Vous êtes sur la page 1sur 7

Proceedings of the Forty Second National Conference on Fluid Mechanics and Fluid Power

December 14-16, 2015, NITK Surathkal, Karnataka, India

FMFP2015PAPER NO.

CFD ANALYSIS OF LIBR+H2O SOLUTION FALLING FILM OVER INLINE


HORIZONTAL TUBES
.
P S. Gunavant , S. S. Gajghate, V. M. Jamadar, G.S. Jadhav
Department of Mechanical Engineering, Dr, Daulatrao Aher College of Engineering,
Karad 415124, Maharashtra (India)
*
Corresponding author (email: pradipgunavant@gmail.com)
*

the form of graphs and flow pattern images for


various geometrical and flow conditions. The
numerical data investigated is compared with data
reported in the literature.

ABSTRACT
The objective of this study is that to
investigate the behaviour of LiBr-H2O solution
falling film over inline horizontal circular tubes by
using commercial Computational Fluid Dynamics
(CFD) code (Ansys-Fluent 13.0.0). Simulations
over horizontal tubes are performed considering
3D configurations. 3D modeling and grid
generation is carried out by using available
commercial software (Ansys-ICEM CFD 13.0.0).
Flow domain is prepared by using all hexahedral
elements. The performance of falling film heat
exchangers is depends upon the behavior of liquid
film between the consecutive tubes. The behaviors
of liquid film like mode transition, film thickness,
departure site spacing is analyzed by varying Film
Reynolds number and different geometrical
parameters i.e. tube diameter and tube spacing.
The Pressure-based solver with Volume-of-fluid
(VOF) multiphase model is used to capture liquidair interface. The results obtained are discussed in

Keywords: CFD, falling film, mode transition,


film thickness
1. INTRODUCTION
In the last decade ozone depletion, global
warming and its effects on the environment have
been of great interest due to environmental issues.
The emission of greenhouse gases (GHG) have
grown over the last decades due to human
activities, thereby increasing the earths average
temperature leading to global warming. The
influence of common refrigerants used in
compression cycles cannot be ignored due to their
very high global warming potential, e.g. R134a
has a GWP of 1300 (IPCC, 2007). Thus,
introduction
of
non
chlorine-containing
refrigerants and the development of new heat

transfer concepts are necessary to achieve the


goals of reduced energy consumption and
environmental impact. . Absorption systems are
such
an
alternative
because
of
their
environmentally friendly working pairs as NH3 /
H2O and H2O / LiBr.

Pise et al. [2] studied in detail the flow transitions


for inline tube arrangement using CFD for 2D
model, and recommend flow transition
correlations. Mitrovic [3][4], experimentally
pointed out that the transition of flow modes was
dependent on the flow rate, liquid thermo-physical
properties and the tube spacing. Armbruster and
Mitrovic [4][5][6], modelled the mode transition
among the droplet, jet and sheet modes, according
to Ref = a Gab, where a is an empirical constant.
They also stated that b = 0.25 found to be
acceptable.

Applied to a refrigeration system, the falling


film evaporator presents several advantages
compared to a shell and tube evaporator,
particularly in terms of higher cycle efficiency,
reduced costs and a smaller environmental impact
from its reduced charge of refrigerant. The
pressure drop is small as the liquid flows only by
gravity, which may imply the use of a recirculation
pump to bring the liquid from the bottom to the
top of evaporator.

Hu and Jacobi [7][8], suggested the flow


modes, based on the observation of flow transition
as droplet mode, droplet-jet mode, unsteady jet
mode, inline jet mode, staggered jet mode, jetsheet mode and sheet mode. They found that, the
flow modes are independent of geometry and
affected by gravitational, inertial, and viscous and
surface tension forces. Roques et al. [9] made flow
visualizations on four types of tubes with three
different fluids under adiabatic flow conditions.
They observed the inter-tube flow modes and
determine what transition takes place as the flow
rate is varied and measured the flow rate to note
the transitions. Jafar et al. [10] [11] used a high
speed CASIO digital video camera to record and
observe the phenomena in slow motion to
facilitate detection of the modes. They observed
the experimental results are in agreement with the
results of Armbruster and Mitrovic [3].

