Available online at www.sciencedirect.

com

ScienceDirect
Solar Energy 131 (2016) 183198
www.elsevier.com/locate/solener

Eect of wind speed and direction on convective heat losses from

solar parabolic dish modied cavity receiver
K.S. Reddy , G. Veershetty, T. Srihari Vikram
Heat Transfer and Thermal Power Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai 600036, India

Received 3 November 2015; received in revised form 10 February 2016; accepted 19 February 2016
Available online 7 March 2016

Communicated by: Associate Editor Robert Pitz-Paal

Abstract

The performance of solar parabolic dish collector is signicantly inuenced by heat losses due to wind speed and direction. In this
article, investigation of convective heat losses from the modied cavity receiver of solar parabolic dish collector is carried out numerically
by considering the wind direction, wind speed, receiver conguration and receiver orientation. The eect of wind on the receiver in var-
ious directions (u =
90 to 90), wide range of operating wind speeds (V = 010 m/s), receiver inclinations (b = 090) and varying
surface temperature on convective heat loss from the receiver are studied. Velocity vectors, velocity contours, temperature contours
are presented to show the eect of wind on the heat loss from the modied cavity receiver. The forced convection is found to have similar
trend of free convection heat loss at lower wind speed. However at higher wind speed, such a pattern is not observed. At lower wind
speeds say less than critical wind speed (<2.5 m/s), the forced convection heat loss is lower than natural convection heat loss for lower
receiver inclinations and wind direction ranging between
90 and 0. The forced convection heat loss is more signicant than free con-
vection heat loss above 5 m/s for all u and b values. For side-on winds, at higher wind speeds above 5 m/s, irrespective of receiver incli-
nation, the variation of forced convection heat loss is marginal (less than 5%). The maximum forced convection heat loss occurs for
partly open receivers (receiver aperture diameter ratio, RAD = 0.4 and 0.6) at u = 0 (side-on wind) for all receiver inclinations and at
u = 30 for RAD = 0.8 and 1. The receiver inclination has less eect on heat loss from the receiver for V > 2.5 m/s due to side-on wind.
The highest convection heat loss occurs for fully open (RAD = 1) receiver as compared to partly open (RAD < 1) modied cavity receiver.
Nusselt number correlation is proposed to calculate combined convection heat losses from the receiver as a function of receiver inclina-
tion, wind direction, wind velocity and aperture diameter ratio.
2016 Elsevier Ltd. All rights reserved.

Keywords: Solar parabolic dish collector; Modied cavity receiver; Combined convection heat losses; Nusselt number correlations; Numerical analysis

1. Introduction uously for ecient energy conversion. The cavity receivers

The parabolic dish collector is one of the most ecient
technologies for utilization of solar energy to produce high
temperature heat. Solar parabolic dish collector is operated
under dierent weather conditions and tracking sun contin-
Corresponding author. Tel.: +91 44 22574702; fax: +91 44 22574652.
E-mail address: ksreddy@iitm.ac.in (K.S. Reddy).
are preferred in parabolic dish collector system for convert-
ing concentrated solar energy to heat due to nominal heat
losses. Several trade-o strategies have been discussed for
design of various components such as parabolic dish collec-
tor, receiver and transport medium/system (Kaushika,
1991). The knowledge of heat loss from the receiver is
one of the prime requirements for ecient design of the
receiver of parabolic dish. As receiver experiences high

