive equations, and apparent phenomena, which are
structural features of the response and should or should not be
included in the constitutive equations depending on the spatial and
temporal resolution of the adopted numerical model.
At least three intrinsic rate-dependent phenomena can be
identied to affect concrete mechanical behavior: 1) creep, 2)
Arrhenius-type behavior of fracture processes, and 3) contribution
to the load carrying capacity of capillary and adsorbed water.
Concrete creep, the increase in time of deformation under
constant applied load, is a complex phenomenon whose fundamental mechanism resides in the cement nano-structure, and it is
hypothesized to be the result of shear slips among Calcium Silicate
Hydrate (CSH) platelets [40,41]. Since the nano-structure of cement
evolves as result of cement hydration and other chemical reactions,
creep behavior changes over time: in general, young concrete features a much more pronounced viscous behavior compared to old
concrete. In addition, concrete creep is affected by moisture content
and temperature and their time rates [42]. In particular, the lower
the relative humidity the less concrete creep deformation is typically observed. Wittmann [43] obtained an 87% decrease in creep
deformation reducing the relative humidity from 100% to 10%. From
the mathematical point of view, concrete creep is typically simulated by aging, temperature, and relative humidity dependent
visco-elastic constitutive equations; this naturally leads to a
dependence of the stress state on strain-rates.
However, creep cannot fully explain the fracturing behavior of
concrete under dynamic loading. For example, creep cannot
simulate experimental data on three point bending specimens
showing almost instantaneous rehardening when subjected to a
sudden increase of loading rate in the softening regime [44]. To
model this phenomenon, Ba
zant analyzed the fracture process
within the classical theory of thermally activated processes, and by
representing the frequency of bond rupture at the nano-scale with
an Arrhenius equation, he derived a dependence of the cohesive
stress (stress across a propagating cohesive crack) upon crack
opening rate. This formulation was successfully used in
Refs. [37,45,46] to simulate the dynamic behavior of concrete under
low to moderate strain rates.
Furthermore, concrete is a porous material featuring a complex
internal structure characterized by pore sizes spanning various
order of magnitudesdfrom few nanometers to several millimeters.
Depending upon the level of relative humidity, water is present in
concrete pores in various forms (capillary, absorbed, or hindered)
and, according to classical poro-mechanics concepts, it can
certainly contribute to the overall carrying capacity. However, under quasi-static loading conditions the amount of load that pore
water can carry is negligible because water is much more
compressible than the solid skeleton. More complicated scenario
arises under high connement quasi-static triaxial loading because
pore collapse occurs and water is squeezed out, even leading to a
certain decrease in carrying capacity due to lubrication effects [47].
On the contrary, at high strain rates, water can contribute signicantly to concrete strength. As far as tensile loading is concerned,
some studies [48] have tried to explain the effect of water at high
strain rates through the so-called Steffan effect [49] according to
which the force needed to pull apart two parallel rigid plates
containing a Newtonian uid in between is proportional to the
relative velocity of the two plates. This approach, although
championed by some authors [50,51], has been criticized because
its application to absorbed and hindered water (forming the
majority of water in real concrete at relative humidity lower that
60e70%) is questionable. For triaxial compressive loading at high
strain rates, water has a signicant effect because, as pores collapse,
water does not have enough time to be released leading to a
signicant build up of pore pressure. Forquin and coworkers [47]

report an increase of 55% of the hydrostatic stress measured in
saturated concrete samples at a strain of 0.08 and a strain-rate of
120 s1 compared to dry samples.
While the intrinsic phenomena discussed above should (if relevant) be included in the formulation of appropriate constitutive
equations for concrete, apparent phenomena should be ltered out
from the experimental evidences and not considered in the
constitutive equations unless they occur at a length scale smaller
than the model resolution. The most important, and often misinterpreted, phenomenon leading to an apparent strain-rate
dependence of concrete behavior is due to inertia effects. During
dynamic events, stress waves travel through concrete and their path
as well as their magnitude are affected by the presence of external
boundaries as well as by concrete internal heterogeneitydthis leads
to overstresses and ctitious connement effects resulting in
higher peak stresses erroneously interpreted as higher material
strength. This aspect was analyzed in detail by the second author in
an earlier publication [37]. As long as inertia is correctly accounted
for in the interpretation of experimental results and in the numerical simulations, no inertia effects should be included in concrete
constitutive equations, especially when they are formulated at ne
length scales and capture material heterogeneity.
Another phenomenon associated with dynamic loading is the
change in crack patterns. On one extreme, under quasi-static
loading, a concrete bar in direct tension always fails with the
propagation of one single macro-crack; on the other extreme, under
ultra-high strain rates in which inertia effects are dominant, the
same bar features a much more distributed crack pattern and
comminution/pulverization might occur [52]. Also the microscopic
characteristics [50,53] of crack patterns change with increasing
strain-rate. For example, it has been observed that at high strainrates, cracks initiate at the same time at multiple locations and in
their subsequent propagation follow shorter paths of higher resistance (for example fracturing aggregate pieces) as opposed to more
tortuous paths of least resistance (for example through the ITZ on
the aggregate boundaries). Also, it is well reported for concrete [54]
and other materials [55] the phenomenon of crack branching under
dynamic crack propagation. Overall, the change in crack patterns
leads to higher energy dissipation resulting in apparent strength
and toughness increase for increasing strain rates.
2. Numerical modeling for UHPC materials and the LDPM-F
formulation
Optimization of cementitious composites, with material properties tailored to a specic application, has been traditionally pursued through experimental investigations that are quite expensive
and are not sustainable for investigation of a large number of
different variables. A less expensive alternative is to optimize the
material through numerical simulations.
Typical constitutive modeling for mode I fracture in concrete
includes: (1) the cohesive crack model (CCM) [56,57], which simulates damage processes through the occurrence of displacement
discontinuities, across which the ability to transfer stresses decreases as function of the discontinuity jump (crack opening); and
(2) the crack band model [58,59], where cracks are assumed to be
distributed over a certain band whose width is deemed a material
property.
For more general triaxial stress states, strain-softening tensorial
constitutive equations have been proposed in the literature [60e63],
the most notable of which are the ones based on the microplane
theory [64e68]. However, these type of models, suffer from pathological mesh sensitivity and spurious energy dissipation upon mesh
renement [69e71]. This problem has been somewhat addressed
15
through the formulation of non-local models with either gradient

formulations or integral formulations, which allow simulation of
micro-crack interaction and size effect [72e76].
A different class of models providing a better representation of
the actual cracking phenomena occurring in the internal structure
of the material is the one consisting of discrete approaches (lattice
and particle models) in which concrete is discretized a priori
according to its meso-structure. Earlier attempts in this direction
are reported in Refs. [77e82], while more recent developments are
discussed in Refs. [83,84].
Cusatis and coworkers formulated, calibrated, and validated a
comprehensive discrete model for concrete, entitled Lattice
Discrete Particle Model (LDPM) [85e87] and extended it to include
the effect of ber reinforcing [88,89]. The resulting model, labeled
LDPM-F, demonstrated an unprecedented ability to model berreinforced concrete and to correctly predict the toughening
mechanisms due to the effect of bers. For this reason, LDPM-F was
selected in this study as the best candidate for the simulation of
UHPC concrete materials in general, and CORTUF in particular.
LDPM represents concrete as a two-phase material composed of
aggregate pieces held together by an embedding cementitious
matrix. In the numerical representation, polyhedral shaped particles (Fig. 1a) simulate individual aggregate pieces (approximated as
spheres) and their surrounding matrix. Geometry and topology of
the model are obtained through: (1) random generation and
placement of aggregate pieces; (2) delaunay tetrahedralization of
aggregate centers; and (3) dual tessellation dening particle interfaces (triangular facets in Fig. 1a) assumed to be the location of
potential cracks in the meso-structure. Fiber reinforcing is included
in such a way to retain the discrete character of LDPM. All the bers
are randomly distributed in the material volume, and each ber is
associated with all intersecting meso-scale facets (Fig. 1b). Further
details of the adopted ber generation algorithm are discussed later
in this paper.
LDPM describes meso-structure deformation through the
adoption of rigid-body kinematics [85] for each individual particle.
Based on this assumption and for given displacements and rotations of the particles associated with a given facet, the relative
displacement EuC F at the centroid of the facet can be used to dene
the following measures of strain.
eN
nT EuC F
;
[
eM
mT EuC F
;
[
eL
lT EuC F
[
(1)
where [ is the length of the straight line connecting two adjacent

particle centers, and n, m, l are unit vectors dening a local system
of reference such that n is orthogonal to the facet, and m, l are
mutually orthogonal as well as tangential to the facet. Cusatis and
Shauffert [90] showed that the strain denitions in Eq. (1) are
consistent with the denition of strain in classical continuum
mechanics. These strains can then be used to compute the LDPM
c m t c l, in accordance with the
facet stress tractions, tc tNc n tM
L
LDPM constitutive law (Appendix A1). In addition, when a tensile
facet deforms beyond the tensile elastic limit, the meso-scale crack
opening can be calculated as w wNn wMm wLl, where

wN [ eN tNc =EN ;

wL [ eL tLc =ET

c
=ET ;
wM [ eM tM
(2)
and EN, ET are the elastic normal and tangential LDPM stiffnesses,
respectively (see Appendix A1).
Furthermore, at the facet level, the formulation assumes a parallel coupling between the bers and the surrounding concrete
matrix. In this case, the total stress traction on each facet can be
calculated as
16
a)
c)
b)
Facet
c
Particle
nf
Fiber
Tetrahedron
Edge
Lf
Ls
p
Triangular
Facet
d)
f) d = 4 mm
a
e)
20 mm
g)
h)
i)
350
Coarse
Fine
1.5
0.4
Force (N)
Fine
Coarse
x 10
Coarse
Fine
250
0.6
20 mm
300
0.8
Stress (MPa)
CDF for Edge Length
da = 0.6 mm
200
150
100
0.5
0.2
50
10
Edge Length (mm)
10
0
0
0.02
0.04
Strain
0.06
0
0
0.1
0.2
Displacement (mm)
0.3
Fig. 1. CORTUF-Plain internal structure at the sub-millimeter scale and particle size comparison. a) Polyhedral cell, b) Fiber-facet interaction, c) Bond splitting mechanism, d)
Undamaged, and e) Damaged cylindrical specimen for Split test, f) Coarse and ne scale aggregate distribution, g) Crack distribution length comparison of coarse- and ne-scaled
aggregates, h) Uniaxial Compression; and i) Split test comparison for coarse and ne aggregate size distribution on 20 mm cube specimen.
t tc eN ; eM ; eL
1 X f
P wN ; wM ; wL
Ac
(3)
f Ac
where Ac is the facet area, and Pf is the crack-bridging force for each
ber crossing the given LDPM facet.
As formally noted in Eq. (3), the ber contribution, Pf, to the load
carrying capacity is assumed to be a function of the meso-scale
crack opening, w, and is formulated according to a precise micromechanical analysis of failure mechanisms affecting the ber
resistance to pull-out. These mechanisms include debonding, friction pull-out resistance, spalling, change in ber orientation during
pull-out, snubbing, and ber rupture. The mathematical description involved is well documented in the literature [91], discussed in
detail by Ref. [88], and summarized in Appendix A2.
The LDPM formulation also simulates rate effect of concrete
strength and toughness at the meso-scale by including creep and
Arrhenius-type behavior. Details of the mathematical description of
the two mechanical phenomena and their amalgamation within
the LDPM framework are reported in Refs. [46,92]. By virtue of this
formulation, LDPM (and its predecessor, the CSL model [37]) has
been applied successfully to the simulation of a variety of experimental data relevant to the dynamic response of regular strength
concrete [46,92]. In addition, LDPM has been recently adopted for
the simulation of penetration, perforation, blast, and impact of
regular strength concrete and it has shown excellent predictive
capabilities [93e95].
3. LDPM-F modeling of CORTUF concrete
This section presents the numerical modeling of CORTUF, an
UHPC recently developed and characterized at the Geotechnical
and Structures Laboratory (GSL), United States Army Engineer
Research and Development Center (ERDC) [98,99].
3.1. Mix-design, basic material properties, and internal structure
CORTUF was designed both with ber reinforcement (CORTUFFiber) and without ber reinforcement (CORTUF-Plain), and its

