 Open Access
 Total Downloads : 9
 Authors : Lakshminarasimha N, Dr. M.S. Rajagopal, M. Vedavyasa
 Paper ID : IJERTCONV3IS17087
 Volume & Issue : NCERAME – 2015 (Volume 3 – Issue 17)
 Published (First Online): 24042018
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Thermal Management of Solar Power Pack using Computational Fluid Dynamics
Lakshminarasimha N 1
1PG Student, Thermal Engg. Department of Mechanical Engg.
Global Academy of Technology Bangalore, India
Dr. M.S. Rajagopal 2
2Professor & Head Department of Mechanical Engg. Global Academy of Technology
Bangalore, India

Vedavyasa 3
3Associate Professor Department of Mechanical Engg. Global Academy of Technology
Bangalore, India
Abstract Solar Power Pack (SPP) is an enclosure which houses heat generating electrical devices such as Battery Bank, Inverter and Controller. The present study involves cooling and ventilation analysis of the enclosure such that the electrical components operate at less than their limiting temperatures. Further, optimization studies are also carried out to determine fan location and locations of inlet and exit vents besides positioning of Baffles for uniform air motion. This study will also help in minimizing the pumping power cost by determining the pressure losses. Thermal Management of SPP has been carried out using CFD software ANSYS Workbench FLUENT for both 2D and 3D geometry. Numerical results obtained from FLUENT agree well with analytical results.
Keywords– Thermal Management, Solar Power Pack, Enclosure, CFD, Optimization, FLUENT, Cooling, Ventilation
Terminology–
L = Length of SPP
W = Width of SPP
H = Height of SPP
Lb = Location of Battery Bank LC = Location of Controller
Li = Location of Inverter
Lf = Location of exhaust fan LV = Location of inlet vent 2D = Twodimensional
3D = Threedimensional TSA = Total surface area CFM = Cubic feet per minute

INTRODUCTION
In todays climate of growing energy needs and increasing environmental concern, alternatives to use of nonrenewable and polluting fossil fuels have to be invested. One such alternative is solar energy. Solar power is the conversion of sunlight
into electricity directly using photovoltaics. The latest and the most cost effective method for integrating solar power to homes/offices/colleges/rural electrification etc. are through the use of Solar Power Pack. Present work is generic and can be used as preliminary literature for studying any different kinds of Solar Power Pack models.
Internal flow and thermal analysis in an enclosure containing heat generating equipments is always a challenge in the field of Heat transfer. Due to high expense of experimental study in recent years, Computational Fluid Dynamics (CFD) is gradually becoming the most efficient tool in thermal enclosure design. Study of Solar Power Pack comprises of; overall evaluation of design involving ventilation and cooling, study of velocity and temperature contours and optimization of design such as; fan size and location, location and size of inlet and exit vents and positioning of Baffles for uniform air motion. Also compute the many derived parameters along with the graphical representation of the interested regions. The main objective of the present work is to undertake a numerical investigation to evaluate the thermal design of SPP and further optimize the design to minimize the cost using simple 2D & 3D models.

LITERATURE SURVEY
In this section, literature survey has been conducted to understand the state of the art in thermal management of Solar Power Packs in particular and CFD in general to know the appropriate boundary conditions, modeling of heat sources and various CFD models. The summaries of most important papers are listed below.
Hoffman A Pentair Company [1], [2003], released technical information data sheet on heat dissipation in an electrical enclosure. This technical data/cut sheet is majorly for electrical enclosures exposed to outdoor conditions. The technical information highlights majorly about, Enclosures temperature
rise/ Determination of temperature rise for an enclosure, Evaluation of solar heat gain and Selection of fans as per cooling requirement for an enclosure. Hence this manual is highly helpful in preliminary design stage, in predicting the cooling requirements for electronic and electrical enclosures.
Bud Industries, Inc., [2], [2007], released technical data hand book on Enclosure design tips. This guide helps in selecting cabinets, enclosures and other packaging system for electronic products. Also this booklet summarizes some of the more important issues of packaging for electronic systems and products, with the goal of helping us evaluate our options quickly, and then selecting the optimum solution for our application. As per project concerned the booklet majorly highlights about Materials and finishes for an enclosure and Basics of cooling for cooling requirement in an enclosure.
It is observed that very few literatures are available on thermal management of SPP using CFD though thermal analysis at PCB and chip level is available. However, industrial data sheets are available as ready reckoner to help designers to choose the right size of fan and number of fans. This study is aimed at optimizing the cooling and ventilation parameters to have better understanding of the system.