Liquid films flows over horizontal tubes are


encountered in several industrial processes such as
in absorption refrigeration, chemical petroleum
refining, desalination, nuclear reactor cooling and
food and dairy industries. In order to understand
the operations and, in particular, the efficiency of
these processes, it is important to have a detailed
study of such a flow and has been the subject of
numerous theoretical and experimental research.
At low flow rate the liquid is usually in the form of
droplets which are generated at a discrete point
along the underside of a horizontal tube. As the
flow rate increases the transition from droplet to
column occur slowly over a range of flow rates,
and eventually forms a sheet over the surface.

In the present study an attempt is made to verify


the behaviours of liquid film like mode transition,
film thickness, departure site spacing by varying
Film Reynolds number and different geometrical
parameters i.e. tube diameter and tube spacing by
computational fluid dynamics (CFD) code
(ANSYS-FLUENT 13) for 3D configurations.

The flow mode transition equations suggested


by different researchers correlating Reynolds
number (Ref) and Galileo number (Ga) in the form
of Ref = aGab, where a and b are empirical
constants.Yung and co-workers [1]derived semiempirical model from above analytical approach
for predicting transitions from the droplet to the jet
mode.
2

2. PROBLEM MODELING

Boundary and initial conditions

In order to understand the flow of falling film


over the horizontal tubes, three horizontal tubes
arranged one above the other in such a way that
tube axes are in plane with the drawing plane.
Physically the problem is considered different by
allowing LiBr+H2O solution to fall on the
horizontal tubes at varying velocities from small
circular openings at inlet jets due to gravity on
tubes. Liquid impacts on the first tube and flows
around the circumference of it and forms a thin
film. After that falls freely in the inter tubes and
impinges on the next tube. Again flows at the next
tube circumference as a film. Occurrence of all the
above steps leads to three distinct flow modes
taking place viz. droplet, jet and sheet along with
transition between those three. The flow is
surrounded by the air which is treated as stagnant
fluid. The liquids are interacting with the
surrounding air. In 3D analysis problem can be
well described by taking plane parallel in
horizontal tube axis.

Wall (Tube surface)


No slip i.e. Velocity v = 0 at wall surface,
Inlet v = v0 where v0 = assumed velocity, free-fall
velocity from the upper tube and volume fraction
= 1 for liquid.
Outlet P = Patm
Operating pressure=101325 Pa and backflow
volume fraction = 0 for liquid.
Symmetry
Zero normal velocity and normal gradients of all
variables at a symmetry plane
3. METHOD OF SOLUTION
The problem was modelled by using ANSYSICEM-CFD 13 software with boundary conditions.
Due to the symmetry about two centre planes only
of the domain is modelled. Problem considered
as two inline cylinders placed one below other
along with inlet at the top and lower cylinder
section at the bottom. Fig.1. shows the mesh
generation techniques performed in this study.

Assumptions
The assumptions are imposed for the numerical
analysis which helps in better understanding of the
realistic problem. The assumptions are like:

Mesh generation is done by hexahedral elements


since this type mesh is better suited for boundary
layer problems because flow is aligned with the
mesh. As hexahedral elements give good CFD
analysis results, hence attempt is made for
generation of only hex elements by using ICEMCFD mesh generation module. The mesh elements
produced near the tube surface are very fine to
capture bounder layer phenomenon and made
coarse while going away from tube. The Problem
is solved for LiBr+H2O liquid for three different
tube diameters with three different tube spacing at
specified inlet jet diameter. The numerical analysis

i. No slip condition at the tube surface


ii. Film surrounding air velocity is assumed to be
zero
iii. No shear stress at the film surface.
iv. Pressure is assumed to be constant and equal to
one atmospheric.
v. Isothermal and incompressible flow.
vi. Air density outside the film is negligible with
respect to that of the liquid itself.
vii. The thermo-physical properties of the fluid are
constant.
3

was performed over tube diameters of 19mm, 30


mm and 50 mm and 10 mm, 15mm and 20 mm
spacing between tubes for fixed inlet jet diameter
of 2 mm.