Table 1
Mixture composition for CORTUF.
Material
Proportion by weight
Cement
Sand
Silica our
Silica fume
Superplasticizer
Water (tap)
Steel bers (CORTUF-Fiber)
1
0.967
0.277
0.389
0.0171
0.208
0.31
average composition is summarized in Table 1. The adopted sand

has a maximum grain size of 0.6 mm. In addition, the basic CORTUF
mix is characterized by cement content, c, equal to 794 kg/m3,
which can be back calculated from the measured wet density
(2442.5 kg/m3) assuming 7.5% entrained air.
It is worth observing here that for the sake of constructing the
LDPM meso-structure, both sand and silica our were considered
aggregate, and both cement and silica fume were considered
binder. This results in water-to-binder and aggregate-to-binder
ratios of w/b 0.15 and a/b 1.03, respectively.
Fibers embedded in the matrix were hooked steel Dramix ZP305
with 0.55 mm diameter, cross sectional area of 0.2376 mm2, 30 mm
nominal length, and a reported Youngs modulus and tensile
strength of 210,000 MPa, and 1100 MPa, respectively.
LDPM-F modeling requires, as a rst step, to provide an
approximated geometric description of concrete internal structure
at the meso-scale, i.e., the scale at which the material can be viewed
as a two-phase composite in which hard and stiff inclusions are
embedded in a somewhat softer and weaker matrix. In addition,
LDPM-F provides realistic representation of material failure if
fracture processes are associated with paths, the tortuosity of
which depends on the size of the major heterogeneities. This
typically happens if aggregate particles and/or inclusions are strong
enough to deviate cracks and make them propagate through the
embedding matrix only.
Fig. 1d shows a CT-scan of an undamaged CORTUF cylindrical
specimen. A voxel intensity histogram was used in the X-ray
microtomography to describe the phases of the material [100]. The
visible circular holes form the pore spaces and surrounding air in
the matrix, the larger darker particles are the aggregates, the
cement hydrates are the smaller lighter particles, and the bright
white inclusions are unhydrated particles. A particulate internal
structure is clearly visible and strongly motivates the adoption of
LDPM-F for CORTUF modeling. Furthermore, Fig. 1e shows a postmortem CT-scan of the same CORTUF specimen subjected to
splitting test. The crack path is affected by the presence of the major
inclusions, which in most cases, are undamaged (see white arrows
in Fig. 1e) and determined the location and propagation path of the
visible cracks.
3.2. Coarse-grained approximation
In the process of approximating the internal structure of concrete, one of the most important parameters is the maximum
aggregate size, da, which for CORTUF is equal to 0.6 mm. The
minimum aggregate size included in LDPM simulations is typically
assumed to be d0 da/2, so if the particle-size distribution is governed by a Fuller curve with exponent nF 0.5, the percentage of
simulated aggregate is about 30%. This ensures, in most cases, a
good resolution of concrete internal structure and a reasonable
representation of meso-scale potential crack paths. However, particle sizes ranging from 0.3 mm to 0.6 mm lead to a very large
number of particles for typical numerical specimens. For example,
for a 50.8 mm (2 in.) diameter and 101.6 mm (4 in.) height cylinder
17
used for the CORTUF characterization in compression, one would

have about 1 million particles; for the CORTUF slabs that were
304.8 mm 304.8 mm 50.8 mm (12 in. 12 in. 2 in.) and tested
under projectile impact at ERDC, the particle number would be
more than 20 million.
To reduce the computational cost of the simulation, a coarsegraining (CG) technique exploiting a force matching approach was
adopted [101]. This coarse-graining technique is based on the
following observations: (1) particle size does not inuence the
macroscopic LDPM response as long as the macroscopic behavior
remains strain-hardening and does not feature damage localization;
(2) particle size does inuence the macroscopic LDPM response in
the case of macroscopic strain-softening and damage localization;
(3) the main parameter governing LDPM strain-softening behavior
is the material characteristic length, dened as [t 2E0 Gt =s2t ,
where E0 normal modulus, Gt meso-scale tensile fracture energy, and st meso-scale tensile strength. For a coarse-graining
factor k dcoarse
=da dcoarse
=d0 and with the same Fuller coefa
0
cient for both the coarse and the ne system, the macroscopic
coarse response can be shown to be a good approximation of the
ne response if the coarse characteristic length is calculated as
coarse
[coarse
[t
t
(4)
[
where [ interparticle distance, which is locally variable according

to the grain size distribution, and the overbar represents the mean
values of the interparticle distance distributions. It must be observed
coarse
that in general, [
=[sk, because the LDPM system is not
composed by close-packed spheres and, consequently, the interparticle distance distribution does not coincide with the particle size
distribution. In this paper, a maximum aggregate size equal to 4 mm
and a Fuller coefcient equal to 0.5 were adopted for the coarsened
system. Fig. 1f shows a 20 mm specimen with visible particles for
maximum aggregate sizes of 4 mm (left) and 0.6 mm (right); Fig. 1g
shows the associated Cumulative Distribution Function (CDF) of the
coarse
inter-particle distance. In this case, [
=[ 5:5 is obtained.
The accuracy and effectiveness of the proposed CG technique is
investigated in Fig. 1h and i, which compare the average response of
three ne (dashed curve) and three coarse (solid curve) LDPM
specimens under unconned compression tests and splitting tensile tests, respectively. The simulated specimens were 20 mm side
cubes comprised of an average of 51,000 and 450 particles for the
ne and coarse system, respectively. Simulations were performed
on the Diamond system at the ERDC Department of Defense
Supercomputing Resource Center, having 2.8 GHz Intel Xeon X5560
Nehalem-EP processors on its login and compute nodes, with 2
processors per node, each with 4 cores, for a total of 8 cores per
node. The simulations utilized 1 node and were characterized by a
run time of 10 h and 20 min (compression) and 5 h and
10 min (splitting) for each run of the ne and coarse systems,
respectively. As can be seen from the results obtained, the coarse
approximation does provide very accurate results with a signicant
saving in terms of computational time; the difference in the peak
response is about 7% for the compression test and 6% for the
splitting test. The observation must be made that the analyzed
examples are characterized by a damage zone whose size is larger
than the coarse maximum aggregate size. The same level of accuracy cannot be expected for cases in which this condition is not
satised.
3.3. Fiber reinforcing
Fiber reinforcing is always used in UHPC to increase toughness
(energy absorption capability) and ductility. Within LDPM, bers
18
are explicitly modeled by generating spatial distributions

mimicking the ones found in real specimens. From a geometric
point of view the physical bers can be described using few
parameters, in the case of straight bers: volume fraction, Vf,
length, Lf, and equivalent diameter, df. Each ber is modeled using a
sequence of one or more segments linked together. Single segments
are sufcient for generating straight bers. Multiple segments are
necessary for generating tortuous bers.
These geometric parameters are used for generating random
bers inside a control volume. For cast ber-reinforced concrete
specimens, the control volume is assumed to be equal to the volume
of the concrete part. In this case, ber location and orientation are
affected in the vicinity of the boundary. On the contrary, for machined
or cored specimens, the control volume is assumed to be larger than
the actual specimen so that bers intersecting the external surface of
the specimens are treated as cut. The portion of a cut ber inside the
specimen is shorter than the length of original ber, and this affects
the mechanical characteristics of the bereconcrete interaction near
the specimen boundaries.
The following algorithm is used for generating random bers
whose number inside a given control volume V can be computed as
N 4Vf V=Lf pd2f . The rst scenario details the case of uniformly
distributed bers. For each ber, a random point with uniform
probability in space inside the control volume is selected as the
initial point of the ber. An initial random direction is then
computed by (1) generating, in a unit cubic volume, a series of
additional random points with a uniform probability in space, (2)
rejecting all points outside a unit sphere, and (3) accepting the rst
point occurring inside the unit sphere. The vector connecting the
ber origin to this point inside the unit sphere identies an unbiased spatial direction. The computed direction is used to initialize
the ber. If the ber consists of multiple segments to simulate ber
tortuosity, then the direction of each additional segment is modied using the following logic.
A random direction, w, is computed as done previously, followed by the normal n to the current ber direction u1 as
n m=jmj, where m w (w$u1)u1. A proportional component
is then added to the tortuosity factor t by v2 u1 t n, and the new
ber direction u2 is normalized using u2 v2 =jv2 j. For t 0, the
ber continues straight in the initial direction. If the ber exits from
the control volume V, the ber is discarded.
In some cases, ber orientation is not random but has a bias
towards a preferential direction. This bias is obtained by a scaling
factor, s, for a stated preferential direction. For s 1, there is no
preferential alignment, while a generic value of s makes alignment
s times more likely for the preferential ber orientation. The most
appropriate scaling factor must be chosen through comparisons
between numerical and experimental specimens. The random ber
generation algorithm is adjusted to include preferentially oriented
bers after the spatially random direction is computed. The dot

product c n$d of random n and preferential orientation d is
computed. The random direction is then extended along the preferential direction and normalized to compute the random ber
direction by v n (s 1)cd and n v=jvj. Figs. 4a and c show
generated ber distributions without and with preferential
orientation.
In the spirit of the discrete multi-scale physical character of
LDPM, the occurrences of ber-facet intersections are determined
by actually computing the locations where bers cross inter-cell
facets. This computation can require signicant resources for
large models in terms of both time and memory. For this reason, an
efcient bin-sorting algorithm was developed for computing these
intersections. For each intersection, the lengths of the ber on both
sides of the facet and the angle at which the ber intersects the
facet are also computed. These parameters are saved in the facet
data structure and used during the simulation for computing the
ber effect.
The modeling of ber reinforcement in CORTUF is then
completed with the simulation of bereconcrete interaction,
which requires the formulation of a local force versus slippage
constitutive equation simulating ne-scale failure phenomena
occurring at the interface between bers and embedding concrete
matrix. The LDPM-F formulation [88,89] of such a law, adopted in
this study and summarized in Appendix A2, includes the description of the following failure mechanisms, (1) ber/matrix
debonding, (2) ber frictional pull-out, (3) local matrix failure due
to stress concentration at the point of exit of bers from the matrix
(micro-spalling), (4) pull-out strength increase due to additional
frictional effects caused by change of direction of ber during
cracking (snubbing), (5) ber rupture, (6) ber strength reduction
due to localized plastic deformations, and (7) ber yielding due to
local ber bending.
In addition to these failure mechanisms, there is evidence for
CORTUF concrete that splitting failure, i.e., a sudden and dynamic
propagation of a crack in a plane containing the ber axis, might
contribute signicantly to the overall tensile resistance in tension
of the material. Let us consider a straight segment of a ber
subject to tension. At the interface between the ber and the
surrounding concrete, shear stresses develop, which in turn lead
to inclined principal stress directions. When the maximum principal stress exceeds the tensile strength, radial cracks initiate, and
the concrete matrix surrounding the ber is forced to transfer
load through an inclined compression eld, see Fig. 1c. The
component of this inclined eld acting orthogonally to the axis of
the ber results in an expansive pressure on the surrounding
concrete of magnitude p Pftan a/(pdfLe). The angle a depends on
both the geometric characteristics of the ber and the remote
stress state. From this pressure, the maximum circumferential
Fig. 2. Effect of high strain rate on compression test of UHPC a) Plot of DIF with varying strain rates; Failure pattern for, b) 101 s1, c) 0.05 s1, d) 10 s1.
b)
1000
Numerical
Experimental
600
400
200
0
0
Axial Stress (MPa)
Load (N)
800
c)
1200
Uniaxial
Strain
0.04 0.06
CMOD (mm)
0.08
0.1
Numerical
600
Experimental
Embed 12.7 mm
Embed 6.35 mm
200
Beta = 0.048
150
Beta = 0.1
100
400
0
0
300
250
800
Triaxial
0 MPa
200
0.02
Triaxial
300 MPa
Load (N)
a) 1000
19
0.05
0.1
0.15
0.2
Axial Strain (Offset = 0.05)
Beta = 0
50
0
0
5
10
Displacement (mm)
15
Fig. 3. Tests results for the parametric identication of LDPM-F for a) 3-point bending, b) uniaxial strain and triaxial compression, c) single ber pull-out.
tensile stress acting at the ber/matrix interface (see Fig. 1c) can
be computed approximately as
smax
p
q
4c2f =d2f 1
(5)
4c2f =d2f 1
where cf is the distance between the ber axis and the closest
external surface of the specimen.
By introducing the expression for p into Eq. (5), setting the
failure condition, smax
st , and solving for the ber force, one can
q
calculate the splitting force
Pfsplit
pdf Le s*t
tan a
pdf Le s*t
tan a
4c2f =d2f 1
4c2f =d2f 1
for
cf =df [1
in CORTUF concrete where ber diameters have similar characteristic sizes of material heterogeneities. Based on this analogy, it
seems logical to observe in CORTUF failure mechanisms that are
similar to the ones of reinforced concrete and that are not observed
in standard ber-reinforced concrete. (2) Splitting failure is
expected to be negligible in the case of ber reinforcement with
low stiffness and high Poissons ratio compared to the one of concrete (e.g., for Polyvinyl Alcohol, PVA, bers), because such properties prevent the build-up of a signicant expansive pressure. (3) If
bond resistance is limited, bond failure occurs before the inclined
cracks can develop. As such, straight bers are less prone to induce
splitting failure than hooked and bent bers.
3.4. Rate effect
(6)
where st can be assumed to coincide with the tensile strength of
the matrix or the composite, depending on ber concentration and
level of ber interaction. At this point, a few observations are in
order. (1) The splitting mechanism highlighted below is typically
observed in reinforced concrete members [102] when concrete
rebars are not provided with proper connement through either
enough concrete cover or transverse reinforcement. It is worth
pointing out that, in regular concrete, rebars have diameter sizes of
the same order of the aggregate particles; the same situation arises
Due to the specic curing protocol used to manufacture