METHODOLOGY
The analysis of SPP is been carried out using Commercial CFD code ANSYS Workbench FLUENT and Modeling and Meshing for 3D SPP enclosure were carried out in ANSYS ICEPAK.
ANSYS WORKBENCH
Combination of
Mahendra Wankhede, et al., [3], [2010], Paper comprises study on evaluation of cooling solutions for outdoor electronics. Both experimental and CFD simulation had been carried out for three different conFig.urations of enclosures including a
Geometry using design modeller
Meshing using ICEM
Boundary condition, solver & postprocesssing using FLUENT
solar radiation shield, doublewall enclosure and a airtoair heat exchanger; using a typical Aluminium enclosure. Each enclosure was consisting 100W generating PCBs. The work showed that solar heat load can increase the internal air temperature by 20%, White coating of an enclosure reduces internal air temperature by around 25% and as the major part of the work highlighted that 5055% can reduce T due to the internal fans compared to a sealed enclosure with no fans and having radiation shield and double walled enclosure with air circulation provided modest improvements of around 25%, where as air toair heat exchanger showed improvement by 75%.
Tom Kowalski and Amir Radmehr, [4], This Paper highlights the significance of using coupled FNM and CFD based analysis in cooling of electronic enclosure. The study demonstrates the application of coupled Fluid Network Method (FNM) and CFD in the analysis of flow behavior and Thermal behavior in a telecommunication cabinet. The cabinet is forced flow air cooled consisting a stack of PCBs. The work showed detailed measurement of airflow velocities in various parts of the cabinet, the predicted variation of the flow rates through card passages and total flow through system were found to be within 10% of their experimentally measured values.
Fig. 1: Combined module of CFD code
ANSYS Workbench delivers innovative, dramatic simulation technology advances in every major physics discipline, along with improvements in computing speed and enhancement to enabling technologies such as geometry handling, meshing and postprocessing (Fig. 1). These advancements alone represent a major step ahead on the path forward in Simulation driven Product Development.
ANSYS ICEPAK provides highly simplified way of modeling and meshing for cubical electronic enclosure probems.

GEOMETRY
The Geometry of SPP consists of Battery Bank, Inverter, Controller, and Inlet vent and outlet/exhaust fan. The most important thing to be noted is that the geometry is planar symmetry. The dimensions specified for the models below in 3.1.1 and 3.1.2 are elaborated in the above terminology.
3.1.1 2D Geometry model
H
L
Fig. 2: 2D SPP Model
Fig. 2, shows the 2D cut plane (XY plane right bottom corner of Fig. 2) taken from planar

MESHING


symmetry 3D model which has been considered for 2D analysis. A simplified 2D analysis is carried out at the midplane of SPP to get an insight of the physics of the problem though this model doesnt capture the overall objective of determination of accurate flow and temperature distributions. 2D geometry does not accurately capture locations of inlet vent, fan location and heat generating devices.
3.1.2 3D Geometry model
The planar symmetry 3D SPP domain with dimensions as per above terminology is shown in Fig. 3. The 3D model captures the problem objective and fetches us real results and provides us the clear visualization of contours and vectors at our interested spots in the domain.
Fig. 3: 3D SPP Model
Meshing is the important criteria as part of analysis considered. Maximum time spent in meshing is the time well spent. Mesh independence studies have been carried out for both 2D and 3D domain meshes. The details of meshing for 2D and 3D domain are as described below.

