Journal of Energy Storage 59 (2023) 106474

Contents lists available at ScienceDirect

Journal of Energy Storage

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

Research papers

Thermal cooling of li-ion cylindrical cells battery module with baffles

arrangement for airflow cooling numerical analysis
Obula Reddy Kummitha
Department of Mechanical Engineering, B V Raju Institute of Technology, Narsapur 502313, India

A R T I C L E I N F O A B S T R A C T

Keywords: The demand for electric cars continues to grow every day due to their benefits, such as zero emissions, and the
Battery thermal management system (BTMS) downsides of traditional internal combustion engines, such as toxic pollutants. There are some complications in
Li-ion batteries the practical use of batteries, like thermal runaway and decreased battery life cycles and usage. Hence, this
Cell holders
research introduced an innovative method to achieve uniform temperature distribution in the battery module
Computational fluid dynamics
and decrease the maximum temperature obtained in the battery cells. The new battery module has been designed
Battery module
by considering the cell holders at different cell heights. These cell holders are like a tray to hold the battery cells
and act as fins, enhancing the uniform temperature distribution within the battery module. Base and new battery
modules are analyzed at different discharge conditions with the same boundary conditions and physical pa­
rameters. The significance of cell holders acting as fins has been evaluated by analyzing the temperature contours
and profiles for every cell of the battery module. The analysis of numerical results found that at 1, 2 and 3C rate,
the maximum temperatures of 301.5, 303.9 and 308.8 K for base model and whereas for new model these are
300.6, 302.1, and 304 K, respectively. At 1C rate discharge, the temperature gradient among the cells (Tcell4-
Tcell6) in the flow directions has been observed as 0.26 ◦ C and 0.04 ◦ C for the base and novel battery model,
respectively.

battery pack can result in electrically unbalanced cells. An electric

imbalance causes the entire battery pack to lose capacity and the over­
1. Introduction charging of the affected cells during charging, resulting in power losses
and higher temperatures [2,3]. The majority of temperature impacts are
Lithium-ion batteries are well suited for fully electric and hybrid- caused by chemical processes inside the batteries and partly by the
electric cars because they have high energy and energy density than materials used in the batteries [4]. The temperature gradient of an
other rechargeable cell chemistries. However, due to safety, expense, electric vehicle's battery system is usually between 40 and 60 ◦ C [5,6].
and poor low-temperature performance, which are all difficulties con­ However, it is widely acknowledged that the optimal operating tem­
nected to battery thermal management, these batteries have not yet perature range for electric vehicle batteries is between 15 and 35 ◦ C
been extensively used commercially in these vehicles. A battery thermal [7–10]. Outside these limits, temperatures have a detrimental effect on
management system (BTMS) is perhaps the essential component of an the thermal performance of batteries, resulting in considerable capacity
electric vehicle (EV) since it ensures that lithium-ion batteries work loss and lifespan decrease. Battery life is projected to be reduced by two
safely and consistently (LiB). To get the most power out of LiBs, it is months for each degree of temperature increase [11]. Simultaneously, it
essential to keep a close eye on their working conditions, especially is advised that the most significant temperature differential between the
temperature, which has been shown to directly influence their perfor­ interior and outside of a battery pack not exceed 5 ◦ C [12].
mance and longevity. The heat-generating processes of lithium-ion High temperatures enhance the reaction rate, resulting in more
batteries during charge and discharge cycles induce the temperature excellent power delivery and increased capacity, but they also
increase near LiBs. An efficient battery thermal cooling system can contribute to higher temperatures and a rise in thermal load [13].
control an electric vehicle's performance and increase safety. Furthermore, due to electrolyte breakdown and non-uniformity of the
The temperature is one of the most critical parameter impacting passivation layer, capacity fading may be an issue caused by high tem­
battery aging, negatively influencing capacity and internal resistance peratures [14]. Temperatures may rise uncontrolled if heat is not
[1]. Furthermore, a non-uniform temperature distribution inside a

E-mail address: [email protected].

https://doi.org/10.1016/j.est.2022.106474
Received 19 April 2022; Received in revised form 24 October 2022; Accepted 17 December 2022
Available online 23 December 2022
2352-152X/© 2022 Elsevier Ltd. All rights reserved.
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Nomenclature λ Thermal conductivity of air (W/m.K)

