Applied Thermal Engineering 106 (2016) 131–140

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Research Paper

Numerical study on the thermal performance of a composite board in

battery thermal management system
Jiajia Yan a, Qingsong Wang a,b,c,⇑, Ke Li a, Jinhua Sun a,b
a
State Key Laboratory of Fire Science, University of Science and Technology of China, Hefei 230026, China
b
Collaborative Innovation Center for Urban Public Safety, Anhui Province, Hefei 230026, China
c
CAS Key Laboratory of Materials for Energy Conversion, University of Science and Technology of China, Hefei 230026, China

h i g h l i g h t s

One composite board with sandwich structure is designed for battery thermal management.

Four working modes are compared under normal operating condition and thermal abuse condition.

The composite board shows good heat dissipation capability and good heat-insulation capability.

Increasing the latent heat of PCM can greatly improve the thermal performance of the composite board.

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

Article history: One kind of composite board based battery thermal management system (BTMS) is proposed and a
Received 15 April 2016 three-dimension battery thermal model is proposed in this work. The composite board consists of three
Revised 29 May 2016 parts with a sandwich structure, which contains a heat conducting shell, an insulation panel and phase
Accepted 30 May 2016
change material (PCM). Then four different modes are compared in detail to verify the thermal perfor-
Available online 30 May 2016
mance of the composite board under normal operating condition and thermal abuse condition. The
results show that the composite board can effectively improve the heat dissipation capability and the
Keywords:
uniformity of the temperature, meanwhile it can enhance the heat-insulation capability of the battery
Lithium ion battery safety
Battery thermal management system
pack to prevent the thermal runaway propagation. In addition, increasing the latent heat of PCM can
Composite board greatly improve the thermal performance of the composite board, thus the PCM with a latent heat of
Heat conducting shell 1125 kJ/kg and the phase change temperature between 303.15 K and 323.15 K are recommend to be used
Insulation panel in the battery thermal management.
PCM Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction temperature rise of the battery. When the temperature of single

Due to the dual pressure of resources and environment, many
countries are developing electric vehicles (EVs) to replace tradi-
tional fuel vehicles, which have less environment pollution, lower
noise and higher efficiency. Lithium ion battery is the dominate
power source of the EVs by now. However, lithium ion battery that
as the main driving energy of the electric vehicles still has some
safety problems. There are many fire accidents of electric vehicles
around the world. This is because that the battery will produce a
lot of heat in the process of charging and discharging, if combined
with the space constraint and time accumulation, leading to the
⇑ Corresponding author at: State Key Laboratory of Fire Science, University of
Science and Technology of China, Hefei 230026, China.
E-mail address: pinew@ustc.edu.cn (Q. Wang).
battery exceeds the normal working temperature, it is tend to ther-
mal runaway, even it will spread to the whole battery pack and
cause the fire of the vehicle. Several studies have shown that the
optimum operating temperature range of lithium ion battery is
from 293.15 K to 313.15 K [1], and Zhao et al. [2] mentioned that
the lifetime of battery will be reduced by two months for each
degree rise in temperature. Therefore, it is of great importance to
design effective battery thermal management system (BTMS) in
order to ensure the battery in the normal operating temperature
range. In addition, enhancing the heat-insulating capability of the
battery pack is also necessary, for it will effectively prevent the
thermal runaway propagation between the batteries.
On the one hand, the battery thermal management system is
essential to control the temperature of batteries in a normal oper-
ating temperature range and to decrease the temperature gradient

http://dx.doi.org/10.1016/j.applthermaleng.2016.05.187
1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
132 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140