CORTUF, it can be assumed condently that capillary and absorbed water content is very small and actually close to none. For
this reason, intrinsic phenomena such creep and moisture effect
can be neglected in the simulations. Arguably one could include
the Arrenyus-type effect but this would require identication of
the relevant parameters, which is possible only through the
analysis of rate-dependent experimental data that is not available
at the moment for CORTUF. Consequently, a choice was made to
neglect any intrinsic phenomena in the LDPM-F formulation for
CORTUF.
Fig. 4. a) Random ber orientation, b) CT scan of ber orientation, c) preferential ber orientation, d) validation test results for 3-point bending, e) failure mode for 3-point bending
specimen on CORTUF-Plain, f) failure mode for 3-point bending specimen on CORTUF-Fiber.
20
However, the model is still able to capture the rate-effect associated with apparent mechanisms. Fig. 2a shows the DIF obtained
by simulating a cylindrical (50.8 mm 101.6 mm) CORTUF specimen xed at the base and subjected, at the top, to various load
rates corresponding to nominal strain rates (velocity of the top side
of the specimen divided by the specimen length) from 0.001 s1 to
10 s1. As one can see the response shows an increase of the
macroscopic compressive strength up to 1.24 times the quasi-static
value. This increase results solely from inertia effects and change in
crack patterns (Fig. 2bed). Finally, it is worth noting that the
adopted formulation cannot capture apparent mechanisms occurring at a length scale smaller than the adopted minimum LDPM
particle size (2 mm in this study). However, introduction in the
constitutive equations of these ne scale effects would be only a
futile academic exercise without relevant and appropriate experimental data needed to calibrate and validate the resulting more
complicated model.
Table 2
Tests performed for calibration and values of LDPM parameters in simulations.
Test
performed
LDPM parameter
UHPC
UHPC-L
UHPC-U
RSC
Uniaxial
strain
1. Compressive
strength
2. Initial hardening
modulus ratio
3. Densication
ratio
4. Transitional
strain ratio
5. Shear strength
ratio
6. Initial friction
7. Softening
exponent
8. Deviatoric strain
threshold ratio
9. Deviatoric
damage parameter
10. Asymptotic
friction
11. Transitional
stress
12. Tensile strength
13. Tensile
characteristic length
14. Bond strength
500
MPa
0.36
425
MPa
0.306
575
MPa
0.414
150
MPa
0.4
2.5
2.5
2.5
17
17
17
2.7
0.5
0.2
0.5
0.2
0.5
0.2
0.2
0.2
300
MPa
4
150
300
MPa
3.4
150
300
MPa
4.6
150
600
MPa
4.03
120
11.5
MPa
4 mm
11.5
MPa
4 mm
11.5
MPa
4 mm
8 mm
Uniaxial
unconned
compression
Triaxial
compression
3.5. Model parameter identication

The LDPM-F constitutive equations at the facet level depend on
a number of model parameters (see Appendix A2) that need to be
identied by tting experimental data. Extensive discussion of
LDPM and LDPM-F calibration procedures is reported in
Refs. [86,89].
In this study, rst, the elastic parameters were calibrated by
matching the macroscopic elastic modulus obtained from the
average slope of the CORTUF-Plain and CORTUF-Fiber unconned
compression curves (51,591 MPa) and Poissons ratio as determined
by Ref. [103] having a value of 0.115. By using the macro-to-meso
formulas reported in Ref. [85], one obtains the normal elastic
modulus, EN 67,000 MPa, and the shear-normal coupling
parameter, a 0.484. Experimental data from both CORTUF-Plain
and CORTUF-Fiber were used, because the LDPM-F formulation
neglects the effect of ber reinforcement on the elastic behavior.
This is a reasonable approximation for low-ber-volume fractions
as the ones typically used in concrete (<10%). Next, identication of
the other concrete static parameters was performed by matching
the experimental results of CORTUF-Plain in four different tests,
namely, uniaxial strain (UX), uniaxial unconned compression
(UC), triaxial compression (TXC) with 300 MPa connement, and
3-point bending (3PBT) on notched specimens. Finally, the LDPM-F
parameters related to concreteeber interaction were identied
by the best-t of single pull-out tests relevant to two ber
embedment lengths, 6.35 mm (0.25 in.) and 12.7 mm (0.5 in.).
Table 2 summarizes the parameters specic to each test as well as
the resulting identied values. Model parameters whose calibration require experimental data not available for CORTUF were
assumed on the basis of standard values available in the literature
[86] and reported in Appendix A.
In order to account for the variation that concrete exhibits in
experiments, the meso-structure for each simulated test was
generated three times, and simulations were performed on each
mesh generation. Similarly, bers in the ber-reinforced CORTUF
specimens were generated three times to reproduce the effect of
ber distribution. The average of the numerical results is plotted for
the numerical data discussed hereinafter. Experimental data are
compared with the numerical ones either with average curves,
circles points in the gures, or with curve ranges, shaded gray areas
in the gures, identied between the minimum and maximum
envelopes of the experimental curves. In addition, unless otherwise
specied, the comparison is done on the basis of simulations performed on the coarse-grained model. The numerical simulations
were performed with the Modeling and Analysis of the Response of
Structures (MARS) software [104].
3-Point test
(Notch)
Single ber
pull-out
15. Coarse
aggregate
Tensile strength and toughness parameters of LDPM were