λb, j Thermal conductivity of battery material in Jth direction
ui, uj Velocity components in X and Y directions (m/s) (W/m.K)
P Pressure (Pa) τij Stress tensor (N/m2)
T Temperature (◦ C) ε Turbulent dissipation rate (m2/s3)
Tb Battery temperature (◦ C) ρ Density (kg/m3)
cp Specific heat at constant pressure (J/kg.K) ρb Density of battery material (kg/m3)
cp, b Specific heat for battery material (J/kg.K) ϕb Source term
k Turbulent kinetic energy (m2/s2) σ k, σε, σT Parameters of the k-epsilon turbulence model
Gk Turbulent kinetic energy generation with gradients of
velocity Acronyms
Gb Turbulent kinetic energy generation with buoyancy BTMS Battery Thermal Management System
YM Fluctuating dilatation contribution EV Electric vehicles
Sk Source term for k LiB Lithium-ion battery
Sε Source term for ε MSMD Multi-Scale Multi-Domain
C1ε, C2ε, C3ε Model constants for k-epsilon turbulence model PISO Pressure-Implicit with Splitting of Operators
ECM Equivalent Circuit Model
Greek symbols SOC State of Charge
μ Dynamic viscosity (Pa.s) OCV Open Circuit Voltage
μt Turbulent dynamic viscosity (Pa.s) PCM Phase change material

dispersed at least at the same rate as it is created in the batteries, the ECM and LSTM models are more precise than the ESPM model.
resulting in the degradation of materials and components [15] or From the available literature review, it is understood that battery
possibly the thermal runaway of the batteries [16]. A thermal runaway pack operating temperature plays a significant role in the efficient
is an event that causes a fast temperature increase, gas production, and operation of electric vehicles. Many researchers discussed different
potentially a battery explosion [17], putting the vehicle's and its occu­ cooling techniques like passive, active, and combinations (hybrid). In
pants' safety in jeopardy [18]. this research, an innovative battery module has been designed and
Quasi temperature distribution within a single cell causes a varied studied its performance at different C rate discharge conditions. The C
rate of electrochemical reactions in different cell regions, resulting in Rate defines the measurement of current at charging and discharging
partial energy usage and shorter battery life [19]. According to studies, a conditions. The cooling performance of the battery module is analyzed
temperature differential of >5 degrees Celsius within a battery cell by assessing the maximum temperature and temperature gradient in the
causes a 25 % increase in thermal aging and a 10 % drop in power ca­ battery module.
pacity [20–22]. At the battery pack level, cell discrepancies in capacity,
voltage, and internal resistance are typical and expected [23]. These 2. Method for geometrical modeling of battery module
differences result in different thermal behaviors, resulting in a temper­
ature gradient throughout the battery pack [24]. A 5 ◦ C temperature The Li-ion battery pack consists of 207 Li-ion cylindrical cells
differential may result in a 1.5–2 % capacity drop in the battery pack (IFR32700); hence, the whole pack is divided into 23 battery modules,
[25], as well as a 10 % reduction in power capability [26]. each consisting of 9 cells. In this research, the nine cells are initially
Aditya et al. [27] studied a phase change material-based passive heat arranged in a matrix form of 3*3 sizes, as shown in Fig. 1. The diameter
management solution for Li-ion cylindrical cells (PCM). The scientists and height of the cell is 32 mm and 70 mm, respectively. All the cylin­
examined battery temperatures for various PCM material thicknesses drical cells are designed with separate cathode and anode parts and
and determined that a minimum thickness of 4 mm is required to keep separate busbars connecting all the positive terminals and negative
the battery's maximum temperature within the range. Jeevan Jaidi et al. terminals, respectively. With these arrangements, the battery module is
[28] have investigated the notion of BTMS for a prismatic cell with a considered in parallel connection. For efficient thermal cooling of bat­
mini-channel cooling flow channel around the cell and distinct coolant teries and uniform temperature distribution, a novel approach is
inlet circumstances. The primary contribution of this study was to considered by designing cell holders' jacket fins, as shown in Fig. 1.
determine the minimal input temperature and mass flow rate of the These jacket fins are designed to improve the heat removal rate from the
coolant required to achieve the desired battery cell cooling. battery surface and ensure uniform thermal distribution within the
Zhao et al. [29] exhibited both experimental and numerical research battery. The heat removal plate (cell holder's jacket) is considered with
for an ultra-fast charging battery module constructed using pouch cell lightweight aluminum with 1 mm thickness, and 10 mm sleeve height
Li-ion battery cells. The battery moducel was placed between foam and around the cell and 4 mm fin extension from the sleeve surface. The
heat sinks on both sides. Several tests were undertaken to determine the battery nominal voltage and capacity are 3.2 V and 6000 mAh, respec­
maximum temperature, surface temperature, and heat flow during ultra- tively. The battery materials and properties are defined in Table 1.
fast charging the battery module at a rate of 5C. At 20 ◦ C inlet coolant
temperature and 5C rate fast charging, the scientists performed further 3. Mathematical modeling and numerical methods
tests to determine the power loss from the battery and determined that it
is 30 % at 20 ◦ C inlet coolant temperature and 5C rate fast charging. The computational flow domains are computed by solving the
Liang et al. [30] did a comparative analysis of the modeling of battery Reynolds-averaged Navier-Stokes (RANS) equations, k-ε turbulence
packs using various voltage techniques. Recurrent neural network model transport equations, and the other governing equations for bat­
model, analogous circuit model, and extended single particle model are tery Multi-Scale Multi-Domain (MSMD) modeling. These governing
the three models. These models were evaluated and compared using four equations' spatial derivatives are discretized using the finite volume
distinct standard driving cycles at six distinct ambient conditions method with a second-order upwind discretization scheme. The PISO
ranging from − 20 ◦ C to 40 ◦ C. The comparative analysis determined that (Pressure-Implicit with Splitting of Operators) algorithm for pressure-