in battery pack. The BTMS can be classified as air system, liquid system has become a new trend, such as the combination of PCM
system, and phase change material (PCM) system, and so on, and liquid cooling, PCM and forced air convection, and PCM and
according to the heat transfer medium. First of all, the air cooling heat pipe.
method is the most simple and most commonly used method in On the other hand, the heat-insulating capability of the battery
reality. And the study of air based thermal management system pack is also of great importance to prevent the thermal runaway
is focused on the battery arrangement (i.e. aligned and staggered propagation. Berdichevsky et al. [12] presented that increasing
cell arrangements) [3], airflow rate, and airflow modes (i.e. two- heat conductive shell on the outside of battery to enhance the heat
directional air flow, reciprocating air flow) [4,5], and so on. How- transfer between the battery and cooling medium, and placing
ever, the heat transfer coefficient of the air is low, and it is very dif- insulation plate and metal plate between different cell layer, which
ficult to satisfy the heat dissipation required at high discharge rate.
Secondly, liquid based thermal management system is more effec- Table 1
tive than air system, for its higher heat transfer coefficient. Tesla Geometric size and thermo-physical parameters used in simulation.
motors [6] adopts the liquid cooling method, and the winding cool- Nomenclature Parameters Value
ing tube is arranged between the battery, so the heat can be taken
Battery length L (mm) 70
away through the liquid flow. But the liquid based system has a Battery width W (mm) 28
biggest drawback that is easy to leak causing the battery short cir- Battery height H (mm) 134
cuit, and it needs additional power consumption. Thirdly, passive Capacity of battery C (Ah) 20
thermal management using PCM has attracted a lot of attention Internal resistance of battery R (mX) 6
Entropy coefficient dE/dT (mV/K) 0.5 [19]
in recent years. Al Hallaj and Selman [7] first put forward that
Thickness of composite board d1 (mm) 10
PCM can be applied to the battery thermal management, which Thickness of heat conducting shell d2 (mm) 1
can use its high latent heat to effectively reduce the maximum Thickness of Insulation panel d3 (mm) 2
temperature and greatly improve the uniformity of temperature Thickness of each PCM layer dPCM (mm) 3
Heat capacity of battery Cb (J/(kg K)) 1605 [20]
of the battery pack. The most commonly used phase change mate-
Specific heat of battery box Cp1 (J/(kg K)) 900
rials is paraffin, but its thermal conductivity coefficient is very low Heat capacity of heat conducting shell Cp2 (J/(kg K)) 385
(about 0.2 W/m K). Then a lot of work have been done to increase Heat capacity of insulation panel Cp3 (J/(kg K)) 800
the thermal conductivity coefficient of PCM through adding high Heat capacity of PCM (solid phase) CPCMS (J/(kg K)) 2150 [21]
heat conductivity materials such as expanded graphite, carbon Heat capacity of PCM (liquid phase) CPCML (J/(kg K)) 2180 [21]
Latent heat of phase change LPCM (J/kg) 225,000
fiber, graphene, aluminum foam, and copper foam [8–11], and so
[21]
on. Besides, the development of composite thermal management Thermal conductivity of battery kb (W/(m K)) 32 [20]
Thermal conductivity of battery box k1 (W/(m K)) 238
Thermal conductivity of heat conducting shell k2 (W/(m K)) 400
Thermal conductivity of insulation panel k3 (W/(m K)) 0.1
Thermal conductivity of PCM (solid phase) kPCMS (W/ 0.358 [21]
(m K))
Thermal conductivity of PCM (liquid phase) kPCML (W/ 0.152 [21]
(m K))
Density of battery qb (kg/m3) 2285 [20]
Density of battery box q1 (kg/m3) 2700
Density of heat conducting shell q2 (kg/m3) 8700
Density of insulation panel q3 (kg/m3) 880
Density of PCM (solid phase) qPCMS (kg/m3) 814 [21]
Density of PCM (liquid phase) qPCML (kg/m3) 724 [21]
Ambient temperature T0 (K) 293.15
Phase change temperature of PCM TPCM (K) 303.15
Heat transfer coefficient of natural air h0 (W/(m2 K)) 10
convection

Fig. 1. The schematic of the Li-ion battery pack.

(a) The stereogram view (b) The vertical view

Fig. 2. The stereogram and vertical view of a composite board.

 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140
can reduce the heat conduction and thermal radiation of different
cell layers in order to prevent the spread of thermal runaway.
Metha et al. [13] proposed that covering a layer of intumescent
material on the surface of the battery, which will absorb heat
and expand to prevent the thermal runaway propagation. These
barrier technologies only considered the enhancement of
heat-insulating capability of the battery pack, but ignored the heat
dissipation capability.
Above all, a good heat dissipation capability and a good heat-
insulating capability are beneficial for battery safety from different
views. However, there is a certain contradiction between the two
capabilities, for a system with a good heat dissipation capability
whose heat-insulation capability is worse, while a system with a
good heat-insulation capability whose heat dissipation capability
is worse in general. Therefore, it is meaningful to design a new bat-
tery thermal management system to balance the two capabilities,
in order to make the heat dissipation and heat insulation of the
system meet the actual requirements.
In this work, a new composite board based BTMS is proposed
and a three-dimension battery thermal model is proposed. The
composite board consists of three parts with a sandwich structure,
which contains a heat conducting shell, an insulation panel and
PCM. This composite board can effectively improve the heat
dissipation capability and uniformity of the temperature, mean-
while enhance the heat-insulation capability of the battery pack.
In this study, four different modes are compared in detail to verify
the thermal performance of the composite board under normal
operating condition and thermal abuse condition.
(a) Mode 1
(c) Mode 3
Fig. 3. Temperature distribution of four modes after the completion of 3 C discharge.
133
2. Modeling
2.1. Geometric construction
A commercial 20 Ah LiFePO4 battery is chosen as the simulation
unit. The schematic of the lithium ion battery pack is shown in
Fig. 1. The system contains three prismatic batteries, and two com-
posite boards are placed in the space between the adjacent batter-
ies, and the outermost is a battery box. In order to verify the
performance of the composite board, the composite board is
replaced by other structure, then four different modes are designed
to test. For convenience, ‘‘Mode 1” stands for that there is no space
for adjacent batteries, and ‘‘Mode 2” represents that there is air gap
between two batteries, and ‘‘Mode 3” stands for that a simple heat
sink board is placed between two batteries, and ‘‘Mode 4” repre-
sents that a composite board is placed between two batteries. In
addition, the number of the batteries are marked as ‘‘1#”, ‘‘2#”
and ‘‘3#” from left to right, respectively.
Fig. 2 illustrates the internal structure of a composite board,
which is similar to ‘‘sandwich structure”. The composite board con-
sists of three parts, which contains a heat conducting shell, an
insulation panel and PCM. Firstly, the outermost layer of the com-
posite board is a heat conducting shell, which is made of high ther-
mal conductivity materials like copper in order to quickly transfer
the heat of batteries. Secondly, an insulation panel is placed in the
middle of the composite board, and it is made of low thermal con-
ductivity materials like asbestos, which is used to slow down the
heat transmission between two batteries and enhance the
(b) Mode 2
(d) Mode 4
134 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140