identied by simulating 3PBTs, which were performed on specimens that had a 50.8 mm (2 in.) width, a 25.4 mm (1 in.) depth, a
203.2 mm (8 in.) span, and a notch dimension of 15.24 mm (0.6 in.)
depth and 5.08 mm (0.2 in.) wide. Load versus CMOD (Crack Mouth
Opening Displacement) curve comparison for the experimental and
numerical results are shown in Fig. 3a.
Fig. 3b shows the best tting obtained for the Uniaxial Strain
(UX) test, characterized by zero lateral expansion, as well as Triaxial
Compression (TXC) tests at 0 MPa (Unconned Compression, UC)
and 300 MPa lateral connement (TXC300), respectively. These
tests were used to identify the compression-related LDPM parameters. In Fig. 3b, the stress versus strain curves are offset by 0.05 for
clarity of representation, and the unloading branches of the UX and
TXC300 are included. Compression tests were all conducted on
cored cylindrical specimens having lengths of 101.6 mm (4 in.) and
diameters of 50.8 mm (2 in.) and were loaded through loading
platens treated with friction reducing materials in order to minimize the effect of friction between the platens and the specimen
ends. Additional details of the experimental procedure can be
found in Refs. [98,99]. As can be seen in Fig. 3b, LDPM-F can
accurately reproduce the behaviors of CORTUF at various levels of
connement, i.e., very brittle strain-softening behavior at no
connement and ductile and strain-hardening behavior at high
connement. The model also nicely captures CORTUF unloading
behavior.
Finally, the calibration of LDPM-F requires the identication of
the material parameters governing the berematrix interaction
model. The basic test used for this task is the single ber pull-out
(SPO) test. When a ber is pulled out of concrete, the initial resistance is due to both debonding fracture energy as well as frictional
resistance [105]. Once the ber debonds completely from the
matrix, the resistance shifts to purely frictional, which can be either
slip-hardening, increasing pull-out force for increasing slippage, or
slip-softening, decreasing pull-out force for increasing slippage.
This feature of the pull-out behavior can be simulated through the
coefcient b (see Ref. [88] and Appendix A) that is >0 in the case of
slip-hardening and <0 in the case of slip-softening.
The shaded portions of Fig. 3c represent the data obtained from
experimental tests with two distinct embedment lengths, 12.7 mm
and 6.35 mm (0.5 in. and 0.25 in.), and direction of pulling coinciding with the ber axis. The shorter embedment length shows
slip-hardening, whereas the longer embedment length displays
more of a slip-softening response, and they can only be tted with
two distinct values of the parameter b (see dashed lines in Fig. 3c).
A decision was made to adopt b 0 (see solid lines in Fig. 3c) that
provides an overestimate of the pull-out capacity of the longer
embedment length and an underestimation of the shorter. The
inability of the model to capture the differences in behavior for
different embedment lengths is due to the presence of hooks at
ber ends that are not accounted for explicitly in the analytical
model [91].
However, this limitation is not crucial for the present study,
since only a small portion of the frictional pull-out resistance
affects the behavior of the overall composite, which is actually
governed by other failure mechanisms. One of these mechanisms is
the splitting failure discussed earlier in this paper (see Eq. (6)). In
absence of specic experimental data, the following assumptions
are employed, st st 4 MPa, Le z Lf/2 15 mm, and a z p/6.
split
This leads to an average splitting resistance of Pf
179 N. Note
that the value a z p/6 is motivated by the likely occurrence of
compressive stresses parallel to the ber axis induced by the ber
hooks and leading to a reduction of the inclined crack orientation,
which would be p/4 in the case of pure tangential stresses (note
that similar behavior is observed for shear cracks in prestressed
concrete members).
Other failure mechanisms include ber rupture with inclined
pulling, micro-spalling, and snubbing, whose governing parameters
(reported in Appendix A2) were assumed on the basis of Ref. [88].
3.6. Model validation and discussion
This section presents the validation of the formulated model.
Validation simulations are necessary to prove that LDPM-F serves
its intended purpose of predicting the mechanical responses of
CORTUF-Plain and CORTUF-Fiber. All prediction simulations were
executed with no further change to the previous calibrated LDPM-F
parameters, and their results are compared to experimental data
that were not utilized during the parameter identication.
3.7. Fracture tests
The same notched 3PBT used for the fracture characterization of
CORTUF-Plain was also used for CORTUF-Fiber. The experiments
were intended to be performed with a random distribution of bers
(Fig. 4a) and with a ber volume fraction equal to Vf 3% corresponding to the ber-to-binder ratio in mass equal to 0.31 reported
earlier in Table 1.
Numerical results of the corresponding numerical simulations
showed lower peaks compared to the experimental data as shown
in Fig. 4d. However, computer tomography (CT) scans [100]
revealed Vf to be between 3.5 and 4% in the vicinity of the notch
tip as opposed to the nominal 3%. In addition, the scans provided
insight into the ber placement and distribution; bers were not
distributed randomly (Fig. 4a) as intended but rather had a preferential orientation in the direction orthogonal to the notch (Fig. 4b
and c).
Fig. 4d shows the load versus mid-span-deection curve for
CORTUF-Fiber with random (thin solid curves) and preferential
(thick solid curves) ber orientation for Vf equal to 3, 4, and 5%.
Simulations with random ber orientation are consistently lower
21
than the experimental results with the curve relevant to the highest
volume fraction barely reaching the experimental lower bound. On
the contrary and in agreement with the CT scans, the predictions
with preferential ber orientation coincides with the experimental
results very well. Comparison of the 3PBT CORTUF-Plain result (see
Fig. 3a) with the 3PBT CORTUF-Fiber result (see Fig. 4d) demonstrates that the effect of bers leads to an increase of the carrying
capacity of more than four times, from about 0.9 kN to 3.8 kN (for
4% ber volume fraction). Furthermore, the capability of the
material to dissipate energy increases by orders of magnitude.
Numerical fracture energy values computed through the work-offracture concept [71] give estimates of 30 N/m and 9800 N/m for
CORTUF-Plain and CORTUF-Fiber (Vf 4%), respectively, at 40% loss
of load carrying capacity.
Figs. 4e and f report the crack distributions at 0.1 mm and
5.9 mm deection for the CORTUF-Plain and CORTUF-Fiber,
respectively. Unlike the unreinforced specimen, the ber addition
creates a more distributed crack pattern with a signicant increase
of the fracture process zone (FPZ) size.
The numerical response of CORTUF-Fiber and the associated
failure mechanisms are further investigated in Fig. 5a. In this gure,
the results of the numerical simulations for Vf 4% obtained by
preventing both ber rupture and bond splitting (curves P3 and R3),
and preventing bond splitting only (curves P2 and R2) are compared
with the response accounting for all failure mechanisms (curves P1
and R1). The prex P indicates preferential ber orientation
whereas R indicates random ber orientation. Additionally, the
CORTUF-Plain results are included for comparison as Plain. As one
can see, the dominating failure mechanism is bond splitting. When
this failure mechanism is prevented, the load carrying capacity
more than doubles, i.e., at 3 mm deection, the ratio between the
load curves P3:P1, and R3:R1 is 2.4 and 2.3, respectively. On the
contrary, ber rupture does not have a signicant role as demonstrated by the comparison between P2 and P3 as well as R2 and R3.
Moreover, results show a slightly more important effect of ber
rupture for the random orientation case. This is due to the fact that,
in this case, there is a higher likelihood for bers to bridge cracks at
an angle leading to ber bending with associated localized damage
at the bent location [91,106e109].
The conclusion that ber bond splitting might be the dominant
mechanism is also conrmed by performed CT-scan analyses,
which did not show any signicant occurrence of ber rupture or
even necking and, at the same time, provides evidence of possible
cracks propagating along the ber axis (see Fig. 6a).
Furthermore, post-mortem evaluation of tested specimens
showed the presence of undamaged bers (with even intact hooks
and bents) across the crack surface (see Fig. 6b). This is not
consistent with ber rupture, leading to the separation of the bers
in two pieces, or pull-out, leading to plastic deformation and
straightening of the ber hooks, but it is completely consistent with
the bond splitting mechanism hypothesized in this paper.
For the case with ber preferential orientation, the onset of
bond splitting failure occurs for a load level of about 3.5 kN as
shown in Fig. 5a by the point at which the P1 and P3 curves start to
diverge. To further investigate this phenomenon, crack opening
distribution with cumulative distribution function plots are shown
in Fig. 5b and c for a centerline displacement of 1.2 mm (dotted line
in Fig. 5a), which is about 75% of the maximum stress. Fig. 5b shows
the comparison of the random ber distributions, while the results
of bers with preferential orientation are shown in 5c. The relative
crack distribution of the plain specimen at 75% of its maximum
stress is also included in the plots. At the onset of failure, P1 and R1
exhibit slightly higher crack distributions than P2, P3 and R2, R3,
respectively. As the crack openings become greater than about
100 mm for the random distribution, there is a change in response
22
Fig. 5. 3-Point bending test results for a) stressestrain curve ber failure mechanisms; crack distribution function for b) random ber orientation, c) preferential ber orientation.
as the coupled R2 and R3 are able to bridge more of the larger cracks.
Lastly, comparisons between the plain and the reinforced specimens clearly show the crack-bridging effect of the bers. For the
plain specimens, the cracks governing failure are magnitudes less
than the reinforced specimens. The inability of cracks to propagate
beyond a few mm results in localized failure.
3.8. Tensile strength tests
The tensile strength of the CORTUF-ber was investigated
through four-point bending tests and splitting (Brazilian) tests.
Three different specimen sizes were used for the 4-point (4pt)
testsdSize A, Size B and Size C (see Fig. 7a). The width for all
specimens was 102 mm, and the other dimensions in units of mm
are as follows: Size A (l1 101, l2 305, l3 356, h 25), Size B
(l1 101, l2 305, l3 356, h 102), and Size C (l1 304, l2 914,
l3 1016, h 102). In the 4pt test, a concrete block was supported
on two supports, and a load was applied through a loading plate at
the top of the specimen. To imitate the loading plate used in the
experimental setup, an additional elastic plate was used to apply
load on top the specimen in LDPM. Unlike the cylindrical and
largest 4pt specimens that were cast, ERDC reported that the
Fig. 6. a) CT scan of ber bond failure on 3-point bending specimen, and b) 4-point
bending specimen showing split ber failure.
smaller size 4pt specimens (A and B) to be cut, and this feature was
also reproduced in the LDPM simulations.
Simulations with both random and preferential orientation of
the bers were performed and, similarly to the 3PBT, results show
that the bers were preferentially oriented in the specimen
in the longitudinal direction. Numerical and experimental
nominal stress, sN 3F(l2 l1)/bh2, versus nominal strain,
2
2
3 N 12hD=3l 4l2 l1 , curves are reported in Fig. 7bed, for
2
the specimen sizes A, B, and C, respectively. For the 4% ber volume
fraction, the peak nominal stress was 55.8, 51.7, and 51.7 MPa for
the three different sizes that corresponds to 24.8% and 23% of the
compressive strength. These values are in very good agreement
with the respective experimental values of as shown in Fig. 7bed,
but it must be noted that both experimental and numerical data
provide a signicant overestimation of the actual tensile strength
because, due to the ber crack bridging effect, the specimens
responses show a pronounced nonlinear behavior prior to the peak
suggesting a stable crack propagation after crack initiation and
before reaching the load carrying capacity. The numerical results
and the experimental data are also compared in Fig. 7e as far as the
crack distribution is concerned, showing the excellent ability of
LDPM in reproducing realistic crack patterns.
The cylindrical specimens used for the splitting tests were
characterized by a radius of 51 mm (2 in.) and length of 204 mm
(8 in.). The test was performed according to the ASTM standard, and
a compressive load was applied along the specimens length with
rectangular rods that prevented crushing of the cylinder underneath the applied load. The rod had dimensions of 10 mm width,
226 mm length, and 5 mm height. In a splitting test, the maximum
stress occurs at the center of the specimen in the direction
orthogonal to the applied load and decreases closer to the supports
until eventually stresses at the support become negative
(compression). This leads to a longitudinal crack that splits the
cylinder into two pieces as Fig. 7f demonstrates by comparing postmortem images of both an experiment and a simulation.
Peak stress numerical results obtained for the CORTUF-Fiber
specimens containing Vf values of 3, 4 and 5 were 10.8 MPa,
12.3 MPa, and 13.8 MPa respectively, and the peak stress for the
CORTUF-Plain specimen was 7.2 MPa. The average experimental
splitting stress for CORTUF-Fiber was 25.3 MPa, and for CORTUFPlain was 10.8 MPa. The discrepancy in these values, especially
for CORTUF-Fiber, can be explained by the fact that a compressive
stress eld exists in the direction parallel to the crack path, and
consequently, transverse to the crack bridging ber during splitting
failure. Such transverse connement can signicantly inuence the
pull-out resistance by preventing certain failure mechanisms (e.g.,
splitting and spalling) and by increasing the frictional resistance.
This phenomenon is not currently handled in LDPM-F, because the
bereconcrete interaction model does not depend on the conning
stress acting transversely to the ber axis.
23
Fig. 7. Validation test results for 4-point bending test on specimens. a) Pictorial size description; stressestrain curve for specimen b) size A, c) size B, and d) size C; failure mode
comparison for e) 4-point bending, and f) splitting tensile test for CORTUF-Fiber specimens.
3.9. Uniaxial, triaxial, and hydrostatic compressive tests

Fig. 8 shows comparisons between experimental data and
numerical simulation results with reference to triaxial tests on
CORTUF-Plain and CORTUF-Fiber that were not used in the calibration phase. For completeness, both experimental and numerical
data, previously simulated for the purpose of parameter identication are reported in light gray.
Figs. 8a and d report CORTUF-plain and CORTUF-ber responses, respectively, under uniaxial strain compression and
b)
400
200
Numerical
Experimental
100
0
0
d)
Volumetric Stress (MPa)
700
600
0.02
0.04
0.06
Volumetric Strain
400
300
100
0
0
Numerical
Experimental
0.02
0.04
0.06
Volumetric Strain
400
300
200
100
0
0
e)
500
200
500
0.08
Uniaxial
Hydrostatic
0.08
1000
Axial Stress (MPa)
500
300
c)
600
Principal Stress Difference (MPa)
600
Uniaxial
Hydrostatic
Principal Stress Difference (MPa)
700
Numerical
Experimental
200
400
600
Numerical
100
Experimental
50
20
400
f)
10
0
0.1
0.2
0.3
0.4
300 MPa
1000
500
400
300
Numerical
200
0
0
200
600
0
0
800
600
100
800
300 MPa
200
Axial Stress (MPa)
a)
hydrostatic compression. The results are plotted in terms of volumetric stress versus volumetric strain, and consequently they
feature the same elastic portion of the curve represented by the
volumetric elastic modulus, EV E/(1 2n) 67000 MPa. In the
inelastic regimes, the two curves are distinctly different. Results
show that LDPM-F can capture such different behavior, unlike most
other constitutive equations available in the literature. LDPM-F can
also capture the behavior for unloading very well. It is worth
noting that the difculties of most models to capture the difference
between the volumetric stress versus volumetric strain curves
800
600
Numerical
200
100
Experimental
50
400
200
20
10
0
Experimental
200
400
600
800
0
0
0.1
0.2
0.3
0.4
Fig. 8. Tests results for the validation of LDPM-F for a) uniaxial strain and hydrostatic compression on CORTUF-Plain, b) uniaxial strain on CORTUF-Plain, c) triaxial compression on
CORTUF-Plain, d) uniaxial strain and hydrostatic compression on CORTUF-Fiber, e) uniaxial strain on CORTUF-Fiber, f) triaxial compression on CORTUF-Fiber.
24
obtained under uniaxial strain and hydrostatic compression are

associated with the interpretation that these curves are an
equation of state. As many other experiments have already
shown, this interpretation cannot be used in the case of concrete,
and frictional materials in general.
For the uniaxial strain tests, the numerical versus experimental
comparison is further investigated in Fig. 8b and e with plots of
principal stress difference versus volumetric stress. Once again,
LDPM-F is able to reasonably simulate the CORTUF response during
both the loading and unloading phases of the test.
Simulation of the stress versus strain curves from the TXC tests
on CORTUF-Plain with varying transverse connement (10, 20, 50,
100, 200 MPa) not used in the parameter identication phase is
shown in Fig. 8c where the curves are offset by 0.05 for clarity of
representation. Similarly, Fig. 8f shows the CORTUF-Fiber TXC test
results with varying transverse connements (0, 10, 20, 50, 100,
200, 300 MPa). As one can see, the experimental data suggests a
very small effect of bers on the overall compressive response of
CORTUF. This is consistent with the LDPM-F assumption that bers
are engaged only for cracks displaying tensile normal opening and
that they do not contribute signicantly to the resistance of the
composites under combined meso-scale compression and
shearing.
Furthermore, CORTUF shows a signicant increase of strength
upon connement. For example, for a transverse connement of
100 MPa, the strength for both CORTUF-plain and CORTUF-ber is
in excess of 550 MPa, corresponding to well over two times the
unconned compressive strength. LDPM-F simulates the strength
increase for all levels of connement considered with excellent
accuracy. In addition to strength, the presence of connement
signicantly affects the ductility of the material. The experimental
data shows a transition from a brittle strain-softening response to
a ductile strain-hardening response for increasing connement,
and the transition seems to occur at a level of connement between 100 and 200 MPa. LDPM-F simulates the brittle-to-ductile
transition but predicts the transition connement to be at about
100 MPa. In addition, for 20 MPa and 50 MPa conning pressure,
the numerical responses seem to overestimate the material
ductility, even though this cannot be known with certainty since
the experimental data does not have the post peak portion of the
curve.
The transition from a brittle-to-ductile behavior in compression
is also clearly seen in Fig. 9aed that display crack patterns at failure
with varying connement. For increasing connement, the failure
evolves from a brittle split crack, to one that resembles the occurrence of shear bands, and then a complete, uniformly distributed
crushing that leads to the observed increased ductility.
4. Numerical simulation of projectile penetration into

