American Journal of Mechanical Engineering, 2013, Vol. 1, No.

8, 470-486
Available online at http://pubs.sciepub.com/ajme/1/8/1
Science and Education Publishing
DOI:10.12691/ajme-1-8-1

Experimental Study and CFD Simulation of Two-Phase

Flow around Multi-Shape Obstacles in Enlarging
Channel
Laith Jaafer Habeeb1,*, Riyadh S. Al-Turaihi2
1
Mechanical Engineering Department, University of Technology, Baghdad, Iraq
College of Engineering/ Department of Mech. Eng., Babylon University, Babil, Iraq
*Corresponding author: laithjaafer@yahoo.com

Received August 01, 2013; Revised November 13, 2013; Accepted December 24, 2013

Abstract Experimental and numerical studies are investigated to study the two-phase flow phenomena around
multi- shape obstacles in a rectangular enlarging channel which has the dimensions (10 3 70 cm) enlarged from
assembly circular tube of the two phases. Experiments are carried out in the channel with air-water flow with
different air and water flow rates. These experiments are aimed to visualize the two phase flow phenomena as well
as to study the effect of pressure difference through the channel with the existence of the obstacle. All sets of the
experimental data in this study are obtained by using a pressure transducer and visualized by a video camera for
different water discharges (20, 25, 35 and 45 l/min) and different air discharges (10, 20, 30 and 40 l/min). While the
numerical simulation is conducted by using commercial Fluent CFD software to investigate the steady and unsteady
turbulent two dimensional flows for different air and water velocities. The results showed that when air or water
discharge increases, high turbulence is appear which generate more bubbles and waves and the mean pressure
difference increases. Also, in a water slug, bubbles move slower than the liquid.

Keywords: steady and unsteady flow, smooth enlargement, two-phase turbulent flow, multi- shape obstacles in
channel, experimental and Fluent CFD software investigation

Cite This Article: Laith Jaafer Habeeb, and Riyadh S. Al-Turaihi, Experimental Study and CFD Simulation
of Two-Phase Flow around Multi-Shape Obstacles in Enlarging Channel. American Journal of Mechanical
Engineering 1, no. 8 (2013): 470-486. doi: 10.12691/ajme-1-8-1.

1. Introduction
The understanding of turbulent two-phase bubbly ows
is important due to the widespread occurrence of this
phenomenon in natural and engineering systems [1]. The
flow patterns of gas-liquid two-phase flow could be
bubble flow, slug flow, plug flow, stratified flow, wavy
flow and disperse flow. There are still many challenges
associated with a fundamental understanding of ow
patterns in multiphase ow and considerable research is
necessary before reliable design tools become available.
Gas-liquid flow was extensively used in industrial systems
such as power generation units, cooling and heating
systems (i.e. heat exchangers and manifolds), safety
valves, etc. Thus two-phase flow characteristics through
these singularities should be identified in order to be used
in designing of the system [2]. In the regimes of bubbly
and slug ows, a spectrum of different bubble sizes is
observed. While dispersed bubbly ows with low gas
volume fraction are mostly mono-dispersed, an increase in
the gas volume fraction leads to a broader bubble size
distribution due to breakup and coalescence of bubbles.
Moreover, the forces acting on the bubbles may depend on
their individual size [3]. In the last decade, the stratified

flows are increasingly modeled with computational fluid

dynamics (CFD) codes. In CFD, closure models are
required that must be validated. The recent improvements
of the multiphase flow modeling in the ANSYS code
make it now possible to simulate these mechanisms in
detail [4]. A general estimates for mean velocities through
and around groups or arrays of xed and moving bodies,
in unbounded and bounded domains, which lie within a
dened perimeter were derived [5]. When the bodies were
well-separated, the interstitial Eulerian and Lagrangian
mean velocities were dened and calculated in terms of
the far-eld contributions from the velocity or
displacement eld within the group of bodies. The
development of numerical code to deal with
incompressible two phase flow around 2D hydrofoil
combined with cavitation model with k- turbulent model
was focused [6]. Also, the calculation of cavitation in
cryogenic fluid was done by implementing the
temperature sensitivity in government equations. Even
though the results showed good agreement with previous
calculation results, further research will be needed to
account for real physics of the formation of cavitation in
cryogenic fluid due to temperature sensitivity. The air
water bubbly ow around two tandem square-section
cylinders with the vortex method proposed by the authors
in a prior study was simulated [7]. Pair of Karman vortices