http://dx.doi.org/10.1016/j.solener.2016.02.039
0038-092X/ 2016 Elsevier Ltd. All rights reserved.
184 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
Nomenclature
A Area, m2
Cp specic heat capacity at constant pressure,
J/(kg K)
d opening diameter of the modied cavity recei-
ver, m
Di diameter of the modied cavity receiver, m
Dx cross-diusion
g acceleration due to gravity, m/s2
Gr Grashof number
Gk generation of k, kg/(m s3)
Gx generation of x, kg/(m s3)
h convection heat transfer coecient, W/(m2 K)
k turbulent kinetic energy, m2/s2
kf thermal conductivity, W/m K
Nu Nusselt number
P pressure, Pa
qcv Heat loss from the receiver surface, W/m2
Qloss Wall heat loss, W
RAD Aperture diameter ratio
Re Reynolds number
S source term
T temperature, K
u velocity component, m/s
V wind speed, m/s
x, y, z cartesian coordinates
temperature and operates in an open atmosphere, all the
modes of heat transfer play a major role. Conduction heat
losses from the receiver can be reduced by covering proper
insulation over the receiver. The radiative loss depends on
the receiver wall temperature, shape factor and emissivity/
absorptivity of the receiver surface. The convective heat
loss depends on the receiver geometry, surface temperature,
ambient conditions and receiver orientation. An accurate
estimation of heat losses from the receiver under actual
conditions is very important to develop an ecient solar
dish collector. Several researchers have proposed dierent
shapes and sizes of the cavity receivers for parabolic dish
collector. Wu et al. (2010) presented an extensive review
on cavity receivers for solar thermal applications and also
presented results on cavities used in electronic cooling
devices, building, etc. Numerical and experimental investi-
gations were carried out to estimate the natural convection
and forced convection heat losses from cavity receiver of
parabolic dish collector system. Kumar and Reddy (2008)
performed numerical investigations on natural convection
heat loss from three types of cavity receivers of a fuzzy
focal solar dish concentrator for various inclinations of
the receivers and area ratios with uniform surface temper-
ature of 673 K. It was reported that the convection heat
losses from the modied cavity receiver were lower than
the other type of receivers considered and it was found that
Yk dissipation of k, N/m2
Yx dissipation of x, N/m2
Greek symbols
b receiver inclination, deg
dij Kronecker delta function
u incident angle of the wind in horizontal plane,
deg
l dynamic viscosity, kg/ms
lt turbulent eddy viscosity, kg/ms
q density, kg/m3
rT eective Prandtl number for T
Ck eective diusivity of k
Cx eective diusivity of x
x specic dissipation rate, s
1
Sux
a ambient
aw area weighted average
b bulk
i, j, k direction components
s surface
t total/combined
the minimum heat loss occurs for an area ratio of 8. The
modied cavity receiver is the preferred receiver for the
fuzzy focal solar parabolic dish collector among cavity,
semi-cavity and modied cavity receivers. Reddy and
Kumar (2008) carried out 2-D numerical studies to deter-
mine the inuence of operating temperature, emissivity of
the surface, orientation and the geometry on combined nat-
ural convection and surface radiation from a modied cav-
ity receiver of solar parabolic dish collector. It was
observed that convective loss is signicantly inuenced by
the receiver orientation and the radiation heat loss is con-
siderably inuenced by the area ratios. An improved 3-D
numerical model was proposed by Reddy and Kumar
(2009) to investigate the natural convection heat loss accu-
rately from the modied cavity receiver. The 3-D model
can be used for an accurate estimation of natural convec-
tion heat loss from the modied cavity receiver of fuzzy
focal solar dish concentrator. Vikram and Reddy (2014,
2015) performed combined natural convection and surface
radiation heat losses from modied cavity receiver of solar
parabolic dish collector considering three dierent temper-
ature proles instead of isothermal wall temperature for
receiver surface.
Cui et al. (2013) analysed the combined natural convec-
tion and surface radiation heat loss from the hemispherical
cavity receiver with a quartz glass cover for the dish
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
concentrating system. A 2-D numerical investigation has
been carried out to study the eect of orientation, temper-
ature and emissivity of inner surface on the heat transfer
and uid ow from the receiver. A signicant reduction
of heat loss from the receiver was observed when it is cov-
ered with quartz glass cover.
Taumoefolau et al. (2004) carried out experimental
investigations to determine natural convection heat losses
from an electrically heated cylindrical cavity receiver and
compared with simulation results. The investigation has
been performed for various orientations of receiver, open-
ing ratios, and wall temperatures ranging from 723 to
923 K. Synthetic Schlieren technique has been used to visu-
alize the ow pattern out of the cavity. Wu et al. (2011) car-
ried out 3-D numerical investigations to study the inuence
of aperture position and size on natural convection heat
loss from the cylindrical cavity type receivers for heat-
pipe application. The investigation was carried out consid-
ering various inclinations, dierent wall temperatures (723
1123 K) and aperture size and positions. The natural con-
vection heat loss varies linearly with wall temperature. A
Nusselt number correlation has been developed to estimate
the heat loss from the cylindrical receiver. Prakash et al.
(2009) carried out experimental and numerical study on
convective heat losses from cylindrical cavity receiver with
wind skirt for various receiver inclinations and uid inlet
temperatures. A Nusselt number correlation has been pro-
posed to estimate the natural convection heat loss based on
experimental and simulation results.
Prakash et al. (2012) carried out 3-D numerical analysis
to determine convection heat losses from dierent cavity
receivers such as cubical, spherical and hemispherical with
dierent opening ratios, various inclinations and isother-
mal wall temperatures ranging from 373 to 573 K. The nat-
ural convection heat loss is found to increase with an
increase in opening ratio and wall temperature. The
increase in natural convection heat loss for dierent incli-
nations varies between 30% and 80% when the opening
ratio is increased from 0.25 to 0.5 for all cavity shapes. A
Nusselt number correlation has been developed consider-
ing wall temperature, opening ratio and receiver inclina-
tion. Xie et al. (2011) carried out experimental, analytical
and numerical investigations on convection heat losses
from conical, spherical and cylindrical cavity receivers of
same aperture size with glass cover facing upwards for col-
lection of solar energy. The average surface temperature
was varied from 363 to 423 K and average air ow rate
along the aperture of the absorber was about 2 m/s. From
the analysis, the conical cavity receiver was found to be of
optimum shape with higher thermal eciency and lower
heat loss than other receivers.
Few studies are available on forced convection heat
transfer from larger cavity receivers that are used in solar
power tower. Leibfried and Ortjohann (1995) carried out
experimental analysis of upward and downward facing
spherical and hemispherical receivers. The eect of heat
losses from the receivers on temperature, tilt angle, geome-
185
try and aperture are studied. Flesch et al. (2014) carried out
numerical analysis of inuence of receiver inclination and
wind on the receivers. The temperature distribution inside
the cavity is presented to understand the eect of wind
on the heat loss from the receiver. Flesch et al. (2015) have
also carried out experimental analysis to study the eect of
wind and wind incidence angle on cavity receivers used in
solar power tower. It has been observed that horizontal
receivers are less inuential under wind conditions whereas
the wind has reasonable eect when the receiver inclination
varies. Also, in some cases, the heat loss due to wind is
lower than the forced convection heat loss from the recei-
ver. It has been suggested that the cavity should be
designed in such a way that when the wind is redirected
parallel to the aperture, it should have positive eect (at
least not negative eect). Liovic et al. (2014) presented
the eect of wind on the receiver used for solar reformer.