2
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 1. Geometrical modeling of battery module a) Base model and b) New model.

The standard k-ε turbulence model is considered with standard wall

Table 1
function as follows:
Thermo physical properties of battery materials [34].
[( ) ]
Component/material Density (kg. Heat Thermal ∂ ∂ ∂ μ ∂k
(ρk) + (ρkui ) = μ+ t + Gk + Gb − ρε − YM + Sk (5)
m− 3) capacity conductivity ∂t ∂xi ∂xj σk ∂xj
(J.kg− 1. (W.m− 1 K− 1)
K− 1 ) [( ) ]
∂ ∂ ∂ μt ∂ε ε ε2
(ρε) + (ρεui ) = μ+ + C1ε (Gk + C3ε Gb ) − C2ε ρ + Sε
LiFePO4 based cathode 1355 1217 1.58 ∂t ∂xi ∂xj σ ε ∂xj k k
Graphite based anode 2217 1437.5 1.05
(6)
Lithium hexafluorophosphate
1124 135.8 0.63
electrolyte (LiPF6) The terms Gk and Gb stand for turbulent kinetic energy generation and
Aluminum (cell holder) 2700 897 167
buoyancy, respectively. The fluctuating dilatation contribution was
represented with YM, and the source terms are defined with S. The model
velocity coupling has been implemented for solving the discredited constants consist of the following default values:
momentum equation and the corresponding velocity flow field. All the C1ε = 1.44, C2ε = 1.92, σk = 1.0, σε = 1.3
numerical simulations are performed at unstable conditions; hence the
time derivatives (transient formulation) are discretized with first-order Modeling a lithium-ion (Li-ion) battery presents a challenge due to
implicit formulation to avoid divergence. The divergence associated its multi-domain, multi-physics character. The Multi-Scale Multi-
with flow physics has been addressed by limiting the under-relaxation Domain (MSMD) approach considers different physics in distinct solu­
factors to less than one. All these multiphysics are numerically tion domains. The vastly varied length scales associated with different
computed by using the ANSYS Fluent 19.2 academic research license physics complicate the challenge. A thermal study aims to ascertain the
tool. The respective flow governing equations for the pre-defined nu­ temperature distribution along the battery's length scale. The physics
merical modeling are as follows [33,34]: governing Li-ion transport occurs in the sandwich layers of the anode,
separator, and cathode (the electrode pair length scale). Li-ion transport
Mass conservation (continuity equation): occurs on an atomic scale in the active material.
The following differential equations are used to solve the thermal
∂ρ ∂ and electrical fields of batteries in the CFD domain at the scale of the
+ (ρui ) = 0 (1)
∂t ∂xi battery cell:
∂ρCp T
Momentum conservation: − ∇ • (k∇T) = σ + |∇∅+ |2 + σ− |∇∅− |2 + q̇Ech + q̇short + q̇abuse (7)
∂t
∂ ∂ ( ) ∂P ∂ ( ) ( )
(ρu ) + ρui uj = − + τ (2)
∂t i ∂xi ∂xi ∂xi ij ∇ • (σ + ∇φ+ ) = − j˙Ech + jshort (8)

Energy equation (flow domain): ∇ • (σ − ∇φ− ) = j˙Ech − jshort (9)

[( ) ]
∂T ∂T ∂ μ ∂T
ρcp + ρcp uj = λ+ t (3) where σ+ and σ− are the effective electric conductivities for the positive
∂t ∂xj ∂xj σT ∂xj
and negative electrodes, φ+and φ− are phase potentials for the positive
and negative electrodes, j̇Ech and q̇Ech are the volumetric current transfer
Energy equation (battery domain): rate and the electrochemical reaction heat due to electrochemical re­
[ ] actions, respectively, jshort and q̇short are the current transfer rate and
∂T ∂ ∂Tb
ρb cp,b b = λb,j + ϕb (4) heat generation rate due to battery internal short-circuit, respectively,
∂xj ∂xj ∂xj
and q̇abuse is the heat generation due to the thermal runaway reactions
under the thermal abuse condition.
Standard k - ε model: The source terms j̇Ech and q̇Ech are computed using an electrochemical
sub-model. If there is no internal short-circuit, jshort and q̇short are equal

