International Communications in Heat and Mass Transfer 39 (2012) 264269

Contents lists available at SciVerse ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

Double-diffusive natural convection in a triangular solar collector

M.M. Rahman a,, Hakan F. ztop b, f, A. Ahsan c, d, M.A. Kalam e, Y. Varol b
a
Department of Mathematics, Bangladesh University of Engineering and Technology, Dhaka-1000, Bangladesh
b
Department of Mechanical Engineering, Technology Faculty, Frat University, Elazig, Turkey
c
Department of Civil Engineering, Faculty of Engineering, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
d
Green Engineering and Sustainable Technology Lab, Institute of Advanced Technology, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
e
Department of Mechanical Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
f
Department of Mechanical Engineering, College of Engineering, King Saud University, Riyadh, Saudi Arabia

a r t i c l e i n f o a b s t r a c t

Available online 9 December 2011 The current study deals with mathematical study for natural convection ow resulting from the combined
buoyancy effects of thermal and mass diffusion inside a triangular shaped solar collector. The solution is per-
Keywords: formed assuming the isothermal and isoconcentration boundary conditions of absorbers and covers of collec-
Solar collector tor. The two-dimensional governing equations for the physical phenomenon are expressed in a velocity
Finite element method pressure formulation, along with the energy and concentration balance equations. A nite element method
Absorbers and covers
is used to solve these equations. Effects of the thermal Rayleigh number and buoyancy ratio are presented
diffusion
by streamlines, isotherms, isoconcentration as well as local and mean heat and mass transfer rates for the
aforesaid parameters. Comparison with the previously published work is made and found to be an excellent
agreement. Particularly, the design for enhancing the performance of the collector is determined by examin-
ing the abovementioned results.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Solar energy is abundant and clean; it is meaningful to substitute
solar energy for conventional energy. Solar energy therefore has an
important role to play in the building energy system [12]. The collec-
tors can be in different shape. They can design rectangular, triangular
or trapezoidal, etc. and they are the main piece of the solar energy
systems.
Double-diffusive natural convection occurs within the solar collec-
tor, owing to the collective thermal and mass diffusion buoyancy ef-
fects as well as the temperature difference between cover and
absorber. Nowadays, the study of solar collectors is one of the most
signicant works to progress their performance with a competitive
price. Numerical solution has the advantage over an experimental in-
vestigation in that the important parameters, such as geometrical di-
mensions, glass thickness and covenant location may be simply
changed. Therefore, its persuading on the overall heat and mass
transfer can be studied at a low price. Numerical investigation is
also helpful in testing the performance of solar collectors for different
components.
In solar collectors, the fresh water manufacture is governed by the
transport phenomena induced by the natural convection of the bulk
moist air enclosed in the area between a transparent cover and
Corresponding author.
E-mail addresses: m71ramath@gmail.com, m71ra@yahoo.com (M.M. Rahman).
water in the collector basin. The inside of the collector is heated by
the solar energy when sunlight enters the cell through a glass top.
This energy allows the water temperature to rise until the water va-
porization is achieved.
Several experiments and theoretic investigations have been un-
dertaken in this region. Boukar and Harmim [3] evaluated the perfor-
mance of one-sided vertical solar still tested under desert climatic
conditions of Algeria, in summer and autumn seasons 2003. The
same authors [4] experimentally investigated design parameters of
an indirect vertical solar still. Omri et al. [5] examined the thermal ex-
change by natural convection and effect of buoyancy force on ow
structure. The authors provided useful information on the ow struc-
ture sensitivity to the governing parameters, the Rayleigh number
and the tilt angle on the thermal exchange. Omri [6] numerically
studied the ow characteristics inside an asymmetrical triangular
still for the conguration optimization. Gao et al. [7] performed a
study on natural convection inside the wavy and inclined solar collec-
tors, but they did not consider ow behavior and thermal elds. A nu-
merical experiment is performed for inclined solar collectors by Varol
and Oztop [8].
Nowadays, a small number of mathematical models of the energy
balance equations describing the heat and mass transfer in a solar still
have been presented. The main envoy of such studies is those pro-
posed by Selcuk [9] and Sodha et al. [10]. Chouikh et al. [11] numeri-
cally investigated the heat mass transfer in the inclined glazing cavity
for solar brackish water desalination. Joudi et al. [12] numerically in-
vestigated the performance of a prism shaped storage solar collector