American Journal of Mechanical Engineering

appeared behind the rst cylinder. The strength of the

Karman vortex downstream of the second cylinder was
larger than that for the single-cylinder. The strength of the
Karman vortex reduced due to the bubble entrainment.
The pressure reduction on the rear surface of the rst
cylinder was relaxed when compared with that on the
single-cylinder. Therefore, this simulation could predict
the reduction of the drag force acting on the rst cylinder,
which was claried by the corresponding experimental
study. Supercavitation around a hydrofoil based on ow
visualization and detailed velocity measurement was
studied [8]. A high-speed video camera was used to
visualize the ow structures under different cavitation
numbers, and a particle image velocimetry (PIV)
technique was used to measure the instantaneous velocity
and vorticity elds. It was shown that the cavitation
structure depends on the interaction of the watervapor
mixture and the vapor among the whole supercavitation
stage. As the cavitation number was progressively
lowered, three supercavitating ow regimes were
observed. Furthermore, in the cavitating region, strong
momentum transfer between the higher and lower ow
layers taken place, resulting in a highly even velocity
distribution in the core part of the cavitating region, and
the lower velocity area became smaller and, as the
cavitation number lowers, moved toward the downstream.
A population balance model in close cooperation between
ANSYS-CFX
and
Forschungszentrum
DresdenRossendorf and implemented into the CFD-Code CFX
was developed [9]. They presented a brief description of
the model principles. The capabilities of this model were
discussed via the example of a bubbly ow around a halfmoon shaped obstacle. For the complex ow around the
obstacle, the general structure of the ow was well
reproduced in the simulations. This test case demonstrated
the complicated interplay between size dependent bubble
migration and the effects of bubble coalescence and
breakup on real ows. However, clear deviations occur for
bubble coalescence and fragmentation. An analytical
solution for local pressure drop due to obstructions in
horizontal air-water two-phase flow was presented [10].
Various obstruction shapes with size were investigated.
The relationship between two-phase multiplier and local
(normalized) pressure drop with the gas superficial
velocity were investigated. The results showed a higher
pressure drops pointed for larger obstructions. A
correlation study on the two-phase multipliers and local
pressure drop for two-phase, air water mixture flows
through obstructions in horizontal channel was made. The
numerical simulation to investigate the gas-liquid twophase flow in the microchannel for the adhesive
dispensing was carried out [11]. A 2-D model of
microchannel with a diameter for 0.04 mm was
established and meshed. The gas-liquid slug flow emerges
after iterating over 1 million steps with the gas flow rate
for 0.1 m/s, the water flow rate for 0.05 m/s. While the gas
flow rate decreased, the length of liquid droplet and gas
bubble increased and decreased, respectively and the total
numbers of bubble and droplet decreased in one period.
This indicated the fluid parameters have a high
relationship with the quantity and volume of droplet. The
vortex structures and particle dispersions in ows around
a circular cylinder by lattice Boltzmann method (LBM),
with non-equilibrium extrapolation method (NEM)

471

dealing with the computational boundaries was

investigated [12]. It was found that both the Reynolds
number and the Stokes number produce signicant
inuences on the particle distribution. The small particles
(St = 0.01) follow the motion of the uid very well and
can disperse into the core regions of the vortex structures.
The particles at intermediate Stokes numbers (St = 0.1 and
1) concentrate on the boundaries of the vortices, and the
large particles (St = 10) also assemble in the outer regions
of the vortices under the inuence of the vortex structures.
Characteristics of adiabatic air-water flow through a
horizontal channel having smooth expansion (enlarging),
which was obtained by inspiring from one of the
singularities existing in safety valves, were investigated
numerically [13]. Numerical simulations are performed
via commercial software, GAMBIT (v. 2.3.16) and
FLUENT (v. 6.3.26), for the flow upstream, through and
downstream the singularity. Eulerian Model was
employed for the analysis. Internal diameter of the
channel enlarges from 32 mm to 40 mm. Flow rate for
water was constant at 3 l/s while that for air was taken as
50 and 61 l/min. The effect of uid properties and
operating conditions on the generation of gasliquid
Taylor ow in microchannels was investigated
experimentally and numerically [14]. Visualization
experiments and 2D numerical simulations have been
performed to study bubble and slug lengths, liquid lm
hold-up and bubble velocities. The results showed that the
bubble and slug lengths increase as a function of the gas
and liquid ow rate ratios. Numerical studies on the
hydrodynamic and heat transfer characteristics of twophase ows in small tubes and channels were reviewed
[15]. The review was then categorized into two groups of
studies: circular and non-circular channels. Different
aspects such as slug formation, slug shape, ow pattern,
pressure drop and heat transfer were of interest. Gaps in
research were found in applications of non-circular ducts,
pressure drop and heat transfer in meandering microtubes
and microchannels for both gasliquid and liquidliquid
two-phase ows.
From the previous review it is denoted that the
recent researches in the two phase flow with the
enlargement and existence of a multi- shape cylinders are
very limited. So, our concern in this study is to study
the effects of wide range of air/water discharge in the
steady and unsteady cases on the flow behavior with the
enlargement from the circular tube of the water phases
which contains the air phase tube, to the rectangular duct
with the existence of a circular cylinder. For the
experimental investigation of co- current air/water flows,
the HAWAC (Horizontal Air/Water Channel) was built.
The channel allows in particular the study of air/water
slug/disperse flow under atmospheric pressure. Parallel
to the experiments, CFD calculations were carried out.
The behavior of slug/disperse generation and propagation
was qualitatively reproduced by the simulation.

2. The Physical Model and Experimental

Apparatus
Figure 1 shows a schematic and photograph of the
experimental Apparatus and measurements system. The
rig is consists of, as shown in Figure 2:

472

American Journal of Mechanical Engineering

Figure 1. The experimental rig and measurements system [18,19]

Figure 2. (a) Water system, (b) Air flow meter, (c) Enlargement connecting part, flange, piping system and pressure transducer sensor [18,19]

1. Main water tank of capacity (1 m3).

2. Water pump with specification quantity (0.08
m3/min) and head (8 m).

3. Valves and piping system (1.25 in)

4. Adjustable volume flow rate of range (10-80 l/min)
is used to control the liquid (water) volume flow
rates that enter test section.