CORTUF slabs
This section discusses the use of LDPM for the simulation of
penetration and perforation experiments.
In the recent past, various researchers have attempted this type
of simulations by using other computational technologies. The
Finite Element Method with element deletion [110] is the simplest
way to simulate arbitrary propagating cracks and it simulates
cracks by removing elements from the mesh when a certain criterion is met. This method requires a very ne mesh to provide
reasonably accurate results and it cannot be used to simulate the
effect of secondary fragments since most of the failed material
disappears from the simulation.
Meshfree methods, such as the Smoothed Particle Hydrodynamics (SPH) method [111,112], have gained some popularity in the
simulation of penetration mechanics for their ability of allowing
very large deformations. Successful examples of application of SPH
to the simulation of perforation through concrete can be found in
Refs. [113e115]. However, the main problem of classical mesh-free
methods is the underlying assumption that concrete can be treated
as an isotropic continuum even during failure and fragmentation.
Consequently, the transition from continuum to discrete is handled
through numerical artifacts as opposed to an actual loss of continuity of the governing mathematical formulation.
A solution to this problem is to describe displacement discontinuities explicitly. Meshless methods enhanced with the description of displacement discontinuities were used successfully in
Refs. [116,117] to simulate dynamic failure of concrete. However,
this method still neglects the heterogeneous character of concrete
and requires, unlike LDPM, additional degrees of freedom to
represent the evolving discontinuities making the computational
cost to increase signicantly during fracture and fragmentation
compared to the unfractured state.
The simulated perforation experiments were performed at
ERDC on three square 304.8 mm (12 in.) CORTUF-Fiber prismatic
slabs. The slabs, labeled A, B, and C, had widths 50.8 mm (2.0 in.),
63.5 mm (2.5 in.), and 76.2 mm (3.0 in.), respectively. A fragment
simulating penetrator (FSP) projectile was used in the experiments.
Although one of each specimen size was tested experimentally,
LDPM allowed for numerous simulations in a cost-effective and
time-efcient manner for a wide range in velocities below, at, and
above the ballistic limit of each panel, as well as various bervolume fractions (3, 4, and 5%).
The projectile was modeled as a plastic high-strength steel
with an elastic modulus of 200 GPa. The classical Flanagan
Belytschko Single Integration Point nite element algorithm with
Fig. 9. Triaxial compression crack pattern comparison for CORTUF-Plain for a) 0 MPa, b) 10 MPa, c) 50 MPa, and d) 200 MPa connements.
hour-glass stabilization was used for its speed and accuracy with a
hexahedral mesh. Additionally, there was no constraint on the FSP
for either translations or rotations. Fig. 10a shows the side, front,
top, and back views of the FSP meshing. For the large contact
forces exerted on the projectile, greater deformation was observed
for the FSP in the simulation compared to experimental results.
However, due to the caliber having a relatively at nose as seen in
Fig. 10a side and top views, the deformations were not excessive
enough to arrest the simulation. A penalty [104] contact algorithm
was adopted between the FSP and the panel, which utilized the
relative displacement of the two nodes lists, i.e., FSP node list and
panel node list. The dimensionless nodes for the panel and bullet
were assumed to be spherical with 1 mm and 2 mm thicknesses
respectively for the contact, which occurs when the two spherical
external surfaces are in contact. Contact conditions between the
nodes of the same list were omitted in the model.
The schematic of the test setup portrayed the panel in a holder
held only at the outer edges. With no denite knowledge about the
exact boundary conditions of the panel during the experimentation, all surface nodes for the panel width zone were held xed
with no possibility of displacement or rotation, and further investigation of the effects of varied boundary conditions were performed. Results of this boundary investigation are discussed later in
the paper.
The average ballistic limit for the projectile impacting three
different unreinforced specimens of Type A, the smallest panel, was
1706 m/s. When a target is impacted, a compressive shock wave
25
passes through the specimen and is reected at the rear face as a

tensile wave. Ultimately, expulsion of the projectile from the front
face causes spalling. In addition, when the reected tensile stress
wave exceeds the tensile capacity of the material, scabbing occurs
on the rear face [19]. Figs. 10bed display the complete evolution of
the phenomenon for an impact velocity greater than the ballistic
limit. The change in velocity of the FSP with time is presented in
Fig. 10b and c shows the penetration depth of the projectile with
time, and Fig. 10d displays the failure patterns observed for the
panel at 0.005 ms time increments in LDPM for each stage of impact
for the side-panel view section. Figs. 10b and c separate three stages
by vertical dotted lines. The initial decrease in velocity slope is due
to the compressive behavior of the impact, the rst change in
velocity slope, and is the rst stage; the second stage, governed by
compression/shear corresponds to a tunneling effect in the panel;
and nally, the rear-face scabbing coincides with the nal change in
velocity slope, the third stage before the curve plateaus upon projectile exit. Since the width of Panel A is only 50.8 mm (2.0 in.), the
tunneling effect is restricted, and the failure observed has limited
tunneling effect, as opposed to Sizes B and C.
Results depicting the change in velocity with time for impact
velocities ranging from 15 m/s to 3353 m/s, below and above the
ballistic limit, are shown for a CORTUF-Plain specimen A in Fig. 11a
and b, respectively. For impact velocities below the ballistic limit,
there was 100% attenuation of the exit velocity until the ballistic
limit, as shown in Fig. 11a. Fig. 11b shows a decrease in velocity
attenuation for increases in projectile impact velocity. Additionally,
Fig. 10. a) Side, front, top and back view of fsp projectile hexahedron mesh; b) Velocityetime plot, c) perforation depthetime plot, and d) Failure patterns at 0.005 ms time intervals
for size A panel above and fsp above ballistic limit (2743 m/s).
b)3500
1600
15 m/s
30
152
305
610
914
1219
1524
1554
Velocity (m/s)
1200
1000
800
600
400
200
0
0
0.05
d)
0.1
0.15
Time (ms)
3000
Velocity (m/s)
1400
2500
2000
1500
1000
1
0.8
0
0
0.2
0.05
e)
0.1
0.15
Time (ms)
0.4
0.2
0
0
0.2
1000
800
600
400
0.1
0.15
Time (ms)
0.2
1
0.8
0.8
Normalized Energy
1200
0.05
f)
1
PLastic
Elastic
Rigid
1400
KE
IE
IEC
IEB
DE
0.6
500
1600
Velocity (m/s)
c)
1554 m/s
1676
1829
2134
2438
2743
3048
3353
Normalized Energy
a)
Normalized Energy
26
KE
IE
IEC
IEB
DE
0.6
0.4
KE
IE
IEC
IEB
DE
0.6
0.4
0.2
0.2
200
0
0
0.05
0.1
0.15
Time (ms)
0
0
0.2
0.6
0.4
0.2
0
0
Penetration Depth (%)
KE
IE
IEC
IEB
DE
DEC
DEF
0.05
0.1
0.15
Time (ms)
0.2
0.05
0.1
0.15
Time (ms)
0.2
i)
100
0.8
0
0
0.2
h)
1
Normalized Energy
0.1
0.15
Time (ms)
80
60
UHPC
UHPC
UHPC
RSC
UHPC
UHPC
1400
1200
L
U
Exit velocity (m/s)
g)
0.05
F3
F5
40
20
0
0
1000
800
600
UHPC
UHPC
UHPC
RSC
UHPC
UHPC
L
U
F3
F5
400
200
500
1000
1500
Velocity (m/s)
2000
0
600
1200 1800 2400 3000

Strike Velocity (m/s)
3600
Fig. 11. Velocityetime curve for a) velocities below the ballistic limit, b) velocities above the ballistic limit; c) energy change of system with time for CORTUF-Plain at ballistic limit
for plastic fsp, d) velocityetime curve for varying fsp at ballistic limit; energy change of system with time at ballistic limit for e) elastic fsp, and f) rigid fsp g) CORTUF-Fiber (ber
ballistic limit); parametric study plot with varied composites of h) penetration depthevelocity, and i) exitestrike velocity.
the variation in energy (normalized) observed throughout the

experiment for the ballistic limit of CORTUF-Plain, is shown in
Fig. 11c. The kinetic energy (KE), the energy given to the system,
decreases as the velocity of the bullet decreases with time since the
energy of the projectile governs and is directly proportional to the
square of the projectile velocity. The internal energy (IE) in the plots
reference the work done by internal forces, which results in
recoverable elastic energy and dissipated energy. The internal
energy in the concrete (IEC) and the bullet (IEB) combine to form
the total internal energy in the system. The dissipated energy in the
concrete (DE) is the unrecoverable energy; it is less than the internal energy in the concrete.
Further investigation was performed on the FSP considering the
effect of using elastic and rigid projectiles for the numerical simulations. Fig. 11d shows a comparison of velocity change with time
for an impact velocity above the ballistic limit for plastic, elastic,
and rigid projectiles. The results show the disparity between
velocities of the elastic and rigid projectiles, and the plastic projectile that absorbs energy. The deformation of the plastic penetrator is more than that in the elastic and the rigid, which results in
an additional 100 m/s residual velocity for the elastic and rigid
projectiles. These values can be explained by comparing the
energies of the three systems. Figs. 11e and f reveal the energy in
the elastic and rigid systems, respectively. Compared to the plastic

FSP (11c), the internal energy the projectile contributes to the
system for the elastic FSP is reduced by 86.6%, and for the rigid
projectile it decreases to zero. Therefore, the internal energy of the
concrete imparts more or all internal energy to the system than the
projectile. For less energy dissipated by the projectile, the residual
velocity of the FSP is increased, and a lower velocity is needed for
complete perforation. Hence, using a projectile that has a greater
deformation results in greater ballistic limit.
The experimental penetration tests were conducted on reinforced CORTUF; therefore, numerical simulations with bers were
also performed to determine the effects for varied ber volumetric
fractions (VF) of 3%, 4% and 5%. Whereas the average ballistic limit
of the plain concrete was 1706 m/s, that of the ber reinforced
specimen was 1783 m/s. The energyetime plot for an impact velocity on a 3% VF ber-reinforced specimen is shown in Fig. 11g. The
energies show a very similar trend to the unreinforced specimens
Fig. 11c with the clear difference being the energy dissipated in the
ber (DEF). However, most of the total dissipated energy (DE) is still
in the concrete panel as the ber dissipates only about 3% of the
total dissipated energy.
Additionally, a parametric analysis was performed on Specimen
A where three LDPM parameters were altered, tensile strength,
compressive strength, and the initial hardening modulus ratio, by