2D Meshing
The 2D mapped fine mesh is shown in Fig. 4. Mesh independence study graphs for the 2D mesh are shown in Fig. 5 and Fig. 6. Based on this study, further analyses are carried out for Quad mesh with 46280 elements and 47204 nodes. The convergence criteria for the continuity, xvelocity, yvelocity, energy, k and epsilon are shown in Fig. 7.
Fig. 4: 2D Mesh
Max. velocity (m/s)
Max. velocity (m/s)
Mesh independence chart Max. Velocity vs. Mesh elements
6.465
6.46
is the optimum mesh which is shown in Fig. 8. The convergence criteria trends for the continuity, x velocity, yvelocity, zvelocity energy, k and epsilon are shown in Fig. 11.
6.455
6.45
6.445
32144 46280 52560
Max. velocity (m/s)
Fig. 5: 2D Mesh independence plot for Maximum Velocity vs.
Mesh elements
Max. temperature (K)
Max. temperature (K)
Mesh independence chart Max. Temp vs. Mesh elements
329.5
329
328.5 328
327.5
max. temperature (k)
Max. Velocuty (m/s)
Max. Velocuty (m/s)
6.6
6.4
6.2
6
5.8
Fig. 8: 3D Mesh
Fig. 6: 2D Mesh independence plot for Maximum Temperature vs. Mesh elements
5.6
5.4
5.2
5
Max Velocity (V)
Number of mesh elements
Fig. 9: 3D Mesh independence plot for Maximum Velocity vs.
Mesh elements
Max Temp (deg. C)
Max Temp (deg. C)
47
46
45
44
43
42
41
40
Fig. 7: 2D Mesh Convergence criteria
Max temperature (T)

3D MESHING
Creating a 3D mapped hexahedral mesh is a big challenge; however Icepak provides an easy platform to create mapped hexahedral mesh for cubical geometries. Icepak also facilitates fluid and solid surface extractions which can be named appropriately. Icepak is used for modeling & meshing process & the mesh has been transferred to fluent for further analysis.
Mesh independence study graphs for the 3D mesh are shown in Fig. 9 & Fig. 10. The study shows that mesh with 272786 elements and 284868 nodes
39424
80249
157005
272786
416174
1059528
39424
80249
157005
272786
416174
1059528
Number of mesh elements
Fig. 10: 3D Mesh independence plot for Maximum Temperature vs. Mesh elements
Page 4 of 10
Fig. 11: 3D Mesh Convergence criteria

SOLUTION METHODOLOGY
In the present investigation, the enclosure exhaust fan of 285 CFM has been selected through analytical calculation [1]. Flow is considered to be incompressible and steady. The internal flow is governed by Reynolds Average NavierStokes equations (RANS). Standard Kepsilon model is used to solve the flow analysis. The heat generating sources wall are at heat flux and no slip condition. Standard Kepsilon model comes under two equation model of turbulence kinetic energy k and its dissipation rate & hence these two equations are solved for obtaining the result. The incoming air through inlet vent is due to negative pressure occurred & removing heat through exhaust fan. Transport equation for momentum and turbulence parameter is solved with SIMPLE discretization scheme.