American Journal of Mechanical Engineering

5. Air compressor and it has a specification capacity

of (0.5 m3) and maximum pressure of (16 bars).
6. Rotameter was used to control the gas (air)
volume flow rates that enter the test section. It has a
volume flow rate range of (6-50 l/min).
7. Valves and piping system (0.5 in) and gages.
8. Pressure transducer sensors which are used to
record the pressure field with a range of (0-1 bar)
and these pressure transducer sensors are located
in honeycombs at the entrance and at the end of
the test section. The pressure sensors with a
distance of (80 cm) between each other are
measuring with an accuracy of (0.1%).
9. The obstacle used is made of stainless steel and its
dimensions are (3 cm) for the length of each
obstacle and; (3 cm) diameter for the circular
cylinder, (3 cm) for the square-section cylinder
sides and (3 cm) for each of the triangle sides and,
which is coated with a very thin layer of black
paint and its center located at (11.5 cm) from the
entrance of the test section.
10. The test section is consisting of rectangular
channel and a cylindrical obstacle. The
rectangular cross sectional area is (10 cm 3 cm)
and has length of (70 cm) which is used to show
the behavior of the two phase flow around the
obstacle and to measure the pressure difference
and records this behavior. The obstacle is mounted
and fixed by screw and nut on a blind panel on the
bottom of the rectangular channel. The three large
Perspex windows of the channel (two lateral sides
with lighting and the top side) allowing optical
access through the test section. Two enlargement
connecting parts are made of steel and
manufactured with smooth slope. The first on is
used to connect the test section with the outside
water pipe in the entrance side while the second
one is used to connect the test section with the
outside mixture pipe in the exit side. The inside air
pipe, in the entrance side, is holed and contained
inside the water pipe by a steel flange.
11. Interface system consists of two parts which are
the data logger and the transformer which contains
in a plastic box. The data logger has a three
connections two of them are connected to the
outside of the box (one connected to the sensors
and the other connected to the personal computer),
the third connection is connected to the
transformer, which is work to receive the signals
as a voltage from the sensors and transmit it into
the transformer and then re-received these signals
after converting it to ampere signals in the
transformer.
12. The interface system which is connected with a
personal computer so that the measured pressure
across the test section is displayed directly on the
computer screen by using a suitable program.
13. A Sony digital video camera recorder of DCRSR68E model of capacity 80 GB with lens of Carl
Zeiss Vario-Tessar of 60 x optical, 2000 x digital
is used to visualize the flow structure. The
visualized data are analyzed by using a AVS video
convertor software version 8.1. A typical sequence

473

snapshots recorded by the camera using a

recording rate of 30 f/s.
The flows of both gas and liquid are regulated
respectively by the combination of valves and bypassages before they are measured by gas phase flow
meter and liquid phase flow meter. The gas phase and
the liquid phase are mixed in the enlargement connection
part before they enter the test section. When the twophase mixture flows out of the test section, the liquid
phase and the gas phase are separated in liquid storage
tank. Experiments were carried out to show the effect of
different operation conditions on pressure difference
across the test section and to visualize the flow around
the obstacle. These conditions are water discharges and
air discharges. The selected experimental values for
water discharges are (20, 25, 35 and 45 l/min) and for air
discharges are (10, 20, 30 and 40 l/min). The experimental
procedures are:
1. Fix the circular cylinder obstacle at the channel
bottom side.
2. Turn on the water pump at the first value (20
l/min).
3. Turn on the air compressor at the first value (10
l/min).
4. Record the pressure drop through the test section and
photograph the motion of the two-phase flow by the
digital camera.
5. Repeat steps (2 to 5) by changing the water
discharge.
6. Repeat step (5) by changing the air discharge.
7. Change the circular cylinder obstacle by the squaresection cylinder obstacle and then by the triangularsection cylinder obstacle.
These give sixteen (16) experimental cases for
volume fraction (Air/Water ratios) for each type of
obstacles.

3. Numerical Modelling
In this study, the computational fluid dynamics (CFD)
software have been applied for the numerical simulation
for adiabatic gas-liquid flow characteristics through a
horizontal channel contain a cylinder (different shape
every time) with smooth expansion from the liquid pipe in
steady, unsteady and 2D cases. In order to compare
numerical results with experimental ones, air-water couple
has been selected as the representative of the gas-liquid
two-phase flow. Construction of the numerical domain
and the analysis are performed via GAMBIT and
FLUENT (ANSYS 13.0) CFD codes, respectively. Twophase flow variables such as void fraction and flow
velocity for liquid (water) and gas (air) at the inlet
condition, and the geometrical values of the system (i.e.
channel length, width and height, pipes and inlet
enlargement connecting part dimensions, and obstacle
dimensions) used in the analysis are selected as the same
variables as the experimental part. Atmospheric conditions
are valid for the experimental facility. Total test rig length
in the experiments, thus in the numerical domain, is (100
cm) including (70 cm) for the test section containing the
obstacle, and (30 cm) for the inlet enlargement part. Water
pipe diameter is (3.175 cm) and air pipe diameter is (1.27
cm) as shown in Figure 3 (for the circular cross
section cylinder).