heat-insulating capability of the system. Thirdly, the space
between heat conducting shell and insulation panel is filled with
PCM, whose high latent heat can absorb lots of heat to improve
the heat dissipation capability and the uniformity of temperature
of the system.
The geometric size and thermo-physical parameters used in this
simulation are summarized in Table 1. These data mainly include
the size, specific heat, thermal conductivity and density of different
parts. The software COMSOL Multiphysics 5.2 was utilized to sim-
ulate the thermal performance of the system.
2.2. Conservation equations
equal to LPCM through integrating the second term during the
whole the phase change stage. Therefore, this method transforms
the effect of latent heat into the equivalent distribution to heat
capacity. On the whole, the specific heat capacity of mushy PCM
is the sum of an equivalent heat capacity and the distribution of
latent heat.
C p ¼ C PCMS ðT < T 1 Þ ð4Þ
1 @ am
Cp ¼ ðhqPCMS C PCMS þ ð1  hÞqPCML C PCML Þ þ LPCM
q @T
ðT 1 6 T 6 T 2 Þ ð5Þ

For simplicity, several assumptions were made in this simula- C p ¼ C PCML ðT > T 2 Þ ð6Þ
tion as follows: where CPCMS, CPCML, LPCM, T, T1, T2 and am stand for heat capacity of
solid PCM, heat capacity of liquid PCM, latent heat of PCM, the tem-
(1) The heat is uniformly generated in the battery. perature of PCM, the onset temperature of phase transition, the end
(2) The thermal conductivity and heat capacity of all the parts temperature of phase transition and the mass fraction, respectively.
are constant value except for the PCM. The meaning of the other parameters is the same as the above.
(3) The electrochemical process during the charging and dis- Here, it should be noted that am represents the mass fraction,
charging is neglected. which is a defined function to simulate the influence of latent heat
on the heat capacity of PCM, the specific formula is described in Eq.
Based on assumptions above, the heat generation per unit vol- (7). Then am is equal to 1/2 before T1 and to 1/2 after T2. Addition-
ume in the battery can be calculated by the following formula [14]: ally, in the ideal case, when (1h) is the Heaviside function (equal

  
I dEoc I dEoc to 0 before T1 and to 1 after T2), dam/dT is the Dirac pulse. There-
q1 ¼ ðEoc  EÞ  T ¼ IR  T ð1Þ fore, the integral of the term (LPCM dam/dT) during the whole phase
Vb dT Vb dT
where q1, I, Vb, Eoc, E and T are volumetric heat generation rate, cell 332
current, battery volume, open circuit potential, cell potential and
328 T1_max
cell temperature, respectively. T1_min
Besides, the energy balance equation in the cell and PCM all can 324
T2_max
be expressed as follows [15]: 320 T2_min
@T
Temperature/K

qC p ¼ krðrTÞ þ q ð2Þ 316

@t
312
where q, Cp, k and q are density, heat capacity, thermal conductivity
308
and volumetric heat generation term, respectively. When in the dif-
ferent region, these thermophysical properties depend on the rele- 304
vant region. 300
It is worth noting that the heat capacity (Cp) of PCM will change
296
between different phases. Kaplan et al. [16] pointed out that the
software COMSOL uses the apparent heat capacity method to 292
model the phase change behavior of PCM. It is acknowledged that 0 200 400 600 800 1000 1200 1400

the phase change of PCM can be divided into three stages: no melt- Time/s
ing, partly melting and totally melting. The first and third stage are (a) Mode 2
relatively simple in modeling, since there is only the solid PCM
before melting and only the liquid PCM after complete melting. 320
Only in the second stage, the PCM becomes a mushy zone, which T1_max
contains both solid PCM and liquid PCM. 316 T1_min
Firstly, the density of PCM can be calculated from the following T2_max
equation: 312 T2_min

q ¼ hqPCMS þ ð1  hÞqPCML
Temperature/K

ð3Þ 308

where q, qPCMS and qPCML stand for the density of PCM, the density
of solid PCM and the density of liquid PCM, respectively. In addition, 304

h represents the fraction of phase before transition, in other word, it

300
is the volume fraction of solid PCM, which is equal to 1 before T1 (no
melting) and to 0 after T2 (totally melting).
296
Then the heat capacity of different phase of PCM can be
described in Eqs. (4)–(6). The primary difference between the sec-
292
ond stage and the others is that the heat capacity of the mushy
0 200 400 600 800 1000 1200 1400
PCM is changed with temperature, while the other two heat capac- Time/s
ity keep at the two constants. In the second stage, the first term on
(b) Mode 4
the right Eq. (5) stands for an equivalent heat capacity of the mix-
ture of solid PCM and liquid PCM. In addition, the second term on Fig. 4. Tmax and Tmin of the batteries in Mode 2 and Mode 4 versus time at 3 C
the right Eq. (5) means the distribution of latent heat, which is discharge.
 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140 135