MATHEMATICAL MODELS
Mass con
+ . (
Mass con
+ . (
FL
ser
um
FL
ser
um
UE
vatio
) =
con
UE
vatio
) =
con
NT a
n e
m
serva
NT a
n e
m
serva
Conservation equations of mass and momentum for all flows are solved in ANSYS FLUENT and an additional equation for energy is solved for flows involving heat transfer. Flow inside a Solar Power Pack involves both fluid flow and fluid flow with heat transfer, hence governing equations [8] that are solved in re as listed below:
Moment
Moment
(
Where,
(
Where,
) +
) +
ti
)
ti
)
o
=
o
=
ua
ua
)
)
+
+
quation:
eq
is
.
eq
is
.
=
Ene
=
Ene
t
[gy c
t
[gy c
. (
stres
+
serva
. (
stres
+
serva
tenso
T)
ion eq
tenso
T)
ion eq
+
give
]
n:
+
give
]
n:
.
n by
.
n by
n tion:
,
,
( +
uati
uati
(
on
+
ove
(
on
+
ove
t
(
equatio
t
(
equatio
o
+ )
ns are
o
+ )
ns are
=
a ge
=
a ge
) +
orm
) +
orm
he s r
compressible and incompressible flows. The governing equations with no time derivative term states steady flow.
be Th Sta
be Th Sta
rbule
rd k
) +
rbule
rd k
) +
to nce
to nce
model
rbulen
)
model
rbulen
)
Also additional transport equations as shown low, need be solved since flow is turbulent.
e tu
nda
(
e tu
nda
(
Tu
(
Tu
(
used for this analysis is
+
+
el.
el.
+
+
ce mod
+
+
An
An
d
(
d
(
) +
) +
(
(
+
)
+
)
+
+
=
(
(
+
+
)
)
+
+
=
y vis
y vis
nera
nera
ion
ion
of
of
+
=
=
Where, turbulent or edd cosity,
he o
,
ues
he o
,
ues
vera
,
s be
vera
,
s be
issip
, a
:
issip
, a
:
p/>
and Gk & Gb represents the ge t turbulence kinetic energy due to mean velocity gradients and buoyancy. YM represents the contribution of the fluctuating dilation in compressible turbulence to t ll d ation rate.
ere ave
ere ave
el c ing
el c ing
=
=
H , the mod onstants C C C, nd
h the follow default val a low
C = 1.44 , C = 1.92 , C = 0.09, 1.0 ,
= 1.3
All the above mathematical models cannot be
solved by analytical method for complex flows. Hence all these equations are solved using FLUENT. Further indepth details regarding mathematical models can be referred in FLUENT Theory guide [8].

RESULTS AND DISCUSSION

2D RESULTS
It is stated in section 3.1.1, that the 2D analysis does not capture the overall problem objective but the main purpose of 2D analysis is to study different solver options, different Boundary conditions and effect of critical geometric parameters and understanding the velocity, pressure and temperature contours.
Figures shown below are the velocity (Fig. 12) and temperature contours (Fig. 13) for the SPP enclosure.
r
( )
The ab
r
( )
The ab
. . ( Sh
neral f of governing equations [8] and are valid for both
Fig. 12: Velocity contour for 2D domain
Fig. 13: Temperature contour for 2D domain
Maximum Velocity (m/s)
6.46
Maximum Temperature ()
57
Enclosure fluid Temperature rise (T) in ()
0.809
Table 1: 2D analysis results
Table1 shows the results obtained from 2D analysis. Maximum temperature is found at the top surface of the Battery Bank and maximum velocity obviously near inlet and outlet. The locations of Battery Bank, Inverter and Controller are actual in 3D, where as in 2D their locations cannot be captured accurately, since 2D mid xy plane is
considered. Further, locations of fan inlet and outlet vents are also approximate. Thus the results are expectedly not accurate, but certainly provide an insight of complex coupled flow and heat transfer problem.