474

American Journal of Mechanical Engineering

Figure 3. The experimental test section dimensions

The 2D physical model is established using a model

of flow focusing channel in CFD. The enlargement
connecting part length consists of: (0.05 m) circular pipe,
m) diverge-link to change the shape from circular to
rectangular and (0.1 m) rectangular duct. Air and water
are selected to be working fluids and their fluid properties
are in Table 1.
Table 1. Property parameters of the gas and liquid in CFD
Fluid

Density (kg/m3)

Viscosity (kg/m.s)

Water

998.2

10.0310

Air

1.225

1.789410-05

Case
number
1
2
3
4
Case
number
9
10
11
12

Surface Tension

-04

Table 2. Air-water flow cases

Air/water
Air/water
Air/water
Case
discharges velocities
discharge
number
(l/min)
(m/sec)
(l/min)
10/20
1.32/0.50
5
20/20
10/25
1.32/0.63
6
20/25
10/35
1.32/0.87
7
20/35
10/45
1.32/1.12
8
20/45
Air/water
Air/water
Air/water
Case
discharge
velocities
discharge
number
(l/min)
(m/sec)
(l/min)
30/20
3.95/0.50
13
40/20
30/25
3.95/0.63
14
40/25
30/35
3.95/0.87
15
40/35
30/45
3.95/1.12
16
40/45

0.072
---

Air/water
velocities
(m/sec)
2.63/0.50
2.63/0.63
2.63/0.87
2.63/1.12
Air/water
velocities
(m/sec)
5.26/0.50
5.26/0.63
5.26/0.87
5.26/1.12

The model geometry structure was meshed by the

preprocessor software of GAMBIT with the Quad/Pave
grids for the circular obstacle and Quad/Submap grids for
square-section and triangular-section obstacles. After
meshing, the model contained (for 2D ) 14978 grid nodes
and 14522 cells for the circular obstacle, 16282 grid nodes

and 32090 faces for square-section obstacle and 14447

grid nodes and 29348 faces for triangular-section obstacle
(as demonstrated in Figure 4 for the square cross section
cylinder) before importing into the processor Fluent for
calculation. This refinement grid provided a precise
solution to capture the complex flow field around the
cylinder and mixing region in the enlargement connecting
part. The boundary conditions are the velocity inlet to the
air and water feeding (Table 2) and the pressure outlet to
the model outlet. In Fluent, the Mixture and Eulerian
Multiphase models were adopted to simulate the flow. The
mixture model is a simplified Eulerian approach for
modeling n-phase flows, while the Eulerian multiphase
model is a result of averaging of NS equations over the
volume including arbitrary particles + continuous phase
and the result is a set of conservation equations for each
phase (continuous phase + N particle media) [16].
Because the flow rates of the air and water in the channel
are high, the turbulent model (k- Standard Wall Function)
has been selected for calculation. The other options in
Fluent are selected: SIMPLE (Semi-Implicit Method for
Pressure-Linked Equations) scheme for the pressurevelocity coupling, PRESTO (pressure staggering option)
scheme for the pressure interpolation, Green-Gauss Cell
Based option for gradients, First-order Up-wind
Differencing scheme for the momentum equation, the
schiller-naumann scheme for the drag coefficient,
manninen-et-al for the slip velocity and other selections
are described in Table 3. The time step (for unsteady case),
maximum number of iteration and relaxation factors have
been selected with proper values to enable convergence
for solution which is about to (0.001) for all parameters.

American Journal of Mechanical Engineering

475

Figure 4. Sample of 2D model geometry structure meshes

Table 3. Other mixture model selections for Fluent
Solver type
Pressure- Based

k- Model

Solution Methods

Cmu=0.09, C1Epsilon=1.44, C2Epsilon=1.92

Volume Fraction and

Turbulent Kinetic
Energy (First-order Upwind)

Starting Solution Controls (Under-Relaxation Factors)

Pressure=0.3, Momentum=0.7, Turbulent Kinetic Energy &
Turbulent Dissipation Rate=0.8
Intensity and Hydraulic Diameter (Turbulent Intensity=3% and
Hydraulic Diameter=0.0127 m)
Solution Initialization
Turbulent Kinetic Energy (m2/s2)=0.0003375, Turbulent Dissipation
Rate (m2/s3)= 0.0007620108 and air-bubble Volume Fraction=0

The hydrodynamics of two-phase flow can be

described by the equations for the conservation of mass
and momentum, together with an additional advection
equation to determine the gas-liquid interface. The twophase flow is assumed to be incompressible since the
pressure drop along the axis orientation is small. For the
incompressible working fluids, the governing equations of
the Mixture Multiphase Model are as following [16,17]:
The continuity equation for the mixture is:
(1)

Where vm is the mass-averaged velocity:

vm =

k =1 k k vk
n

(2)

and m is the mixture density:

m = k =1 k k

(3)

k is the volume fraction of phase k.