transformation is equal to the latent heat LPCM as shown in Eq. (8), @2T
kn ¼ hðT  T 0 Þ ð12Þ
which is mainly used to transform the effect of latent heat into the @n2
equivalent distribution to heat capacity.
where r is Stefan-Boltzmann constant, n represents three axes (x, y,
1 ð1  hÞqPCML  hqPCMS z), and T0 is the temperature of ambient.
am ¼ ð7Þ
2 q
Z Z 3. Results and discussion
T2
@ am T2
@ am
LPCM dT ¼ LPCM dT ¼ LPCM ð8Þ
T1 @T T1 @T A three-dimensional battery thermal model of the battery pack
The heat exchange between battery and the environment is was studied in this work. With the purpose of comparing the heat dis-
mainly achieved by convection and radiation, so the heat transfer sipation capability and heat-insulating capability of different modes,
coefficient h is mainly composed of two parts, one is convection the batteries were operated under normal work condition and thermal
heat transfer coefficient h1, and the other is radiation transfer coef- abuse condition, respectively. The cooling effects and heat-insulating
ficient h2. The value of convection heat transfer coefficient h1 was capability of different modes were discussed in the following section.
assumed to be 10 W/(m2 K) between the battery and the ambient.
The radiation transfer coefficient h2 can be calculated from the fol- 3.1. Temperature distribution of normal working condition
lowing formula:
In order to study the heat dissipation capability of composite board
h ¼ h1 þ h2 ð9Þ
under normal working condition, three batteries discharged at the
same rate, and the discharging rate of 1 C, 3 C, 5 C and 10 C were con-
rðT 4  T 40 Þ
h2 ¼ ð10Þ sidered. The thermal effects and temperature distribution of battery
T  T0 pack with four kinds of modes were compared in this section.
Therefore, the initial condition and boundary conditions can be Fig. 3 shows the temperature distribution of four modes after
written as follows: the completion of 3 C discharge. The thermal effects of Mode 1
and Mode 2 are basically the same, and their maximum tempera-
Tðx; y; z; 0Þ ¼ T 0 ð11Þ
ture are around 327 K at 1200 s. So it indicates that changing the

400 400

T_max (1 C) T_max (1 C)
380 T_max (3 C) 380 T_max (3 C)
T_max (5 C) T_max (5 C)
T_max (10 C) T_max (10 C)
360 360
Temperature/K

Temperature/K

340 340

320 320

300 300

280 280
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time/s Time/s
(a) Tmax of Mode 2 (c) Tmax of Mode 4
35 35

ΔT (1 C) ΔT (1 C)
30 30
ΔT (3 C) ΔT (3 C)
ΔT (5 C) ΔT (5 C)
25 ΔT (10 C) 25 ΔT (10 C)

20 20
ΔT/K

ΔT/K

15 15

10 10

5 5

0 0
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time/s Time/s
(b) ΔT of Mode 2 (d) ΔT of Mode 4
Fig. 5. Tmax and DT of the batteries in Mode 2 and Mode 4 versus time at different discharge rate.
136 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140

distance of the batteries can’t improve heat dissipation capability
effectively. This is because that the air convection heat transfer
coefficient is small in enclosed space, then the quantity of heat car-
ried away by convection is small. The maximum temperature of
Mode 3 and Mode 4 are significantly lower than that of Mode 1
and Mode 2, and the Tmax decreases by 7 K and 9 K, respectively.
All the Tmax appear on the upper part of the battery pack, and the
temperature of the bottom part are lower. This is due to the heat
conduction between the bottom part of batteries and battery
box, which takes away the heat of batteries. In addition, the tem-
perature of the heat sink board of Mode 3 and the composite board
of Mode 4 are higher than that of battery, so heat sink board and
composite board can effectively remove the battery’s heat, further-
more, composite board shows better performance.
On the account of the symmetry of the whole structure, the fol-
lowing study is mainly focused on the 1# and 2# batteries. Fig. 4
displays the Tmax and Tmin of the batteries in Mode 2 and Mode 4
versus time at 3 C discharge. In Fig. 4a, the maximum temperature
and minimum temperature of 1# battery in Mode 2 rises almost
linearly with time, and 2# battery’s temperature is slightly higher
than 1# battery. Because both sides of 2# battery in Mode 2 is air,
which is heated by adjacent battery, the thermal exchange
between 2# battery and air is smaller. The maximum temperature
difference DT increases with time, and DT is 9.6 K when time is
1200 s. The circumstances of Mode 1 and Mode 3 are similar to
Mode 2. However, the regulation of Mode 4 is much different from
Mode 2. As is shown in Fig. 4b, the temperature rise curve of Mode
4 can be divided into three stages. The temperature of the battery
rises rapidly when it is below or above the phase change temper-
ature of PCM, while the temperature rises slowly when it falls in
the phase change temperature range. In the second stage
(300 s < t < 700 s), Tmax of Mode 4 increases slowly with time, for
the temperature rises to the melting point of PCM, and the phase
change of PCM absorbs lots of heat. It was found that the maxi-
mum temperature of 2# battery in Mode 4 is slightly lower than
1#, which is due to the good heat conduction between 2# battery
and two composite board. Moreover, the DT of Mode 4 decreases to
5.8 K at 1200 s. Therefore, the composite board in Mode 4 can
improve the heat dissipation capability of battery pack greatly,
meanwhile, it can improve the uniformity of battery temperature.
Fig. 5 reveals the Tmax and DT of the batteries in Mode 2 and
Mode 4 versus time at different discharge rate. In Mode 2, with
the increase of the discharge rate, the Joule heat increases obvi-
ously, so the Tmax and DT of the batteries increase rapidly. When
the discharge rate changes from 1 C to 10 C, the Tmax and DT of
the batteries in Mode 2 increase from 307.2 K, 3.7 K to 391.4 K
and 33.3 K, respectively. Analogously, the variation trends of the
Tmax and DT of the batteries in Mode 4 are similar to Mode 2, only
the maximum value is different. As Ling et al. [8] summarized, the
maximum temperature of Li-ion batteries should not exceed