plus and minus 15% to capture the variations exhibited in the experiments for a composite 15% stronger and weaker than the one
calibrated. This value was selected because it closely captured the
post-peak variation of the 3-point tests, and the other parameters
were altered by the same quantity for consistency. Fig. 11h shows
the effect of the varied LDPM parameters in terms of penetration
depth versus velocity plots for the ultra-high-performance CORTUF
specimen (UHPC), CORTUF 15% stronger (UHPC-U), CORTUF 15%
weaker (UHPC-L), CORTUF with 3% and 5% VF (UHPC-F3, UHPC-F5),
and a regular strength concrete (RSC). A typical RSC specimen was
used with an experimental compressive strength of 38.35 MPa, an
experimental tensile strength of 3.85 MPa, and a coarse aggregate
size of 8 mm. The parameters used for these specimens are given in
Table 2.
Fig. 11h demonstrates the variations in penetration depth for
velocities below the ballistic limit of UHPC-F5 (1798 m/s). For the
various CORTUF compositions, the penetration depth was only
signicantly altered for velocities close to the ballistic limit. The
addition of bers did not signicantly affect the penetration depth
of the specimen as the reinforced CORTUF consistently had a
penetration depth within the scatter of the unreinforced CORTUF
with strength 15% higher and lower than calibrated. Additionally,
the weaker specimen was more affected by the increase in velocity, and the RSC had a greater penetration depth than the
CORTUF. Moreover, a plot of exit versus strike velocities is shown
in Fig. 11i for velocities above the ballistic limit. This plot follows a
similar trend with greatest effect on the different CORTUF composites around the ballistic limit. We can conclude that, according
to these simulations, varying material strength by plus/minus 15%
has only an appreciable effect at about the ballistic limit, and the
addition of bers reduces the exit velocity. However, going from
3% to 5% bers does not incur signicant retardation on the exit
velocity.
To determine which conguration best captured the failure
patterns observed as well as to investigate the sensitivity of the
penetration depth and perforation velocity to the applied constraints, the slabs were modeled using various boundary conditions. Three states were studied for specimen size A.
1. Fixed Edge d only nodes along the edges of the panel were
selected, and they were fully constrained for displacement and
rotation in all directions (see Fig. 12a).
2. Fixed Side d all nodes along the width of the panel sides were
xed to prevent any displacement or rotation (see Fig. 12b). This
is the boundary condition applied to the panel for the previous
simulations performed.
3. Free d no constraint was applied on the panel.
27
Fig. 12c shows the depth of penetration with time plot for an
impact velocity of 1676 m/s. Penetration is rst completed by the
projectile impacting the free panel, whereas it takes the longest for
the xed-side. The xed-edge condition, most similar to the given
boundary conditions, closely follows the xed side. However, for
the time given, the difference in penetration depth of the free- and
the xed-sided panel is only about 5%. The reason the penetration
depth is least in the xed-side panel is due to the connement the
boundary provides for the panel, which cause resistance to the
projectile. In reality, the actual effects on the boundary are nonuniform, and the xed-side and completely free scenarios represent the limits bounding the true values of the boundary
conditions.
The crack pattern for the RSC, UHPC, and UHPC with increasing
ber content were also explored using LDPM-F, and the results are
displayed in Fig. 13. The rst row of results shows the scabbing for
the xed-side boundary, the second row is the scabbing for the
completely free-boundary condition, and the third row shows the
side view for the xed condition. Additionally, the rst column
represents the views for the RSC, the second column shows the
UHPC, and the third and fourth columns show the UHPC 3% and 5%
VF, respectively. The crater size increased for the UHPC in comparison to the RSC since the higher strength material with a smaller
aggregate size has less inclusions to prevent the propagation of the
crack through the specimen. Also, the scabbing effects are greatly
reduced by the addition of bers, but there is not much change as
the VF increases from 3 to 5%. For the free-edged panels, splitting
radial cracks, which propagate from the main damage zone outward as expected, are intensied for the more brittle UHPC and are
reduced by the effect of the bers.
Comparison of the panel failure with the actual experimental
data is shown in Fig. 14. Experimental and numerical results of the
impact face for the RSC is displayed in Fig. 14a and b, respectively.
The red zone of the numerical results represents complete disintegration of the panel as seen in the experimental result. Figs. 14c
and d show the experimental and numerical results for the
UHPC-F3 impact face as well as the FSP penetrator damage. The
crater size is reduced by the addition of the bers in the numerical
prediction and experimental results. However, the FSP penetrator
obtained more damage than observed in the experimental results
by modeling it as plastic. As previously investigated, this additional
damage accrued to the projectile results in an overestimate of the
ballistic limit and a decrease in residual velocity.
This information is also useful in evaluating the disparity
between the numerical and experimental ballistic limit. The numerical ballistic limit of the plain concrete Size A was 1706 m/s,
while the ballistic limit of the ber reinforced specimen, tested by
ERDC, was 1100 m/s and had a percent difference of 47% when
Fig. 12. Pictorial description of boundary conditions on panel for a) xed edge, b) xed side; c) penetration depthetime plots for varied panel boundary conditions.
28
Fig. 13. Comparison of scabbing for xed-side boundary, free boundary, and side crater view of (a) RSC (b) UHPC (c)UHPC-VF3 (d)UHPC-5.
compared to the numerical ballistic limit for UHPC-F3 of 1783 m/s.

Experiments performed on the panel resulted in perforation of the
thinnest specimen, Size A, with 92.1% attenuation of the velocity,
and 100% attenuation for the larger slabs, B and C, for an impact
velocity of 1100 m/s. Based on experimental data on only one
specimen, it is not possible to judge whether or not the difference
between the numerical and experimental values are within the
always-existing statistical scatter of experimental results. Additionally, many other effects contribute to the disparity of experimental and numerical ballistic limits, some of which were
previously discussed; among them are the effects of a plastic
projectile, boundary conditions of the panel, and the appreciable
difference in penetration depth observed for impact velocities close

to the ballistic limit for variations of CORTUF (see Fig. 11h and i). The
effect of coarse graining the panels and modeling a coarse-grain
size greater than 0.6 mm can also be a factor in determining the
ballistic limit.
Further investigation was performed on the CORTUF-Plain
specimen Sizes B and C to determine the effect on increasing
panel depth, and average ballistic limits of 2042 m/s and 2469 m/s
were obtained for these specimens, respectively. Compared to
Specimen A with an average ballistic limit of 1706 m/s, increasing
the depth of the panel by 25% resulted in a 20% increase in ballistic
limit, and a panel with 50% increase in width leads to a 45% increase
Fig. 14. Projectile perforation comparison on impact face of NSC for a) experimental, and b) numerical tests; CORTUF specimens with deformed projectile for c) experimental, and d)
numerical tests.
29
Fig. 15. Parametric analysis of panels A, B, and C showing a) penetration depthevelocity plot b) exitestrike velocity plot; comparison of c) scabbing, and d) side crater view of panel
B for xed-side boundary; comparison of e) scabbing, and f) side crater view of panel C for xed-side boundary.
in ballistic limit. Fig. 15 shows a comparison of the three sizes, A, B,

and C, for penetration depth versus velocity plots. For very low
impact velocities, the penetration depth of each specimen is the
same. However, as the velocities increase to the ballistic limit, the
thinner specimens, are able to perforate with a lower impact
velocity. Additionally, Fig. 15b shows that the exit velocity for the
thinnest specimen is greater than the larger specimens and have a
similar trend as Panel B and C. Figs. 15cef show the failure patterns
observed for Panels B and C at their respective ballistic limits.
5. Conclusions
1. This paper successfully used LDPM-F to simulate the responses
of the novel ultra-high-performance concrete, CORTUF, by rst
identifying the parameters through quasi-static experiments,
single pull-out, uniaxial compression, uniaxial strain, triaxial
compression and 3-point bending tests on CORTUF-Plain.
2. Without any change to previously identied parameters, LDPMF was able to predict the following validation experiments for
both plain and ber-reinforced CORTUF: uniaxial compression,
uniaxial strain, hydrostatic compression, triaxial compression
(connements 10, 20, 50, 100, 200 and 300 MPa), 4-point
bending, and 3-point bending (notched).
3. LDPM-F was also able to simulate vital testing conditions
necessary to accurately reproduce the experimental results
including the ability to capture both random and preferentially
oriented bers and cast or cored specimens.
4. The model remarkably captures the intrinsic transition of
compressive behavior from a very brittle failure for zero or low
connement to a ductile failure for high connement in the
triaxial tests. Additionally, the model has the essential ability to
capture localized cracks and characteristic crack patterns for
plain concrete and a crack-bridging effect through ber reinforcement for the CORTUF-Fiber model.
5. In-depth analysis of the failure mechanisms affecting CORTUF
behavior revealed the importance of accounting for micro-
splitting failure induced by the presence of hooks at ber ends

and the brittleness of the embedding matrix.
6. LDPM-F can be used condently to investigate the behavior of
CORTUF under impact loading, penetration and perforation
events.
7. Simulations of projectile perforation showed the effect of
increasing ber content in the target specimens. Fibers added to
the concrete reduce brittleness and hence the area of the crater,
but it has only a secondary effect on reducing the exit velocity
associated with strike velocities above the ballistic limit. Also, an
increase in ber content in excess of 3% does not seem to produce an appreciable improvement on the panel performance.
8. The reduced damaged area observed in CORTUF-Fiber compared
to CORTUF-Plain is an important feature associated with postimpact residual strength of CORTUF panels and their performance under multiple hit scenarios.
Acknowledgment
This effort was sponsored by the U.S. Army Engineer Research
and Development Center. Permission to publish was granted by the
Director, Geotechnical and Structures Laboratory. The work of the
rst and second authors was also supported under DTRA grant
No. HDTRA1-09-1-0029 to Rensselaer Polytechnic Institute (with
subcontracts to Northwestern University and Sandia National Labs).
A. LDPM-F Constitutive Equations
A.1. LDPM Constitutive Equations
Normal and shear stresses in the elastic regime of LDPM are:
sN EN3 N; sM ET3 M; sL ET3 L, where EN E0, effective normal

modulus, ET aE0, and a shear-normal coupling parameter.
For stresses and strains beyond the elastic limit, LDPM mesoscale failure is characterized by three mechanisms as described
below.
30
Fracture and cohesion due to tension and tension-shear. For

tensile loading (3 N > 0), the fracturing behavior is formulated
q
2
2 3 2 , and stress,
3 a3
through an effective strain, 3
N
M
L
q
s s2N sM sL 2 =a, which dene the normal and shear
stresses as sN 3 N(s/3 ); sM a3 M(s/3 ); sL a3 L(s/3 ). The effective

stress s is incrementally elastic s_ E0 3_ and must satisfy the
inequality 0ssbt(3 ,u) where sbt s0(u)exp[H0(u)h3 3 0(u)i/
p
s0(u)], hxi max{x,0}, and tanu 3 N =a3 T sN a=sT . The post
peak softening modulus is dened as H0 u Ht 2u=pnt , where
Ht is the softening modulus in pure tension (u p/2) expressed as
Ht 2E0/(lt/le 1); lt 2E0 Gt =s2t ; le is the length of the tetrahedron edge; and Gt is the meso-scale fracture energy. LDPM provides
a smooth transition between pure tension and pure shear
(u 0) with parabolic variation for strength given by
q
s0 u st rst2 sinu sin2 u 4acos2 u=rst2 =2acos2 u,
where rst ss/st is the ratio of shear strength to tensile strength.

Compaction and pore collapse from compression. Normal stresses
for compressive loading (3 N < 0) are computed through the
inequality sbc(3 D,3 V)sN0, where sbc is a strain-dependent
boundary function of the volumetric strain, 3 V, and the deviatoric
strain, 3 D. Beyond the elastic limit, sbc models pore collapse as a
linear evolution of stress for increasing volumetric strain with
stiffness Hc for 3 V3 c1 kc03 c0 : sbc sc0 h3 V 3 c0iHc(rDV);
Hc(rDV) Hc0/(1 kc2hrDV kc1i); sc0 is the meso-scale compressive
yield stress; and kc1, kc2 are material parameters. Compaction and
rehardening occur beyond pore collapse (3 V3 c1). In this case one
has sbc sc1(rDV) exp[(3 V 3 c1)Hc(rDV)/sc1(rDV)] and
sc1(rDV) sc0 (3 c1 3 c0)Hc(rDV).
Friction due to compression-shear. The incremental shear stresses
p
p
are computed as s_ M ET _3 M 3_ M and s_ L ET _3 L 3_ L , where
p
p
_
_
3_
lv4=vs , 3_ lv4=vs , and l is the plastic multiplier with
M
loading-unloading conditions 4l_ 0 and l_ 0. The plastic potential

q
s2M s2L sbs sN , where the nonlinear fricis dened as 4
tional law for the shear strength is assumed to be

sbs ss (m0 mN)sN0[1 exp(sN/sN0)] mNsN; sN0 is the transitional normal stress; m0 and mN are the initial and nal internal
friction coefcients.
A.2. FibereMatrix Interaction Behavior
The pull-out resistance of the ber is related to the load, P and
the relative displacement/slippage, v, between concrete and ber.
Full debonding is represented by a critical slippage value
vd 2s0 L2e =Ef df 8Gd L2e =Ef df 1=2 where Le is a generic embedment length, Gd is the bond fracture energy, and s0 is a constant
frictional
stress.
When
v
<
vd
(debonding),
Pv p2 Ef d3f s0 v Gd =21=2 . After full debonding (v > vd),
P(v) P0[1 (v vd/Le)][1 (b(v vd)/df)] where P0 pLedfs0 and
b 0;>0;<0 if the interfacial friction is independent, increases,
decreases with slippage, respectively.
When the orientation of the embedded segment and the crack
bridging segment of ber are different, localized fracture and fragmentation can occur reducing the embedded ber length by a spalling length, sf PfNsin(q/2)/kspstdfcos2(q/2), where Pf is the crackbridging force, q is the angle between the normals to the embedded
segment and the crack face, st is the meso-scale tensile strength, 40f is
the change in ber orientation, and ksp is a material parameter.
The snubbing model adopted in LDPM-F assumes that the ber,
in a perfectly exible manner, wraps around the exit of the tunnel
crack that was shortened by spalling. The change in the tensile load
for the exible tendon is expressed as Pf expksn 40f Pv where P
is the summation of all slip-friction and debonding forces parallel