The momentum equation for the mixture can be

obtained by summing the individual momentum
equations for all phases. It can be expressed as:

( m vm ) + ( m vm vm )
t

T
= p + m vm + vm

n
+ m g + F + k =1 k k vdr ,k v

(4)

where n is the number of phases, F is a body force and

m is the viscosity of the mixture:

m = k =1 k k
n

vdr ,k is the drift velocity for secondary phase k:

vdr ,=
vk vm
k

p p + p p vm
t

n
= p p vdr , p + q =1 m qp m pq

)
(

(7)

The relative velocity (also referred to as the slip

velocity) is defined as the velocity of a secondary phase (p)
relative to the velocity of the primary phase (q):

v pq
= v p vq
(8)

Specification Method for turbulence

( m ) + ( m vm ) = 0
t

From the continuity equation for secondary phase p, the

volume fraction equation for secondary phase p can be
obtained:

(5)
(6)

The mass fraction for any phase (k) is defined as:

ck =

k k
m

(9)

The drift velocity and the relative velocity ( vqp ) are

connected by the following expression:

n
vdr=
, p v pq k =1 ck vqk

(10)

ANSYS FLUENTs mixture model makes use of an

algebraic slip formulation. The basic assumption of the
algebraic slip mixture model is that to prescribe an
algebraic relation for the relative velocity, a local
equilibrium between the phases should be reached over a
short spatial length scale. The form of the relative velocity
is given by:

p p m

v pq =

f drag p

(11)

where p is the particle relaxation time:

p =

p d 2p
18q

(12)

d is the diameter of the particles (or droplets or bubbles)

of secondary phase p , is the secondary-phase particles

acceleration. The default drag function f drag :

1 + 0.15 Re0.687 Re 1000

f drag =
Re > 1000
0.0183 Re

and the acceleration is of the form:

=g ( vm ) vm m .
t

(13)

(14)

The simplest algebraic slip formulation is the so-called

drift flux model, in which the acceleration of the particle
is given by gravity and/or a centrifugal force and the
particulate relaxation time is modified to take into account
the presence of other particles.

476

American Journal of Mechanical Engineering

In turbulent flows the relative velocity should contain a

diffusion term due to the dispersion appearing in the
momentum equation for the dispersed phase. ANSYS
FLUENT adds this dispersion to the relative velocity:

v pq
=

( p m ) d 2p t p q
t p

18q f drag

(15)

where t is a Prandtl/Schmidt number set to 0.75 and t

is the turbulent diffusivity. This diffusivity is calculated
from the continuous-dispersed fluctuating velocity
correlation, such that:

=
t C

k2 y

1 + y

1 2
1 + C 2

v pq

(16)

4. Experimental Results
(17)

2 3k

v pq v p
Where C =
1.8 1.35cos , cos =

and y
v pq v p
2

is the time ratio between the time scale of the energetic

turbulent eddies affected by the crossing-trajectories effect
and the particle relaxation time. If the slip velocity is not
solved, the mixture model is reduced to a homogeneous
multiphase model.
And the governing equations of the Eulerian
Multiphase Model are as following [16,17]:
- The continuity equation is:

q q
t

) + u = n m
( q q q ) pq

(18)

p =1

Where q is the volume fraction for the qth phase:

- The momentum equation for qth phase:

q q uq

) + u u
( q q q q)

t
= q p + q q g + q
=

(19)

( R pq + m pq uq ) + q q ( Fq + Flift ,q + Fvm,q )
p =1

Where the first left hand side term is referring to the

transient and the second one is refer to the convection. In
the right hand side, the terms are referring to the pressure,
body force, shear stress, (interphase forces exchange and
interphase mass exchange) and (external, lift, and virtual
mass forces) respectively. Hence, solids pressure term is
included for granular model. The inter-phase exchange
forces are expressed as:

=
R pq K pq u p uq

pipe in the experiments, therefore, the gas flow through

the air pipe and the mixture occurred outlet of it are
modeled in 3D (Figure 3). According to the simulation, air
with known mass flow rate flows through air pipe and
then disperses into the water at the exit of the pipe. At air
flow rates (thus volumetric void fraction), phase inlet
velocity and void fraction profiles obtained at the air and
water pipes outlet are extracted from the experimental
calculations in order to be introduced as the inlet condition
for the flow analysis regarding the numerical 2D domain.
In the present study bubble diameter is equal to (1 cm).
Assuming the bubbles are in spherical shape and
neglecting the coalescence between them along the
channel.

(20)

Where K pq is the exchange coefficient and in general

Fpq = Fqp .
In FLUENT application, boundary conditions like
velocity inlet is taken as the inlet condition for water
and air while interior and outflow are employed as the
water-air mixture. Air is injected to the water via an air

The experimental results are represented of a gas-liquid

flow through channel with the existence of the obstacle for
different water discharges (20, 25, 35 and 45 l/min) and
different air discharge (10, 20, 30 and 40 l/min) as
photographs and as pressure graphs.

