Ali 1

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/340731475
Investigation of Forced Convection Boiling Heat Transfer for R-134a Flow in a

Vertical Tube Filled by Metal Foam
Article · June 2019
CITATION READS
1 241
2 authors:
Ali Samir Ihsan Y. Hussain

University of Baghdad University of Baghdad - College of Engineering
6 PUBLICATIONS 8 CITATIONS 196 PUBLICATIONS 170 CITATIONS
SEE PROFILE SEE PROFILE
All content following this page was uploaded by Ali Samir on 18 April 2020.
The user has requested enhancement of the downloaded file.

International Journal of Mechanical & Mechatronics Engineering IJMME-IJENS Vol:19 No:03 96
Investigation of Forced Convection Boiling Heat

Transfer for R-134a Flow in a Vertical Tube
Filled by Metal Foam
Ali Samir #1, Ihsan Y. Hussain #2
#1 Refrigeration & Air-Conditioning Technical Engineering Department, college of Technical Engineering. The Islamic
University ,Najaf, Iraq, ali.samir.aneed@gmail.com
#2 Mechanical Engineering Department, College of Engineering, University of Baghdad, Baghdad, Aljadria, 47024, Iraq,
drihsan@uobaghdad.edu.iq , dr.ihsanyahya1@gmail.com
Abstract— forced convection boiling heat transfer in vertical employed in the model owing to high flow velocity including
tube filled by metal foam for Refrigerant(R-134a) (the working Forchheimer regime (laminar flow). Hence, extension of
fluid was used because it is zero ozone depletion) is numerically
investigated using the modified multiphase mixture model for
Darcy law, which is called Forchheimer’s law, was used
various heat flux. The effect of changing mass flux, porosity and instead of Darcy’s law. The inertia effect on the flow becomes
PPI was analysed. It was found that, both of pressure drop and significant. Taking into account Forchheimer's correlation is
heat transfer increase as mass flux increased and higher values the only difference from the formulation of Wang's model,
are obtained with higher PPI and lower porosity. Vapour
dryness reaches its maximum value at the exit (about 30%) with and the other derivation processes for the fluid phase are
higher PPI, lower porosity, and low mass flux. Results also show completely the same as Wang's ones [4].
that the local Nusselt number increases more rapidly along the [11] and [12] carried out analytical study for heat transfer
heated wall when heat flux increased which leads to form a two-
(forced convection), fully developed in pipe filled by metal
phase zone adjacent to the heated wall, while the average Nusselt
number increases as mass flux increases for constant heat flux foam and annulus of tube-in-tube heat exchangers. They
value and will be higher for high PPI metal foam. concluded that the existence of metal foams could
significantly enhance the coefficient of heat transfer and high-
Index Term-- Metal Foam; Forced Convection; Boiling;
pressure drop. “(Du et al., 2011)” investigated analytically
Multiphase Mixture Model; Forchheimer; Vertical Tube.
(based on a modified analytically model) the metal-foam filled
I. INTRODUCTION tube flow boiling. Microstructure model of metal foams was
Interest in using metal foam in heat transfer and fluid found, by which fiber diameter and density of surface area
flow applications motivate the need for fully characterizing were completely predicted. The flow boiling heat transfer
them more imperious. Examples for application include model in metal foams was established on annular pattern, in
thermally enhanced oil recovery, heaters, compact heat which two phase fluid was assumed that the vapour zone in
exchangers, silencers, flame arrester, mine heat exchanger, the center of the tube and liquid zone annularly near the wall.
and electrochemical applications and geothermal energy Due to that, nucleate boiling performed only in the liquid
exploitation, [1]. Boiling in metal foam is analytically zone. To obtain the coefficient of heat transfer at interface, for
intractable and poorly understood because it is complicated metal foams, fin analysis and heat transfer network were
process, which involves heat transfer and multi-phase flow. integrated. The verification of analytical solution was showed
This means, it should be dealt with moving and irregular a good agreement with experimental data. [6] investigated the
interfaces between the single and two-phase sub regions and single and two-phase mechanism through a vertical channel
nonlinear and coupled nature of the governing equations for (10 *50*200 mm) filled with copper foam (40 PPI and 95%
two-phase flow. To eliminate these difficulties, an alternative porosity) experimentally, the working fluid was n-pentane
approach, Multiphase Mixture Model (MMM), was proposed flowing upward. Two methods of inserting copper foam were
by [3] in which the governing equations for the single- and used; one was soldered to the wall, while the other is just
two-phase regions are a single equation. Hence, it is popularly incorporated in the channel. For single-phase zone, the
used in the study of two-phase flow in metal foam. Numerical Forchheimer model was used to evaluated the permeability
solution had been obtained by [1] to investigate boiling with and inertial the coefficient, then concluded that the heat
thermal convection in a porous layer heated from below using transfer coefficient was increased by with limited increase in
MMM. The main assumption of creeping flow (Darcy flow) pressure drop twice with the presence of metallic foam. In
through the porous media as used in Wang’s model was not two-phase zone, the heat flux was strongly increased by
191303-2828-IJMME-IJENS © June 2019 IJENS