324
324 T_max(Tm=303K)
T_max ( LPCM = 112.5 kJ / kg ) 320 T_max(Tm=313K)
320 T_max ( LPCM = 225 kJ / kg )
T_max(Tm=323K)
T_max ( LPCM = 450 kJ / kg ) 316
316 T_max ( LPCM = 1125 kJ / kg )
T_max ( LPCM = 2250 kJ / kg )
Temperature/K

312
Temperature/K

312
308
308

304
304

300
300

296
296

292
292
0 200 400 600 800 1000 1200 1400
0 200 400 600 800 1000 1200 1400
Time/s
Time/s
(a) Tmax of Mode 4
(a) Tmax of Mode 4
7

ΔT ( CPCM = 112.5 kJ / kg ) ΔT ( Tm = 303 K )

6 ΔT ( CPCM = 225 kJ / kg ) 6 ΔT ( Tm = 313 K )
ΔT ( CPCM = 450 kJ / kg ) ΔT ( Tm = 323 K )
ΔT ( CPCM = 1125 kJ / kg ) 5
ΔT ( CPCM = 2250 kJ / kg )

4 4
ΔT/K
ΔT/K

2 2

0 0
0 200 400 600 800 1000 1200 1400 0 200 400 600 800 1000 1200 1400
Time/s Time/s
(b) ΔT of Mode 4 (b) ΔT of Mode 4

Fig. 6. Tmax and DT of the batteries in Mode 4 versus time under different latent Fig. 7. Tmax and DT of the batteries in Mode 4 versus time with different phase
heat of PCM. change temperature of PCM.
 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140 137

333.15 K, and Pesaran [17] presented that the maximum tempera- 400
ture difference in the battery pack should not exceed 5 K. 390
T_max (Mode 1)
T_max (Mode 2)
Therefore, only the 1 C discharge in Mode 2 and Mode 4 meet T_max (Mode 3)
the conditions, and the 3 C discharge in Mode 4 almost satisfies 380 T_max (Mode 4, LPCM = 225 kJ / kg )
the conditions. Considering that the normal discharge rate of elec- 370 T_max (Mode 4, LPCM = 1125 kJ / kg )

tric vehicle is ranged from 1 C to 3 C, the following part presents 360

Temperature/K
some improvement measurements on the basis of Mode 4 to
350
enhance its performance.
340
3.2. Effect of latent heat and phase change temperature of PCM in 330
composite board
320

As Fig. 6 shows, the Tmax and DT of the batteries in Mode 4 versus 310
time under different latent heat of PCM. The latent heat were cho- 300
sen to be 112.5 kJ/kg, 225 kJ/kg, 450 kJ/kg, 1125 kJ/kg and 2250 kJ/
290
kg, respectively. The phase change temperature of PCM is 303.15 K. 0 1 2 3 4 5 6 7 8 9 10
It can be seen that Tmax decreases with the increase of latent heat. If Discharge rate/C
the latent heat of PCM increases from 112.5 kJ/kg to 1125 kJ/kg, the
(a) Tmax of different modes
Tmax decreases from 320.4 K to 306.1 K. However, when the latent
heat of PCM increases from 1125 kJ/kg to 2250 kJ/kg, the Tmax only
decreases from 306.1 K to 304.5 K. Thus the latent of 1125 kJ/kg is ΔT (Mode 1)
enough to absorb the battery’s heat in such condition, the Tmax 30
ΔT (Mode 2)
ΔT (Mode 3)
decreases little with higher latent heat. Similar trend can be found ΔT (Mode 4, LPCM = 225 kJ / kg )
for the curve of DT. When the latent heat of PCM increases from ΔT (Mode 4, LPCM = 1125 kJ / kg )
112.5 kJ/kg to 1125 kJ/kg, the DT drops from 6.2 K to 2.9 K, which
meets the condition of the maximum temperature difference in
20
the battery pack should not exceed 5 K. Therefore, 1125 kJ/kg is rec-
ΔT/K