4.1. Effect of Air and Water Discharges

Figure 5, Figure 6, Figure 7 and Figure 8 show
photographs for the two phase flow behavior around the
obstacles for water discharge (Qw=20, 25, 35, 45 l/min)
respectively and air discharges (Qa = (a)10, (b)20, (c)30
and (d)40 l/min) for the three cylinders (circular,
triangular and square). It shows that the number (amount)
of bubble is few and has small size at low water
discharge. These photographs describe the flow behavior
and it appears that it is near to slug or plug region when
the discharge is low. This is due to the low velocity of
water at low water discharge. Also when increase the air
discharge the size and number of bubbles increases and
the bubble cavities develops to cloud cavitations
especially at high air discharge. This is due to the high
velocity of water at high water discharge which leads to
more turbulence in the flow and the flow becomes
bubbly as shown. It is clear that the flow becomes
unstable and unsymmetrical around the cylinders and the
number and size of bubble become higher compared
with the low velocities cases. It appears that the vortices
behind and beside the cylinders becomes more strong
compared with the low discharge cases. When the two
phases increases, more unsteady behavior is noticed and
the flow oscillates between bubble and disperse regions.
Also, when water discharge increases with increase air
discharge, flow becomes unsteady, vortices developed
around the cylinders surface and most bubbles
transformed to cloudy flow, then a disperse region and
strong vortex shedding is observed. This is due to the
important effect of the cylinders existence in the
rectangular channel which effect on pressure difference
across the inlet and outlet of the channel. The
experimental data shows that the average number of
bubbles generally increases with increasing mixture
velocities. Independently of the inlet velocities, the
highest number of bubbles is found in the mixing region.
Moreover, higher gas velocities have a higher number of
bubbles in the mixing region and the smooth obstacle
(circular cylinder) generates less bubble and turbulence.

American Journal of Mechanical Engineering

Figure 5. Photographs for the two phase flow behavior for Qw=20 l/min and Qa=10, 20, 30, 40 l/min respectively

Figure 6. Photographs for the two phase flow behavior for Qw=25 l/min and Qa=10, 20, 30, 40 l/min respectively

477

478

American Journal of Mechanical Engineering

Figure 7. Photographs for the two phase flow behavior for Qw=35 l/min and Qa=10, 20, 30, 40 l/min respectively

Figure 8. Photographs for the two phase flow behavior for Qw=45 l/min and Qa=10, 20, 30, 40 l/min respectively

American Journal of Mechanical Engineering

Figure 9. Mean pressure difference with air discharge for different values of water discharge

Figure 10. Mean pressure difference with water discharge for different values of air discharge

479

480

American Journal of Mechanical Engineering

4.2. Effect of Pressure Difference

Figure 9 represents the mean pressure difference with
air discharge for different values of water discharge for
the three cylinders. While Figure 10 represents the
mean pressure difference with water discharge for
different values of air discharge for the three cylinders.
When air or water discharge increases, the mean
pressure difference increases. This is due to the increase
of air or water discharge resulting in velocity increases. It
is already noticed that the circular cylinder has a pressure
difference values more than that of the other two obstacles.
Also, the mean pressure difference has a significant
influence on two-phase flow behavior. Therefore, it is

expected that the flow instability will also depend upon

the pressure difference.

4.3. Effect of Time Evolution of Pressure

Figure 11 represents effect of time evolution of
pressure across the (a) circular, (b) triangular and (c)
square cylinder respectively obtained by experiments
for water discharge Qw=20 l/min and air discharge
Qa=10 l/min at inlet and outlet of the rectangular
channel. The pressure sensor at the inlet -after
honeycombs- and outlet of the test section are record
pressures that fluctuating as a function of time due to
two-phase effect.

Figure 11. Effect of time evolution of pressure for circular, triangular and square cylinder respectively

5. Numerical Results
The numerical results are represented as colored
contours and center line distributions for the same airwater discharges cases in the experimental part. As
mentioned above, the 2D inlet (line) air or water velocities
are calculated from the 3D experimental inlet (surface)
area from the air or water discharge. These give sixteen
(16) numerical cases for volume fraction (Air/Water
ratios) for each type of obstacles.

5.1. Steady State

Figures (12-a, b, c, d, f and g) depict volume
fraction (water) contours for selected cases (case1, 3, 5,
6, 10, 11 and 12 respectively) for the circular cylinder.
While Figures (13-a, b, c, d, e, f and g) depict volume

fraction contours for selected cases (1, 2, 4, 5, 11, 12

and 13) respectively for the triangular cylinder. Also,
Figures (14-a, b, c, d, e, f and g) depict volume fraction
(water) contours for selected cases (case 4, 5, 6, 7, 10,
14and 15 respectively) for the square cylinder. The
differences between the experimental snapshots and
numerical Figures are due to two reasons; the first is the
differences in the overall flow rates of air and water for
the same inlet velocities from the inlet regions (small
lines in 2D numerical cases and big square and
annulus areas in 3D experimental cases), and the second
reason is that the snapshots are taken roughly from the
experimental movies for each case and may be for another
snapshot from the same case movie, the differences will
be less. From these figures it is appear that a slug to
disperse regions flow pattern is achieved. The flow rates
of air and water have a large range and it show the
increase in water phase and with the decrease of the gas

American Journal of Mechanical Engineering

481

flow rate, the volume fraction of the gas decreased and the
volume fraction of the water increased simultaneously.
According to the figures, stratified water-air mixture
enters the singularity section and begins to decelerate due
to the smoothly enlarging cross-section and it show how
the volume fraction affected the flow behavior. A uniform
dispersed two-phase flow, in which the dispersed phase
(either air bubbles or water droplets) moves with their
carrier fluid (water or air), approaches to a circular
obstacle. Due to strong changes of both magnitude and
direction of local velocities of the fluid flow (i.e. local
fluid velocity gradients) and density difference between
the dispersed phase and the fluid, the local phase
distribution pattern changes markedly around the obstacle.
Strong air flows are induced and a strong vortex is created
as a result of the entered air and small vortices are also
produced. A recirculation zone in the wake, a ow
separation at the edge of the obstacle and a wavy motion
are noticed. It is appear that maximum turbulent viscosity
and high turbulence regions depends on the volume
fraction ratio. Also, when air velocity increases, separation
area is detected after the cylinder.