IJE NS
increasing the superheat (Tw-Tsat). With the mass flux the experimental data for different metallic-foam inserted in
velocity (10– 40 kg /m2 s) there is no effect on the heat flux, channels for the developed correlation which is taking account
even at low superheat. Finally, they compared performance of the effect of diameter on pressure drop characteristics for
copper foam to other materials. For the same conditions the working fluid flow boiling. Then a new heat transfer
boiling curve for bronze sintered fibers is similar to the foam correlation is taking account of tube diameter influence was
inserted at low superheat for the same conditions. Moreover, developed. Enhancement of heat transfer characteristics with
for the same mass and heat transfer fluxes, the pressure drop presence of oil by a 50% under intermittent and low vapor
in the fibers is 10 times larger than in the foam. A modified qualities, whilst it deteriorates the characteristics of the heat
two-phase mixture model was developed by [7] to simulate transfer by a 35% under high quality of vapor and
high velocity flow boiling under high heat flux in a porous concentration of high oil. [16] studied experimentally the
medium, which is a cylindrical bronze-particle-sintered one characteristics of boiling heat transfer for R410A flow in
with a porosity of 42% and particle size of 160 μm in tubes (diameter of 23.4 or 13.8 mm) filled with PPI (10 or 5)
diameter. These modifications include adding the inertial copper foam and porosity 95%, at a saturation temperature of
force in the momentum equation by using Ergun's Law instead 7 ˚ C, mass flux ranges 10 - 90 kg /m2 s, vapor quality (0.1 -
of Darcy's Law and adopting the LTNE model (Local Thermal 0.8) and heat flux range (3.1 to 18.3) kW /m 2. The results
Non-Equilibrium) in the energy equation. One of their major showed that: Metallic-foam cause that the coefficient of heat
findings are that there was no significant temperature transfer was enhanced by a maximum of 220%. The heat
difference between the LTE and LTNE models unless the heat transfer coefficient decreases with vapor quality when the
flux is above 1 MW/m2. Although the authors argued that, mass flux is low, whilst the opposite is true. A model for
more realistic and reasonable results could be obtained for the predicting the coefficient of heat transfer was developed with
two-phase flow under high heat flux and high velocity by respect to flow pattern. It covers the working condition of tube
using their models, the involvement of a large number of diameter of 13.8-26 mm, metal foam of 5-40 PPI and 90%-
empirical equations in the unknown needs to be further 95% porosity, mass flux of 10-106 kg /m2 s and heat flux of
verified by experiments. [13] investigated with experimental 3.1-19 kW/ m2. The heat transfer model agrees with 98% the
work heat transfer and the boiling of fluid flow in tubes filled experimental data in this work, and 95% of the data from
by metal-foam (20-40 PPI). The effects of mass and heat experiments in previous work by a deviation of ±20% for all.
fluxes, microstructures of foam and pressure on pressure drop [9] investigated experimentally the boiling of two phase flow
and boiling heat transfer were presented: first, for higher mass in metallic-foam filled a plate heat exchanger. The working
flux, the coefficient of heat transfer becomes larger when the fluid were pure R245fa and a zeotropic mixture of
quality of vapour increased. Approximately, the coefficient of R245fa/R134a (molar ratio is 0.6/0.4). The metal foam
heat transfer for tubes filled by foam is larger than plain tubes samples were (20, 30, 60 PPI) with constant porosity. The
by three times. The coefficients of heat transfer could be results indicated that both zeotropic mixture and R245fa
approximately doubled as the PPI changed (from 20 to 40), refrigerant were presented. Compared to a conventional
because of the increasing surface area, flow mixing and evaporator; the (20 PPI) foam causes an increase of 230% in
density. The main cause of limit enhancing effect is due to, the coefficient of heat transfer, while for (30 and 60 PPI) the
broken foam structures and thermal contact resistance increases were 200% and 130%, respectively. Numerical
between the metal foam and tube surface. [14]and [15] carried solutions using the new model were obtained by (Zhao and
out experimental investigation of flow boiling for (R410A)-oil Liao,[5]) to study boiling heat transfer in a vertical packed bed
mixture in metal-foam (95 % porosity with 5PPI and 10 PPI) subjected to either aiding or opposing flows. They examined
filled tubes with diameters of (13.8 and 7.9 mm). The effect of the effect of particle diameters on the minimum liquid
tube diameter on pressure drop and characteristic of heat saturation, also for aiding and opposing flows, their results
transfer of refrigerant-oil mixture flow boiling was showed that both flow field and temperature distribution
investigated with respect to get experimental data of 13.8 mm greatly affected by the direction of the incoming fluid. Finally,
and 7.9 mm diameter both with other previous studies of (26.0 (Ali Samir and Ihsan Y.Hussain,2019)[19] investigated
and 23.4) mm tubes. The main conclusions were: Increasing natural convection boiling heat transfer for Refrigerant R-
of PPI lead to increase both pressure drop and coefficient of 134a flow in a metal foam filled vertical tube is numerically
heat transfer. The coefficient of heat transfer decreased in by using the Multiphase Mixture Model MMM at
range (24-32%) as the diameter of tube decreased (from 23.4 constant heat flux for both transient and steady state
to 13.8 mm), if the decreasing in the diameter of tube (from behaviour under different parameters. The effect of changing
13.8 to 7.9 mm), it is within 10-42%. A good agreement with imposed heat flux, porosity, PPI (pore per inch), and tube