3D RESULTS
3.5.2a. Velocity contours and Temperature contours
Fig. 14: Velocity contour for 3D domain
Fig. 15: Temperature contour for 3D domain
Maximum Velocity (m/s)
6.0
Maximum Temperature ()
42.5
Enclosure fluid Temperature rise (T) in ()
4
Table 2: 3D analysis results
Figures 14 & 15 show the velocity and temperature contours in 3D domain. The temperature contours near the Battery is shown in Fig. 16 and near the Inverter and Controller in Fig. 17. The maximum temperatures at the Inverter, Battery and Controller are shown in table3. Though the overall velocity distribution is relatively uniform, the flows are relatively lesser at the Inverter resulting in higher temperature. Hence flows need to be optimized such that the temperatures of the devices are uniform.
Fig. 16: Temperature in Battery
Fig. 17: Temperature in Inverter and Controller
Table2 shows the maximum velocity and temperatures in 3D analysis. Comparison of Table
1 and Table2, show that there is considerable difference in the values of maximum temperature, maximum velocity and temperature difference of fluid temperature. This is expected since 2D and 3D geometries are different.
The velocity stream line plot is as shown in Fig. 18.
Fig. 18: Velocity stream line plot
From the stream line plot it is quite clear that the flow is channeled in all four directions (Top, right, left & bottom) and the most predominant flow being in the top direction. Considering this there is scope to optimize the flow and a suitable Baffle at the right position can be provided to direct more air towards the Inverter. These are discussed in section 3.5.2c. Further the locations of inlet and outlet also are studied in section 3.5.2b to understand its effect on temperature distribution.
3.5.2b. Optimization of the location of inletvent and exhaust fan
Two different locations for the inlet vent have been studied (Fig. 19) and the results are shown in Table4
Heat generating devices
Temperature obtained in
Battery
39
Inverter
42.5
Controller
34.61
Volume 3, IssuTeab1l7e 3: Temperature of Heat generating devicePsublished by, www.ijert.org 7
Fig. 19: Inletvent optimization
Devices
Temperature in
for inlet location1
Temperature in
for inlet location2
Battery
39
37.64
Inverter
42.5
47
Controller
34.61
38.2
Table 4: Results obtained by optimizing the location of inlet vent
Changing the location of inlet vent from location1 to location2, the Battery temperature is reduced
3.5.2c. Positioning of Baffle
As stated in section 3.5.2a, the flow is predominant in top direction, hence decided to provide Baffle at position1 and position2.
In the first analysis, Baffles without perforation are provided near the inlet to direct more air towards the Inverter (Fig. 21). Fig.22 gives the comparison of temperatures of heat devices. It is observed that position2 is better since the maximum temperature in all devices is limited to 38.
but the temperatures of Inverter and Controller
have increased. Hence location1 is optimum. Similarly, effect of two different outlet locations on temperatures has also been carried out (Fig. 20). It is found that there was no considerable change in
Fig. 21: Positioning of Baffle with no perforated vents
Devices temperature at different baffle
the temperatures of heat devices. For location2,
position for baffle with no perforated vents
the distance between inlet to exhaust fan is lesser compared to location1 and considerable amount of air escapes without cooling the heating devices. And hence is not preferred.
39.5
Temperature in deg.C
Temperature in deg.C
39
38.5
38
37.5
37
36.5
36
35.5
35
34.5
Devices
Temperature for Baffle position1
Temperature for Baffle position2
Fig. 20: Optimization of exhaust fan
Fig. 22: Devices temperature at different Baffle position for Baffle with no perforated vents
In the next analysis, perforated Baffles with rectangular holes are provided at position 2 as shown in Fig. 23. The openings provided on the Baffle are 50% of the total area of the Baffle.
Enclosure Pressure drop & Devices temperature for different SPP analysis type
45
40
35
30
25
20
15
10
5
0
With no Baffle Baffle with no Baffle with
perforated vents at rectangular position2 perforated vents at
position2
Enclosure Pressure drop & Devices temperature for different SPP analysis type
45
40
35
30
25
20
15
10
5
0
With no Baffle Baffle with no Baffle with
perforated vents at rectangular position2 perforated vents at
position2
Fig. 23: SPP with Baffle with perforated rectangular vents
Fig. 24 shows the comparison results for three different cases viz. i. With no Baffle ii. Baffle with no vents iii. Baffle with rectangular vents. It is evident that results are better with Baffe compared
Fig. 24: Comparison graph for Temperature and Pressure drop for SPP with Baffle & with no Baffle
n
n
Battery
Inverter
Battery
Inverter
Controller
Controller
Pressure drop
Battery
Pressure drop
Battery
Inverter
Inverter
Controller
Pressure drop
Controller
Pressure drop
Battery
Battery
Inverter
Controller
Inverter
Controller
Pressure drop
Pressure drop