The geometry of the receiver, inner cavity wall temperature
and wind tend to inuence the heat loss from the receiver.
Very limited studies are available on forced convection
heat loss from the cavity receivers used in dishes. Ma
(1993) performed experimental studies to estimate the con-
vective heat loss from the cavity receiver of a paraboloidal
dish collector for dierent wind directions and receiver
inclinations. The convection heat loss from the receiver is
found to increase by the presence of side-on wind for all
receiver tilt angles. The convective heat losses from the
receiver with tilt angle 90 for head-on and side-on winds
are approximately same. The forced convection heat losses
from the receiver are about three times higher than natural
convection. The eects of wind on convective heat loss
from the receiver are the greatest for side-on wind and
the smallest for head-on wind. Experiments on combined
natural-forced convective heat loss of a fully open cylindri-
cal cavity with constant heat ux boundary was conducted
by Wu et al. (2015). The eect of wind incidence angle, cav-
ity tilt angle and wind speed has been studied and devel-
oped models for predicting heat loss from fully open
cylindrical cavity.
Paitoonsurikarn and Lovegrove (2003) investigated
combined free-forced convection studies on open cavity
receivers for paraboloidal dish applications considering
winds parallel and normal to aperture and wind velocities
up to 20 m/s. The free convection heat loss dominates the
forced convection heat loss at certain wind speeds. Also,
Prakash et al. (2009) estimated convection heat loss from
cylindrical receiver under wind velocities, 1 and 3 m/s
and two wind directions, head-on and side-on winds. The
eect of receiver inclination, uid inlet temperature, wind
speed and direction are studied. The heat loss due to
head-on wind is higher than the side-on wind. Xiao et al.
(2012) carried out a numerical study to estimate combined
free-forced convection heat loss from open cavity receiver.
The eect of parameters such as wind speed, wind inci-
dence angle and receiver inclinations on combined free-
forced convection has been performed. The combined con-
vection heat loss of receiver becomes indistinguishable at
186 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
dierent inclinations as wind speed increases. Forced con-
vection heat losses from dierent cavity geometries have
been analysed by Jilte et al. (2014) considering dierent
wind speeds, receiver inclination, wind directions and
isothermal wall temperatures. It was found that convective
heat loss due to head-on wind remains high for all cavity
shapes. A generalised Nusselt number equation has been
developed to estimate the forced convection heat loss due
to head-on wind from dierent cavity geometries.
It is observed from literature that the heat losses due to
wind speed and direction are crucial for estimating the per-
formance of solar dish collector. The studies carried out in
this area are limited as far as information available in the
open literature. Most of the researchers (Prakash et al.
(2009), Ma (1993), Paitoonsurikarn and Lovegrove
(2003), Jilte et al. (2014)) focused their studies on head-
on and side-on wind except Wu et al. (2015) and Xiao
et al. (2012). It is noticed that, the studies on forced convec-
tion heat losses are less consistent and may not be gener-
alised as it depends on various parameters such as
geometry (shape and size), operating conditions etc. The
results are specic to the conditions considered especially
the geometry. Hence, in this paper, an attempt has been
made to study the eect of receiver aperture size (aperture
diameter ratio), wind directions and speed on convective
heat loss from cavity receiver of parabolic dish system by
considering more accurate temperature proles and actual
prevailing wind conditions. The present model can be used
for estimating the heat losses from the receiver which is
used for high temperature process heat applications (about
600 C) and medium temperature applications (between
290 C and 590 C).
2. Investigation of convection heat losses
2.1. Geometrical model
A circular parabolic dish collector of 20 m2 has been
considered to investigate forced convection heat losses
from the modied cavity receiver. The modied cavity
receiver is a hemispherical dome with varying aperture
made of the metallic tubes wound spirally in aperture
and helically in hemispherical dome. The dome is covered
with ceramic thermal insulation to reduce the heat loss
from receiver to the ambient. The concentrated solar
energy is focussed onto the modied cavity such that most
of the reected energy is received by it. The heat transfer
uid (uid owing through the receiver) ows through
the tubes and gains heat energy as it ows and the temper-
ature of uid increases. The air is considered as the medium
surrounding the receiver. The schematic view of the para-
bolic dish collector with modied cavity receiver indicating
wind direction is shown in Fig. 1a. Based on the focal ux
distribution, the receiver diameter (Di) is considered as
305 mm (Reddy et al., 2015). The receiver inclination varies
from sideward facing (b = 0) to downward facing
(b = 90). The wind direction (u) is assumed to be in XY
plane (Fig. 1a).
The wind can blow over the receiver at any direction (0
360), but the dish system operates tracking the sun with
the help of 2-axis tracking system (EW and NS). The wind
direction is considered with respect to Y axis in counter
clockwise direction (Fig. 1a). For simulation purpose, only
one half of the orientations say (
90 to 90) is considered.
The receiver is assumed to be facing towards +ve X-axis
direction such that aperture of the receiver is in YZ plane.
When the wind is blowing perpendicular to the receiver
aperture, it is called head-on wind, which corresponds to
u = 90. When the wind blows parallel to the receiver aper-
ture, it is called side-on wind and it corresponds to u = 0.
The wind blowing over the backside of the receiver is
referred as back-on wind, which corresponds to u =
90.
In the present study, three dimensional numerical anal-
ysis of modied cavity receiver has been carried out consid-
ering wind directions ranging from
90 to 90, various
orientations of the receiver (b = 090), wind speeds
(V = 010 m/s) and aperture diameter ratios (RAD = 0.41)
to study the eect of these parameters on the convection
heat loss from the modied cavity receiver. The maximum
value of 10 m/s (36 kmph) corresponds to the design wind
speed. The eect of receiver aperture size on forced convec-
tion heat loss has been investigated by varying the diameter
ratios from 0.4 (partly closed cavity) to 1 (open cavity)
keeping the receiver diameter (Di) xed based on the focal
image characteristics. The value less than 0.4 may not be
practically relevant to the solar applications, as the aper-
ture diameter is small to allow the concentrated solar radi-
ation to focus onto the inner surface of the receiver. Based
on the process heat applications, the receiver temperature is
considered to increase linearly from 300 to 800 K from inlet
to outlet. The ambient temperature is assumed to be 300 K
which corresponds to the average ambient conditions. The
temperature prole of receiver is considered to match the
actual conditions to cater the process heat and power
requirement.
The dish considered in the present study is deep dish
with rim angle of 70 and focal length of 1.8 m. When there
is back-on wind (u =
90), the wind may not be deected
towards the receiver because of the higher focal length,
higher rim angle and the less curvature of the dish. Hence
the eect of concentrator for back-on wind may be
neglected (Fig. 1b). For the side-on wind (u = 0, parallel
to the receiver aperture), concentrator may not inuence
the forced convection heat loss from the receiver, as the
wind blows parallel to the aperture and concentrator aper-
ture (Fig. 1c). Similarly, for the head-on wind (u = 90,
normal to the receiver aperture), the concentrator may
not have impact on the receiver for large portion of eleva-
tion (Fig. 1d). Hence, it can be concluded that the eect of
concentrator on heat loss estimation is very marginal;
therefore the eect of concentrator may be neglected.
Paitoonsurikarn and Lovegrove (2006) carried out the
study on eect of dish structure convection heat loss from
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198 187