IJE NS
diameter was analysed. It is found that liquid saturation will Assumptions
reach its minimum value at the exit, and the dry out point will 1. The metal foam is considered as a porous media
be seen always at the 2nd half of the tube length. Results also and is rigid, uniform, isotropic and fully saturated
show that smaller tube diameter and higher porosity for the with fluid.
same PPI caused a higher temperature variation over the entire 2. Forchheimer’s law is introduced instead of
metal foam tube. Darcy’s law for high velocities and high heat flux
In the present work, forced convection boiling heat and it is applicable to the both liquid and vapor
transfer in a vertical metal foam filled tube subjected to a phases.
range of heat flux, as shown in Fig. 1, will be numerically 3. At the inlet the flow is uniform and at the outlet
solved based on modified MMM, taking into account that the flow is fully develop flow.
Forchheimer's correlation is the only difference from the 4. Axisymmetric laminar flow
formulation of Wang's model, and the other derivation 5. The thermophysical properties of both solid and
processes for the fluid phase are completely the same as fluid are assumed constant.
Wang's ones [4]. The main objective is to study the influence 6. The vapor, liquid and solid phases are in local
of forced convection heat transfer on the flow field and the thermal equilibrium (LTE).
associated process in such system. The influence of various 7. No internal heat generation (Heat source).
parameters such as the porosity, PPI, mass flux also examined. 8. Boundaries are impermeable.
The working fluid used in the present work is R-134a due to 9. Thermal dispersion and chemical reactions in the
its use in wide range of applications, and due to its low system are negligible.
evaporation enthalpy.
B. Governing Equations
The governing equations obtained by adopting the
multiphase mixture model MMM [2] are:
Conservation of mass:
(1)
Conservation of momentum:
(2)
Where
K: is the absolute permeability
Fig. 1. Schematic of the physical problem.
f: “is the inertial coefficient which are strongly related to the
II. MATHEMATICAL FORMULATIONS structure of metal foam” [10].
Conservation of energy:
A. PROBLEM DESCRIPTION
A metal foam filled vertical tube, initially saturated

with a liquid refrigerant supplied from a pump at various
mass flux, is considered as two-dimensional geometric (3)
configuration with the coordinate system shown in Fig. 1. The mixture variables and properties in Eqs. (1) -(3) are
The boundary for bottom is liquid with inlet velocity defined as:
while the top boundary is permeable. The applied heat flux Density:
is uniform and constant. As the tube heated from outside (4)
the continuously flowing liquid absorb the heat and caused
the liquid to evaporate. When the heat flux increases Velocity:
sufficiently, boiling occurs near the vertical heated wall (5)
and extends to the neighbouring region. The flow velocity Enthalpy:
is within the laminar regime with ReK > 1(Forchheimer (6)
regime).
(7)

IJE NS
Kinetic density: With the hindrance function f (s) given by:
(21)
(8)
The following constitutive relations are used for the
Viscosity:
relative permeabilities [17]:
(9) (22)
Advection correction coefficient: (23)
(10) The effective thermal conductivity was calculated based on [8]
as:
Effective heat capacitance ratio:
(24)
(11)
Where,
Effective thermal diffusion coefficient:
(12)
Capillary diffusion coefficient:

(13)
Relative mobility:
, (14)
And e=0.189
The temperature and the liquid saturation can be calculated
from the enthalpy as: The porosity for copper metal foam are 98% and 95 with
80,60,40 PPI for each porosity while other thermophysical
properties of the water-borosilicate glass beads are listed in
Table I.
TABLE I
Thermophysical Properties of the Water-Borosilicate Glass Beads System.
R134a at 0.6 Mpa
Property Symbol Solid Liquid Vapor
(15) Copper R-134a R-134a
foam
,
Density 8940 1219.68 29.16
Thermal k,
391 0.0 84 ---
conductivity W/m K
(16) Specific C,
385 1411 1010.36
Subscripts l and v refer to the liquid and vapour heat J/kg K
respectively. The liquid and vapour velocity can be calculated
Kinetic
from the mixture velocity as follows: ---
viscosity
(17) Expansion β,
---
(18) coefficient
Where j is a mass diffusion flux: Interfacial σ
--- 0.0079
tension N/m
(19) Latent heat
of ---
Where is the capillary diffusion: J/kg
evaporative
(20)