Figure 13. Volume fraction (water) contours for (case4, 5, 6, 7, 14, 15

and 16) respectively

5.2. Unsteady State

Figure 12. Volume fraction (water) contours for cases (1, 3, 5, 6, 10, 11
and 12) respectively

Figures (15-a, b, c, d, e and f) represent volume

fraction (water) contours development for randomly
selected unsteady case8 for the circular cylinder. Figures
(16-a, b, c, and d) represent volume fraction (water)
contours development for selected unsteady case3 for the
triangular cylinder. Figures (17-a, b, c, d, e and f)
represent volume fraction (water) contours development
for randomly selected unsteady case1 for the square
cylinder. It show how the volume fraction develops with
time. As can be seen, there are two unsteady
asymmetrical pattern recirculating zones behind the
cylinder when the volume fraction increases or when the
two-phase velocity (Reynolds number) increases.
Figures (18-a, b, c, d, e, f, g, h, i, j, k and l)
represent center line (x=0.05 m) of the static and
dynamic pressure distributions of the mixture along the
test rig (1 m) for unsteady case8 at different time steps
for the circular cylinder. It show how the pressures are
fluctuating along the channel and the fluctuation is

482

American Journal of Mechanical Engineering

increases with time due to the effects of mixing, twophase flow, circular cylinder existence and density
difference. Region that the line does not passing through it

(i.e. empty at position (z) = 0.115 m) is the circular

cylinder area, as well as at the beginning (z = -0.3 to z=0.28 cm) there is no mixing.

Figure 14. Volume fraction (water) contours for cases (1, 2, 4, 5, 8, 11 and 12) respectively

Figure 15. volume fraction (water) contours development for unsteady case8

Figure 16. Volume fraction (water) contours development for unsteady case3

American Journal of Mechanical Engineering

Figure 17. Volume fraction (water) contours development for unsteady case1

483

484

American Journal of Mechanical Engineering

Figure 18. Static and dynamic pressure distributions of the mixture along the test rig (1 m) for unsteady case8 at different time steps

6. Conclusions
The study has focused on phase distributions in low
quality dispersed two phase flows around obstacle. It
consists of a theoretical part of a more general nature and
an experimental part highlighting bubbly flows around a
cylinder in horizontal channel. Concluding remarks are
summarized below1. A novel approach for fluid dispensing with high
consistency and accuracy had been proposed based
on the fluid dynamics of the gas-liquid two-phase
flow.
2. Random-like manner, inducing pressure, velocity and
phase fraction fluctuations: the flow is unsteady,
even when the flow rates of gas and liquid are kept
constant at the channel inlet.

3. When air discharge increases, high turbulence is

appear which generate more bubbles and waves.
4. The pressure sensor at the inlet and outlet of the test
section are record pressures that fluctuating as a
function of time due to two-phase effect. Also,
when air or water discharge increases, the mean
pressure difference increases.
5. Due to strong changes of both magnitude and
direction of local velocities of the fluid flow and
density difference between the dispersed phase and
the fluid, the local phase distribution pattern changes
markedly around the obstacle.
6. It should be noted that the prediction on the bubble
size does not correctly describe the size observed
in experiments. This is due to the difference in the
numerical definition of vapor bubble and visual
bubble boundary.

American Journal of Mechanical Engineering

7. In a water slug, bubbles move slower than the

liquid. The average velocity of the bubbles is
slightly slower than the slug tail velocity. This
means that the dispersed bubbles in the liquid
slug will be caught up by the arriving elongated
bubble.
8. Realistic bubble trajectories, with a number of
bubble trajectories entering the wake of a cylinder,
are only obtained if the effect of liquid velocity
fluctuations (or turbulence in the liquid) is simulated
and some kind of sliding phenomenon for colliding
bubbles is taken into account.
9. The effect of the existence of a cylinder is clear in
dividing the two-phase flow, generate vortices and
finally enhance mixing and the smooth obstacle
(circular cylinder) generates less bubble and
turbulence.
In this study, diameter of the bubbles is considered
constant and coalescence between the bubbles is
neglected. However, bubbles in the actual flow break
down and unite as the flow develops along the channel
and this gives a varying diameter distribution which
causes lift and drag forces to be calculated locally.
Therefore, a simulation considering the effects of differing
bubble diameter and interfacial forces is suggested for
better modeling of the flow investigated.

vm

vdr ,k

m
Subscripts
a
life
k,p,q
m
vm
w

Air
Lift force
Secondary phase
Mixture
Virtual mass force
Water

References
[1]

[2]
[3]

Nomenclature
[5]

Mass fraction

(-)

f drag

Drag function

(-)

g,g
K pq
m
m pq

Gravity acceleration
Exchange coefficient

(m/s2)
(-)

n, q
Q

Number of phases
Flow discharge

R pq

Interphase forces exchange (kg/m2.s2)

t
uq

Time
(sec)
Velocity for the qth phase (m/s)

Diameter of the particles (m)

Body force
(N)
External force
(N)

Mass flow rate

(kg/m3.s)
Interphase mass exchange (kg/m3.s)

q
t
p

[6]

[7]

[8]
[9]

(-)
(l/min)

Greek Symbols
k ,q
Volume fraction of phase k,q(-)

Secondary-phase p
(m/s2)

Turbulent
diffusivity
articles
t
2
acceleration
(N/m .s)
Turbulent diffusivity
(kg/m3)
m

[10]

[11]

[12]

[13]

Mixture density for the qth phase

Prandtl/Schmidt number

(kg/m3)
(-)

Particle relaxation time

(sec)

th

Shear stress for the q phase

(N/m2)

(-)

Drift velocity for secondary phase k

(velocity of an algebraic slip component
relative to the mixture)
(-)
Viscosity of the mixture (N/m2.s)

[4]

ck
d

Fp

Mass-averaged velocity

485

[14]

[15]

Igor A. Bolotnov, Kenneth E. Jansen, Donald A. Drew, Assad A.

