Journal Pre-proof

Effect of finned networks on PCM based battery thermal management system for
cylindrical Li-ion batteries

Hrishav Dey, Sukumar Pati, Pitambar R. Randive, László Baranyi

PII: S2214-157X(24)00603-8
DOI: https://doi.org/10.1016/j.csite.2024.104572
Reference: CSITE 104572

To appear in: Case Studies in Thermal Engineering

Received Date: 30 January 2024

Revised Date: 9 May 2024
Accepted Date: 17 May 2024

Please cite this article as: H. Dey, S. Pati, P.R. Randive, L. Baranyi, Effect of finned networks on PCM
based battery thermal management system for cylindrical Li-ion batteries, Case Studies in Thermal
Engineering, https://doi.org/10.1016/j.csite.2024.104572.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2024 Published by Elsevier Ltd.

 Effect of finned networks on PCM based battery thermal
management system for cylindrical Li-ion batteries
Hrishav Dey1, Sukumar Pati1, Pitambar R. Randive1, László Baranyi*2
1
Department of Mechanical Engineering,

National Institute of Technology Silchar, Silchar-788010, India

2
Department of Fluid and Heat Engineering, Institute of Energy Engineering and
Chemical Machinery, University of Miskolc, 3515 Miskolc-Egyetemváros, Hungary

f
r oo
-p
re
lP
na
ur
Jo

*Corresponding author

Email: [email protected] (S. Pati), [email protected] (L. Baranyi)

1
Abstract

The present numerical work investigates the effect of finned structures on the passive thermal
management of cylindrical Lithium-ion batteries. Novel hexagonal thermal conduction networks
using metal fins are proposed. The electrochemical-thermal coupling for the numerical modeling
is achieved using the Newman-Tiedemann-Gu-Kim (NTGK) semi-empirical method extended
with the enthalpy model to incorporate phase change materials. Firstly, the effect of the different
structures on the thermal behavior and the delivered capacity for a high discharge rate of 4 C is
analyzed. The addition of base PCM (phase change materials) decreases the overall increase in
temperature by about 71.9 %, and up to 75.58% for the optimal thermal conduction network. The
finned structures also show more temperature uniformity in the battery module. Next, the effect
of the ambient temperature on the thermal performance of the proposed designs is shown. Further,
the behavior of the enhanced system is studied under a stressed discharge rate of 6 C and an abuse

f
oo
discharge rate of 8 C. The study is further extended to examine the viability of the proposed battery
thermal management system (BTMS) when the cell arrangement is externally shorted with a small

r
resistance of 0.5 ohm. Based on the numerical analysis the proposed designs perform well.
-p
Keywords: Battery thermal management systems, finned networks, Li-ion batteries, phase
re
change materials
lP

1. Introduction
The energy density of lithium-ion batteries (LIBs) is very high compared to most available
na

batteries, along with high discharge rates, fast charging capacity, longer life, low self-discharge
rates, and high electrochemical potential [1]. However, the performance of the LIB is highly
ur

susceptible to the working temperature. The optimal operating temperature range of LIBs is
288~313K, and prolonged exposure to a higher temperature can endanger the operation by causing
Jo

several adverse effects like degradation of capacity, thermal runaway, fire, etc. The implementation
of proper thermal management strategies becomes of utmost importance in this regard to check any
unwanted temperature surge.
A thorough literature review tells us that a great deal of work has been performed both
numerically and experimentally on the lines of battery thermal management [2-5]. It finds its roots
in the pioneering work of Bernardi et al. [6] where the authors proposed a mathematical model for
the heat generation rates in batteries, taking into account both entropic and ohmic factors. Hong et
al. [7] inferred that for the safer operation of LIBs, 4 K is the limiting maximum temperature
difference between cells. In the literature, several types of battery thermal management system
(BTMS) have been reported illustrating utilities, including air-cooled systems [8], liquid-cooled
systems [9], hybrid cooling [10, 11]. For the successful development of any BTMS, the
mathematical modeling explores the heat dissipation techniques and the coupled electrochemical-
thermal behavior. Kwon et al. [12] proposed a model describing the relationship between the rate
of discharge, aspect ratio of the electrodes, current collecting tab location and uniformity of Depth
of Discharge (DOD). In that model, formulation developed by Tiedemann and Newman [13] was
utilized to express the polarization of the electrodes. Later, Kim et al. [14] extended the same model
incorporating non-isothermal signatures of the battery. Furthermore, multi-scale multi-dimensional

2
(MSMD) model suggested by Kim et al. [15] has been widely adopted in modeling the thermal-
electrochemical behavior of battery packs.
Phase change materials (PCMs) have gained popularity in their usage for BTMSs in the
last decade owing to their ease of use and structural simplicity. However, the low thermal
conductivity of pure PCM offers a hindrance to their full-fledged use. Mills and Al-Hallaj [16]
reported an enhancement in the thermal conductivity with the addition of expanded graphite (EG).
An enhancement in thermal conductivity was also reported with the use of porous foam [17].
Kaabinejadian et al. [18] showed that increasing the size of the pores resulted in a decrease in the
maximum temperature attained by the battery. Ranjbaran et al. [19], after performing an
optimization between the thermal conductivity and available latent heat of enthalpy for different
PCM-based copper foam additives based on their volume percentage, concluded that the 6 percent
volume copper foam performed the best in terms of cooling efficiency. The use of cooling plate
and PCM together [20], encapsulation of the PCM [21], the addition of nanoparticles to the PCM

f
[22], etc., have also been proven to be efficient passive thermal conductivity enhancement

oo
techniques. Huang et al. [23] successfully studied the cooling effectiveness of phase change
materials during overcharging. They found that a critical thickness of up to 3 mm provides the most

r
effective thermal management system.
-p
Researchers have also investigated the use of integrated heat network systems combining
re
additional active cooling techniques like liquid cooling [24], forced air cooling [25], nanofluids
[26, 27], etc. Yang et al. [28] proposed a novel efficient bionic mini-cooling and PCM-based
lP

BTMS. It was shown that the nature-inspired hexagonal shape of the BTMS proved to be better
than rectangular systems. Mondal et al. [29] discuss the enhancement of heat transfer in active
thermal management methods with vortex generators. Lopez et al. [30] found that cell spacing is
na

crucial in disseminating thermal gradients inside a battery module to prevent local premature
capacity fade.
ur