IJE NS
C. Initial and Boundary Conditions Where;
The initial and boundary conditions for the present problem
are given as follows: B=
1) Initial Conditions
At t=0, It can be written for two-dimensional polar coordinate
system as;
(24)
u=v=0 (25) r- Momentum equation
S=1 (26)
2) Boundary Conditions
At Z = 0 (inlet) (36)
(27)
z- Momentum equation
(28)
At Z = L (outlet)
(37)
(Fully developed) (29)
The above equations will be discretized by using the
(30)
finite volume method. By finite difference, eqs. (36) and
At axisymmetric axis, r = 0 (37) will be;
(31)
(38)
(32)
At the wall, r = R
(impermeable) (33) (39)
(34) And for Z direction;
D. Numerical Procedure
(40)
because of the nonlinearity of the momentum equation
(2) and pressure and velocity distributions are strongly
interlinked, the direct method cannot be used and have to (41)
be solved eq. (1) and eq. (2) have to be solved
simultaneously with iterative method by SIMPLE The procedure of the SIMPLE algorithm will be
algorithm. The SIMPLE algorithm starts with guessed started with guess value for the pressure field P*, then we
pressure and velocity fields that are first input to the obtain the starred values of the velocities u*and v* from;
momentum equations to obtain improved values for the
velocities. The improved velocities are used in the
(42)
pressure correction equation, which is obtain by
combining the mass and momentum conservation
equations. The pressure corrections are used to yield (43)
corrected pressures and velocities, which are again used in
the momentum equations at the next cycle. Multiple
iterations are needed to reach a converged solution. In (44)
addition, the finite volume method is used to derive the
discretization equations.
At first, the momentum equation (2) in general form as; (45)
(35)

IJE NS
Now it will be defined the correction P' as the
difference between the correct pressure field P and the
(53)
guessed pressure field P*, as shown below;
Table II
P=P*+P' (46) Grid Size with Relative Error.
Similarly, the velocity corrections u' and v' will be defined Relative
Grid size
as; error
u=u*+u' (47a) 1005 0.35

1641 0.12
v=v*+v' (47b)
2771 0.01
Subtraction eqs. (42) - (45) from eqs. (38)- (41), with
3277 0.00425
some mathematical manipulation, respectively gives;
3981 0.00156
(48) 5543 0.0006321

6681 0.000011
(49)
III. RESULTS AND DISCUSSION

(50) For all cases studied, the inner diameter of the tube is
16 mm with length 140 mm. Six tubes filled with copper
foam with different specifications (10,20 and 40 PPI with
(51) porosity 95 % and 98% for each). The working fluid is
used R-134a at pressure 0.6 MPa which enters the domain
From continuity equation (1) for two dimensional with a uniform velocity and a constant temperature of 10
o
coordinates; C. Different values of imposed heat flux are used (10-80
kW/m2).
(52) A. Pressure Drop Characteristics

At first, the main concern about using metal foam in
To discretize this equation, integrate it over the control flow boiling heat transfer application is the pressure drop.
volumes and over a time interval from t to t+Δt, substitute eqs. Fig. (2) shows the variation of average pressure drop as a
(48) - (52), which represent the velocity correction formulas in function of mass flux for six copper foams with different
this equation, then pressure equation is solved by TMDA specifications (10,20 and 40 PPI with porosity 95 % and
algorithm (a line-by-line tri-diagonal matrix). Then the 98 % for each). It can be seen that as expected if the mass
velocity field can be obtained from the momentum equation, flux rise, the drop of pressure will have a remarkable
Eq.(2) with the results of pressure field. The individual increase. For each porosity, as the cell size becomes
velocities of vapour and liquid are then calculated by Eqs. (17) smaller (80 PPI), the pressure drop significantly increases
and (18), and they are stored at the interfaces of the control and this indicates that the flow observes much more flow
volumes. A fully implicit control volume-based finite resistance for higher PPI due to the presence of metal-
difference formulation with the use of the power law scheme foam structures. The flow resistance in the channels of
of Patankar [18] developed a solution to the energy equation, foam mainly involve three parts, viscous resistance of
to discretize the combined convective and conductive terms. Darcy, drag resistance of channel surface and form
The temperature profile and the saturation of liquid are resistance due to metal-foam structures. Foam 98%
calculated from the enthalpy based on Eqs. (15) and (16). A porosity and 40PPI presents the best performance, due to
relative error of less than is required for both the the lowest pressure drops. Comparing the six foams the
velocity and temperature fields between successive iterations pressure drop at low mass flow rate almost was with less
as shown in Table II , and the normalized residual for the difference, while at high mass flow rate foam 80PPI and
pressure and enthalpy field are less than to achieve 95% porosity presents the highest pressure drops.
convergence.