ommended as the best latent heat for PCM in this case, which can
effectively improve the heat dissipation capability of the battery
pack and the uniformity of the temperature.
Fig. 7 reveals the Tmax and DT of the batteries in Mode 4 versus
time with different phase change temperature of PCM. The phase
change temperature of PCM were chosen to be 303.15 K,
313.15 K and 323.15 K, respectively. The latent heat of PCM is
225 kJ/kg. When the phase change temperature increases from
303.15 K to 313.15 K, the Tmax is almost unchanged, only the time
of phase change is different, which is due to the complete melting
of PCM and the full use of the latent heat. In contrast, when the
phase change temperature increases from 313.15 K to 323.15 K,
the Tmax increases from 318.6 K to 322.1 K. This is because the
maximum temperature of battery just reaches the highest phase
change temperature (323.15 K), resulting in the useless of the
latent heat. The curve of DT with different phase change tempera-
ture is similar to the former. During the stage of phase change, Tmax
and DT all increase slowly with time. As the phase change temper-
ature changes from 303.15 K to 313.15 K, DT nearly keeps at 5.8 K
at 1200 s. Instead, when the phase change temperature increases
from 313.15 K to 323.15 K, DT has a little rise from 5.8 K to 6.2 K.
So in order to make full use of the latent heat of PCM, the phase
change temperature should not be too high. Ramadass et al. [18]
proposed that the capacity fade of 18,650 lithium ion battery could
be 70% after 490 cycles when the temperature of battery was over
323.15 K. Therefore, the phase change temperature of PCM should
not exceed 323.15 K, and the phase change temperature range
between 303.15 K and 323.15 K is recommended.
3.3. The comparison of thermal performance between different modes
Fig. 8 reveals the Tmax and DT of the batteries in different modes
versus discharge rate. As shown in Fig. 8a, the Tmax of the batteries
in Mode 1 and Mode 2 are basically the same, which demonstrates
that increasing the distance between batteries in enclosed space is
useless to enhance the capacity of heat dissipation. Otherwise, the
thermal performance of Mode 3 and Mode 4 is much better than
Mode 1 and Mode 2. It is noted that when the discharge rate
changes from 1 C to 5 C, the Tmax of Mode 4 (LPCM = 225 kJ/kg) is
10
0
0 1 2 3 4 5 6 7 8 9 10
Discharge rate/C
(b) ΔT of different modes
Fig. 8. Tmax and DT of the batteries in different modes versus discharge rate.
slightly lower than that of Mode 3, but when the discharge rate
increases to 10 C, the Tmax of Mode 4 (LPCM = 225 kJ/kg) is slightly
higher than that of Mode 3 instead. When the discharge rate is
lower, the high latent heat of PCM in Mode 4 can absorb lots of
heat, leading to the lower Tmax. However, when the discharge rate
rise to 10 C, the batteries generate more heat, resulting in that the
latent heat of PCM in Mode 4 runs out, additionally the heat con-
ductivity coefficient of composite board in Mode 4 is smaller than
the heat sink board in Mode 3, so Tmax of Mode 4 (LPCM = 225 kJ/kg)
become higher. Moreover, increasing the latent heat of PCM can
obviously improve the thermal performance of Mode 4, as the
latent heat of PCM increases from 225 kJ/kg to 1125 kJ/kg, the Tmax
drops from 334.9 K to 316.5 K that meets the safe operating tem-
perature range. Therefore, the heat dissipation capability of differ-
ent modes can be sorted as follows on the base of the maximum
temperature: Mode 4 (LPCM = 1125 kJ/kg) > Mode 4 (LPCM = 225 kJ/
kg)  Mode 3 > Mode 2  Mode 1.
As shown in Fig. 8b, the DT of the batteries in Mode 1 and Mode
2 are basically the same like the former one, and the DT of the bat-
teries in Mode 3 and Mode 4 are lower. By contrast, the DT of Mode
4 (LPCM = 225 kJ/kg) is higher than Mode 3. The smaller heat con-
ductivity coefficient of composite board in Mode 4 is a dominate
factor, for the insulation panel in the composite board suppresses
the direct heat conduction between two batteries. It can be found
that when the discharge rate changes from 1 C to 5 C, the DT of
138 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140

Mode 4 (LPCM = 1125 kJ/kg) is slightly lower than Mode 3, but when
the discharge rate increases to 10 C, the DT of Mode 4
(LPCM = 1125 kJ/kg) is slightly higher than Mode 3 instead. This is
because the transformation of the main influence factor from the
high latent heat to low heat conductivity coefficient. Therefore,
the capability of keeping the uniformity of temperature of different
modes can be sorted as follows according to DT: Mode 4
(LPCM = 1125 kJ/kg) > Mode 3 > Mode 4 (LPCM = 225 kJ/kg) > Mode
2  Mode 1.
Above all, Mode 4 has the best thermal performance that effec-
tively improves the heat dissipation capability and uniformity of
the temperature, especially when the PCM in Mode 4 has higher
latent heat. In addition, Mode 3 is better, Mode 1 and Mode 2 are
the worst.