Compared to the methods described above for augmentation of the thermal conductivity of
PCM, fin-enhanced systems offer the exclusiveness of simplicity, structural stability, and low cost
Jo

in addition to bypassing the expense of externally-driven energy requirements in the system [31].
Lv et al. [32] first used a combination of polyethylene, paraffin/EG, and fins and observed a check
in temperature up to 323 K even in high discharge rates. Wang et al. [33] experimentally studied
the effect of the number of fins against different constant heat rates. They showed the efficient use
of the different numbers of vertical fins for a cylindrical mock-up battery. Arshad et al. [34] studied
the effect of pin fin diameter, with the best performance observed with a diameter of 3 mm. Yazici
et al. [35] experimentally examined the effect of the inclination of PCM-based heat sinks for battery
modules. Sun et al. [36] investigated a fin intensified thermal conduction network, shedding light
on the effect of fin length on different ambient temperatures, and showed that such networks
increased the working time by nearly two folds. Ping et al. [37] also studied the interplay of PCM
species and thickness, fin thickness and spacing on the thermal performance of a BTMS.
Although several studies have been conducted on the evaluation of fin intensified
performance of phase change materials, relatively few works are available on the evaluation of fin
intensified passive BTMS. No work is available where detailed electrochemical-thermal modeling
of Li-ion batteries with fin-intensified PCM-based BTMS is evaluated to get an insight into the
battery performance during charging and discharging and the effect of different BTMS parameters.
This is the novelty of this work. The present study utilizes the multi-scale multi-dimensional NTGK

3
(Newman, Tiedermann, Gu and Kim) method for a semi-empirical thermal-electrochemical
coupling of the Li-ion chemistry.
2. Problem Definition
The main aim of the present work is to study the effect of compatibility and viability of two
different novel bionic hexagonal designs of fin-enhanced passive BTMS at fast discharge rates. To
do that, a hexagonal novel design of a thermal management module for a battery pack of 6
cylindrical Li-ion batteries arranged in a 6S1P manner is considered. The cylindrical cells are
chosen to be of the 18,650 configurations, with NMC 811 cathode. The cell specifications are given
in Table S1 in the Supplementary material.
The thermal conduction network is arranged in a hexagonal manner inspired by the
honeycomb structure as proposed in [28]. In the base design, the PCM is contained in a hexagonal
aluminum container of thickness 1 mm and diagonal 88 mm, as shown in Fig. 1(a) and (b), up to a
height of 60 mm just to wrap the active zones of the cells. The next design referred to as design 1

f
oo
(D1) (see Fig. 1(c)), is proposed with the consideration of aluminum cylindrical rings, of diameter
18 mm and thickness of 0.5 mm, around each cell aiming for a more uniform heat transfer
interaction between the heat-generating source and the heat dissipation sink with vertical fins of

r
-p
height 60 mm and width 1 mm inter-connecting the six cells. The second proposed network,
referred to as design 2 (D2) (see Fig. 1 (d)), connects the outer aluminum housing with each of the
re
cylindrical rings with vertical fins of a height of 60 mm and a width of 1 mm. The PCM used in
this study is PEG (Polyethylene glycol) 1000, the properties of which are listed in Table S2 in the
lP

Supplementary material, with the properties for aluminum (used as the container, fins, rings, and
the positive tabs) and copper (used as the negative tabs) [28]. Poly-Ethylene Glycol (PEG) based
PCMs are excellent candidates for BTMS applications owing to their ideal phase transition
na

temperature between 293K to 313K [38]. PEG is an organic phase change material with high latent
enthalpy and a relatively wider temperature transition range. It also possesses qualities like high
ur

thermal stability, non-toxicity, non-corrosive nature, and low flammability. Despite having a higher
viscosity than their paraffin-based counterparts, they excel in performance in terms of heat transfer
Jo

efficiency and energy storage per volume [39].

The onset of natural convection is completely neglected in the current study owing to the
high density of the PCM used [28]. Further, virtual connections are used rather than busbars, thus
neglecting the ohmic heat generation in the connections [28]. Also, the radiation heat transfer is
neglected in all cases, and contact resistance is neglected for all contact surfaces.

(a)

4
 (b) Base Design (c) Design D1

f
r oo
-p
re
(d) Design D2 (e) Outer wall regions in a cell
lP

Fig. 1 Schematic: (a) Li-ion cell, (b)–(d) Proposed design configurations considered in the
present investigation, (e) Outer wall regions of a single cell
na

3. Mathematical modeling and theoretical formulation

In order to obtain an enhanced passive BTMS, the novel design with finned networks in a battery
ur

module of 6 Li-ion cylindrical cells each of 2.6 Ah capacity is proposed and compared with the
base design. The numerical modeling of the battery packs is performed using the NTGK method
Jo

of MSMD module [40]. The active zone is where the reactions occur, and the passive zones are the
current collecting tabs. Aluminum and copper are chosen as the positive and negative tabs,
respectively.
In the NTGK method, the layered structure of the Li-ion cell (the jelly roll, to be specific)
is left unresolved and instead treated as a continuum with orthotropic properties [15]. The
electrochemistry is dictated by the charge conservation equations (in the form of Poisson equations)
in each of the positive and negative zones of the continuum, as proposed by Kwon [12]:
𝜎+ ∇2 ∅+ = −𝐽𝑒𝑐ℎ , (1)
2
𝜎− ∇ ∅− = 𝐽𝑒𝑐ℎ . (2)
Here the conductivities (electrical) are represented by 𝜎− and 𝜎+ for the negative and positive
electrodes, respectively. 𝐽𝑒𝑐ℎ is the current transfer rate (volumetric), and the electrode potential
fields are represented by ∅− and ∅+ .
Newman and Tiedemann [41] observed a linear relation between the potential difference
across the cell and the current transfer rate and formulated a mathematical model, dependent on the
polarization curve, given as
𝐽𝑒𝑐ℎ = 𝛼𝑌[𝑈 − (∅+ − ∅− )]. (3)

5
Here 𝛼 is the specific area of the electrode; Y and U are the curve-fitting parameters. It should be
noted that U and Y are the open-circuit voltage and the inverse slope of the V-I curve respectively,
the functional dependence of which was given by Gu [42] as