Fig. 1a. Schematic view of parabolic dish collector showing wind directions and cross section of the receiver.

Fig. 1c. Eect of dish structure on side-on wind.

Fig. 1b. Eect of dish structure on back-on wind.

the receiver due to wind. It has been reported that, compli-
cated velocity pattern results from the interaction of ow
with dish and receiver. However it has been concluded that,
the results cannot be generalized and it is specic to the
receiver that has been considered. So there exists uncer-
tainty in ascertaining the eect of dish structure on the con-
vection heat loss due to wind. However, the eect of dish
structure on the forced convection heat loss from the recei-
ver will be carried out as a part of our future work.
2.2. Numerical procedure
The 3-D modelling of modied cavity receiver and cor-
responding grid generation has been carried out using
GAMBIT 2.4.6. To account for the eect of surrounding
temperature and wind eects on heat transfer; dierent
sizes of computational domains have been considered. It
is found that the ow and heat transfer characteristics
inside the cavity and its immediate surroundings does not
change signicantly for the computational domain of
188 K.S. Reddy et al. / Solar Energy 131 (2016) 183198

Computational domain and grid generation for 3-D mod-

elling of the cavity receiver is shown in Fig. 2.
As the receiver is in the open atmosphere, the wind con-
dition is considered as turbulent state. Shear Stress Trans-
port (SST) kx turbulent model has been used for
simulation, as it is appropriate for both near-wall and
far-eld zones (Christo (2012) and Fluent Inc (2005)).
The kx model does not involve any complex non-
linearity damping functions and therefore the simulation
is more accurate and robust (Christo, 2012) for conditions
of ow separation under adverse pressure gradients which
is expected in case of wind blow over a cavity receiver,
which is kept in open atmosphere of wind conditions
(Xiao et al., 2012). The 3-D dierential governing equa-
tions can be written in tensor form in Cartesian system
are as follows (Fluent Inc, 2005):
Continuity equation:
@quj
0 1
@xj
Momentum equation:
Fig. 1d. Eect of dish structure on head-on wind.
@qui uj @p


@xj @xi

 
@ @ui @uj 2@uk
l lt
dij qgi
@xj @xj @xi 3@xk
2

Energy equation:

 
@quj C p T @ lt C p @T
k
@xj @xj rT @xj


@ui @ui @uj 2@uk
l lt
dij
@xj @xj @xi 3@xk i
@p
uj 3
@xj

k-equation:


@qui k @ @k
Ck Gk
Y k S k 4
@xi @xj @xj
x-equation:


@qui x @ @x
Cx Gx
Y x Dx S x 5
@xi @xj @xj

where
Gk denotes the generation of turbulent kinetic energy
due to mean velocity gradients and is given by,


Fig. 2. 3-D modelling of modied cavity receiver (a) computational @uj
domain, (b) grid generation. Gk min
qu0i u0j ; 10qb kx 6
@xi
Gx denotes the specic dissipation rate and is given by,
10D in both horizontal and vertical planes and 20D in

downwind direction. There are about 1,432,962 number a 0 0
@uj
Gx
qui uj 7
of tetrahedral cells generated in the simulation space. mt @xi
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198 189

Yk denotes the dissipation of turbulence and given by, temperature other than stagnation conditions of the recei-
 ver is primarily dependent on the ow rates of the working
Y k qb kx 8
uid. The outer hemispherical portion of the receiver tubes
where b bi 1 f F M t ; is covered with ceramic wool insulation to minimise the

0 for M t 6 M to heat loss from outer surface to ambient. Hence, the outer
F M t , M t cRT
2k
Yx denotes
M 2t
M 2to for M t > M to hemispherical surface of the receiver is considered as adia-
the dissipation of x and given by, batic surface.
To consider the wind eect, a rectangular domain is con-
Y x qb0 x2 9 sidered surrounding the receiver. One of the square faces
Dx denotes the cross diusion term and is given by, (near receiver) of the computational domain (Fig. 2) is con-
sidered as velocity inlet boundary conditions and the oppo-
1 @k @x site side of velocity inlet is considered as pressure outlet
Dx 21
F 1 qrw;2 10
x @xj @xj boundary condition. The external walls of the rectangular
domain are considered as wall except for inlet and outlet.
In the above equation, F1 is the blending function  and is
p Hence no-slip boundary condition has been considered
given by F 1 tanh/1 where, /1 min max0:09xy
4 k
; on the surrounding wall of the rectangular domain. The


10 boundary conditions of the receiver are shown in Fig. 3
500l
; 4qk , and D
qy 2 x rw;2 D y 2
1 1 @k @x
w max 2q rw;2 x @xj @xi ; 10
w (a). The variation of temperature along the receiver surface
Sk and Sx are user dened source terms; The values of is shown in Fig. 3(b). The governing equations were solved
the various constants used in the above equations (Fluent using the FLUENT 6.3.36 (Fluent Inc, 2005) using implicit
Inc, 2005) are given as: Mto = 0.25, f* = 1.5, b1 = 0.09, solver. The Semi-Implicit Pressure Linked Equation (SIM-
bi,1 = 0.075, bi,2 = 0.0.0828, rk,1 = 1.176, rw,1 = 1, rk,2 = PLE) scheme has been considered for pressure velocity
2, rw,2 = 1.168, a1 = 1, a1 = 0.52. coupling and second order scheme has been used for