3
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

to zero. The Equivalent Circuit Model (ECM) based on the mimic of an Reynolds number (Re) based on the local properties like dynamic vis­
electrical circuit with three resistors, two capacitors, and open-circuit cosity (μf), density of fluid (ρf) and the velocity of fluid (U) as shown
voltage based on Chen [31] has been considered for battery electrical- below. The mass flow rate of air at a velocity of 0.5 m/s corresponds to a
thermal modeling. Reynolds number of 1008.
The relation between the voltage and current for the considered
ρf UD
circuit model is obtained by solving the following set of equations: Re = (28)
μf
V(t) = VOCV (SOC) − V1 − V2 − Rs (SOC)I(t) (10)
4. Grid-independent study and experimental validation
dV1 1 1
= V1 − I(t) (11)
dt R1 (SOC)C1 (SOC) C1 (SOC)
The mesh generation is the crucial step in the CFD problem for ac­
curate numerical results; hence, utmost care has been taken for gener­
dV2 1 1
= V2 − I(t) (12) ating the mesh. As the computational domain consists of solid battery
dt R2 (SOC)C2 (SOC) C2 (SOC)
and surrounding flow domain, the unstructured mesh has been gener­
d(SOC)
/ ated using both the hexahedral and tetrahedral elements, as shown in
= I(t) 3600QAh (13) Fig. 2. Different mesh size elements 0.10 mm, 0.05 mm and 0.01 mm are
dt
considered for the coarser, medium, and fine mesh settings, respectively.
For the considered MSMD battery model and ECM sub-model, the Their impact on the numerical results is analyzed. The meshing was
circuits open voltage, capacitors capacitances, and resistors resistances generated by considering the inflation option to limit the y+ value
are the state of charge (SOC) functions. In ANSYS Fluent, these relations within the acceptable range. For this case, based on the nearest wall
are defined in two ways as follows: distance (y) from the battery cell, the y+ value is considered as 1.0. A
grid independence study has been conducted for the base model with a
Fifth order polynomial form: different number of mesh elements by observing the cell5 temperature at
1C discharge rate as shown in Fig. 2 (b). From the temperature profile
Rs = a0 + a1 (SOC) + a2 (SOC)2 + a3 (SOC)3 + a4 (SOC)4 + a5 (SOC)5 (14)
analysis, it has been identified that the temperature does not vary much
after the medium mesh refinement unless with increased computational
R1 = b0 + b1 (SOC) + b2 (SOC)2 + b3 (SOC)3 + b4 (SOC)4 + b5 (SOC)5 (15)
time and cost. Hence, the complete numerical analysis for the base and
new models are performed with medium mesh elements.
C1 = c0 + c1 (SOC) + c2 (SOC)2 + c3 (SOC)3 + c4 (SOC)4 + c5 (SOC)5 (16)

R2 = d0 + d1 (SOC) + d2 (SOC)2 + d3 (SOC)3 + d4 (SOC)4 + d5 (SOC)5 (17) 4.1. Experimental validation of the numerical model