0735-1933/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.icheatmasstransfer.2011.11.008
 M.M. Rahman et al. / International Communications in Heat and Mass Transfer 39 (2012) 264269 265

Nomenclature

Br buoyancy ratio
c concentration of species
ch high species concentration (source)
cL low species concentration (sink)
C dimensionless species concentration
D species diffusivity
g gravitational acceleration
H enclosure height
Le Lewis number
Nu average Nusselt number
p dimensional pressure
P non-dimensional pressure
Pr Prandtl number
RaT thermal Rayleigh number
Sh average Sherwood number
T temperature
Th hot wall temperature (source)
TL cold wall temperature (sink)
u horizontal velocity component
U dimensionless horizontal velocity component Fig. 1. Schematic diagram for the problem.
v vertical velocity component
V dimensionless vertical velocity component
x horizontal coordinate
X dimensionless horizontal coordinate collector. The enclosed space consists mostly of an absorber plate
y vertical coordinate and two inclined glass covers that form a cavity. The absorber plate
Y dimensionless vertical coordinate is represented by a horizontal bottom wall kept to a constant temper-
ature Th while the inclined walls are considered transparent and
maintained at a constant temperature TL with Th > TL.
Greek symbols
2.2. Mathematical formulation
thermal diffusivity
T thermal expansion coefcient
The salty water on the bottom side of the glazing enclosure is va-
c compositional expansion coefcient
porized from the liquidvapor interface. The vapor moves through
kinematic viscosity
the air and condenses at the cooled side walls. Conservation equa-
non-dimensional temperature
tions are written using the velocitypressure formulation and consid-
density
ering constant properties except for the density in the body force
stream function
term. For steady two-dimensional laminar convection the coupled
general dependent variable
transport equations for U, V, and C written in Cartesian coordinates
in the dimensionless form as:

with a right triangular cross sectional area. Kalogirou [13] offered an

in general review of a huge variety of systems used in desalination U V
0 1
processes. The author gave some universal guidelines for choice of X Y
desalination systems motorized by renewable energy. The author !
concluded on the major design parameters that require to be mea- U U P 2 U 2 U
U V Pr 2
sured in this eld. Different types of solar still available in the litera- X Y X X 2 Y 2
ture are conventional solar stills, single-slope solar still with passive
condenser, double condensing chamber solar still [14], vertical solar !
2 2
still [15], the inverted absorbers solar still [16] and multiple effect V V P V V
U V Pr RaT Pr BrC 3
solar still [17]. The heat transfer inside the isosceles triangular enclo- X Y Y X 2 Y 2
sure is studied by Varol et al. [1820] for different thermal boundary
conditions, including entropy generation.
It is clear from the above literature survey that double diffusive con-
vection is not studied in earlier works for triangular types of solar collec-
tors. Thus, the main aim of the present work is to present numerical
results on natural convection heat and mass transfer in a triangular
shaped enclosure for different buoyancy ratio and Rayleigh number.