IJE NS
B. Heat transfer Characteristics various mass. At first, Fig. (8) shows the liquid saturation
Generally, the wall temperatures are higher than those in for six foam sizes at constant heat flux (10Kw/m2) for
foam cavity due to heat transfer from the wall to the fluid. mass flux (2.5,8, and 11 kg/s.m2), its noticeable that at
As expected, both wall and foam cavity temperature constant mass flux as the cell size increases the liquid
increased with increased heat flux and the thermal saturation decreases, also (expand in two-phase zone),
boundary layer thickness increased upward on the hot wall. while increasing the porosity increases the liquid
The temperature distribution for the case (40PPI with saturation. It is obvious from Fig. (8a) that the liquid
ε=95%) for two mass fluxes (11and 15 kg/s.m2) and the saturation decreases drastically and becomes least along
same heat flux value are shown in Figs. (3) and (4), the heated wall with the lowest mass flux (Fig. (8a)), it
respectively. As shown in Fig. (3) for low mass flux reaches a minimum value at the outlet of the tube for a
(G=11 kg/s.m2) buoyancy force is dominant, by increasing specific heat flux. This observation means that the dry-out
the imposed heat flux, the energy conducted to the region point always occurs at the outlet of the tube for the
near the heated tube wall increases also and yield an geometry under consideration. As the mass flux increases
increase in the fluid temperature and take enough time to (as in Fig. (8) b and c)) the liquid saturation along the
reach to the temperature of evaporation and extension to heated wall for the same imposed heat flux decreases,
two phase region. It is noticed that at 40 kW/m2, as shown because high mass flux causes a reduction in the
in Fig. (3c), the evaporation in the copper foam region temperature of the heated wall, and as a consequence a
become so vigorous and most tube heated covered with reduction in the liquid saturation along the heated wall.
two phase region. In Fig. (4) for higher mass flux than Thus, the tube that subjected to a larger mass flux gives a
those in Fig. (3) and at same heat fluxes, the evaporation is higher dryout heat flux. The minimum liquid saturation
delayed due to reduction in the fluid temperature nearby progressively increases slightly with the increase of the
the heated wall because of higher fluid velocity and the pore density of foam.
fluid temperature attains the evaporation point late As the heat flux increases, the liquid phase zone
compared with lower mass flux. Generally, with increase contracts whilst the two-phase zone expands. As shown in
of the inlet velocity, the heat transfer from the heated wall Figs. (9) and (10) the liquid saturation becomes smaller
to the fluid increased and a smaller amount of vapor is along the heated wall with the increase of the imposed heat
generated, and the two-phase zone shrinks substantially. flux and for q=40 kW/m2, reaches a minimum value at the
The temperature distribution for the case (60PPI with exit of the channel for a specific heat flux.
ε=95%) for mass flux (11kg/s.m2) and heat flux values Fig. (11) presents the steady local Nusselt number on
(10-40 Kw/m2) are shown in Fig. (5) compared with Fig. the heated wall for the 40 PPI and 95% porosity cell size
(3) case (40PPI with ε=95%), the two- phase zone appears with mass flux 15kg/m2, at various heat fluxes (10,20 and
earlier for 60 PPI because of decreasing of permeability as 40 kW/m2). It can be seen from this figure that the local
the pore size decreased (increase in PPI) which cause Nusselt number have the same behavior over three applied
higher flow resistance (increases the drag force over the values of the imposed heat flux. As increasing heat rate to
whole metal foam tube will increase) and allowed to fluid the tube wall, the superheated temperature difference
to absorbed more energy. decreases upstream, and causes an increase in the heat
The temperature distribution for cases (80PPI with transfer rate depending on the value of the imposed heat
ε=95%) and (80PPI with ε=98%) for mass fluxes (20 flux. For imposed heat flux of 10 kW/m2, the local Nusselt
kg/s.m2) and same heat flux values are shown in Figs. (6) number increases slightly and reaches its maximum value
and (7), respectively. When the porosity decreases the drag at 53% of tube length. As the imposed heat flux further
force over the whole metal foam tube increase. This increases for the highest investigated value (40 kW/m 2),
perhaps because of the fact that the metal foam tube with a the local Nusselt number increases more rapidly along the
smaller porosity has a lower permeability, which leads to heated wall and reaches maximum value at 0.07% of tube
higher resistance for the mass flux. This can be explaining length. At this point, a two-phase zone is formed adjacent
why at the same heat flux higher fluid temperatures to the heated wall, then a discontinuity in the temperature
difference for higher porosity, and boiling will began gradient at the heated wall occurs and the heat transfer
earlier in the higher porosity tube. coefficient reaches its maximum defined value just before
Figs. (8) to (10) present the variation of the liquid the two-phase zone existence adjacent to the heated wall.
saturation along the completely heated wall for the In Fig. (11), a general trend was observed for the variation
different cell sizes of foams at various heat flux and of the local heat transfer coefficient and the local Nusselt