1 1
𝑌 = [∑5𝑛=0 𝑎𝑛 (𝐷𝑂𝐷)𝑛 ] exp (−𝐶1 ( − )), (4)
𝑇 𝑇𝑟𝑒𝑓

𝑈 = [∑5𝑛=0 𝑏𝑛 (𝐷𝑂𝐷)𝑛 ] − 𝐶2 (𝑇 − 𝑇𝑟𝑒𝑓 ), (5)

where 𝑎𝑛 , 𝑏𝑛 , and 𝐶𝑛 are the correction coefficients of the empirical fitting as given below: a0=
49.96104; a1= -339.4632; a2= 2198.016; a3= -5272.467; a4= 5523.023; a5= -2091.518; b0=
4.153698; b1= -1.485645; b2= 6.259681; b3= -23.8404; b4= 35.7476; b5= -18.44826; C1= -700; C2=
-0.00095. The DOD is calculated using [14]:

f
𝑉 𝑡

oo
𝐷𝑜𝐷 = ∫ 𝐽 𝑑𝑡, (6)
3600𝑄 0 𝑒𝑐ℎ
where 𝑄 is the theoretical nominal capacity of the cell and V is the volume of the cell.

r
This semi-empirical electrochemical framework is then coupled to the thermal response of
-p
the cell using the energy equation, considering several heat generation factors including the ohmic
heat (given by 𝜎|∇∅|2 ) and the heat generated during the electrochemical reactions taking place at
re
the electrode-electrolyte interfaces, as given by Kim et al. [14]
𝜕𝑇
𝜌𝑐 𝐶𝑐 − 𝑘𝑐 ∇2 𝑇 = 𝜎+ |∇∅+ |2 + 𝜎− |∇∅− |2 + 𝑞̇ 𝑒𝑐ℎ .
lP

(7)
𝜕𝑡
Here 𝐶𝑐 , 𝑘𝑐 , and 𝜌𝑐 are the specific heat capacity, thermal conductivity, and density for the Li-ion
cells. Moreover, 𝑞̇ 𝑒𝑐ℎ is the electrochemical volumetric heat generation rate and it consists of two
na

parts; an irreversible part, taking into account the energy loss due to the difference between the cell
voltage and the open circuit voltage because of the polarization (given by 𝑈 − (∅+ − ∅− )), and a
ur

𝑑𝑈
reversible part, considering the entropic changes in the system (given by ) given as [14]
𝑑𝑇
Jo

𝑑𝑈
𝑞̇ 𝑒𝑐ℎ = 𝐽𝑒𝑐ℎ [𝑈 − (∅+ − ∅− ) − 𝑇 ]. (8)
𝑑𝑇
The enthalpy method [43] is employed here for modeling the PCM, using the same governing
equations for modeling both the liquid and solid phases. We assume a finite melting and
temperature range of 2 K for solidification and by this an artificial mushy zone is developed and
thereby the total material undergoes a transition in its melt fraction 𝛽, which is given as [43]
0 𝑇 < 𝑇𝑠
𝑇−𝑇𝑠
𝛽={ 𝑇𝑠 ≤ 𝑇 ≤ 𝑇𝑙 , (9)
𝑇𝑙 −𝑇𝑠
1 𝑇𝑙 < 𝑇
where 𝑇𝑙 and 𝑇𝑠 are the temperatures of the liquidus and solidus, respectively. The energy equation
for the phase change material (PCM) domain is given as [43]:
𝜕𝐻
𝜌𝑃𝐶𝑀 = 𝑘𝑃𝐶𝑀 ∇2 𝑇. (10)
𝜕𝑡
Here 𝜌𝑃𝐶𝑀 and 𝑘𝑃𝐶𝑀 are the density and thermal conductivity of the PCM, respectively. Note that
𝐻 indicates specific enthalpy, which is the sum of the specific sensible heat 𝑠, and specific latent
heat ∆𝐻𝑓 (see eq. (11)) [28]
𝐻 = 𝑠 + ∆𝐻𝑓 , (11)

6
 𝑇
𝑠 = 𝑠𝑟𝑒𝑓 + ∫𝑇 𝐶𝑃𝐶𝑀 𝑑𝑇, (12)
𝑟𝑒𝑓

∆𝐻𝑓 = 𝛽𝐿. (13)

Here 𝐶𝑃𝐶𝑀 and L are the specific heat capacity and latent heat of fusion of the PCM, respectively.
Furthermore, ℎ𝑟𝑒𝑓 is the reference enthalpy at a reference temperature 𝑇𝑟𝑒𝑓 .
The outer walls of the cells are divided into four mutually exclusive outer wall domains
defined as Γ− (negative tab face), Γ+ (positive tab face), Γ𝑇𝑀𝑆 (in contact with the thermal
management system (TMS)), and Γ𝑜 (in contact with the ambient air) on the cell [12], as shown
in Fig. 1(e). The boundary conditions for the electric potential ∅+ and ∅− are given as
𝜕∅+ 𝐼𝑜
−𝜎+ = on Γ+ , (14)
𝜕𝑛 𝐴
𝜕∅+ (15)
=0 on Γ0 , Γ− and Γ𝑇𝑀𝑆 ,
𝜕𝑛

f
where 𝐼𝑜 is the charge/discharge current (constant) through area A and 𝑛 is the normal to the

oo
boundary. Equation (14) gives a constant current density across the positive tab face implying
a constant-current discharge operation, whereas Eq. (15) shows that there is no current

r
component due to the positive potential gradient flowing through those faces. Similarly, for ∅− :
∅− = 0
-p on Γ− , (16)
𝜕∅−
re
on Γ0 , Γ+ and Γ𝑇𝑀𝑆 . (17)
=0
𝜕𝑛
lP

where Eq. (16) gives a reference potential for the cells and (17) expresses that the current
component due to the negative potential gradient is absent on all faces other than the negative tab
face. For temperature T the conditions are
na