2.3. Boundary conditions

The receiver is subjected to various boundary condi-

tions. The receiver consists of the tubes through which
the working uid gets heated when solar ux falls on it.
Therefore, working uid temperature increases uniformly
from inlet to outlet and hence the surface temperature of
the receiver. The surface temperature of the receiver tubes
is assumed to be equal to that of uid temperature, as the
thermal conductivity of the receiver (copper tubes) is high.
In the present study, the central plane temperature of the
receiver tubes is considered as the receiver wall tempera-
ture. This has been considered to accommodate the actual
ow conditions, which is realistic under steady state condi-
tions of the receiver. Hence the temperature of the receiver
wall surface is assumed as that of the working uid which is
attained due to the incident concentrated solar radiation.
Under steady state conditions, the temperature of the uid
is proportional to the given solar ux. The surface temper-
ature of the receiver depends on solar ux (Zapata et al.,
2013). Therefore, linear variation in temperature along
the length of the receiver tube has been considered (from
inlet to outlet).
The parabolic dish considered in the present study cor-
responds to 20 m2 parabolic dish (Reddy et al., 2015). Con-
centration ratio of the dish collector is 274. When solar ux
of 1000 W/m2 is considered, the concentrated solar ux at
the receiver surface corresponds to 274,000 W/m2. This
ux results in the stagnation temperature of 640 C
(Sendhil Kumar, 2008). The continuous loss of heat from
the receiver to the ambient occurs due to conduction, con-
vection and radiation heat losses. Therefore the tempera- Fig. 3. Modied cavity receiver: (a) Boundary conditions of modied
ture of the receiver cannot increase innitely. The cavity receiver, (b) variation of temperature along the receiver surface.
190 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
pressure discretization. The second order upwind scheme is
considered for the momentum, energy, turbulent kinetic
energy and specic dissipation rate. The convergence crite-
rion of 10
6 is imposed on the residuals of the continuity,
momentum, turbulent kinetic energy and specic dissipa-
tion equations, whereas for energy equation it is considered
as 10
8.
The ambient air properties are considered as a function
of temperature. The polynomial relationships have been
considered for specic heat (Cp), dynamic viscosity (l)
and thermal conductivity (kf) at constant pressure with
temperature (Holman, 2002).
C p 965:7 0:091  T 9  10
5  T 2 300 6 T
6 1000 K 11
l 4  10
6 5  10
8  T
10
11  T 2 300 6 T
6 1000 K 12
k f
3:82  10
3 1:2  10
4  T
7:3  10
8  T 2
2:38  10
11  T 3 300 6 T 6 1000 K 13
2.4. Grid Independent study and numerical validation
The grid independent study has been carried out to
study the eect of the grid size on simulation. The grid with
850,675, 1,432,962 and 22,408,116 are considered for simu-
lation. The deviation of the heat loss from the receiver for
850,675 and 22,408,116; 1,432,962 and 22,408,116 are
about 9.15% and 2.79% respectively. Hence, the grids with
1,432,962 cells are considered for further numerical simula-
tion. The numerical procedure is validated with experimen-
tal data (Ma, 1993) for head-on wind speed of 3.6 m/s and
receiver tubing surface temperature of 550 K and is shown
in Fig. 4. The deviation between the present numerical
model and experimental data (Ma, 1993) is found to vary-
ing from +21% to
14% which is comparable.
Fig. 4. Validation of present numerical model with experimental data
(Ma, 1993).
3. Results and discussion
3.1. Estimation of the heat losses due to wind eects
The heat losses from the modied cavity receiver have
been estimated considering various parameters such as
wind speed (V), wind directions (u), receiver inclinations
(b) and receiver aperture diameter ratio (RAD). The heat
losses from the receiver surface is expressed in terms of heat
losses per unit area of the receiver inner wall in order to
compare the heat losses from the receiver with dierent
opening ratios. This is given by:
Qloss
qcv 14
As
3.1.1. Eect of receiver inclination and wind direction for
given wind speed
Fig. 5(a) presents the temperature contours of the recei-
ver at dierent inclination for no-wind (natural convection)
case. It can be seen that, when the receiver inclination var-
ies from 0 to 90, the stagnation zone increases and the
convective zone reduces. It is clear that natural convective
heat loss from the receiver greatly depends on the receiver
inclination. Fig. 5(b) presents the velocity contours of the
receiver at head-on wind speed of 5 m/s for dierent recei-
ver inclinations. When the receiver inclination varies from
0 to 90, the amount of air entering the increases due to
pressure developed inside the receiver due to the receiver
inclination. It can be seen that the variation of the velocity
contours on back side of the receiver are dierent for dier-
ent receiver orientations. The adverse pressure gradient
refers to the pressure developed in and around the receiver
due to the wind blowing over the receiver. The adverse
pressure gradient is formed due to the receiver inclination,
thus forming one of the important reasons for variation of
heat loss from the receiver. The velocity contours shown in
Fig. 5 corresponds to the velocity contours at the plane,
z = 0 with size of the domain about 4D  8D. Hence the
velocity near wall of the rectangular domain is not visible.
The image shows the close view (4D  8 D) focusing the
receiver.
The eect of receiver inclinations and wind directions
for a given wind speed is shown in Fig. 6. The variation
of the heat losses from the receiver surface with various
receiver inclinations for RAD = 0.4 and V = 1 m/s, 5 m/s
and 10 m/s for dierent wind directions are shown in
Fig. 6(a)(c) respectively. When b = 0, it is sunrise condi-
tion (altitude angle = 0) and b = 90 corresponds to the
noon condition (altitude angle = 90). Hence, the receiver
inclination is considered varying from 0 to 90. The heat
losses from the receiver at b = 0 may be lesser as the system
just started to operate, however it is more appropriate for
b = 180 as it corresponds to the sunset condition and
hence the receiver is still at higher temperature, when the
receiver is operated throughout the day. The receiver
inclinations from b = 90 to 180 is the reverse direction
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198 191

Fig. 5. (a) Temperature contours of the receiver with dierent receiver

inclination for natural convection case (b) velocity contours of the receiver
at 5 m/s head-on wind speed for dierent inclinations (a) b = 0, (b)
b = 45 and (c) b = 90.

of 090. The receiver inclination of b = 0 represents the

limiting conditions for both b = 0 and 180. From Fig. 6
(a), it can be seen that as receiver inclination varies from
0 to 90, heat losses decrease for all wind directions and fol-
lows a similar trend like natural convection heat loss as
wind speed of 1 m/s is not strong enough to cause the eect
of forced convection. However, at higher wind speeds,
there is no denite trend in heat loss. From Fig. 6
(b) and (c), for wind directions, u =
90,
60 and

30, the wall heat losses decreases for receiver inclination
of 0 to 60 and then increases at 90. For side-on wind
(u = 0) and at higher wind speed, say, V > 5 m/s, the vari-
ation of wall heat loss is marginal (less than 5%) irrespec-
tive of receiver inclination. For head-on wind, at
Fig. 6. Variation of heat loss from the receiver surface with receiver
inclination and wind direction for dierent wind speeds (a) V = 1 m/s (b)
V = 5 m/s (c) V = 10 m/s.
V = 5 m/s, the heat loss reduces and then increases to a
maximum of 10%. However for V = 10 m/s, the variation
of heat loss is about 4%, 27.5% and 56% respectively for
receiver inclinations 30, 60 and 90 when compared to
0. Similarly for back-on wind, at V = 5 m/s, variation
ranges between
36% and 30%. At higher wind speed
192 K.S. Reddy et al. / Solar Energy 131 (2016) 183198

i.e., at 10 m/s, the variation is about
25% and 60%.The

heat losses from the receiver is higher at wind directions,
u = 0 to 90 than wind directions u =
90 to 0. There
is no denite trend in heat loss variations with receiver
inclinations and wind directions except at side-on wind
condition. From these gures, it can be concluded that,
heat loss remains unchanged at receiver inclination of 90
irrespective of the wind direction but varies for dierent
wind speeds i.e., at 1 m/s, the heat loss is about 400 W/
m2 whereas at 5 m/s and 10 m/s, it is about 1300 W/m2
and 2200 W/m2 respectively. At higher wind speeds greater
than critical wind speed, the maximum heat loss occurs at
90 receiver inclination irrespective of the wind direction.
The wind blowing from any direction (u =
90 to 90)
has the same eect. When the receiver is facing down, there
is less scope for the air to sweep away the hot air, which
aects the heat losses from the receiver. Hence, the forced
convection heat losses end up in same point irrespective of
the wind direction.