IJE NS
number with the increase of the heat flux, that the drag force near the exit extremely large and a vapor
maximum defined value of the local heat transfer choking take place in the upstream and as a consequence a
coefficient increases as the heat flux increases till at a drop in the heat transfer process that can be seen as a
certain value of the imposed heat flux where the two-phase decrease in the magnitude of the defined value Nusselt
zone covers the upper section of the tube. This causes a number. As the mass flux increase the value Nusselt
drag force near the exit extremely large and a vapor number will increase slightly till reach to max value (118)
choking take place in the upstream and as a consequence a at 20kg/s.m2.
drop in the heat transfer process that can be seen as a Fig. (14) demonstrates the variation of the average
decrease in the magnitude of the maximum defined value Nusselt number with Reynolds number at constant
Nusselt number as shown in Fig. (11). imposed heat flux for different cell sizes of foams. In this
The effect of heat flux on the local Nussult number for figure, temperature gradient increase as the Reynolds
various cell size of metal foam is shown in Fig. (12) for number increasing and hence the Nusselt number increases
mass flux of 15 kg/s.m2. It can be seen from Fig. (12a) that also. This, of course, is owing to the definition of the
for lowest imposed heat flux (10 kW/m2) the local Nusselt Nusselt number, which indicates a direct proportionality
number increased as the pore density increased (lower between the Nusselt number and temperature gradient on
permeability) for the same heat flux. This may be referred the tube wall. Also, it is obvious from Fig. (14) that the
to the fact that batter temperature distribution due to cell sizes variation does make variation on the Nusselt
increase in the drag force (because of the low permeability) number at certain imposed heat flux values and a reduction
over the hole domain at same heat flux. As the temperature in the simulation time of the foam zone because of the
increased in the foam tube for a lower permeability, the increasing of the liquid flow rate especially when
temperature difference between the local wall temperature convective flow boiling is invoked and convection
Tw and the saturation temperature Tsat is decreased and currents become vigorous.
caused an increased in the local Nusselt number. As the
heat flux is further increased to 20 kW/m2 and expanded IV. COMPARISON WITH PREVIOUS THEORETICAL
the two-phase zone, the maximum defined value of the RESULTS
local Nusselt number is increased as the permeability In literature, no numerical solution is found that adapts
increased as shown in Fig. (12b). This behaviour is related the modified (MMM) for forced boiling of R134a to compare
to the size of the two-phase zone that formed in the foam with. Therefore, the present work is compared with a closet
region at each permeability. It can be seen from Fig. (12c) work [9] where the working fluid was zeotropic mixture of
that the two-phase zone is retreat to the tube exit section as R245fa/R134a (0.6/0.4 molar ratio) and 60 PPI. For
the permeability increased, and this caused a reduction in comparing pressure drop with mass flux refer to Fig. (15). In
the drag force and as a consequence, an improvement in general, it can be seen that the comparison indicates similar
overall the heat transfer process that can be seen as an behavior for same mass flux range. Also, both works indicate
increase in the magnitude of the maximum defined value that metal foam with higher both of PPI and mass flux causes
of the local Nusselt number as the permeability increased a higher pressure drop.
for the same imposed heat flux. For heat transfer side, the comparison was with "(Y. P.
Fig. (13) demonstrates the variation of steady state Du et al, 2011)", in this work the working fluid was R-134a
local Nusselt number on the heated wall for the 40 PPI and flow in horizontal tube but with different mathematical model.
98% porosity cell size with heat flux 10kW/m2, at various As shown in Fig. (16), the heat transfer coefficient had same
mass fluxes (2.5,8,11,15 and 20 kg/s.m2). It can be notice behavior as the vapor quality change. But the difference in
that for lowest mass flux (2.5 kg/s.m2) the local Nusslet values can be explain by two reason, first the difference in two
number will be at lowest value. This may be referred to the cases the present work with vertical flow while "(Y. P. Du et
fact that at low inlet velocity the fluid will take enough al, 2011)" horizontal flow, the 2nd is due to difference in mass
time to reach for saturation temperature and two phase flux and PPI.
zone expanded earlier, this will lead to lower Nusselt V. CONCLUSIONS
number for two reasons; first, because of the fluid reach to In the present work, a numerical investigation was
saturation temperature near the entrance the wall done for refrigerant (R-134a) forced convection boiling
temperature will be increased and reducing the magnitude heat transfer in a vertical tube filled with metal foam. The
of local Nusselt number while the second, the two-phase effect of variation of the applied heat flux, PPI, porosity,
zone covers the upper section of the tube. This causes a

IJE NS
and mass flux were investigated. The following are the
conclusions drawn from the present investigation;
1- As mass flux rises, the pressure drop dramatically
increases. The average pressure gradients along the
tube linearly decrease and high PPI of foam causes
largest pressure losses.
2- Enhancement in heat transfer occurred as mass flux
increases and more enhancement is obtained with
higher PPI, and lower porosity
3- Liquid saturation attains its minimum value
(about70%) at the exit with higher PPI, lower
porosity and /or low mass flux.
4- The local Nusselt number increases more rapidly
along the heated wall when heat flux increased which
leads to form a two-phase zone adjacent to the heated
wall, then a discontinuity in the temperature gradient
at the heated wall occurred and the local Nusselt
Number reaches its maximum defined value just
before the two-phase zone exists. lower permeability qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2
(a) (b) (c)
will lead to higher local Nusselt number. Fig. 3. Steady State Temperature Contours for the Mass Flux
5- The average Nusselt number increases as mass flux G =11 kg/s.m2 and Foam Cell Size 40PPI, ε =0.95
increases for constant heat flux value and will be
higher for high PPI metal foam.
qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

(a) (b) (c)
Fig. 2. Pressure Drop Through Tube with Mass Flux Fig. 4. Steady State Temperature Contours for the Mass Flux
for Different Cell Sizes of Foams. G =15 kg/s.m2 and Foam Cell Size 40PPI, ε =0.95.