2. Problem statement

2.1. Physical description

The geometry for the conguration is shown by a schematic dia-

gram in Fig. 1. The model considered here is a triangular type solar Fig. 2. Grid independency study for Ra = 106, Pr = 0.71 and Br = 5.
266 M.M. Rahman et al. / International Communications in Heat and Mass Transfer 39 (2012) 264269

2 2
(a)
U V 4
X Y X 2 Y 2
!
2 2
C C 1 C C
U V 5
X Y Le X 2 Y 2

where the dimensionless variables are introduced as:

x Y uH vH p gyH2 TT L ccL
X ;Y ;U ;V ;P ; and C
H H 2 T h T L ch cL
6

The variables have their usual sense in uid mechanics and heat
transfer as listed in the nomenclature. The dimensionless parameters
appearing in the above equations are the Prandtl number (Pr), the
thermal Rayleigh number (RaT), Lewis number (Le) and buoyancy
ratio (Br), which are dened respectively as

g T T L H 3 c c
(b) Pr

; RaT T h

; Le and Br c h L
D T T h T L
7

The dimensionless boundary conditions corresponding to the con-

sidered problem are

on the bottom wall : U V 0; C 1

on the inclined walls : U V 0; C 0

The local heat and mass transfer rates on the surface of heat and
contaminant sources are dened respectively as

 C 
Nux  and Shx 
Y Y0 Y Y0

The average heat and mass transfer rates on the surface of heat
and contaminant sources can be evaluated by the average Nusselt
and Sherwood numbers, which are dened respectively as
Fig. 3. Comparison of the (a) present model with (b) the results of Kent [21] for natural
convection inside a isosceles triangular enclosure for ve different base angles. 1

Nu dX 8
0
Y

(a) (b) (c)

Br = 20
Br = 10
Br = 5
Br = 0.5

Fig. 4. Effect of buoyancy ratio on streamlines for (a) Ra = 104, (b) Ra = 105 and (c) Ra = 106.
 M.M. Rahman et al. / International Communications in Heat and Mass Transfer 39 (2012) 264269 267

(a) (b) (c)

Br = 20
Br = 10
Br = 5
Br = 0.5

Fig. 5. Effect of buoyancy ratio on isotherms for (a) Ra = 104, (b) Ra = 105 and (c) Ra = 106.

and 3. Numerical Solution

1 3.1. Solution procedure

C
Sh dX 9
Y
0 The Galerkin weighted residual method of nite element formula-
tion is used as a numerical procedure in this work. The nite element
The stream function is calculated from method begins by the partition of the continuum area of interest into
a number of simply shaped regions called elements. These elements
may be different shapes and sizes. Within each element, the depen-
dent variables are approximated using interpolation functions. In
U ;V 10
Y X the present study erratic grid size system is considered especially

(a) (b) (c)

Br = 20
Br = 10
Br = 5
Br = 0.5

Fig. 6. Effect of buoyancy ratio on isoconcentration for (a) Ra = 104, (b) Ra = 105 and (c) Ra = 106.
268 M.M. Rahman et al. / International Communications in Heat and Mass Transfer 39 (2012) 264269

near the walls to capture the rapid changes in the dependent vari- (a)
ables. The coupled governing Eqs. (2)(5) are transformed into sets
of algebraic equations using nite element method to reduce the con-
tinuum domain into discrete triangular domains. The system of alge-
braic equations is solved by iteration technique. The solution process
is iterated until the subsequent convergence condition is satised:
| m + 1 m| 10 6 where m is number of iteration and is the gen-
eral dependent variable.

3.2. Grid renement check

A grid independence analysis is carried out in this section to make

sure the accuracy of the numerical results and to nd out a proper
size of grids for the case of RaT = 10 6, Pr = 0.71 and Br = 5.0. In
these study ve different non-uniform grids of triangular elements: (b)
1486, 2808, 3186, 4490 and 5212 are used. The value of average Nus-
selt and Sherwood numbers inside the cavity are used as a sensitivity
measure of the accuracy of the solution and are selected as the mon-
itoring variables for the grid independence study. Fig. 2 shows the
convergence of average Nusselt and Sherwood numbers with grid re-
nement. The grid size of about 3186 elements was therefore, chosen
for all cases studied.

3.3. Code validation

The study is compared with an earlier work on natural convection

in a triangular enclosure was performed by Kent al. [21]. Compariso-
nal results are illustrated in Fig. 3 for different inclination angle which
is studied by Kent [21]. He used Fluent commercial code as numerical
analysis. The comparisonal results showed a good agreement with lit-
(c)
erature on heat transfer and uid ow. Then, we jumped to next step
to add mass transfer to simulate solar collectors.