(a) Mode 1 (b) Mode 2

(c) Mode 3 (d) Mode 4(LPCM=225 kJ/kg)

(e) Mode 4(LPCM=1125 kJ/kg) (f) Mode 4(LPCM=2250 kJ/kg)

Fig. 9. The cross section temperature of battery pack in different modes.
 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140
3.4. Results of thermal abuse condition
In order to study the thermal runaway blocking ability of com-
posite board under thermal abuse condition, one battery was set as
a thermal runaway battery, the others were set to non-operation.
The simulation condition of thermal runaway is that the heat gen-
erate rate of battery is equivalent to the heat of 10 C discharge and
the time of duration is set as 2000 s. Then the thermal runaway
blocking ability of battery pack with different kinds of modes were
compared in this section.
Fig. 9 displays the cross section temperature of battery pack in
different modes. The height of cross section is equal to half of the
height of the battery. As shown in Fig. 9a–d, the distribution of
temperature field in Mode 1 and Mode 3 are relatively uniform,
indicating that the thermal conduction between batteries is good
in both cases, which is advantageous to thermal runaway propaga-
tion. In addition, Mode 2 has the best heat-insulating capability,
but the temperature of the thermal runaway battery is higher than
the others, causing higher risk of battery fire. By comparison, Mode
4 has the better heat-insulating capability, and it can be seen that
the temperature of both sides of the border between two batteries
have an obvious difference. In Fig. 9d–f, the cross section temper-
ature of battery pack in Mode 4 with different latent heat are sim-
ilar, only the bulk temperature decreases with the increasing latent
heat. So properly increasing the latent heat of PCM is useful to
enhance the heat-insulating capability and decrease the tempera-
ture of the system.
With the purpose of quantitatively analyzing the heat-insulating
capability of different modes, the interval time Dt between two bat-
teries occurring thermal runaway is set as a metric. Khateeb et al.
[15] presented that the thermal runaway onset temperature range
of the lithium ion battery is 343.15 K to 373.15 K, and the simula-
tion adopts 373.15 K as the onset temperature of thermal runaway.
All the interval time of different models are listed in Table 2. There-
fore, the heat-insulating capability of different modes can be sorted
as follows according to the interval time Dt:
Mode 2 > Mode 4 > Mode 3 > Mode 1:
It is clear that the heat-insulating capability of Mode 2 is the
best one, for the air convection heat transfer coefficient is very
small, so the transmission of thermal runaway to the next battery
needs a longer time (838 s). However, the temperature of the first
thermal runaway battery in Mode 2 is significantly higher than
that of the other modes, leading to the higher risk of battery fire.
While the interval time Dt of Mode 1 and Mode 2 is shorter, which
is due to their good heat conduction. Above all, Mode 4 has a better
enhancement in the heat-insulating capability of battery pack. In
addition, when the latent heat of PCM in Mode 4 increases from
225 kJ/kg to 2250 kJ/kg, the interval time Dt increases from 451 s
to 674 s, thus the presence of higher latent heat of PCM extends
the interval time Dt by 49%. Extending the interval time is very
useful to detect the thermal runaway of battery, thus there is
enough time to take safety measures to prevent thermal runaway
propagation in battery module.
Table 2
The interval time Dt between two batteries occurring thermal runaway in different
modes.
Different modes Interval time Dt (s)
Mode 1 106
Mode 2 838
Mode 3 168
Mode 4 (LPCM = 225 kJ/kg) 451
Mode 4 (LPCM = 1125 kJ/kg) 550
Mode 4 (LPCM = 2250 kJ/kg) 674
139
4. Conclusions
In this work, the thermal performance of a composite board
based battery thermal management system is evaluated and a
three-dimension battery thermal model is proposed. The compos-
ite board consists of three parts with a sandwich structure, which
contains a heat conducting shell, an insulation panel and PCM.
Then four different modes are compared in detail to verify the ther-
mal performance of the composite board under normal operating
condition and thermal abuse condition. The numerical results are
included:
(a) The Tmax and DT of batteries in Mode 1 (no space) and Mode
2 (air gap) is basically the same, indicating that increasing
the distance between batteries in enclosed space is useless
to enhance the capacity of heat dissipation.
(b) Mode 3 (heat sink board) has a better heat dissipation capa-
bility, but its heat-insulation capability is the second worst,
the interval time Dt between two batteries occurring ther-
mal runaway only increases from 106 s to 168 s compared
to Mode 1.
(c) The Mode 4 (composite board) generally improves the heat
dissipation capability and uniformity of the temperature,