𝜕𝑇
=0 on Γ− , Γ+ , (18)
𝜕𝑛
ur

𝜕𝑇
−𝑘𝑐 = ℎ(𝑇 − 𝑇𝑤 ) on Γ0 , (19)
𝜕𝑛
𝜕𝑇 𝜕𝑇𝑇𝑀𝑆
Jo

−𝑘𝑐 = −𝑘 𝑇𝑀𝑆 on Γ𝑇𝑀𝑆 . (20)

𝜕𝑛 𝜕𝑛
Here ℎ, 𝑇𝑤 and 𝑘 𝑇𝑀𝑆 are the convective heat transfer coefficient, ambient temperature and thermal
conductivity of the TMS in contact with the cell. Adiabatic boundary conditions on the positive
and negative tab faces are shown in Eq. (18), while Eq. (19) represents a convective boundary
condition on the faces in contact with the ambient air. Equation (20) reflects the continuity of
conductive heat flux across the TMS and the cell boundaries.
4. Numerical procedure
The numerical computations are carried out for the battery thermal and electric fields and also for
the electrochemical parameters like the volumetric current transfer rate, the DOD, the open circuit
voltage, the temperature dependent apparent current density and the volumetric heat generation
rate. The physical domain of each Li-ion cell is divided into two zones, active and passive. The
active zone is where the electrochemical reactions occur and the passive zone are the current
collecting zones. Thus, the electric fields and the electrochemical parameters are solved only in the
active zone.
For the numerical simulations the Finite Volume Solver ANSYS Fluent [40] was used and
the physical domain is divided into segments and then using the second order upwind scheme of

7
discretization, the governing equations namely the energy and the electric field equations are
transformed into algebraic equations. The solver is then initialized with the initial temperature, the
discharge rate, the nominal capacity and the initial DOD value and is solved with appropriate
boundary conditions. After each time step, the electrochemical parameters are updated using the
temperature and the electric field values and the new values of the electrochemical sub-model
parameters are again fed back into the solver for the next time step. The time step is taken to be 5
s for the calculations of parameters without any thermal management while for the base design, D1
and D2 calculations with thermal management, the time step is set to be 1 s [28].
4.1. Grid independence test
A grid independence test has been carried out to determine if the mesh used is sufficiently resolved
to capture the physics associated with the battery. To do this, the variation in temperature with time
for the base design has been tested as shown in Table 1, with mesh sizes of 61032 (M1), 162874
(M2), 470232 (M3), 902400 (M4) and 1172566 (M5) elements. The relative percentage

f
temperature difference is just 0.04% while transiting from mesh M3 to M5. Hence the simulations

oo
are carried out using mesh M3 with accuracy and computational cost in mind. A similar grid
independence test has also been conducted for designs D2 and D1 before carrying on with the

r
solution.
-p
Table 1: Comparison of 𝑇 for different grid sizes for the base design in a 4 C discharge rate
re
Number of elements 𝑇, K % Difference
lP

61032 309.832 2.7%

162874 309.071 0.6%
na

470232 308.883 0.04%

902400 308.880 0.02%
ur

1172566 308.874 0%
Jo

4.2. Model validation

Before continuing with the investigation of the problem, we first confirm the accuracy of the
numerical solver used to solve the governing equations. First the accuracy of the NTGK model
with the correction parameters (see Table S3 in the Supplementary material) is validated for a single
cell of the specifications as mentioned earlier (see Table S1 in the Supplementary material) with
the experimental data provided by Yang et al. [28] for different discharge rates of 4 C, 3 C, and 2
C, as shown in Fig. 2. The variation of average temperature with time is shown in Fig. 2(a), and
the cell voltage is shown in Fig. 2(b). As can be observed, both figures show good agreement of
the numerical model against the experimental data from [28].
The enthalpy porosity method used for simulating the fin-enhanced phase change
mechanism is also validated with the established experimental data provided by Wang et al. [33].
The validation was done for a cylindrical battery, replicated with an 8 W heat generator embedded
with a cylindrical PCM housing. The results obtained numerically shown in Fig. 2(c) were within
a 1% difference from the experimental data, thus confirming the reliability of the model. The close
agreement between the current results and the reported experimental results in the literature
establishes the accuracy and validity of the methods employed herein.

8
 (a) (b)

f
r oo
-p
re
lP

(c)
Fig. 2 Model validation: comparison of the time evolution of (a) temperature and (b) voltage for a single
18650 NMC cell with Yang et al. [28], (c) time evolution of battery surface temperature with the
na

experimental results of Wang et al. [33].

ur

5. Results and discussion

The present work discusses the scope of passive enhancement of PCM-based thermal management
Jo

systems for cylindrical Li-ion batteries with the use of finned networks. First, the effect of the fast
discharge operating conditions of 2 C, 3 C, and 4 C are tested on the 6S1P cell arrangement. Then
the performance evaluation is carried out for the proposed designs to reflect on the optimal design
at a discharge rate of 4 C. Further, the optimal design is tested for a stressed condition of 6 C
discharge and an abuse condition of 8 C discharge to demonstrate the efficacy of the proposed
system. Following that, a comparative assessment is also drawn between active and the proposed
passive cooling strategies, where different heat transfer coefficient values mimic the active strategy
in the range of 5– 100 Wm-2 K-1. This is carried out on the grounds of the ability of the studied
systems to keep the battery module from running into a possible risk of thermal runaway. The
effectiveness of the optimal proposed design in an extreme case of an external short-circuit is also
tested using a low external resistance of 0.5 ohm.
5.1. Effect of finned network designs
Despite their excellent heat storage capacity as heat sinks, pure phase change materials suffer from
intrinsically low thermal conductivities [44]. Thus, they mostly display localized heating when
used as a heat sink due to the lack of heat distribution throughout the sink. This brings a need for
thermal conductivity enhancers to increase the heat distribution rate. In this regard, implementing
finned networks to increase the overall area of heat interaction between the heat generation source