4. Results and discussion

In this work, numerical analysis of convection inside an isosceles

solar collector has been made using nite element technique. Air
was used as working uid inside the cavity with Pr = 0.71. The Lewis
number was held xed throughout this investigation at Le = 2.0.
Fig. 4 illustrates the effects of the buoyancy ratio on streamlines at dif-
ferent Rayleigh numbers. Rayleigh number is calculated based on the
temperature difference and buoyancy ratio denotes the relative
Fig. 7. Effect of buoyancy ratio on local heat transfer rate at (a) Ra = 104, (b) Ra = 105
strengths of two buoyancy forces namely mass and thermal buoyancy
and (c) Ra = 106.
forces. As given from the gures two circulation cells were formed in dif-
ferent rotation directions. The left cell rotates counterclockwise and the
right one clockwise. In other words, symmetric ow distribution is ob-
served according to the middle axis of the triangle. Both Rayleigh num-
ber and buoyancy ratio enhance the ow strength. The maximum
stream function value is obtained for Br =20 and Ra =106. For these
cases, isotherms are plotted in Fig. 5. As seen from the gure, conduction
mode of heat transfer is dominant inside the cavity for Ra= 104 but
plume like temperature distribution has become stronger from Br =0.5
to 20 and Ra= 105. However, temperature distribution showed similar
behavior for Ra = 106 and Br= 5, 10 and 20 due to strong convection in-
side the cavity. Isoconcentration is given in Fig. 6. Distribution of isocon-
centration exhibits a similar distribution with temperature distribution.
Fig. 7 illustrates the distribution of local heat transfer rate along
the bottom wall. It is an interesting result that heat transfer becomes
a minimum at X = 0.5 for all parameters. The reason of this value is
the stagnation point and minimum velocity at this point. Trend of
local heat transfer rate exhibits a symmetric distribution according
to the mid-axis. As an expected result, highest heat transfer is formed
for Ra = 10 6. Heat transfer rate decreases from the left and right edge
to the mid-axis. However, it has maximum values around X = 0.25
and 0.85 for higher values of the buoyancy ratio. If we look at the var-
iation of average heat transfer rate with the buoyancy ratio, it shows Fig. 8. Effect of buoyancy ratio on average heat transfer rate.
 M.M. Rahman et al. / International Communications in Heat and Mass Transfer 39 (2012) 264269
increasing trend with buoyancy ratio and Rayleigh number (Fig. 8).
The highest value is formed at Ra = 10 6 and Br = 20. Local mass trans-
fer rate is given in Fig. 9. Its trend is similar with the local heat trans-
fer rate, and it decreases with decreasing of the buoyancy ratio for all
Rayleigh numbers. Finally, average mass transfer is shown in Fig. 10
that a linear increasing is obtained according to the buoyancy ratio.
Average value of mass transfer rate increases with increasing of Ray-
leigh number.
5. Conclusions
This study illustrates a numerical work concerning the uid ow,
heat transfer and mass transfer inside triangular shaped solar collec-
tors. The results are obtained according to different values buoyancy
ratio and Rayleigh number. Some important observations can be
listed as follows:
Both heat transfer and mass transfer increase with increasing of
buoyancy ratio and Rayleigh number.
Symmetric ow eld, temperature distribution and mass distribu-
tion are found according to the middle axis of the triangular cavity.
(a)
(b)
(c)
Fig. 9. Effect of buoyancy ratio on local mass transfer rate at (a) Ra = 104, (b) Ra = 105
and (c) Ra = 106.
269
Fig. 10. Effect of buoyancy ratio on average mass transfer rate.