IJE NS
qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2 qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

(a) (b) (c) (a) (b) (c)
Fig. 5. Steady State Temperature Contours for the Mass Flux Fig. 7. Steady State Temperature Contours for the Mass Flux
G =11 kg/s.m2 and Foam Cell Size 60PPI, ε =0.95. G =20 kg/s.m2 and Foam Cell Size 80PPI, ε =0.98.
qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

(a) (b) (c)
Fig. 6. Steady State Temperature Contours for the Mass Flux
G =20 kg/s.m2 and Foam Cell Size 80PPI, ε =0.95.
G=2.5 kg/s.m2
(a)

IJE NS
G=11 kg/s.m2
(c)
Fig. 8. Steady State Liquid Saturation Distribution Along
the Heated Wall for qw=10KW/m2, and Different Cell Sizes.
G=8 kg/s.m2
(b)
G=8 kg/s.m2
(a)

IJE NS
G=11 kg/s.m2 G=15 kg/s.m2

(b) (c)

IJE NS
G=15 kg/s.m2
(b)
G=11 kg/s.m2
(a)
G=20 kg/s.m2
(c)

IJE NS
Fig. 11. Steady State Local Nusslet Number for 40 PPI with Qw=20KW/m2
95% Porosity, G=15 kg/s.m2 and Different Heat Fluxes.
(b)
qw=10KW/m2 Qw=40KW/m2
(a) (c)
Fig. 12. Steady State Local Nusslet Number for Different Cell Sizes, G=15
2
kg/s.m , and Different Heat Fluxes.

IJE NS
Fig. 13. Steady State Local Nusslet Number for 40 PPI with
98% Porosity Cell Sizes, qw=10 kW/m2, and Different Mass Fluxes.
Fig. 16. Comparison of Heat Transfer Coefficient with vapour quality

between present work and (Y. P. Du et al ,2011).
REFERENCES
[1] C. Y. Wang and P. Cheng, "Multiphase flow and heat transfer in
porous media", Advances in Heat Transfer, vol. 30, pp. 93–196, 1997.
[2] C.Y. Wang and C. Beckermann "A two-phase mixture model of
liquid–gas flow and heat transfer in capillary porous media–I.
Formulation" Int. J. Heat Mass Transfer, vol. 36, No. 11, pp. 2747–
2758, 1993.
[3] C.Y. Wang, C. Beckermann and C. Fan "Numerical study of
boiling and natural convection in capillary porous media using the
two-phase mixture model" Numer. Heat Transf., Part A, vol. 26, pp.
375-398, 1994.
[4] C.Y. Wang "A fixed-grid numerical algorithm for two-phase
flow and heat transfer in porous media" Numer. Heat Transf., Part
B, vol. 32, pp. 85-105, 1997.
Fig. 14. Steady State Average Nusslet Number for Different Cell Sizes, qw=10
kW/m2. [5] T.S. Zhao and Q. Liao "Mixed convective boiling heat transfer in
a vertical capillary structure heated asymmetrically"J. Thermophys.
Heat Transf. vol. 13, pp. 302-307, 1999.
[6] Topin, F., Bonnet J., Madani, B. and Tadrist L "Experimental
Analysis of Multiphase Flow in Metallic foam: Flow Laws" Heat
Transfer and Convective Boiling Advanced Eng. Materials, Vol.8,
No.9, pp.890-899, (2006).
[7] Yuki, K., Abei , J., Hashizume, H., and Toda, S. "Numerical
Investigation of Thermofluid Flow Characteristics with Phase
Change Against High Heat Flux in Porous Media" ASME Journal
of Heat Transfer, Vol. 130, pp. 1-12 ,(2008).
[8] Dai, Z., Nawaz, K., Park, Y., Bock, J., Jacobi, A.: "Correcting and
extending the Boomsma–Poulikakos effective thermal conductivity
model for three-dimensional, ﬂuid-saturated metal foams." Int.
Commun. Heat Mass Transf. 37(6), 575–580. 2010
Fig. 15. Comparison of Pressure Drop with mass flux between present work
and [9]. [9] Gholamreza Bamorovat Abadi, Kyung Chun Kim "Enhancement
of phase-change evaporators with zeotropic refrigerant mixture
using metal foams" International Journal of Heat and Mass Transfer
Vol. 106, pp. 908–919. 2017