9
and the sink is a simple yet promising technique. Hence, the finned networks are proposed and
studied under different discharge conditions.
Fig. 3 (a) depicts the overall average temperature profile for the 6S1P arrangement without
any TMS for the normal discharge rates of 2 C, 3 C and 4 C. The End of Discharge (EoD) average
temperature for the 6 cells reached about 336 K for the 4 C discharge, while for 3 C, the EoD
temperature was about 325 K, and for 2 C, it was about 313 K. It is obvious that the normal
conditions of 3 C and 4 C discharges lead to higher-than-optimal temperatures for the Li-ion cells,
as a temperature above 313 K can accelerate the formation of the solid electrolyte interface (SEI)
interface, jeopardizing the operation of the battery module, thus bringing into the need for a well-
designed thermal management system. The curves initially exhibit a steeper gradient which
gradually decreases. The heat generation profiles can explain this for all the discharge cases taken
so far, as illustrated in Fig. 3(b). The heat generation rate keeps on increasing the discharge rate as
expected; however, interestingly, for all the given discharge rates, the value of total heat generation

f
peaks is almost the same DOD value, reported here to be 0.138. This also explains the variation of

oo
the maximum temperature difference profiles across the six cells for all three discharge cases, as
shown in Fig. 3(c). Quantitatively, the peak temperature difference for both 2 C and 3 C discharge

r
cases is small, with values not more than 0.5 K and 1 K, respectively. In comparison, the maximum
-p
temperature difference among the cells for the 4 C discharge is around 1.8 K, which falls in the
re
acceptable range of within 4 K.
lP
na
ur
Jo

(a) (b)

(c)
Fig. 3 Effect of C-rate without any thermal management: variation of (a) time evolution of
temperature (b) time evolution of total heat generation rate and (c) time evolution of maximum
temperature difference at 4 C, 3 C and 2 C discharge rates.

10
 To assess the comparative performance of the different proposed designs, the rapid
discharge rate of 4 C is undertaken. Fig. 4(a) shows the time evolution of the temperature in the
different thermal management systems. The plateau-like behavior is attributed to the melting of the
PCM and the absence of a steeper rise in temperature in the completion time range indicates that
the whole PCM has not melted yet. As expected, the base design has the highest temperature
throughout the discharge. Design 2 shows the best behavior in terms of EoD temperature and
indicates a slower onset of melting of the PCM by about 54 % compared to the base design. It
should be noted that just the addition of base PCM decreases the overall increase in temperature,
compared to the absence of any TMS, by about 71.9 %, and this value increases up to 75.58% with
the use of D2. As observed in the case of Design 1, the behavior does not improve much in terms
of EoD temperature, with a decrease of only about 0.15 K and a delay in the onset of melting by
only 10 %. It has also been found that the average temperature of the cells during the onset of

f
melting of PCM is higher for both D1 and D2 compared to the base design, thus again highlighting

oo
their effectiveness in spreading the heat throughout the sink.
Fig. 4(b) shows the maximum temperature difference across the 6S1P battery module with

r
different arrangements. The peaks of the maximum difference for all the arrangements are observed
-p
near the onset of the melting range, with the highest for the base PCM up to about 1.24 K. D1
shows the best results with the minimum difference at the peak value and also throughout. This is
re
attributed to the interconnection of the cells cancelling the gradients along the fin length. Fig. 4(c)
lP

depicts the melt fraction plots for the different arrangements. The maximum melt fraction can be
observed for D2 and is about 0.146. This further suggests the scope for reduction of the volume of
the TMS. It is also found from Fig, 4(d) that the cell voltage behavior with the DOD for all the
na

different arrangements of the PCM-enhanced systems resulted in a negligible capacity degradation.

Thus, all the studied systems can be used without hampering the power delivered by the cell
ur

arrangement.
Jo

(a) (b)

11
 (c) (d)
Fig. 4 Effect of finned networks during 4 C discharge: (a) time evolution of temperature, (b) time

f
evolution of maximum temperature difference, (c) time evolution of melt fraction of the PCM and

oo
(d) time evolution of voltage

r
5.2. Effect of ambient temperature
-p
re
It is well understood that the working of a Li-ion battery is highly sensitive to the ambient
temperature. At higher ambient temperatures, battery life can decrease drastically on continuous
lP

discharge without any thermal management system [45]. On the opposite spectrum, long-term
operation in a cold ambient can lead to dendrite formation in the anode of the battery, resulting in
a decreased life. Hence, a study is undertaken to understand the effect of the ambient temperature
na

on the proposed design. These Li-ion batteries are expected to deliver power safely and provide
stable performance under varying ambient conditions pertaining to different seasonal variations.
ur

In the present study, the effect of variation in ambient temperature from 288.15 K to 308.15
K is studied for the optimal design, i.e., D2, in terms of delivered capacity, overall temperature,
Jo

and the maximum temperature difference. Fig. 5(a) confirms the robustness of the design because
at all the considered ambient temperatures; the delivered power is negligibly reduced. The
temperature variation for the case with an ambient temperature of 303.15 K, as illustrated in Fig.
5(b), also exhibits a plateau-like variation with time, indicating a thermal control of the system
throughout the discharge, just like the cases with the ambient temperatures of 293.15 K and 298.15
K. It should be noted that the lower the ambient temperature, the slower the onset of the plateau-
like behavior, i.e., the slower the onset of melting. This illustrates that for lower ambient
temperatures, the thermal control by the PCM is achieved more by the sensible heat of enthalpy
than the latent counterpart. Fig. 5(c) illustrates the maximum temperature difference for different
ambient temperatures, and it is safe to conclude that the temperature gradients for all the undertaken
cases are within safe limits.
5.3. Effect of stressed C rates
For electric vehicles to accelerate to high speeds, sometimes-abrupt increase in the cell discharge
rates can occur. Abusive discharge rates always bring enormous heat generation rates within the
cells, leading to temperatures much higher than the optimal range. It should also be understood that
the higher the discharge rates are, the faster they deplete the capacity of the battery, which ensures