Increasing of the buoyancy ratio at the same Rayleigh number en-
hances the local heat and mass transfer rate.
Local heat transfer and mass transfer have minimum value at the
middle of the bottom wall due to stagnation point and motionless
uid at that point.
References
[1] X.Q. Zhai, R.Z. Wang, Y.J. Dai, J.Y. Wu, Y.X. Xu, Q. Ma, Solar integrated energy sys-
tem for a green building, Energy Building 39 (2007) 985993.
[2] G.A. Florides, S.A. Tassou, S.A. Kalogirou, L.C. Wrobel, Review of solar and low
energy cooling technologies for buildings, Renewable and Sustainable Energy
Reviews 6 (2002) 557572.
[3] M. Boukar, A. Harmim, Performance evaluation of a one-sided vertical solar still
tested in the Desert of Algeria, Desalination 183 (2005) 113126.
[4] M. Boukar, A. Harmim, Design parameters and preliminary experimental investi-
gation of an indirect vertical solar still, Desalination 203 (2007) 444454.
[5] A. Omri, J. Or, S.B. Nasrallah, Natural convection effects in solar stills, Desalina-
tion 183 (2005) 173178.
[6] A. Omri, Numerical investigation on optimization of a solar distiller dimensions,
Desalination 206 (2007) 373379.
[7] W. Gao, W. Lin, E. Lu, Numerical study on natural convection inside the channel
between the at-plate cover and sin-wave absorber of a cross-corrugated solar
air heater, Energy Conversion Management 41 (2000) 145151.
[8] Y. Varol, H.F. Oztop, Buoyancy induced heat transfer and uid ow inside a tilled
wavy solar collector, Building and Environment 42 (2007) 20622071.
[9] K. Selcuk, Thermal and economic analysis of the overlapped glass plate solar-air
heaters, Solar Energy 13 (1971) 165191.
[10] M.S. Sodha, A. Kumark, G.N. Tiwary, R.C. Tiagi, Simple multiple wick solar still:
analysis and performance, Solar Energy 26 (1981) 127131.
[11] R. Chouikh, L. Ben Snoussi, A. Guizani, Numerical study of the heat and mass
transfer in inclined glazing cavity: Application to a solar distillation cell, Renew-
able Energy 32 (2007) 15111524.
[12] K.A. Joudi, I.A. Hussein, A.A. Farhan, Computational model for a prism shaped
storage solar collector with a right triangular cross section, Energy Conversion
Management 45 (2004) 391409.
[13] S.A. Kalogirou, Seawater desalination using renewable energy sources, Prog,
Energy and Combustion Science 31 (2005) 242281.
[14] G.N. Tiwari, A. Kupfermann, S. Agrawal, A new design of double condensing
chamber solar still, Desalination 114 (1997) 153164.
[15] J.P. Coffey, Vertical solar distillation, Solar Energy 17 (1975) 375378.
[16] S. Suneja, G.N. Tiwari, Optimization of number of effects for higher yields from an
inverted absorber solar still using the RungeKutta method, Desalination 120
(1999) 197209.
[17] H. Tanaka, T. Nosoko, T. Nagata, Parametric investigation of a basin type multiple
effect coupled solar still, Desalination 130 (2000) 295304.
[18] Y. Varol, A. Koca, H.F. Oztop, Natural convection in a triangle enclosure with ush
mounted heater on the wall, International Communications in Heat and Mass
Transfer 33 (2006) 951958.
[19] Y. Varol, H.F. Oztop, I. Pop, Entropy generation due to natural convection in
non-uniformly heated porous isosceles triangular enclosures at different posi-
tions, International Journal of Heat and Mass Transfer 52 (2009) 11931205.
[20] Y. Varol, H.F. Oztop, I. Pop, Inuence of inclination angle on buoyancy-driven con-
vection in triangular enclosure lled with a uid-saturated porous medium, Heat
Mass Transfer, 44, Springer, 2008, pp. 617624.
[21] E.F. Kent, Numerical analysis of laminar natural convection in isosceles triangular
enclosures for cold base and hot inclined walls, Mechanics Research Communica-
tions 36 (2009) 497508.