C2 = e0 + e1 (SOC) + e2 (SOC)2 + e3 (SOC)3 + e4 (SOC)4 + e5 (SOC)5 (18) The validity of the current numerical model has been verified by
solving the experimental test conditions available in the open literature.
VOCV = f0 + f1 (SOC) + f2 (SOC)2 + f3 (SOC)3 + f4 (SOC)4 + f5 (SOC)5 (19) Wu's [32] conducted experimental research for a prismatic cell at a 5C
discharge rate and observed the cell temperature concerning discharge
Proposed functions based on the Chen [31] model:
time. The same work has been solved with the present CFD model with
Rs = a0 + a1 exp[ − a2 (SOC) ] (20) the same experimental conditions like battery properties, discharge
condition, discharge time etc., predicted the surface temperature of the
R1 = b0 + b1 exp[ − b2 (SOC) ] (21) cell and compared the same with the experimental results as shown in
Fig. 2. c). The comparison results found that the present CFD model
C1 = c0 + c1 exp[ − c2 (SOC) ] (22) results are in good matching with the experimental with a minimum and
maximum error of 0.1 K and 0.32 K, respectively. Therefore, this model
R2 = d0 + d1 exp[ − d2 (SOC) ] (23) has been implemented for this research work and analyzed the flow field
and battery thermal distribution.
C2 = e0 + e1 exp[ − e2 (SOC) ] (24)
5. Results and discussion
VOCV = f0 + f1 (SOC) + f2 (SOC)2 + f3 (SOC)3 + f4 exp[ − f5 (SOC) ] (25)
The source terms for the electrical potential governing equations are Numerical simulation has been performed for the base and new
derived from the following expressions: model battery modules with the ECM battery sub-models at different
discharge conditions. The numerical results are analyzed by evaluating
jECh = I/Vol (26)
the temperature contours and profiles for individual cells and the bat­
[ ] tery module. For easy observation of heat generation vs. airflow stream
I dU
q̇ECh = VOCV − (φ+ − φ− ) − T (27) effect on battery cooling, a horizontal plane has been designed on XY
Vol dT
plane at cell middle height. The cell temperatures at different discharge
Proper boundary conditions and initial conditions play a significant conditions are shown in Fig. 3. From observing temperature contours for
role in solving the computational fluid dynamics problem. Hence, the the base and new models at a 1C rate, the temperature has been
battery module airflow computational domain inlet and outlet bound­ distributed uniformly with significantly less temperature gradient
aries are defined with the mass flow inlet and pressure outlet boundary among the battery module for the new model. If we observe the tem­
conditions, respectively. The interface between the battery cell, cell perature of cells in the flow direction, the base model has recorded the
holder and the surrounding flow area is defined with temperature con­ highest temperature difference of (Tcell4-Tcell6) 0.26 ◦ C, whereas the new
tinuity boundary conditions. The remaining surrounding walls are model with cell holders acting as fins has recorded 0.04 ◦ C, which is a
assigned with adiabatic boundary wall conditions. As the simulations significantly less temperature gradient for the battery module. Therefore
are unsteady, initial battery and flow domain conditions are assigned even if extrapolated for the whole battery pack, it will be less than
with the environmental temperature at the beginning of the solution. desired temperature limits (<5 ◦ C as per standards, but for our model, it
The airflow rate into the battery module is calculated from the local will be <2 ◦ C).
With an increase in discharge current, the heat generation has been

4
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 2. a) Mesh generation and b) Grid independence study and c) Validation [32].

increased, and the same has been observed with the temperature con­ is observed that comparatively among three cells, cell 5 consists of low
tours, as shown in Fig. 3. At 1C rate, the maximum temperatures of temperature due to better conduction and convection of heat transfer.
301.5 and 300.6 K have been recorded for the base and new models, There is no considerable variation in the temperature gradient as the
respectively. With further increase in discharge current, it is found that discharge current increases unless increased maximum temperature.
at 2C rates 303.9 and 302.1 K and with 3C rates, these are 308.8 and The analysis of these temperature profiles among the multi cells reveals
304.9 K respectively for the base and new models. that the temperature profiles for new models are uniform, i.e., negligible
Two different straight lines are considered; one is horizontal along temperature gradient due to better heat distribution among the cells and
the X-axis, which covers cells 4, 5, and 6, and another straight-line along increased conduction and convection with cell holders acted as baffles
the Y-axis covers cells 2, 5, and 8. These straight lines are considered and fins.
strictly at cell half of the height and used to export the post-processing For efficient analysis of cell holders' impact on better cooling, the
data. The impact of airflow distribution on thermal cooling for the individual cell behavior has been analyzed by observing the cell's core
base model and the new model has been further assessed by observing temperature along its height, as shown in Fig. 5. Initially, cells 6, 5, and
the temperature profiles for the considered straight lines along with the 4, which are in-line with the flow direction, have been monitored, as
cells, as shown in Fig. 4. shown in Fig. 5. As per the given cell arrangement and its numbering,
From the analysis of temperature profiles for cells 4, 5, and 6, it is initially, cell 6 was exposed to the fresh air and then cell 5 and lastly cell
observed that the temperature profile (Fig. 4) increases gradually in the 4 as shown in Fig. 1. Hence, first cell 6 has been plotted and observed
flow direction for the base model, whereas for the new model with fins, that at 1C rate, the temperature is uniform along with its height for the
it is constant along the flow direction (X-axis). As the discharge current new model, whereas slightly increased temperature has been observed
increases, the temperature gradient (dT/dx) also increases in the base at the middle of the cell for the base model.
model's flow direction, whereas it is very less almost negligible for the As the discharge current increased to 2C and 3C rate, the base model
new model. The temperature profiles for another straight-line covering cell temperature and its slope and height also increased due to less heat
cell 2, 5, and 6 reveal that the temperature profiles are uniform for all dissipation. Whereas with new models, it is observed that the cell tem­
cells but recorded higher temperature with the base model. And also, it perature is uniform throughout its height, and with increased discharge

5
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 3. Base model and new model battery modules temperature contours on XY plane at the middle height of the cell in the z-direction.