3.1.2. Eect of wind speed and receiver inclination for given

wind direction
The variation of heat losses from the receiver for various
wind speeds and receiver inclinations due to back-on, side-
on and head-on wind conditions for RAD = 0.4 are shown
in Fig. 7. The heat loss increases linearly for wind speeds
greater than 2.5 m/s for dierent receiver inclinations.
But at lower wind speeds say, no wind and 1 m/s, the vari-
ation of the receiver heat loss is less for side facing cavity
(b = 0). This is due to insucient wind movement to take
away the hot air from inner portion of the receiver. For
the receiver inclination of 60 and 90, the geometry of
the receiver facilitates easy movement of hot air from the
receiver. Hence, a linear increase in heat loss is observed.
For side-on wind, and at higher wind speed, the variation
of heat loss remains marginal, irrespective of receiver
inclination.
A detailed view of the velocity and temperature con-
tours in and around the receiver (RAD = 0.4, b = 0) for
back-on, side-on and head-on wind at V = 5 m/s are shown
in Figs. 810 respectively. The velocity vectors and con-
tours in and around the receiver in XZ/XY plane respec-
tively are shown in (a) and (b), (c) shows the velocity
contour at the receiver aperture, the temperature contours
inside the receiver at dierent planes: horizontal plane
(XY plane) and vertical plane (XZ plane) respectively
are shown in (d) and (e) respectively. The heat loss from
the receiver depends on the wind speed, amount of air that
enters and leaves the receiver, pattern of air recirculation Fig. 7. Variation of heat loss from the receiver with dierent wind speeds
for dierent wind direction (a) back-on (b) side-on (c) head-on wind.
and stagnation zone (stagnation air pocket) inside the
receiver. The ow pattern due to back-on wind condition
is shown in Fig. 8(a) and it is observed that there is recircu-
lation, adverse pressure gradient and ow separation The pattern of air recirculation and stagnation zone inside
behind the receiver in the ow direction. The wind blow the receiver in the vertical plane (XZ plane) can be
around the receiver, for side-on wind is shown in Fig. 9 observed from the velocity vectors (Figs. 8(b) and 10(b)).
(a). The recirculation is observed at the aperture of the It can be seen that the recirculation patterns are more or
receiver for head-on wind, which is shown in Fig. 10(a). less similar in the case of the head-on and back-on wind
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
Fig. 8. Velocity and temperature contours for back-on wind for 5 m/s of wind speed (a) velocity vector around the receiver at middle plane (Y = 0), (b)
velocity vector in the receiver at middle plane (Y = 0), and velocity magnitude at aperture (X = 0), (c) temperature contour in horizontal plane (Z = 0) and
(d) temperature contour in vertical plane (Y = 0).
and whereas in case of side-on wind, it is non-uniform
(Fig. 9(b)). This may lead to variation in the heat loss for
dierent wind directions and receiver inclinations for a
given wind speed.
From the velocity contours at the aperture of the mod-
ied cavity receiver, positive values in the velocity contour
indicates that the hot air is leaving from the receiver and
negative values indicates that ambient air is entering the
receiver and zero indicates the plane of separation between
the air which enters and leaves the receiver. The heat loss
from the receiver depends on the speed of the wind entering
and leaving the receiver due to the wind from various direc-
tions. In case of back-on wind (Fig. 8(c)), these velocities
are lesser than that of head-on (Fig. 9(c)) and side-on wind
conditions (Fig. 10(c)). The entry and exit wind speed in
case of the side-on wind are higher than other wind direc-
tions and therefore higher heat loss may be expected when
compared to other wind directions.
The temperature contours at XY and XZ plane for dif-
ferent wind directions say head-on (Fig. 8(d) and (e)), side-
on (Fig. 9(d) and (e)) and back-on (Fig. 10(d) and (e)) wind
are also shown. It can be seen that temperature variation
inside the receiver in case of head-on and back-on wind is
high as compared to that of side-on wind. There is a clear
stagnation zone that can be identied because of higher
temperature zone at the top in case of vertical plane (XZ
plane) for head-on and back-on wind conditions.
193
3.1.3. Eect of wind speed and direction for given receiver
inclination
The variations of the heat losses from the receiver sur-
face with wind speed and back-on, side-on and head-on
winds for RAD = 0.4 and receiver inclinations b = 0, 30
and 60 are shown in Fig. 11. For the wind speed upto
2.5 m/s, there is no denite trend in the variation of heat
losses with wind direction and for the range of wind
directions say,
90 to 40, heat loss is lesser than the
natural convection (no wind conditions) at b = 0. This
indicates that there is a reduction of heat loss due to
suppression of hot air, which tries to escape from the
receiver. For the wind speed, say above 5 m/s, there is
a uniform trend in the variation of heat ux and this
indicates that forced convection is dominating. It is also
observed that maximum heat loss occurs for side-on
wind (u = 0). The velocity contours show that the
amount of air entering the cavity varies with receiver
inclination and hence the temperature distribution in
the receiver. For back-on wind, when the receiver incli-
nation is 0, the amount of air entering the receiver is
less due to recirculation in-front of the receiver aperture
and when the receiver inclination is 90, the hot air
enters through aperture due to the parallel movement
of wind towards the receiver aperture. This can be
armed with the help of velocity and temperature con-
tours. This variation occurs due to suppression of the
194 K.S. Reddy et al. / Solar Energy 131 (2016) 183198

Fig. 9. Velocity and temperature contours for side-on wind for 5 m/s of wind speed (a) velocity vector around the receiver at middle plane (Z = 0), (b)
velocity vector in the receiver at middle plane (Z = 0) and velocity magnitude at opening of receiver (X = 0), (d) temperature contour in horizontal plane
(Z = 0) and (e) temperature contour in vertical plane (Y = 0).

Fig. 10. Velocity and temperature contours for head-on wind for 5 m/s of wind speed (a) velocity vector around the receiver at middle plane (Y = 0), (b)
velocity vector inside the receiver at middle plane (Y = 0), (c) velocity magnitude at opening of the receiver (X = 0), (d) temperature contour horizontal
plane (Z = 0) and (e) temperature contour at vertical plane (Y = 0).
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198 195

inclinations for a wind speed of V = 5 m/s is shown Fig. 12.

The maximum heat loss is observed at side-on wind for
aperture diameter ratios 0.40.8. However, for RAD = 1,
maximum heat loss is observed at 30 for the case of recei-
ver inclination b = 0 as shown in Fig. 12(a). When the
receiver inclination changes to 30, the maximum heat loss
occurs at side-on wind direction for RAD = 0.4 and 0.6,

Fig. 11. Variation of heat loss from the receiver with wind directions for
dierent wind speeds and receiver inclinations (a) b = 0, (b) b = 30, (c)
b = 60.

hot air inside the receiver because of the recirculation

zone in front of the receiver, which directly opposes ow
of wind.