Operational variables of liquids for numerical


investigation are taken as per standard values and
are shown in Table 1.
Table 1 : Thermo physical Properties of
different liquids

The mesh is generated in domain with two


symmetry planes. This domain have four half
circular openings with equidistant arrangement.

a.

S. No.

Property

1
2
3
4

Ga*

50 % Conc. H2oLiBr
1676
0.088
0.005
116.827

The flow mode transitions are observed for inlet


velocity ranging from 0.04 m/s to 1.2 m/s in a
velocity step of 0.04 m/s. These parameters are set
to run for a time step of 0.0005 sec. The pressurebased solver with Volume-of-fluid (VOF) model is
selected for analysis. Wall adhesion is taken into
the multiphase model. Explicit scheme for
volume fraction is used with Geo-Reconstruct
formulation. Discretization for Pressure and
Momentum are respectively set to PRESTO and
Power law. Standard Pressure-velocity coupling
SIMPLEC algorithm and for all discretization
schemes first-order method is used. Under
relaxation factors are used for pressure, density,
body forces and momentum is assigned a value of
0.5 to get the solution for the desired accuracy.
The grid formed used for solution settings are
checked for grid and time step, which shows the
same profile. For the flow mode study, video clips
of the contour plots of volume fraction are created
in FLUENT. These clips were proved to be very
helpful in observing the flow modes and transition
in between them. For the detailed study these
videos are again converted in images form.

b.

c.

Figure 1: Mesh generation (a) computational


domain (b) mesh created in domain and
(c) full view of domain mesh.
4

The flow mode transition is shown in Fig.2. As


the diameter increases, mode transition takes place
at lower film Reynolds number and as the tube
spacing increases, mode transition takes place at
higher film Reynolds number, is shown in Fig.3
and Fig, 4 respectively.

4. RESULTS AND DISCUSSION


Numerical analysis of aqua. Lithium bromide
solution falling film is carried out for different
flow rates and varying geometrical parameters i.e.
tube diameter and tube spacing.

Film Reynolds number,Re

4.1 Effect on flow mode transition of falling


film

DROP MODE

160
140
120
100
80
60
40
20
0
15

35
Tube diameter,D

drop to drop_jet
jet to jet_sheet

DROP_JET MODE

55

drop_jet to jet
jet_sheet to sheet

Film Reynolds number,Re

Figure 3: Tube diameter effect on mode


transition

JET MODE

JET_SHEET MODE

160
140
120
100
80
60
40
20
0
10

15

20

25

Tube spacing,S

drop to drop_jet

drop_jet to jet

jet to jet_sheet

jet_sheet to sheet

Figure 4: Tube spacing effect on mode


transition

SHEET MODE

Figure 2: Flow mode transition of aqua.


Lithium bromide falling film

Same five flow modes was observed by transition


effect is observed by Hu and Jacobi [7][8].
5

4.2 Effect on film thickness of falling film


vel = 0.08m/s

It is observed that as diameter minimum, the


film thickness is maximum and when diameter is
maximum film thickness is maximum but at
optimum combination of diameter and spacing
thickness is minimum. For small diameter tubes,
as spacing increases thickness decreases but big
diameter tubes, as spacing increases thickness
increases, is shown in Fig.5.

film thickness in mm

0.500

0.24v

0.32v

vel = 0.24 m/s

vel = 0.40 m/s

vel = 0.56 m/s

0.40v

0.450
vel = 0.72 m/s
0.400

Figure.6: Flow velocity effect on departure


wavelength

0.350
0.300

5. CONCLUSIONS

0.250

An extensive numerical model is developed to


study the behaviour of falling film of aqua.
Lithium bromide solution on horizontal tubes. The
3 D model developed is useful to predicting the
flow mode transition, film thickness and departure
wavelength behaviour for geometrical variations
like tube diameter and tube spacing.

Figure 5: Film Thickness variation of aqua.