meanwhile it can enhance the heat-insulation capability of
the battery pack to prevent the thermal runaway propagation.
(d) Increasing the latent heat of PCM can greatly improve the
thermal performance of the composite board, and the PCM
with a latent heat of 1125 kJ/kg and the phase change tem-
perature between 303.15 K and 323.15 K is recommend to
be used in the battery thermal management.
To summarize, the composite board has a good heat dissipation
capability and a good heat-insulation capability, which provides a
new idea for the design of the battery thermal management sys-
tem. There are still a lot of work need to be investigated further
to improve the thermal performance of the whole system.
Acknowledgements
This work is supported by the External Cooperation Program of
BIC, Chinese Academy of Sciences – China (No.
211134KYSB20150004) and the Fundamental Research Funds for
the Central Universities – China (No. WK2320000034). Dr. Q.S
Wang is supported by Youth Innovation Promotion Association
CAS – China (No. 2013286).
References
[1] Y. Ye, L.H. Saw, Y. Shi, A.A.O. Tay, Numerical analyses on optimizing a heat pipe
thermal management system for lithium-ion batteries during fast charging,
Appl. Therm. Eng. 86 (2015) 281–291.
[2] R. Zhao, J. Gu, J. Liu, An experimental study of heat pipe thermal management
system with wet cooling method for lithium ion batteries, J. Power Sources 273
(2015) 1089–1097.
[3] N. Yang, X. Zhang, G. Li, D. Hua, Assessment of the forced air-cooling
performance for cylindrical lithium-ion battery packs: a comparative
analysis between aligned and staggered cell arrangements, Appl. Therm. Eng.
80 (2015) 55–65.
[4] K. Yu, X. Yang, Y. Cheng, C. Li, Thermal analysis and two-directional air flow
thermal management for lithium-ion battery pack, J. Power Sources 270
(2014) 193–200.
[5] R. Mahamud, C. Park, Reciprocating air flow for Li-ion battery thermal
management to improve temperature uniformity, J. Power Sources 196
(2011) 5685–5696.
[6] D.T. Adams, G. Berdichevsky, T.E. Colson, Battery Pack Thermal Management
System, U. S. patent No. 11, 779, 583, 2007-7-18.
[7] S. Al Hallaj, J. Selman, A novel thermal management system for electric vehicle
batteries using phase change material, J. Electrochem. Soc. 147 (2000) 3231–3236.
[8] Z. Ling, J. Chen, X. Fang, Experimental and numerical investigation of the
application of phase change materials in a simulative power batteries thermal
management system, Appl. Energy 121 (2014) 104–113.
140 J. Yan et al. / Applied Thermal Engineering 106 (2016) 131–140
[9] Z. Wang, Z. Zhang, L. Jia, L. Yang, Paraffin and paraffin/aluminum foam
composite phase change material heat storage experimental study based on
thermal management of Li-ion battery, Appl. Therm. Eng. 78 (2015) 428–436.
[10] A. Babapoor, M. Azizi, G. Karimi, Thermal management of a Li-ion battery using
carbon fiber-PCM composites, Appl. Therm. Eng. 82 (2015) 281–290.
[11] W.Q. Li, Z.G. Qu, Y.L. He, Y.B. Tao, Experimental study of a passive thermal
management system for high-powered lithium ion batteries using porous
metal foam saturated with phase change materials, J. Power Sources 255
(2014) 9–15.
[12] E.M. Berdichevsky, P.D. Cole, A.J. Hebert, Mitigation of Propagation of Thermal
Runaway in a Multi-Cell Battery Pack, U. S. patent No. 7, 433, 794, 2008-10-7.
[13] V.H. Mehta, W.A. Hermann, N.R. Kalayjian, Cell thermal Runaway Propagation
Resistant Battery Pack, U. S. patent No. 7, 820, 319, 2010-10-26.
[14] M.R. Giuliano, S.G. Advani, A.K. Prasad, Thermal analysis and management of
lithium-titanate batteries, J. Power Sources 196 (2011) 6517–6524.
[15] S.A. Khateeb, S. Amiruddin, M. Farid, J.R. Selman, S. Al-Hallaj, Thermal
management of Li-ion battery with phase change material for electric
scooters: experimental validation, J. Power Sources 142 (2005) 345–353.
[16] F. Kaplan, C. De Vivero, S. Howes, Modeling and analysis of phase change
materials for efficient thermal management, 32nd IEEE Comp. Des. (2014)
249–256.
[17] A.A. Pesaran, Battery thermal models for hybrid vehicle simulations, J. Power
Sources 110 (2002) 377–382.
[18] P. Ramadass, B. Haran, R. White, B.N. Popov, Capacity fade of sony 18650 cells
cycled at elevated temperatures Part I. Cycling performance, J. Power Sources
112 (2002) 606–613.
[19] Q. Sun, Q. Wang, X. Zhao, J. Sun, Z. Lin, Numerical study on lithium titanate
battery thermal response under adiabatic condition, Energy Convers. Manage.
92 (2015) 184–193.
[20] A. Greco, X. Jiang, D. Cao, An investigation of lithium-ion battery thermal
management using paraffin/porous-graphite-matrix composite, J. Power
Sources 278 (2015) 50–68.
[21] N. Javani, I. Dincer, G.F. Naterer, B.S. Yilbas, Heat transfer and thermal
management with PCMs in a Li-ion battery cell for electric vehicles, Int. J. Heat
Mass Trans. 72 (2014) 690–703.