3.1.4. Eect of aperture ratio and wind direction for a given

receiver inclination Fig. 12. Variation of heat loss from the receiver surface with wind
The variation of heat losses from the receiver with wind direction for dierent receiver inclinations (a) b = 0, (b) b = 30 and (c)
direction for dierent aperture diameter ratios and receiver b = 60.
196 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
whereas for RAD = 0.8 and 1, maximum heat loss occurs at
a wind direction of 30 (Fig. 12(b)). As the receiver inclina-
tion changes, it is more convenient for the hot air to sweep
away the heat from the receiver due to the geometry of the
receiver. When the receiver inclination is 60, the heat loss
values are higher for wind directions, u = 60 and 90
(Fig. 12(c)), as the sweeping of hot air from the cavity is
facilitated easily.
The variation in heat losses with dierent wind speeds
and aperture diameter ratios due to side-on wind (u = 0)
is shown in Fig. 13. As receiver opening increases, heat
losses from the receiver also increases and this may be
due to the following reasons: increase in the air movement,
reduced stagnation zone and increased amount of sweeping
of the hot air present in the receiver due to the external
wind. For no wind condition, the increase of heat loss is
low as compared to windy conditions. An increase of
28% is observed for aperture diameter ratio (RAD) chang-
ing from 0.4 to 1 for no wind condition. As wind speed
increases, the heat loss also increase i.e. 35% and 43% for
wind speed of 1 m/s and 2.5 m/s respectively. At high wind
speed, the increase in heat loss remains same and average
increase of heat loss is about 48% for wind speed of
510 m/s.
The comparison of natural and forced convection heat
losses for various aperture diameter ratios and receiver
inclinations due to head-on wind (u = 90) with V = 5 m/s
is shown in Fig. 14. As expected, natural convection heat
loss decreases with increase in receiver inclination from
0 to 90. In addition, larger opening of the receiver leads
to increase in natural convection heat loss. However, same
trend is not observed for forced convection. The maximum
forced convection heat loss is found to be at b = 60 and
RAD= 0.8 and 1. As the opening ratio of the receiver
increases, trapping of hot air within the cavity is reduced.
The inclination of the receiver is such that it allows to
sweep away the heat tending to increase in the heat loss.
Fig. 13. Variation of heat loss from the receiver surface with dierent
wind speeds for dierent receiver aperture ratio.
Fig. 14. Comparison of variation in wall heat ux with receiver
inclination for dierent openings of the cavity receiver for the case
V = 5 m/s due to head- on wind and natural convection.
3.2. Correlations for evaluating the combined convection heat
losses from the modied cavity receiver
The 3-D numerical simulations have been carried out
for wind (combined/total) and no wind conditions (natural
convection) considering the eect of parameters such as:
inclination (b), wind direction (u), wind speed (V) and
aperture diameter ratio (RAD) on the combined free-
forced convection from modied cavity receiver. Nusselt
number (Nut) for wind case is dened as Nut = htDi/kf
The heat transfer coecient is calculated by employing
the wall heat loss (qcv) divided by temperature dierence
(Taw,s
Ta). All the properties are taken at bulk mean tem-
perature which is dened as Tb = (Ta + Taw,s)/2, where
temperature, Taw_s is dened as the area weighted average
surface temperature
RA . Rof the receiver surface which is given
A
by T aw s 0 TdA 0 dA where, T is the temperature of
the surface in K and A corresponds to the area in m2.
A correlation has been developed to estimate the com-
bined convection heat transfer from the modied cavity
receiver. A non-linear regression analysis is carried out
considering to estimate the Nusselt number for combined
convection heat loss from the receiver as a function of b,
u, d/Di, Re (to account the velocity eect) and Gr (to
account the buoyancy eects). The Nusselt number correla-
tion is given as:

s
t
q r d Gr
Nut p1 cosb 3 sinu sin2u sin3u
Di Re2
15
Ranges of parameters are: 0 6 b 6 90;
90 6 u 6 90;
0.4 6 d/Di 6 1; 1.2  104 6 Re 6 9  104; 1.97  108 6
Gr 6 1.98  108.
The present form for the receiver inclination has been
chosen to avoid zero when the receiver inclination is 90.
It can be seen from the index term q, the importance of
the receiver inclination on the forced convection heat loss.
 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
Table 1
Values of coecients (p, q, r, s and t) in Nut equation (Eq. (15)) for dierent wind directions.
Wind direction (u) p q

90 to
30 35.112
0.548
0 (side-on) 2.613
0.545
3090 87.138
0.220
Fig. 15. Parity plot for Nusselt number correlation.
Similarly to avoid a zero value of the term for wind direc-
tion, the present form is selected. The monotonous relation
between wind direction and heat loss is avoided by consid-
ering the present form. These present forms for receiver
inclination and wind direction terms are chosen based on
several combinations to get a good t. As heat loss from
the receiver under various wind directions is complex and
does not follow any denite trend, the correlations have
been developed for three dierent wind direction ranges:
(i) Range 1: u =
90 to
30, (ii) Range 2: u = 0 and
(iii) Range 3: u = 3090. The coecients (p, q, r, s, and
t) in Eq. (15) for dierent range of the wind directions
are given in Table 1, with corresponding coecients of cor-
relations (R2). In Eq. (15), the non-dimensional numbers
are dened by taking Di as the characteristic length equal
to diameter of the receiver. The wind velocity eect is given
in terms of ratio of Reynolds number and Grashof number.
To know the importance of natural/forced convection, (Gr/
Re2) can be calculated. If the value of the term is less than
0.1, ow is forced convection dominated and if it is greater
than 10, it can be concluded that the ow is natural convec-
tion dominant. When the value lies between 0.1 and 10,
then both natural and forced convection play role in the
heat loss from the receiver. Based on proposed Nusselt
number correlation, the heat losses from the cavity recei-
vers can be estimated eectively which will help in under-
standing the eect of wind and various other parameters
on solar dish collector system.
The parity plot between the numerical and correlated
Nusselt number is shown in Fig. 15.
197
r s t R2