Lithium bromide falling film

The specific conclusions from the above study are


as follows:

4.3 Effect on departure site spacing of falling


film

i.

It is observed that as droplet frequency increases,


departure wavelength decreases. Same result is
observed by Jafer [10][11].As flow velocity
increases, departure wavelength decreases upto
drop to jet mode transition. After jet mode,
departure wavelength remains constant but jet
diameter increases upto jet to sheet mode
transition. In sheet mode jet are trying to mix each
other, hence departure wavelength decreases to
form sheet, is shown in Fig.6.

ii.

Mode transition is sensitive to tube


diameter and tube spacing. But effect of
tube spacing is dominant as compare to
tube diameter.
Film thickness is also sensitive to tube
diameter and tube spacing. For equal
distribution of flow over the horizontal
tubes proper combination of tube diameter
and tube spacing is necessary.

iii.

tubes, ASME Journal of Heat Transfer, Vol.


124, (2002), pp. 491-499.
10.
F. Jafar, G. Thorpe and O.F. Turan Liquid
film falling on horizontal circular cylinders
16thAustralasian Fluid Mechanics Conference,
December (2007), pp.1193- 1200.
11.
F. Jafar, G. Thorpe and O. F. Turan Flow
mode characterization of liquid films falling on
horizontal
plain
cylinders
Seventh
International Conference on CFD in Minerals
and Process Industries CSIRO, December
2009, pp.1- 6
12.
ANSYS 13.0 Release Guide Help
documentation, November 2010.

In jet mode, departure wavelength always


remains same for different flow velocities.

Further efforts are needed to investigate the


correlations for flow mode transition and film
thickness calculation.
REFERENCES
1. D. Yung, J. J. Lorentz, and E. N. Ganic,
Vapor/Liquid Interaction and Entrainment in
Falling Film Evaporators ASME Journal of
Heat transfer,Vol.102, (1980), pp. 20-25.
2. Pise, A. T., Korde, N. U., and Salunkhe, R. H.,
Numerical Simulation of Falling Film
Droplets/Jets over Horizontal Tubes, Int. J.
Fluid Mech. Res., 2013, 40, No. 3, pp. 266
279.
3. J. Mitrovic, Influence of tube spacing and flow
rate on heat transfer from a horizontal tube to a
falling liquid film, 8th International Heat
Transfer Conference, San-Francisco, (1986).
4. R. Armbruster, J. Mitrovic, Patterns of Falling
Film Flow over Horizontal Smooth Tubes, 10th
International Heat Transfer Conference,
Brighton, (1994).
5. R. Armbruster and J. Mitrovic, Evaporative
cooling of a falling water film on horizontal
tubes. Experimental Thermal and Fluid
Science, Vol.18, (1998), pp.183-194.
6. J. Mitrovic Flow structures of a liquid film
falling on horizontal tubes Chemical
Engineering and Technology, Vol. 28, (2005),
pp. 684- 694
7. X. Hu, A.M. Jacobi, The Intertube Falling
Film Part 1- Flow characteristics, mode
transition and hysteresis. ASME Journal of
Heat Transfer, Vol. 118, (1996), pp.616-625.
8. X. Hu, A.M. Jacobi, Departure site spacing for
liquid droplets and jets falling between
horizontal circular tubes, Experimental
Thermal and Fluid Science, Vol. 16, (1998), pp.
322-331.
9. J. F. Roques, J. R. Thome and V. Dupont,
Falling film transitions on plain and enhanced

Nomenclature
d

Horizontal tube diameter (mm)

Acceleration due to gravity, m/s2

Mass flow rate of the liquid film per unit


length of
Cylinder (kg/s-m)

Total absolute pressure (Pa)

Time (sec.)

Velocity (m/s)
Greek Symbols

Density, Kg/m3

Dynamic viscosity, kg/ms

Surface tension, N m

Capillary constant
Dimensionless Number

Ref

Film Reynolds number Ref = 4/

Ga

Galileo number, Ga = 3/4g

Ga* Modified Galileo number, Ga1/4

Instability wavelength, m

Normalized wavelength, *=/

Vous aimerez peut-être aussi