12
a shorter time span for the temperature increase. Thus, it becomes of utmost importance to ensure
that the thermal management system incorporated into use must be able to prevent any such
hazardous phenomena, if not be able to keep the battery in the optimal range. In this study, a
stressed condition of 6 C discharge is taken with an abusive discharge rate of 8 C and the
performance of the D2 under both the circumstances is examined.
Fig. 6 illustrates the thermal performance of D2 in terms of the average temperature and
the maximum temperature difference of the cell arrangement during the discharge rates of 4 C
(normal), 6 C (stressed), and 8 C (abusive). In Fig. 6(a), the plots for 6 C and 4 C rates show a
plateau-like behavior till EoD, which the design fails to maintain for 8 C. Two points are marked,
viz-A and B, with a square and a circle, respectively. The square represents the onset of melting,
and the circle represents the end of the plateau-like behavior. The plots for 4 C and 6 C do not
display any end to the plateau-like behavior. Point B represents the point of loss of thermal control
by design, as the temperature increases again. This can be explained by looking at Fig. 7, which

f
illustrates the melt fraction contours for the 8 C discharge at the different instants of A, B, and the

oo
EoD. At about t = 98 s, the vicinity of the cells starts melting, thus onsetting the streak of very
minute temperature changes as most of the heat from here is taken up by the latent part of the heat

r
of enthalpy. At about t = 190s, this behavior changes as the vicinity of the cell melts down
-p
completely (melt fraction of 1), and again only the sensible part of the heat can take up heat from
re
the cells and diffuse this heat to the outer un-melted part of the PCM. The low thermal conductivity
value of PEG 1000 cannot immediately conduct this heat to the outer part. Thus, during this time,
lP

the heat keeps accumulating, leading to an increase in the temperature, departing the plot from a
plateau-like behavior. Fig. 6(b) depicts the maximum temperature difference for the different
stressed C-rates, and it is observed that for all the variations, the parameter does not go beyond 4K
na

at any point of time.

ur
Jo

(a) (b)

13
 (c)
Fig. 5 Effect of ambient temperature on Design 2 during a 4 C discharge: (a) time evolution of

f
cell voltage, (b) time evolution of temperature, and (c) time evolution of temperature difference

r oo
-p
re
lP
na

(a) (b)
ur

Fig. 6 Effect of stressed conditions on Design 2: variation of (a) temperature and (b)
temperature difference with time
Jo

(a)
t = 98 s t = 190 s t = 397 s

14
 (b) (c) (d)
Fig. 7 (a) Melt fraction contour distribution at the mid-section in Design 2 and (b)-(d)
distribution of melt fraction at 8 C at various time steps.

In this study, an attempt has also been taken to illustrate the performance of D2 against
active thermal management methods for the abuse discharge condition of 8 C. For this purpose, the
cell arrangement has been studied without any thermal management under natural convective
boundary conditions, mimicked by a heat transfer coefficient (HTC) of 5 W/(m2K). The HTC was

f
then increased up to 100 W/(m2K) to mimic forced cooling conditions and was compared to the

oo
results exhibited by the fin design D2. It should be noted that beyond a temperature of 353.15 K,
the risk of a thermal runaway is alarmingly high.

r
Fig. 8(a) shows that the temperature rise for natural cooling conditions profoundly initiates
-p
a thermal runaway in the module as the temperature is way beyond the extreme limits, and the limit
is reached within 225 s. Even an HTC value of 25 Wm-2K-1 is not sufficient to prevent a thermal
re
runaway. To design an active forced cooling, the parameters chosen should be such that the HTC
is at least 50 Wm-2K-1, to dissolve any possibility of the initiation of a thermal runaway. It can be
lP

observed in Fig. 8 (b) that even an HTC value of 100 Wm-2K-1 is not sufficient to keep the
temperature difference within 4 K. This proves the unmatched candidature of the PCM enhanced
na

systems for thermal management of batteries.

ur
Jo

(a) (b)
Fig. 8 Variation in (a) temperature and (b) temperature difference at 8 C discharge rate for the
cell arrangement to illustrate the effectiveness of the proposed design (passive cooling approach)
over active cooling approach (forced convection).

5.4. Effect of the external short-circuit

The performance evaluation of thermal management systems under any short circuits is also crucial
as short circuits are the de-facto thermal runaway initiators. External short circuits are more
probable than their internal counterparts and must be tested in the design stage of the BTMS. For

15
modeling an external short circuit, it is made sure the ratio between the voltage across the cell and
the discharge current is kept equal to the load resistance.

(a) (b)

f
oo
Fig. 9 Performance assessment of proposed Design 2, compared to the case without any
thermal management, in case of an external short circuit condition as compared to: (a)
variation of cell voltage and (b) variation of temperature.

r
-p
In this study, D2 is tested for an external short circuit caused by a very small resistance of 0.5
re
ohm. As shown in Fig. 9(a), the average cell voltage reaches the nominal minimum of 2.8 V within
just 50 s for both with and without design D2, and then goes on decreasing far beyond this value at
lP

a slower pace till about 300 s. The voltage loss is large because enormous current density values
arise during external short circuits, forcing the electrodes to go far beyond their nominal
characteristics. It can be seen in Fig. 9(b) that the temperature of the cells, without any thermal
na

management, reaches the critical thermal initiation temperature just within 70 s. On the other hand,
D2 has successfully kept a check-in temperature up to 339 K, ruling out the possibility of a thermal
ur

runaway even under an external short circuit.

Jo

6. Conclusions
In the present study metal fin intensified hexagonal PCM-based battery thermal management
designs D1 (consisting of aluminum rings and vertical interconnecting fins), and D2 (consisting of
aluminum rings and vertical fins diagonally spread out in the heat sink) are proposed and compared.
The thermal-electrochemical coupling is achieved using the NTGK (Newman, Tiedermann, Gu
and Kim) semi-empirical model. The effect of discharge rates and ambient temperature on the
thermal management of the 6S1P cylindrical Li-ion cell arrangement is studied. Further, the
viability of the system in keeping the battery module safe even in case of external short circuits is
also investigated. The main findings are:
➢ On studying various discharge rates, it is found that the heat generation rate increases
with the increase in C-rate and reaches its peak when the Depth of Discharge (DOD) is
around 0.138 for all the studied C-rates.
➢ The peak temperature during the 4 C discharge for the cell arrangement without any
TMS is observed to be 336 K, whereas the peak temperature for the base design is
around 308.9 K indicating a reduction in the increase in temperature from the initial
value by almost 71%. This reduction in the increase in temperature is further enhanced