rate, there is no considerable increase in temperature gradient compared same flow conditions of the air. The temperature profiles show that cells
to the base models. Cell 5 is located in the middle of the battery module, 2 and 8 have similar temperature profiles because of their geometrical
as shown in Fig. 1. Its temperature profiles are predicted along with its similarity and are exposed to similar airflow conditions. Whereas cell 5
height at different discharge currents, as shown in Fig. 5. These profiles registers slightly fewer temperatures as its exchanges, its heat with
are almost similar to that of cell 6 except for an increased overall tem­ surrounding cells in all the directions.
perature as it is subjected to the downstream air of cell6. Similarly, cell 4 As the simulations are performed with transient conditions, the
is also subjected to already used air; hence cell4 temperature profiles temperature development in the cells with discharge time has been
consist of higher temperature gradients at high discharge currents. monitored for almost one hour, as shown in Fig. 6. The transient tem­
Similarly, the row of cells 2, 5, and 8 temperatures are also predicted perature profiles for each cell of the battery module reveal that the cells'
at different discharge conditions, as shown in Fig. 5. These three cells are temperature increases with an increase in discharge rate and time. This
arranged on a single straight-line along the Y-axis at the same distance is due to continuous heat generation and accumulation in the battery
from the air inlet position such that all the cells are experienced with the itself. Hence, the cooling of batteries plays a significant role in the

6
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 4. Temperature profiles for a) a horizontal line along 4, 5, 6 cells b) a vertical line along 2, 5, and 8 cells.

efficient operation battery pack. From the comparison of temperature cell holders for the new model. This leads to a decrease in the maximum
profiles for the base and new models at a 1C rate, it is found that the base temperature of the cells and temperature gradient among the battery
model has registered the highest temperature of 301 K and the tem­ module and maintains within the desirable range.
perature gradient also increases with increases in discharge time. For a
new model at a similar discharge rate (1C), the maximum temperature is 6. Conclusion
300.22 K, and the temperature gradient among the different cells is
negligible. In this research article, a novel technique has been introduced to cool
Similarly, the base and new models are studied at 2C and 3C the battery module and maintain the whole battery pack with a uniform
discharge rates and predicted temperature profiles, as shown in Fig. 6. temperature. Initially, a base model battery module was designed with
The maximum temperatures of 304 K and 309 K are registered for the nine cells, and after concerning this model, a new model was designed
base model at 2C and 3C discharge conditions, respectively. Whereas for by considering baffle cell holders that act like fins and heat conduction
new models, these are 300.9 and 302 K, these temperatures are smaller media. The primary intention of this research is to reduce the battery
compared to the base model due to the presence of baffles, which also operating temperature and temperature gradient among the battery
act like fins and increase the heat dissipation from the battery cells. It is cells. Hence, two models are analyzed at different discharge conditions
also identified that the temperature gradient among different cells in­ with similar boundary conditions. Temperatures of both models are
creases for the base models with an increase in discharge rate and time. analyzed at different discharge conditions using plane contours and
In contrast, the temperature gradient is negligible for the new model due temperature profiles for each cell of the battery module. The tempera­
to the uniform temperature among the battery module with baffle cell ture contours for both models represent that the base model has recor­
holders. ded the highest cell temperature and temperature gradient. The
The voltage and depth of discharge profiles are observed for the base numerical analysis finds the following quantitative conclusions:
and new models at different discharge conditions and plotted in Fig. 7.
The analysis of voltage profiles shows that the voltage drop increases • At 1C discharge rate, the temperature gradient among the cells
with an increase in discharge rate and time and found no variation for (Tcell4-Tcell6) in the flow directions was found to be 0.26 ◦ C for the
base and new models voltage profiles. Depth of discharge profiles is standard battery model and 0.04 ◦ C for the innovative battery model.
found that with the increase in discharge rate and time, the depth of • Under various discharge settings (1C, 2C, and 3C rate), the base
discharge also increases but no variation between the base and new model's maximum temperatures are 301.5 K, 303.9 K, and 308.8 K,
models. but the new battery model's maximum temperatures are 300.6 K,
The most critical parameter during battery discharge conditions is 302.1 K, and 304.9 K.
the heat generation source; hence, this parameter has been monitored • Comparatively, under the same discharge condition of 1C, the basic
for the two models at different discharge conditions, as shown in Fig. 8. model (cell 9) reached a maximum temperature of 301 K in 66 min.
The analysis of profiles shows that the heat generation source is signif­ However, the highest temperature reached by the new battery type
icantly less for 1C and 2C discharge conditions than for the 3C discharge (cell 9) at the same period and 1C rate was only 300.2 K. Identical
condition. In addition, the heat generation rate is higher for the first 500 performance is seen at both 2C and 3C rate.
s of the discharge time, and after that, it is uniform as the cooling per­
formance activates the heat dissipation. Comparatively, a little higher CRediT authorship contribution statement
heat generation is observed for the base model with increases in
discharge rate (Fig. 8) due to poor heat dissipation and the cells' accu­ The author performs the following tasks:
mulation of generated heat energy. Even though heat generation is I am the sole author of this manuscript.
similar in the base and new models, its storage energy is less due to Obula Reddy Kummitha: Problem statement, Introduction and
increased heat dissipation to the surrounding fluid due to the presence of literature, Methodology, Numerical Modeling, and Simulations, Data