to the embedment length, and ksn is the snubbing parameter. For a
Pf leading to ber rupture, there will be no contribution from the
bers to structural strength and ductility. Single ber pull-out tests
on polyvinyl alcohol (PVA) revealed lower rupture loads for
increasing 40f . This strength reduction (due to bending) is adopted
0
in LDPM-F, and expressed as sf suf ekrup 4f where sf is the axial
stress in the crack-bridging segment, krup is a material parameter
and suf is the ultimate tensile strength of the ber.
A detailed description of the LDPM idealization of concrete
meso-structure can be found in Cusatis et al. [85] and more information on the ber stiffness and unloading adopted in LDPM-F is
provided in Ref. [88].
References
[1] Yazici H, Yardimici MY, Aydin S, Karabulut A. Mechanical properties of
reactive powder concrete containing mineral admixtures under different
curing regimes. Constr Build Mater 2009;23(3):1223e31.
[2] Ming-Gin L, Yung-Chih W, Chui-Te C. A preliminary study of reactive powder
concrete as a new repair material. Constr Build Mater 2007;21(1):182e9.
[3] Zhang M, Shim V, Lu G, Chew C. Resistance of high-strength concrete to
projectile impact. Int J Impact Eng 2005;31(7):825e41.
[4] Wittmann FH. Crack formation and fracture energy of normal and high
strength concrete. Sadhana 2002;27(4):413e23.
[5] Habel K, Viviani M, Denari E, Bruhwiler E. Development of the mechanical
properties of an ultra-high performance ber reinforced concrete (UHPFRC).
Cem Concr Res 2006;36(7):1362e70.
[6] Meddah MS, Zitouni S, Belabes S. Effect of content and particle size distribution of coarse aggregate on the compressive strength of concrete. Constr
Build Mater Apr. 2010;24:505e12.
[7] Corinaldesi V, Moriconi G. Mechanical and thermal evaluation of ultra high
performance ber reinforced concretes for engineering applications. Constr
Build Mater 2012;26(1):289e94.
[8] Sajedi F, Razak HA. Effects of thermal and mechanical activation methods on
compressive strength of ordinary Portland cement slag mortar. Mater Des
Feb. 2011;32:984e95.
[9] Sajedi F. Effect of curing regime and temperature on the compressive
strength of cement-slag mortars. Constr Build Mater Nov. 2012;36:549e56.
[10] Razak HA, Sajedi F. The effect of heat treatment on the compressive strength
of cement-slag mortars. Mater Des Sept. 2011;32:4618e28.
[11] Kang S-T, Lee Y, Park Y-D, Kim J-K. Tensile fracture properties of an ultra high
performance ber reinforced concrete (UHPFRC) with steel ber. Compos
Struct 2010;92:61e71.
[12] Zhang MH, Sharif MSH, Lu G. Impact resistance of high-strength bre-reinforced concrete. Mag Concr Res 2007;9831(3):199e210.
[13] Naaman AE, Reinhardt HW. Strain hardening and deection hardening ber
reinforced cement composites HPFRCC-4. Mater Struct 2003;30:757e9.
[14 Dancygier AN, Savir Z. Flexural behavior of HSFRC with low reinforcement
ratios. Eng Struct 2006;28:1503e12.
[15] Liang NH, Liu XR, Sun J. Experimental study of splitting tensile for multi-scale
polypropylene ber concrete. Adv Mater Res Jan 2012;450e451:168e73.
[16] Yi N-H, Kim J-HJ, Han T-S, Cho Y-G, Lee JH. Blast-resistant characteristics of
ultra-high strength concrete and reactive powder concrete. Constr Build
Mater 2012;28(1):694e707.
[17] Corbett GG, Reid ISR, Johnson W. Impact loading of plates and shells by freeying projectiles: a review. Int J Impact Eng 1996;18(2):141e230.
[18] Li QM, Tong DJ. Perforation thickness and ballistic limit of concrete target
subjected to rigid projectile impact. J Eng Mech 2003;0(September):
1083e92.
[19] Tai Y. Flat ended projectile penetrating ultra-high strength concrete plate
target. Theor Appl Fract Mech Apr. 2009;51:117e28.
[20] Nystrm U, Gylltoft K. Comparative numerical studies of projectile impacts on plain and steel-bre reinforced concrete. Int J Impact Eng
2011;38:95e105.
[21] Dancygier AN, Yankelevsky DZ. High strength concrete response to hard
projectile impact. Int J Impact Eng 1996;18(6):583e99.
[22] Neil EFO, Neeley BD, Cargile JD. Tensile properties of very-high-strength
concrete for penetration-resistant structures. Shock Vib 1999;6:237e45.
[23] Mainstone R. Properties of materials at high rates of straining or loading.
Matr Constr 1975;8(2):102e16.
[24] Dilger WH, Koch R, Kowalczyk R. Ductility of plain and conned concrete
under different strain rates. J Am Concrete Inst 1984;81(1):73e81.
[25] Bischoff PH, Perry SH. Compressive behaviour of concrete at high strain rates.
Mater Struct 1991;24:425e50.
[27] Watstein D. Effect of straining rate on the compressive strength and elastic
properties of concrete. ACI 1953;49(4):729e44.
[28] Hughes BP, Gregory R. Concrete subjected to high rates of loading in
compression. Mag Contr Res 1972;24(78):25e36.