0.292 0.180
0.323 0.98
2.394
0.089
0.324 0.98

0.394
0.049
0.322 0.98
4. Conclusion
A comprehensive heat loss study of modied cavity
receiver has been carried out considering geometry,
orientation and wind characteristics. The wind eects
play signicant role in characterising heat losses which is
due to variations in the ow conditions inside the receiver
and the adverse pressure gradient and ow separation at
backside of the receiver. It is found that there exists a
critical wind speed for each operating conditions below
which there is no forced convection eect. This critical
speed was found to be less than 5 m/s, but varies with u
and b. At higher wind speeds and all receiver inclinations,
the variation in heat loss due to side-on wind is negligible.
The heat loss from the receiver is higher for side-on wind
when compared to head-on and back-on wind conditions.
The natural convection heat loss decreases when the recei-
ver inclination varies from 0 to 90. The forced convection
heat loss from the receiver remains unchanged and remains
maximum at receiver inclination of 90 irrespective of wind
direction. The forced convection heat loss is found maxi-
mum at receiver inclination of 60 for receiver opening
ratio in the range of 0.81.0. Generalized Nusselt number
correlation has been developed to estimate the convection
heat losses from the modied cavity receiver due to wind
eects. The correlation can be used to estimate the forced
convection heat losses from solar dish collectors used for
high temperature applications.
Acknowledgement
The authors wish to acknowledge the nancial support
provided by Department of Science and Technology
(DST), New Delhi, Government of India through research
project (Grant No.: DST/TM/SERI/2k12/16(G)).
References
Christo, F.C., 2012. Numerical modelling of wind and dust patterns
around a full-scale paraboloidal solar dish. Renew. Energy 39, 356
366.
Cui, F., He, Y., Cheng, Z., Li, Y., 2013. Study on combined heat loss of a
dish receiver with quartz glass cover. Appl. Energy 112, 690696.
Flesch, R., Stadler, H., Uhlig, R., Pitz-Paal, R., 2014. Numerical analysis
of the inuence of inclination angle and wind on the heat losses of
cavity receivers for solar thermal power towers. Sol. Energy 110, 427
437.
Flesch, R., Stadler, H., Uhlig, R., Hoschmidt, B., 2015. On the inuence
of wind on cavity receivers for solar power towers: an experimental
analysis. Appl. Therm. Eng. 87, 724735.
Fluent Inc., Fluent Users guide. Version 6.2.16, Lebanon, USA, 2005.
198 K.S. Reddy et al. / Solar Energy 131 (2016) 183198
Holman, J.P., 2002. Heat Transfer, eighth ed. Tata McGraw-Hill
Publishing Company Limited, New Delhi.
Jilte, R.D., Kedare, S.B., Nayak, J.K., 2014. Investigation on convective
heat losses from solar cavities under wind conditions. Energy Procedia
57, 437446.
Kaushika, N.D., 1991. Viability aspects of paraboloidal dish solar
collector systems. Renew. Energy 3 (6/7), 787793.
Kumar, N.S., Reddy, K.S., 2008. Comparison of receivers for solar dish
collector system. Energy Convers. Manage. 49, 812819.
Leibfried, U., Ortjohann, J., 1995. Convective heat loss from upward and
downward-facing cavity solar receivers: measurements and calcula-
tions. ASME J. Solar Energy Eng. 117, 7584.
Liovic, P., Kim, J.S., Hart, G., Stein, W., 2014. Wind dependence of
energy losses from a solar gas reformer. Appl. Therm. Eng. 63, 333
346.
Ma, R.Y., 1993. Wind eects on convective heat loss from a cavity receiver
for a parabolic concentrating solar collector. Sandia National Labo-
ratories Report, SAND92-7293.
Paitoonsurikarn, S., Lovegrove, K., 2003. On the study of convection loss
from open cavity receivers in solar paraboloidal dish applications. In:
Proceedings of 41st Conference of the Australia and New Zealand
Solar Energy Society (ANZSES), Melbourne, Australia.
Paitoonsurikarn S, Lovegrove K. 2006. Eect of paraboloidal dish
structure on the wind near a cavity receiver. In: Proceedings of the
44th Annual Conference of the Australian and New Zealand Solar
Energy Society, Canberra, Australia.
Prakash, M., Kedare, S.B., Nayak, J.K., 2009. Investigations on heat
losses from a solar cavity receiver. Sol. Energy 83, 157170.
Prakash, M., Kedare, S.B., Nayak, J.K., 2012. Numerical study of natural
convection loss from open cavities. Int. J. Therm. Sci. 51, 2330.
Reddy, K.S., Kumar, N.S., 2009. An improved model for natural
convection heat loss from modied cavity receiver of solar dish
concentrator. Sol. Energy 83, 18841892.
Reddy, K.S., Kumar, N.S., 2008. Combined laminar natural convection
and surface radiation heat transfer in a modied cavity receiver of
solar parabolic dish. Int. J. Therm. Sci. 47, 16471657.
Reddy, K.S., Natarajan, S.K., Veershetty, G., 2015. Experimental
performance investigation of modied cavity receiver with fuzzy focal
solar dish concentrator. Renew. Energy 74, 148157.
Sendhil Kumar, N., 2008. Investigation of cavity receivers for fuzzy focal
solar parabolic dish concentrator. Ph.D. thesis, IIT Madras.
Taumoefolau, T., Paitoonsurikarn, S., Hughes, G., Lovegrove, K., 2004.
Experimental investigation of natural convection heat loss from a
model solar concentrator cavity receiver. ASME J. Solar Energy Eng.
126, 17.
Vikram, T.S., Reddy, K.S., 2014. Estimation of heat losses from modied
cavity mono-tube boiler receiver of solar parabolic dish for steam
generation. Energy Procedia 57, 371380.
Vikram, T.S., Reddy, K.S., 2015. Investigation of convective and radiative
heat losses from modied cavity based solar dish steam generator using
ANN. Int. J. Therm. Sci. 87, 1930.
Wu, S.Y., Shen, Z.G., Xiao, L., Li, D.L., 2015. Experimental study on
combined convective heat loss of a fully open cylindrical cavity under
wind conditions. Int. J. Heat Mass Transfer 83, 509521.
Wu, S.Y., Xiao, L., Cao, Y., Li, Y.R., 2010. Convection heat loss from
cavity receiver in parabolic dish solar thermal power system: a review.
Sol. Energy 84 (8), 13421355.
Wu, S.Y., Xiao, L., Li, Y.R., 2011. Eect of aperture position and size on
natural convection heat loss of a solar heat-pipe receiver. Appl. Therm.
Eng. 31, 27872796.
Xiao, L., Wu, S.Y., Li, Y.R., 2012. Numerical study on combined free-
forced convection heat loss of solar cavity receiver under wind
environments. Int. J. Therm. Sci. 60, 182194.
Xie, W.T., Dai, Y.J., Wang, R.Z., 2011. Numerical and experimental
analysis of a point focus solar collector using high concentration
imaging PMMA Fresnel lens. Energy Convers. Manage. 52, 2417
2426.
Zapata, J.I., Pye, J., Lovegrove, K., 2013. A transient model for the heat
exchange in a solar thermal once through cavity receiver. Sol. Energy
93, 280293.