7
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 5. Temperature profiles along with cell height for the middle horizontal three cells (cell6, cell5, and cell4) and the middle vertical three cells (cell2, cell5,
and cell8).

8
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 6. Transient temperature profiles at different discharge conditions.

9
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

Fig. 7. Voltage and depth of discharge for whole battery pack at different discharge conditions.

[5] S. Goutam, J.M. Timmermans, N. Omar, P. Van den Bossche, J. Van Mierlo,
Comparative study of surface temperature behavior of commercial Li-ion pouch
cells of different chemistries and capacities by infrared thermography, Energies 8
(2015) 8175–8192.
[6] I. Buchmann, Batteries in a Portable World: A Handbook on Rechargeable Batteries
for Non-engineers, 3rd ed., Cadex Electronic Inc, Canada, 2011.
[7] I. Naik, M. Nandgaonkar, Review of the approaches and modeling methodology for
lithium-ion battery thermal management systems for electric vehicles, in:
C. Pandey, V. Goyat, S. Goel (Eds.), Advances in Material and Mechanical
Engineering, Select Proceedings of ICFTMME 2020, Springer Nature Singapore Pte
Ltd., Singapore, 2021, pp. 75–112.
[8] C. Roe, X. Feng, G. White, R. Li, H. Wang, X. Rui, C. Li, F. Zhang, V. Null, M. Parkes,
Y. Patel, Y. Wang, H. Wang, M. Ouyang, G. Offer, B. Wu, Immersion cooling for
lithium-ion batteries – a review, J. Power Sources 525 (2022), 231094, https://doi.
org/10.1016/j.jpowsour.2022.231094.
[9] 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.
[10] J.P. Rugh, A. Pesaran, K. Smith, Electric Vehicle Battery Thermal Issues and
Thermal Management Techniques, Alternative Refrigerant and System Efficiency
Symposium, SAE, Arizona USA, 2011. September 27–29.
[11] 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.
[12] A.A. Pesaran, Battery thermal models for hybrid vehicle simulations, J. Power
Sources 110 (2002) 377–382.
[13] J. Jaguemont, N. Omar, M. Abdel-Monem, P. Van den Bossche, J. Van Mierlo, Fast-
Fig. 8. Heat generation source at different discharge conditions. charging investigation on high-power and high-energy density pouch cells with 3D-
thermal model development, Appl. Therm. Eng. 128 (2018) 1282–1296.
[14] Q. Wang, B. Jiang, B. Li, Y. Yan, A critical review of thermal management models
analysis, and Discussions. and solutions of lithium-ion batteries for the development of pure electric vehicles,
Renew. Sust. Energ. Rev. 64 (2016) 106–128.
Declaration of competing interest [15] J. Jaguemont, L. Boulon, Y. Dubé, A comprehensive review of lithium-ion batteries
used in hybrid and electric vehicles at cold temperatures, Appl. Energy 164 (2016)
99–114.
We wish to confirm that there are no known conflicts of interest [16] J. De Hoog, J. Jaguemont, M. Abdel-Monem, P. Van Den Bossche, J. Van Mierlo,
associated with this publication. N. Omar, Combining an electrothermal and impedance aging model to investigate
thermal degradation caused by fast charging, Energies 11 (2018) 804.
[17] A. Hammami, N. Raymond, M. Armand, Lithium-ion batteries: runaway risk of
Data availability forming toxic compounds, Nature 424 (2003) 635–636.
[18] J. Kim, J. Oh, H. Lee, Review on battery thermal management system for electric
vehicles, Appl. Therm. Eng. 149 (2019) 192–212.
No data was used for the research described in the article.
[19] H. Cai, C. Xu, Y. Liao, L. Su, Z. Weng, Mass maldistribution research of different
internal flowing channels in the cooling plate applied to electric vehicle batteries,
References Appl. Sci. 9 (2019) 636.
[20] X. Han, L. Lu, Y. Zheng, X. Feng, Z. Li, J. Li, M. Ouyang, A review on the key issues
[1] L. Spitthoff, P.R. Shearing, O.S. Burheim, Temperature, ageing and thermal of the lithium ion battery degradation among the whole life cycle, eTransportation
Management of Lithium-ion Batteries, Energies 14 (2021) 1248. 1 (2019), 100005.
[2] T.M. Bandhauer, S. Garimella, T.F. Fuller, A critical review of thermal issues in [21] F. Larsson, B.-E. Mellander, Abuse by external heating, overcharge and short
lithium-ion batteries, J. Electrochem. Soc. 15 (2011) 1–25. circuiting of commercial lithium-ion battery cells, J. Electrochem. Soc. 161 (2014)
[3] L. Ianniciello, P.H. Biwolé, P. Achard, Electric vehicles batteries thermal A1611–A1617.
management systems employing phase change materials, J. Power Sources 378 [22] X. Feng, S. Zheng, D. Ren, X. He, L. Wang, H. Cui, X. Liu, C. Jin, F. Zhang, C. Xu, et
(2018) 383–403. al., Investigating the thermal runaway mechanisms of lithium-ion batteries based
[4] S. Ma, M. Jiang, P. Tao, C. Song, J. Wu, J. Wang, T. Deng, W. Shang, Temperature on thermal analysis database, Appl. Energy 246 (2019) 53–64.
effect and thermal impact in lithium-ion batteries: a review, Prog. Nat. Sci. Mater. [23] L. Lu, X. Han, J. Li, J. Hua, M. Ouyang, A review on the key issues for lithium-ion
Int. 28 (2018) 653–666. battery management in electric vehicles, J. Power Sources 226 (2013) 272–288.