[29] Malvern LE, Tang T, Jenkins DA, Gong JC. Dynamic compressive strength of
cementitious materials. Mater Res Soc 1985;64(1):119e38.
[30] Abrams DA. Effect of rate application of load on the compressive strength of
concrete. ASTM 1917;17(2):364e77.
[31] Lundeen RL. Dynamic and static tests of plain concrete specimens: report 1.
Tech. rep. Vicksburg, Mississippi: US Army Engineers Waterways Experiment
Station; 1963. p. 1e38
[32] Mechtcherine V, Millon O, Butler M, Thoma K. Mechanical behavior of SHCC
under impact loading. In: Parra-Montesinos G, Reinhardt H, Naaman A, editors. High performance ber reinforced cement composites 6. Vol. 2 of
RILEM state of the art reports. Netherlands: Springer; 2012. p. 297e304.
[33] Wang S, Zhang M-H, Quek ST. Effect of high strain rate loading on
compressive behaviour of bre-reinforced high-strength concrete. Mag
Concr Res 2011;63(11):813e27.
[34] Wang Y, Wang Z, Liang X, An M. Experimental and numerical studies on
dynamic compressive behavior of reactive powder concretes. Acta Mech Sol
Sin 2008;21(5):420e30.
[35] Kim DJ, El-Tawil S, Naaman AE. Rate-dependent tensile behavior of high
performance ber reinforced cementitious composites. Mater Struct May
2008;42:399e414.
[36] Caverzan A, Cadoni E, di Prisco M. Dynamic tensile behaviour of high performance bre reinforced cementitious composites after high temperature
exposure. Mech Mater Apr. 2013;59:87e109.
[37] Cusatis G. Strain-rate effects on concrete behavior. Int J Impact Eng Apr.
2011;38:162e70.
[38] O
zbolt J, Rah KK, Mestrovi&cacute D. Inuence of loading rate on concrete
cone failure. Int J Fract May 2006;139:239e52.
[39] O
zbolt J, Sharma A. Numerical simulation of dynamic fracture of concrete through uniaxial tension and l-specimen. Eng Fract Mech 2012;85:
88e102.
[40] Powers TC. Mechanism of shrinkage and reversible creep of hardened
cement paste. Cem Concr Assoc 1965:319e44.
[41] Powers TC. The thermodynamics of volume change and creep. Mat Struct
1968;1(6):487e507.
[42] Ba
zant Z, Cusatis G, Cedolin L. Temperature effect on concrete creep modeled
by microprestress-solidication theory. J Eng Mech June 2004;130:691e9.
[43] Wittmann FH. Interaction of hardened cement paste and water. J Am Ceram
Soc 1973;56(8):409e15.
[44] Ba
zant ZP, Gettu R. Rate effects and load relaxation in the static fracture of
concrete. ACI Mater J 1992;89(5):456e8.
[45] Ba
zant Z, Caner F, Adley M, Akers S. Fracturing rate effect and creep in
microplane model for dynamics. J Eng Mech 2000;126(9):962e70.
[46] Cusatis G, Ba
zant Z, Cedolin L. 3D Lattice model for dynamic simulations of
creep, fracturing and rate effect in concrete. In: Ulm FJ, Bazant ZP,
Wittmann FH, editors. Creep, shrinkage and durability mechanics of
concrete and other quasi-brittle materials. Proceedings of the 6th international conference CONCREEP-6. MIT, Cambridge, MA. USA: Elsevier; Aug
2001. p. 113e8.
[47] Forquin P, Safa K, Gary G. Inuence of free water on the quasi-static and
dynamic strength of concrete in conned compression tests. Cem Concr Res
Feb. 2010;40:321e33.
[48] Zheng D, Li Q. An explanation for rate effect of concrete strength based on
fracture toughness including free water viscosity. Eng Fract Mech Nov.
2004;71:2319e27.
[49] Cottrell A. The mechanical properties of matterIn Wiley series on the science
and technology of materials. NY: Krieger Publishing Company; 1981. p. 300e
430.
[50] Vegt I, Weerheijm J. Inuence of moisture content on the dynamic behaviour
of concrete. In: Gdoutos EE, editor. Fracture of nano and engineering materials and structures, (Greece). Springer Netherlands. July 2006. p. 501e2.
[51] Cadoni E, Labibes K, Albertini C, Berra M, Giangrasso M. Strain-rate effect on
the tensile behaviour of concrete at different relative humidity levels
2001;34(February):21e6.
k P, Caner Ferhun C. Comminution of solids caused by kinetic
[52] Ba
zant Zdene
energy of high shear strain rate, with implications for impact, shock, and
shale fracturing, PNAS; 2013 [Epub ahead of print November 11, 2013],
http://dx.doi.org/10.1073/pnas.1318739110.
[53] Cadoni E, Solomos G, Albertini C. Concrete behaviour in direct tension tests
at high strain rates. Mag Concr Res June 2013;65:660e72.
[54] O
zbolt J, Bosnjak J, Sola E. Dynamic fracture of concrete compact tension
specimen: experimental and numerical study. Int J Sol Struct December
2013;50(25e26):4270e8.
[55] Ramulu M, Kobayashi AS. Mechanics of crack curving and branchingda
dynamic fracture analysis. Int J Fract 1985;27:187e201.
[56] Barenblatt GI, Salganik RL, Cherepanov GP. On the nonsteady propagation of
cracks. PMM 1962;26(2):328e34.
[57] Hillerborg A, Moder M, Petersson PE. Analysis of crack formation and crack
growth in concrete by means of fracture mechanics and nite elements. Cem
Concr Res 1976;6.
[58] Rashid Y. Ultimate strength analysis of prestressed concrete pressure vessels.
Nucl Eng Des 1968;7(4):334e44.
[59] Ba
zant ZP, Oh BH. Crack band theory for fracture of concrete. Matr Constr
May 1983;16:155e77.
[60] Bicanic N, Pearce CJ. Computational aspects of a softening plasticity model
for plain concrete. Mech Cohesive-frictional Mater 1996;1(1):75e94.
31
[61] Comi C, Perego U. Fracture energy based bi-dissipative damage model for
concrete. Int J Sol Struct 2001;38(3637):6427e54.
[62] Grassl P, Jirsek M. Damage-plastic model for concrete failure. Int J Sol Struct
2006;43:7166e96.
[63] Mazars J, Pijaudier-Cabot G. Continuum damage theory e application to
concrete. J Eng Mech 1989;115(2):345e65.
[64] Ba
zant ZP, Caner FC, Carol I, Adley MD, Akers AA. Microplane model M4 for
concrete. I: formulation with work-conjugate deviatoric stress. J Eng Mech
2000;126(9).
[65] Di Luzio G, Cusatis G. Solidication-microprestress-microplane (smm) theory
for concrete at early age. Theory, validation and application. Int J Sol Struct
march 2013;50:957e75.
[66] Caner F, Ba
zant Z. Microplane model M7 for plain concrete: I. Formulation.
J Eng Mech 2012:1e32. http://dx.doi.org/10.1061/(ASCE)EM.1943-7889.
0000570.
[67] O
zbolt J, Li Y, Kozar I. Microplane model for concrete with relaxed kinematic
constraint. Int J Sol Struct 2001;38(16):2683e711.
[68] Di Luzio G. A symmetric over-nonlocal microplane model m4 for fracture in
concrete. Int J Sol Struct 2007;44(13).
[69] Cedolin L, Ba
zant ZP. Effect of nite element choice in blunt crack band
analysis. Comput Methods Appl Mech Eng 1980;24:305e16.
[70] De Borst R, Sluys LJ, Muhlhaus HB, Pamin J. Fundamental issues in nite
element analyses of localization of deformation. Eng Comput 1993;10:
99e122.
[71] Ba
zant ZP, Planas J. Fracture and size effect in concrete and other quasibrittle
materials. Boca Raton: CRC Press LLC; 1997. p. 489e560.
[72] Ba
zant ZP, Jirsek M. Nonlocal integral formulations of plasticity and damage: survey of progress. J Eng Mech 2002;128(11):1119.
[73] Walsh P. Fracture of plain concrete. Indian Concr J 1972;46.
[74] Ba
zant ZP, Kazemi MT. Determination of fracture energy, process zone length
and brittleness number from size effect, with application to rock and concrete. Int J Fract 1990;44.
[75] Ba
zant ZP, O
zbolt J. Nonlocal microplane model for fracture, damage, and
size effect in structures. J Eng Mech 1990;116(11):2485e505.
[76] Leong HP, Somsak S. Over-nonlocal gradient enhanced plastic-damage
model for concrete. Int J Sol Struct Dec 2009;46:4369e78.
[77] Herrmann H, Hansen A, Roux S. Fracture of disordered, elastic lattices in two
dimensions. Phys Rev B 1989;39(1):637e47.
[78] Ba
zant ZP, Tabbara MR, Kazemi MT, Pijaudier-Cabot G. Random particle
model for fracture of aggregate or ber composites. J Eng Mech 1990;116(8):
1686.
[79] Schlangen E, Van Mier J. Experimental and numerical analysis of micromechanisms of fracture of cement-based composites. Cem Concr Compos
1992;14(2):105e18.
[80] Cusatis G, Ba
zant ZP, Cedolin L. Connement shear lattice model for concrete
damage in tension and compression. I: Theory. J Eng Mech 2003;129(12):
1439e48.
[81] Cusatis G, Ba
zant ZP, Cedolin L. Connement-shear lattice model for concrete
damage in tension and compression. II: numerical implementation and
validation. J Eng Mech 2003;129(12):1449e58.
[82] Gianluca C, Luigi C. Two-scale study of concrete fracturing behavior. Eng
Fract Mech 2007;74(1e2):3e17. Fracture of Concrete Materials and
Structures.
[83] Elis J, Stang H. Lattice modeling of aggregate interlocking in concrete. Int J
Fract Mar. 2012;175:1e11.
[84] Cusatis G, Nakamura H. Discrete modeling of concrete materials and structures. Cem Concr Compos 2011;33(9):865e6.
[85] Cusatis G, Pelessone D, Mencarelli A. Lattice discrete particle model (LDPM)
for failure behavior of concrete. I: Theory. Cem Concr Compos 2011;33(9):
881e90.
[86] Cusatis G, Mencarelli A, Pelessone D, Baylot J. Lattice discrete particle model
(LDPM) for failure behavior of concrete. II: Calibration and validation. Cem
Concr Compos 2011;33(9):891e905.
[87] Alnaggar M, Cusatis G, Di Luzio G. Lattice discrete particle modeling of alkalisilica-reaction (ASR) deterioration of concrete structures. J Eng Mech
2013;41:45e59.
[88] Schauffert EA, Cusatis G. Lattice discrete particle model for ber reinforced
concrete (LDPM-F): I. Theory. ASCE J Eng Mech July 2012;138:826e33.
[89] Schauffert EA, Cusatis G, Pelessone D, ODaniel JL, Baylot JT. Lattice
discrete particle model for ber reinforced concrete (LDPM-F): II. Tensile
fracture and multiaxial loading behavior. ASCE J Eng Mech 2012;138(7):
834e41.
[90] Cusatis G, Schauffert EA. Discontinuous cell method (DCM) for cohesive
fracture propagation. In: Oh BH, editor. 7th international conference on
fracture mechanics of concrete and concrete structures (FraMCos 7), (Jeju,
South Korea). Korea Concrete Institute. May 2010. p. 23e8.
[91] Yang E-H, Wang S, Yang Y, Li VC. Fiber-bridging constitutive law of
engineered cementitious composites. J Adv Concr Technol 2008;6(1):
181e93.
[92] Cusatis G, Mencarelli A, Pelessone D, Baylot J. Lattice discrete particle model (
LDPM ) for fracture dynamics and rate effect in concrete. ASCE 2008;315(42):
1e11.
[93] Mencareli A. Numerical simulation of the effect of blast and penetration on
reinforced concrete structures [PhD thesis]. Troy, NY: Rensselaer Polytechnic
Institute; December 2010. p. 1e147.
32
[94] Cusatis G. The lattice discrete particle model (LDPM) for the numerical
simulation of concrete behavior subject to penetration. Hoboken, NJ, USA:
John Wiley and Sons, Inc; 2013. p. 369e87.
[95] Cusatis G, Diluzio G, Cedolin L. Meso-scale simulation of concrete: blast
and penetration effects and AAR degradation. In: Cadoni E, di Prisco M,
editors. Performance, protection and strengthening of structures under
extreme loadingdapplied mechanics and materials. Switzerland: Trans
Tech Publications; 2011. p. 75e80.
[98] Williams EM, Graham SS, Reed PA, Rushing TS. Laboratory characterization
of cortuf concrete with and without steel bers. Technical Report. Vicksburg,
MS: US Army Corps of Engineers, Engineer Research and Development
Center; May 2009. p. 1e83.
[99] Roth MJ, Rushing TS, Flores OG, Sham DK, Stevens JW. Laboratory investigation of the characterization of cortuf exural and splitting tensile properties. Technical Report. Vicksburg, MS: US Army Corps of Engineers,
Engineer Research and Development Center; Sep 2009. p. 1e57.
[100] Landis EN, Keane DT. X-ray microtomography. Mater Charact 2010;61(12):
1305e16.
[101] Alnaggar M, Cusatis G. Automatic parameter identication of discrete
mesoscale models with application to the coarse-grained simulation of
reinforced concrete structures. In: Carrato J, Burns J, editors. 2012 structures
congress, (Chicago, IL). ASCE. 2012. p. 406e17.
[102] Maeda M, Otani S, Aoyama H. Effect of connement on bond splitting
behavior in reinforced concrete beams. IABSE, Struct Eng Int 1995;5(3):
166e71.
[103] Persson B. Poissons ratio of high-performance concrete. Cem Concr Res Oct.
1999;29:1647e53.
[104] Pelessone D. MARS, modeling and analysis of the response of structures d users
manual. San Diego, CA: ES3; 2011. p. 1e237. http://es3inc.com/marssolver/.
[105] Mumm DR, Faber KT. Interfacial debonding and sliding in brittle-matrix
composites measured using an improved ber pull out technique. Acta
Metall Mater 1995;43(3):1259e70.
[106] Li V, Wang Y, Backer S. Effect of inclining angle, bundling and surface

treatment on synthetic bre pull-out from a cement matrix. Composites Mar.
1990;21:132e40.
[107] Lee Y, Kang S-T, Kim J-K. Pullout behavior of inclined steel ber in an ultrahigh strength cementitious matrix. Constr Build Mater Oct. 2010;24:
2030e41.
[108] Laranjeira F, Molins C, Aguado a. Predicting the pullout response of inclined
hooked steel bers. Cem Concr Res Oct. 2010;40:1471e87.
[109] Kang S-T, Kim J-K. The relation between ber orientation and tensile
behavior in an ultra high performance ber reinforced cementitious composites (UHPFRCC). Cem Concr Res Oct. 2011;41:1001e14.
[110] Hibbit, Karlsson, Sorensen. ABAQUS/standard analysis users manual. USA:
Hibbit, Karlsson, Sorensen Inc; 2007. p. 1e1051.
[111] Lucy L. A numerical approach to the testing of ssion hypothesis. Astron J
1977;82:1013e24.
[112] Gingold RA, Monaghan JJ. Smoothed particle hydro-dynamics: theory and
applications to non-spherical stars. Mon Not R Astron Soc Nov 1977;181:
375e89.
[113] Rabczuk T, Eibl J. Modelling dynamic failure of concrete with meshfree
methods. Int J Impact Eng Nov. 2006;32:1878e97.
[114] Rabczuk T, Xiao SP, Sauer M. Coupling of meshfree methods with
nite elements: basic concepts and test results. Commun Numer Methods
Eng. 2006. 1031e65. John Wiley & Sons, Ltd.
[115] Rabczuk T, Eibl J. Simulation of high velocity concrete fragmentation using
SPH/MLSPH. Int J Numer Methods Eng Mar. 2003;56:1421e44.
[116] Rabczuk T, Belytschko T. A three-dimensional large deformation meshfree
method for arbitrary evolving cracks. Comput Methods Appl Mech Eng May
2007;196:2777e99.
[117] Rabczuk T, Belytschko T. Cracking particles: a simplied meshfree method
for arbitrary evolving cracks. Int J Numer Methods Eng 2004;61(13):
2316e43.

Discrete Modeling of Ultra High Performance Concrete With Application To Projectile Penetration - 2014 - International Journal of Impact Engineering PDF

Transféré par

Informations du document

Description originale:

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

Discrete Modeling of Ultra High Performance Concrete With Application To Projectile Penetration - 2014 - International Journal of Impact Engineering PDF

Transféré par

Droits d'auteur :

Formats disponibles

International Journal of Impact Engineering 65 (2014) 13e32

Contents lists available at ScienceDirect

International Journal of Impact Engineering

Discrete modeling of ultra-high-performance concrete

* Corresponding author. McCormick School of Engineering and Applied Science,

uses including its adoption in high-rise structures, long-span

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

Although there are many methods focused on increasing the

between intrinsic phenomena, which should be included in the

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

signicant build up of pore pressure. Forquin and coworkers [47]

through the formulation of non-local models with either gradient

where [ is the length of the straight line connecting two adjacent

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

CDF for Edge Length

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

average composition is summarized in Table 1. The adopted sand

used for the CORTUF characterization in compression, one would

where [ interparticle distance, which is locally variable according

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

are explicitly modeled by generating spatial distributions

bers after the spatially random direction is computed. The dot

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

Axial Stress (MPa)

Due to the specic curing protocol used to manufacture

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

3.5. Model parameter identication

Tensile strength and toughness parameters of LDPM were

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

3.9. Uniaxial, triaxial, and hydrostatic compressive tests

Principal Stress Difference (MPa)

Principal Stress Difference (MPa)

Volumetric Stress (MPa)

Axial Stress (MPa)

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

obtained under uniaxial strain and hydrostatic compression are

4. Numerical simulation of projectile penetration into

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

passes through the specimen and is reected at the rear face as a

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

Penetration Depth (%)

Exit velocity (m/s)

1200 1800 2400 3000

the variation in energy (normalized) observed throughout the

the elastic and rigid systems, respectively. Compared to the plastic

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

compressive strength, and the initial hardening modulus ratio, by

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

compared to the numerical ballistic limit for UHPC-F3 of 1783 m/s.

difference in penetration depth observed for impact velocities close

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

in ballistic limit. Fig. 15 shows a comparison of the three sizes, A, B,

splitting failure induced by the presence of hooks at ber ends

sN EN3 N; sM ET3 M; sL ET3 L, where EN E0, effective normal

J. Smith et al. / International Journal of Impact Engineering 65 (2014) 13e32

Fracture and cohesion due to tension and tension-shear. For

stresses as sN 3 N(s/3 ); sM a3 M(s/3 ); sL a3 L(s/3 ). The effective

where rst ss/st is the ratio of shear strength to tensile strength.

loading-unloading conditions 4l_  0 and l_  0. The plastic potential

tional law for the shear strength is assumed to be

loading-unloading conditions 4l_ 0 and l_ 0. The plastic potential