Oberai, Richard T. Lahey Jr. and Michael Z. Podowski, Detached
Direct Numerical Simulations of Turbulent Two-Phase Bubbly
Channel Flow, International Journal of Multiphase Flow, 37,
(2011), 647-659.
Brennen and Christopher Earls, Fundamentals of Multiphase
Flow, Cambridge University Press, 2005.
Eckhard Krepper, Matthias Beyer, Thomas Frank, Dirk Lucas and
Horst-Michael Prasser, CFD Modelling of Polydispersed Bubbly
Two-Phase Flow Around an Obstacle, Nuclear Engineering and
Design 239 (2009) 2372-2381.
Thomas HHNE, Experiments and Numerical Simulations of
Horizontal Two Phase Flow Regimes, Seventh International
Conference on CFD in the Minerals and Process Industries,
CSIRO, Melbourne, Australia, December 2009.
I. EAMES, J. C. R. HUNT and S. E. BELCHER, Inviscid Mean
flow through and Around Groups of Bodies, J. Fluid Mech., vol.
515, pp. 371-89, Cambridge University Press, 2004.
Seyoung Lee, Changjin Lee, and Soohyung Park, Unsteady
Cavitation and Cryogenic Flow Cavitation around 2D Body,
IEEE computer society, Fifth International Conference on
Computational Science and Applications, 49. pp. 306-312, 2007.
T Degawa and T Uchiyama, Numerical Simulation of Bubbly
Flow Around Two Tandem Square-Section Cylinders by Vortex
Method, J. Mechanical Engineering Science, Proc. IMechE Vol.
222 Part C, 2008, pp. 225-234.
Xiangbin Li, Guoyu Wang, Mindi Zhang and Wei Shyy,
Structures of Supercavitating Multiphase Flows, International
Journal of Thermal Sciences 47 (2008), pp. 1263-1275.
Eckhard Krepper, Matthias Beyer, Thomas Frank, Dirk Lucas and
Horst-Michael Prasser, CFD Modelling of Polydispersed Bubbly
Two-Phase Flow Around an Obstacle, Nuclear Engineering and
Design 239 (2009) 2372-2381.
Hameed Balassim Mahood, Hala A. Kadim and Ali N. Salim,
Effect of Flow-Obstruction Geometry on Pressure Drops in
Horizontal Air-Water Two-Phase Flow, Al-Qadisiya Journal For
Engineering Sciences, Vol. 2, No. 3, 2009, pp. 641-653.
Peng Peng, Jianhua Zhang and Jinsong Zhang, Simulations for a
Novel Fluid Dispensing Technology Based on Gas-liquid Slug
Flow, International Conference on Electronic Packaging
Technology & High Density Packaging, IEEE 2011.
Hao Zhou, Guiyuan Mo, Kefa Cen, Numerical Investigation of
Dispersed GasSolid Two-Phase flow Around a Circular Cylinder
Using Lattice Boltzmann Method, Computers & Fluids, vol. 52,
pp. 130-138, 2011.
Emrah Deniz and Nurdil Eskin, Numerical Analysis of Adiabatic
Two-Phase Flow through Enlarging Channel, Istanbul Technical
University, Mechanical Engineering Faculty, Istanbul, Turkiye,
2011.
Thomas Abadie, Jolle Aubin, Dominique Legendre, Catherine
Xuereb, Hydrodynamics of GasLiquid Taylor Flow in
Rectangular Microchannels, Springer-Verlag, Microfluid
Nanofluid (2012) 12:355-369.
V. Talimi, Y. S. Muzychka, S. Kocabiyik, A Review on
Numerical Studies of Slug flow Hydrodynamics and Heat Transfer
in Microtubes and Microchannels, International Journal of
Multiphase Flow, vol. 39, pp. 88-104, 2012.

486

American Journal of Mechanical Engineering

[16] Introductory FLUENT Notes, FLUENT v6.3, Fluent User Services

Center, December 2006.
[17] ANSYS 13.0 Help, FLUENT Theory Guide, Mixture Multiphase
Model.
[18] Esam M. Abed and Riyadh S. Al-Turaihi, Experimental Study of
Two-Phase Flow around Hydrofoil in Open Channel, Journal for

Mechanical and Materials Engineering, Iraq, 2012, accepted and

submitted for publication.
[19] Riyadh S. Al-Turaihi, Experimental Investigation of Two-Phase
Flow (Gas Liquid) Around a Straight Hydrofoil in Rectangular
Channel, Journal of Babylon University, Iraq, 2012, accepted and
submitted for publication.