16
 to 72% by the use of Design D1 and 75.58% for Design D2, respectively, demonstrating
that the finned structure network improves the thermal management performance.
➢ The temperature differences in the cell arrangement (∆𝑇) in the case of fin-intensified
designs (D1, D2) decrease from the case without any TMS by about 60% and 42%
respectively, thus showing a significant influence of fin intensification on establishing
temperature uniformity in cell arrangements.
➢ The onset of melting of the PCM is found to be slowed by 54% with the use of Design
D2 compared to the base PCM case, proving its ability to diffuse heat throughout the
sink and thus preventing any localized heating, which would otherwise lead to higher
temperature gradients across the cells.
➢ For the tested ambient temperatures on Design D2, varying from 288.15 K to 308.15
K, the interplay between the thermal management offered by the latent and the sensible

f
heat of enthalpy was captured. Up to 303.15 K, the design demonstrates excellent

oo
thermal management performance in terms of the peak temperature as well as the
temperature uniformity in the cell arrangement, as the latent heat of the PCM is

r
available for the design to use, which is not the case for 308.15 K, where only the
-p
sensible heat of enthalpy could be used for dissipating the heat.
➢ Even at higher C-rates (stressed condition of 6 C), the performance of D2 remains
re
unhindered, demonstrating its effectiveness at managing stressed conditions with a
temperature rise controlled up to 310 K.
lP

➢ The point of initiation of thermal runaway is observed when the cell arrangement (without
any thermal management) is operated at 8 C discharge rate with peak temperature reaching
na

the critical value of 353.15 K within 225 s.

➢ D2 appears to cross the point of thermal control when operated at 8 C (abuse conditions),
ur

as evident from the complete melting of PCM in the vicinity of the cells. The temperature
difference is under a safe limit of 4 K with a maximum temperature of 315 K, thus
Jo

preventing possible thermal runaway.

➢ The study on the effect of an external short circuit also reconfirmed that D2 will not undergo
thermal runaway as it sustained temperatures up to 339 K for up to 280 s.
The findings of the present study can help in designing PCM-enhanced battery packs.

17
References
[1] S.S. Madani, M.J. Swierczynski, and S.K. Kaer, The discharge behavior of lithium-ion batteries
using the Dual-Potential Multi-Scale Multi-Dimensional (MSMD) battery model, 2017 Twelfth Int.
Conf. Ecological Vehicles Renewable Energies. (2017) 1-14.
[2] Z.Y. Jiang, Z.G. Qu, Lithium–ion battery thermal management using heat pipe and phase change
material during discharge–charge cycle: A comprehensive numerical study, Appl. Energy 242
(2019) 378–392.
[3] C. Zhao, W. Cao, T. Dong, F. Jiang, Thermal behavior study of discharging/charging cylindrical
lithium-ion battery module cooled by channeled liquid flow, Int. J. Heat Mass Tran. 120 (2018) 751–
762.
[4] X. Xu, G. Tong, R. Li, Numerical study and optimizing on cold plate splitter for lithium battery
thermal management system, Appl. Therm. Eng. 167 (2020), 114987.
[5] Y. Li, Z. Zhou, W.T. Wu, Three-dimensional thermal modeling of Li-ion battery cell and 50 V Li-

f
ion battery pack cooled by mini-channel cold plate, Appl. Therm. Eng. 147 (2019) 829–840.

oo
[6] D. Bernardi, E. Pawlikowski, J. Newman, A general energy balance for battery systems. J.
Electrochem. Soc. 132 (1985) 5–12.

r
[7] J. Hong, H. Maleki, S. Al Hallaj, L. Redey, J.R. Selman, Electrochemical‐Calorimetric studies of
-p
Lithium‐ion cells, J. Electrochem. Soc. 145 (1998) 1489–1501.
[8] K. Kirad, M. Chaudhari, Design of cell spacing in lithium-ion battery module for improvement in
re
cooling performance of the battery thermal management system, J. Power Sources 481 (2021),
229016.
lP

[9] L. Sheng, L. Su, H. Zhang, K. Li, Y. Fang, W. Ye, Y. Fang, Numerical investigation on a lithium
ion battery thermal management utilizing a serpentine-channel liquid cooling plate exchanger, Int.
na

J. Heat Mass Tran. 141 (2019) 658–668.

[10] Md. Faizan, S. Pati, P. R. Randive, Implications of novel cold plate design with hybrid cooling on
thermal management of fast discharging Lithium-ion battery, J Energy Storage 53 (2022) 105051
ur

[11] Md. Faizan, S. Pati, P. R. Randive, Effect of channel configurations on the thermal management
of fast discharging Li-ion battery module with hybrid cooling, Energy 7 (2023) 126358
Jo

[12] K.H. Kwon, C.B. Shin, T.H. Kang, C.S. Kim, A two-dimensional modeling of a lithium-polymer
battery, J. Power Sources 163 (2006) 151–157.
[13] W.H. Tiedemann and J. Newman, Current and potential distribution in lead-acid battery plates,
Proceedings of the Symposium in Battery Design and Optimization, The Electrochemical Society,
(1979) 39-49.
[14] U.S. Kim, C.B. Shin, C.S. Kim, Effect of electrode configuration on the thermal behavior of a
lithium-polymer battery, J. Power Sources 180 (2008) 909–916.
[15] G.-H. Kim, K. Smith, K.-J. Lee, S. Santhanagopalan, A. Pesaran, Multi-domain modeling of
Lithium-ion batteries encompassing multi-physics in varied length scales, J. Electrochem. Soc. 158
(2011), A955.
[16] A. Mills, S. Al-Hallaj, Simulation of passive thermal management system for lithium-ion battery
packs, J. Power Sources 141 (2005) 307–315.
[17] T. Rehman, H.M. Ali, M.M. Janjua, U. Sajjad, W.M. Yan, A critical review on heat transfer
augmentation of phase change materials embedded with porous materials/foams, Int. J. Heat Mass
Tran. 135 (2019) 649–673.
[18] A. Kaabinejadian, H.A. Ahmadi, M. Moghimi, Investigation of porous media effects on lithium-
ion battery thermal management, J. Therm. Anal. Calorim. 141 (2020) 1619–1633.