IJE NS
[10] Nield, D. A. and Bejan A. "Convection in Porous Media" New Oil Mixture Flow Boiling in Metal-Foam Filled Tubes" Int. J.
York: Springer-Verlag (2006). Refrigeration, Vol.41 , pp.121-136, (2014b).
[11] Lu, W., Zhao, C.Y. and Tassou, S.A. "Thermal Analysis on [16] Zhu, Y., Hua , H., Sun, S. and Ding,G. "Heat Transfer
Metal-Foam Filled Heat Exchangers, Part I: Metal-Foam Filled Measurements and Correlation of Refrigerant Flow Boiling in
Pipes." Int. J. of Heat Mass Transfer, Vol.49, No.15-16, pp. 2751- Tube Filled with Copper Foam" Int. J. Refrigeration, Vol.38, pp.215-
226,(2014).
2761, 2006.
[17] K.S. Udell "Heat Transfer in Porous Media Considering Phase
[12] Zhao, C.Y., Lu, W. and Tassou, S.A."Thermal Analysis on
Change and Capillarity–The Heat Pipe Effect" Int. J. Heat and
Metal-Foam Filled Heat Exchangers, Part II: Tube Heat
Mass Transfer, Vol.28 (2), pp.485-495, 1985.
Exchangers" Int. J. of Heat Mass Transfer, Vol.49, No.15-16, pp.
[18] S. V. Patankar "Numerical heat transfer and fluid flow. " New
2762-2770, 2006.
York: Hemisphere, 1980.
[13] Zhao,C.Y., Lu,W.and Tassou,S.A, "low Boiling Heat Transfer in
[19] Ali Samir and Ihsan Y. Hussain "Simulation of natural
Horizontal Metal Foam Tubes" ASME J. Heat Transfer ,Vol.131 ,
pp.121002-1-121002-7, (2009). convection boiling heat transfer for refrigerant R-134a flow in a
metal foam filled vertical tube" Case Studies in Thermal
[14] Hua ,H., Zhu, Y., Peng, H.,Ding,G. and Sun, S., "Effect of Tube Engineering Volume 13, March 2019, 100390
Diameter on Pressure Drop Characteristics of Refrigerant Oil
Mixture Flow Boiling Inside Metal-Foam Filled Tubes"Applied
Thermal Engineering,Vol.62, pp.433-443,(2014a).
[15] Hua ,H., Zhu, Y., Peng, H.,Ding,G. and Sun, S., "Influence of NOMENCLATURE
Tube Diameter on Heat Transfer Characteristics of Refrigerant-
Letter Description Units H Specific enthalpy J/kg
 Thermal diffusivity m2/s hfg Latent heat of liquid /vapor J/kg
eff. Effective thermal m2/s phase change
conductivity of the porous J Diffusive mass flux kg/(m2 s)
media kr Relative permeability -
 Volumetric thermal 1/K keff Effective thermal W/(m K)
expansion coefficient conductivity
Two-phase advection - kf Thermal conductivity of W/(m K)
correction coefficient the fluid
Effective thermal diffusion m2/s ks Thermal conductivity of W/(m K)
coefficient the solid
∆ρ Density difference (=ρl-ρv) kg/m3 K Absolute permeability m2
ε Porosity - L Length in Z-direction M
ζ Under-relaxation factor - P Mixture pressure Pa
λ Relative mobility - q” Imposed heat flux W/m2
μ Dynamic viscosity kg m/s R Radial coordinate M
V Kinetic viscosity m2/s S Liquid saturation -
ρ Density kg/m3 T Temperature °C
Surface tension N/m t Time S
Ω Effective heat capacitance - Superficial or Darcian m/s
ratio velocity vector
Del operator - u Superficial or Darcian m/s
velocity vector in
Δh Hydrostatic head Mm
r-direction
Δr, ΔZ Distance between the grid - v Superficial or Darcian m/s
points
velocity vector in
C Specific heat J/kg K Z-direction
D(s) Capillary diffusion m2/s Z Axial coordinate M
coefficient
D Diffusion term
F Convection term
f(s) Hindrance function -
G Gravitational acceleration m/s2

IJE NS
View publication stats

Ali 1

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Ali 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ali 1

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Investigation of Forced Convection Boiling Heat Transfer for R-134a Flow in a

Article · June 2019

Ali Samir Ihsan Y. Hussain

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

Investigation of Forced Convection Boiling Heat

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

A metal foam filled vertical tube, initially saturated

191303-2828-IJMME-IJENS © June 2019 IJENS

Capillary diffusion coefficient:

191303-2828-IJMME-IJENS © June 2019 IJENS

(34) And for Z direction;

At first, the momentum equation (2) in general form as; (45)

191303-2828-IJMME-IJENS © June 2019 IJENS

u=u*+u' (47a) 1005 0.35

(48) 5543 0.0006321

III. RESULTS AND DISCUSSION

(52) A. Pressure Drop Characteristics

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

191303-2828-IJMME-IJENS © June 2019 IJENS

qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2 qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

qw = 10kW/m2 qw = 20kW/m2 qw = 40kW/m2

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

G=11 kg/s.m2 G=15 kg/s.m2

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

Fig. 16. Comparison of Heat Transfer Coefficient with vapour quality

191303-2828-IJMME-IJENS © June 2019 IJENS

191303-2828-IJMME-IJENS © June 2019 IJENS

View publication stats

You might also like