10
O.R. Kummitha Journal of Energy Storage 59 (2023) 106474

[24] U. Iraola, I. Aizpuru, I. Gorrotxategi, J.M.C. Segade, A.E. Larrazabal, I. Gil, Battery Module Used for EVs, in: SAE Technical Paper 2022-01-0711, 2022,
Influence of voltage balancing on the temperature distribution of a li-ion battery https://doi.org/10.4271/2022-01-0711.
module, IEEE Trans. Energy Convers. 30 (2015) 507–514. [30] Y. Liang, A. Emadi, O. Gross, C. Vidal, et al., A comparative study between physics,
[25] X. Feng, C. Xu, X. He, L. Wang, G. Zhang, M. Ouyang, Mechanisms for the evolution electrical and data driven lithium-ion battery voltage modeling approaches, in:
of cell variations within a LiNixCoyMnzO2/ graphite lithium-ion battery pack SAE Technical Paper 2022-01-0700, 2022, https://doi.org/10.4271/2022-01-
caused by temperature non-uniformity, J. Clean. Prod. 205 (2018) 447–462. 0700.
[26] A. Olabi, H.M. Maghrabie, O.H.K. Adhari, E.T. Sayed, B.A. Yousef, T. Salameh, [31] M. Chen, G.A. Rincon-Mora, Accurate electrical battery model capable of
M. Kamil, M.A. Abdelkareem, Battery thermal management systems: recent predicting runtime and I-V performance, IEEE Trans. On Energy Conversion 21 (2)
progress and challenges, Int. J. Thermofluids 15 (2022), 100171, https://doi.org/ (2006) AI54–AI61.
10.1016/j.ijft.2022.100171. [32] W. Wu, W. Wu, S. Wang, Thermal management optimization of a prismatic battery
[27] A.R. Bais, D.G. Subhedhar, N.C. Joshi et al., Numerical investigation on thermal with shape-stabilized phase change material, Int. J. Heat Mass Transf. 121 (2018)
management system for lithium-ion battery using phase change material, Materials 967–977.
Today: Proceedings, 10.1016/j.matpr.2022.05.269. [33] Kai Chen, Yiming Chen, Zeyu Li, Fang Yuan, Shuangfeng Wang, Design of the cell
[28] J. Jaidi, S.D. Chitta, C. Akkaldevi, S. Panchal, M. Fowler, R. Fraser, Performance spacings of battery pack in parallel air-cooled battery thermal management system,
study on the effect of coolant inlet conditions for a 20 ah LiFePO4 prismatic battery Int. J. Heat Mass Transf. 127 (2018) 393–401.
with commercial Mini Channel cold plates, Electrochem 3 (2022) 259–275, [34] K. Yu, X. Yang, Y. Cheng, C. Li, Thermal analysis and two-directional air flow
https://doi.org/10.3390/electrochem3020018. thermal management for lithium-ion battery pack, J. Power Sources 270 (2014)
[29] Z. Zhao, S. Panchal, P. Kollmeyer, A. Emadi, et al., 3D FEA Thermal Modeling with 193–200, https://doi.org/10.1016/j.jpowsour.2014.07.086.
Experimentally Measured Loss Gradient of Large Format Ultra-fast Charging

11