18
[19] Y.S. Ranjbaran, S.J. Haghparast, M.H. Shojaeefard, G.R. Molaeimanesh, Numerical evaluation of
a thermal management system consisting PCM and porous metal foam for Li-ion batteries, J. Therm.
Anal. Calorim. 141 (2020) 1717–1739.
[20] K.S. Kshetrimayum, Y.G. Yoon, H.R. Gye, C.J. Lee, Preventing heat propagation and thermal
runaway in electric vehicle battery modules using integrated PCM and micro-channel plate cooling
system, Appl. Therm. Eng. 159 (2019), 113797.
[21] Y. Konuklu, M. Unal, H.O. Paksoy, Microencapsulation of caprylic acid with different wall
materials as phase change material for thermal energy storage, Sol. Energ. Mat. Sol. C. 120 (2014)
536–542.
[22] S.L. Tariq, H.M. Ali, M.A. Akram, M.M. Janjua, M. Ahmadlouydarab, Nanoparticles enhanced
phase change materials (NePCMs)-A recent review, Appl. Therm. Eng. 176 (2020), 115305.
[23] P. Huang, A. Verma, D.J. Robles, Q. Wang, P. Mukherjee, J. Sun, Probing the cooling effectiveness
of phase change materials on lithium-ion battery thermal response under overcharge condition, Appl.
Therm. Eng. 132 (2018) 521–530.

f
oo
[24] E. Jiaqiang, D. Han, A. Qiu, H. Zhu, Y. Deng, J. Chen, X. Zhao, W. Zuo, H. Wang, J. Chen, Q.
Peng, Orthogonal experimental design of liquid-cooling structure on the cooling effect of a liquid-
cooled battery thermal management system, Appl. Therm. Eng. 132 (2018) 508–520.

r
-p
[25] Z. Ling, F. Wang, X. Fang, X. Gao, Z. Zhang, A hybrid thermal management system for lithium
ion batteries combining phase change materials with forced-air cooling, Appl. Energy 148 (2015)
re
403–409.
[26] M. Mashayekhi, E. Houshfar, M. Ashjaee, Development of hybrid cooling method with PCM and
lP

Al2O3 nanofluid in aluminium minichannels using heat source model of Li-ion batteries, Appl.
Therm. Eng. 178 (2020), 115543.
[27] B. Mondal, C.F. Lopez, P.P. Mukherjee, Exploring the efficacy of nanofluids for lithium-ion
na

battery thermal management, Int. J. Heat Mass Tran. 112 (2017) 779–794.
[28] W. Yang, F. Zhou, Y. Liu, S. Xu, X. Chen, Thermal performance of honeycomb-like battery
ur

thermal management system with bionic liquid mini-channel and phase change materials for
cylindrical lithium-ion battery, Appl. Therm. Eng. 188 (2021), 116619.
Jo

[29] B. Mondal, C.F. Lopez, A. Verma, P.P. Mukherjee, Vortex generators for active thermal
management in lithium-ion battery systems, Int. J. Heat Mass Tran. 124 (2018) 800–815.
[30] C.F. Lopez, J.A. Jeevarajan, P.P. Mukherjee, Evaluation of combined active and passive thermal
management strategies for Lithium-ion batteries, J. Electrochem. Energy Convers. Storage 13
(2016), 031007.
[31] R. Baby, C. Balaji, Experimental investigations on phase change material based finned heat sinks
for electronic equipment cooling, Int. J. Heat Mass Tran. 55 (2012) 1642–1649.
[32] Y. Lv, X. Yang, X. Li, G. Zhang, Z. Wang, C. Yang, Experimental study on a novel battery thermal
management technology based on low density polyethylene-enhanced composite phase change
materials coupled with low fins, Appl. Energy 178 (2016) 376–382.
[33] Z. Wang, H. Zhang, X. Xia, Experimental investigation on the thermal behavior of cylindrical
battery with composite paraffin and fin structure, Int. J. Heat Mass Tran. 109 (2017) 958–970.
[34] A. Arshad, H.M. Ali, S. Khushnood, M. Jabbal, Experimental investigation of PCM based round
pin-fin heat sinks for thermal management of electronics: Effect of pin-fin diameter, Int. J. Heat
Mass Tran. 117 (2018) 861–872.
[35] M.Y. Yazici, M. Avci, O. Aydin, Combined effects of inclination angle and fin number on thermal
performance of a PCM-based heat sink, Appl. Therm. Eng. 159 (2019), 113956.

19
[36] Z. Sun, R. Fan, F. Yan, T. Zhou, N. Zheng, Thermal management of the lithium-ion battery by the
composite PCM-Fin structures, Int. J. Heat Mass Tran. 145 (2019), 118739.
[37] P. Ping, R. Peng, D. Kong, G. Chen, J. Wen, Investigation on thermal management performance
of PCM-fin structure for Li-ion battery module in high-temperature environment, Energy Convers.
Manag. 176 (2018) 131–146.
[38] A.S. Fini, A. Gharehghani, Experimental investigation of pressure effect on the PCM performance
in Li-ion battery thermal management system, J Energy Storage, 79 (2024) 110273.
[39] Y. Kou, S.Y. Wang, J.P. Luo, K.Y. Sun, J. Zhang, Z.C. Tan, Q. Shi, Thermal analysis and heat
capacity study of polyethylene glycol (PEG) phase change materials for thermal energy storage
applications, J. Chem. Thermodynamics 128 (2019) 259–274.
[40] ANSYS Fluent advanced addon modules, 17.0, Section 2: ANSYS Fluent battery module (2016).
[41] J. Newman, W. Tiedemann, Potential and current distribution in electrochemical cells
interpretation of the half-cell voltage measurements as a function of reference-electrode location, J.
Electrochem. Soc. 140 (1993) 1961–1968.

f
oo
[42] H. Gu, Mathematical analysis of a Zn/NiOOH Cell, J. Electrochem. Soc. 130 (1983) 1459–1464.
[43] C.R. Swaminathan, V.R Voller, On the enthalpy method, Int. J. Numer. Method. H. 3 (1993) 233-
244.

r
-p
[44] R. Fan, N. Zheng, Z. Sun, Evaluation of fin intensified phase change material systems for thermal
management of Li-ion battery modules, Int. J. Heat Mass Tran. 166 (2021), 120753.
re
[45] S. Yang, C. Ling, Y. Fan, Y. Yang, X. Tan, H. Dong, A review of lithium-ion battery thermal
management system strategies and the evaluate criteria, Int. J. Electrochem. Sci. 14 (2019) 6077–
lP

6107.
na
ur
Jo

20
Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:

of
ro
-p
re
lP
na
ur
Jo