COMPARISON OF RESULTS WITH ANALYTICAL RESULTS
Considering flow to be incompressible and steady, Applying simple mass balance i.e. flow at inlet = flow at outlet. Since same areas provided at both
to the case without Baffles. In the cases with Baffles, Baffle without vents gives reduced temperatures but with increased pressure drops.
3.5.2d. Pressure losses and Pumping power cost
Pressure drop is directly proportional to the Pumping power. Among the various cases studied, pressure losses are minimum for the case without Baffle and case with Baffle having rectangular
inlet and outlet, the velocity at both inlet and outlet
should be 6.0 m/s as per manual calculations. It is observed to be same from CFD as well. This is applicable for both 2D & 3D cases.
Similarly, considering heat balance, the heat dissipated by the heat devices should be equal to heat carried away by the fluid. This is verified in both 2D & 3D cases.
2D case:
Fluid temperature rise by analytical method: 0.88
vents though the later is having
slightly lesser
Fluid temperature rise by CFD: 0.809
pressure drop (Fig. 24). It is obvious that Baffles result is higher drops at the entry, but pressure losses further downstream are also important. In general when the ventilation design is void of constrictions, pressure losses can be reduced which is also evident from CFD results. However, CFD plays a major role in determining the exact pressure drops which will be very useful to the designer.
3D case:
Fluid temperature rise by analytical method: 4.0 Fluid temperature rise by CFD: 4.0
Thus it is evident that the CFD results are justified as both mass and heat balances are maintained.
3.7 CONCLUSIONS
The following conclusions can be drawn based on
CFD analysis carried out on Solar Power Pack
(SPP) enclosure consisting Battery Bank, Inverter and Controller.
1, CFD is very powerful tool to determine the exact pressure drops, maximum velocities and temperatures.

CFD also provides insight of the cooling and ventilation problem, by determining velocity and
temperature distributions. designer.
This will help the

Parametric studies have been carried out to understand the effect of following parameters:

Location of Inlet vent

Location of exhaust fan (outlet vent)

Provision of Baffles

Effect of Baffles with and without vents


It can be concluded that for minimizing temperatures, Baffles without vents are recommended though there is a higher pressure drop.

Considering both temperature and pressure drop, Baffle with rectangular holes is recommended.
It must be concluded that the present study is by no means exhaustive and there is further scope for optimization. However, the CFD results are very useful to the designer and these results will help him to make decisions faster. Further, they will aid in preparing a ready reckoner for cooling and ventilation design.
ACKNOWLEDGMENT
We would like to thank the Global Academy of Technology college management and Department of Mechanical Engineering, (VTU Research Centre) for providing us the facility to successfully complete this project and encouraging us in publishing paper. We also thank M/s. SunEdison for providing data at critical junctures and their overall support.
REFERENCES

Hoffman A Pentair Company, Heat dissipation in Electrical enclosure, Technical information on Thermal Management of Electrical enclosures, Â©2003 Hoffman Enclosures Inc.

Bud Industries, Inc., Enclosure Design Tips Handbook, July 2007,
Â© Bud Industries Inc.

Mahendra Wankhede, Vivek Khaire, Dr. Avijit Goswami and Prof. S. D. Mahajan, Evaluation of cooling solutions for outdoor electronics, Journal on electronics cooling from electronicscooling.com, Volume 16, No. 3, Fall 2010

Tom Kowalski and Amir Radmehr, Thermal Analysis of an Electronics Enclosure: Coupling Flow Network Modeling (FNM) and Computational Fluid Dynamics (CFD), no details

SÃ¼ddeutscher Verlag onpact (Rittal GmbH & Co. KG), Project Planning Manual: Enclosure Heat Dissipation, Â© 2009 by SÃ¼ddeutscher Verlag onpact

Anetis Stylianos, The process of heat transfer and fluid flow in CFD problems, American Journal of Science and Technology, 2014, 1(1): 3649

A practical formula for aircooled boards in ventilated enclosures, from link: http://www.electronics cooling.com/1997/09/apracticalformulaforaircooled boards inventilatedenclosures/

ANSYS Fluent User Guide, Release 15.0, Nov. 2013 and ANSYS Fluent 12.0, Theory guide, April 2009

H K Versteeg and W Malalasekara, Computational Fluid Dynamics Finite Volume Method, Text book of CFD, edition 1995

Younus. A. Cengel, Text book of Heat transfer chapter 15 cooling of electronic equipments