World Electric Vehicle Journal Vol.

6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0175

EVS27
Barcelona, Spain, November 17-20, 2013

Modeling and thermal simulation of a PHEV battery

module with cylindrical LFP cells
Cicconi P.1, Germani M., Landi D.
1
Università Politecnica delle Marche, via Brecce Bianche, 60131 – Ancona – Italy, p.cicconi@univpm.it

Abstract
Generally a part of electric vehicle diffusion is still based on marketing of cars and vans suitable for
specific use like work vehicles. A flexible design methodology is required to support rapid prototyping and
product customization in the market of tailored EV/PHEV. The research focuses the cooling simulation for
a PHEV Li-Ion battery. The thermal analysis is based on the physical parameters of the single cell and on
the experimental data. The proposed methodology concerns firstly an analytical approach which evaluates
the average heat generated by a single cell during working condition. Then the proposed virtual prototyping
analysis has been divided into two levels: the thermal simulation of one cell, and the CFD analysis of a
battery module. This workflow has been applied to support the design of a battery pack for a prototypal
ecological hybrid vehicle. That test case vehicle is a small van, used for the curbside collection, which has
in parallel an internal combustion engine and an electric motor supplied by a LFP battery with small
cylindrical cells. The analysis concerns one of the four module which constitutes the whole battery pack.
The virtual model has been parameterized and the behavior of air cooling system has been evaluated
through virtual tools.

Keywords: lithium battery, cooling, modeling, simulation, PHEV

testing are essential phases to save time and avoid

1 Introduction many prototyping iterations.
Generally, the Virtual Prototyping (VP) techniques
Nowadays, the development and use of electric are widespread in engineering design. These kinds
vehicles are strongly encouraged to reduce the of solutions, which are suitable for large
environmental impact of transport issues. Apart companies, have to be customized for use in
from large companies, which tend to develop producing small lots [2]. In particular, in the
general purpose solutions, small and medium context of PHEV, these tools are used to analyze
sized companies aim to produce customized lots different issues, such as the co-simulation of the
of vehicles for specific customer needs (i.e. real driving cycle, the computational fluid dynamic
municipalities). In this case, the design of (CFD) study of the cooling effect, and the
vehicles and sub-systems has to be rapid and computing of the electrochemical formulas to
sufficiently flexible in order to analyze the estimate the voltage curves and the heat generated
specific applicative context and translate it into a at different current rates. A particular limit is
successful solution [1]. Simulation and early present in the battery cooling analysis where
commercial tools provide an electrochemical

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 1
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0176

characterization which requires a deep proposed research activities is focalized on

knowledge of the many parameters related to the customized production of a typical European SME
cell materials. The issue concerns the lack of (Small Medium Enterprise).
knowledge related to the compositions and The traditional design approach, commonly used
dimensions of the internal cell material layers. in many SMEs, is mostly based on the experience
The proposed approach is outlined in of engineers. So that designers use basic
collaboration between an Italian company which calculation tools to analyse and elaborate new
elaborates and markets medium-small ecological solutions. The engineer calculates the electrical
vehicles powered by natural gas, electricity or layout configuration, determines the geometrical
hybrid powertrain (gasoline and electric in shapes and defines the cooling system size. Then,
parallel mode). This small-medium enterprise the corresponding physical prototype is
works in the industry of customized items, where manufactured and physical tests give him/her the
the lead motive is the reducing of time and cost. performance quality. Possible design errors
The paradigm is the just in time (JIT) production introduce expensive iterations, thus time to market
which requires flexible and agile tools to support gets longer. The main activities are repetitive tasks
the designing and manufacturing processes. In and designers are limited to use own experience on
the design of customized vehicle is very past project data (such as empirical tables,
important to get smart design tools and methods drawings, reports and feedbacks). The main
to support engineers in their tasks. parameters in battery designing are: the operative
Next sections explain our research work on the conditions, the single cell type, the cell number,
thermal issue for the automotive battery pack. the battery layout, and finally the evaluation of
The proposed methodology provides an heat released by the electrochemical reactions.
analytical valuation of heat generated by each Li- According to the temperature problems in Li-ion
ion cell and the integration of this calculation in cells, a battery pack needs successful thermal
the virtual prototyping analysis based on FVM energy management and a cooling system to favor
(Finite Volume Methods) tools. A test case has heat dissipation and thermal runaway. The two
been exposed to explain the defined workflow in required aspects for an automotive lithium ion
the cooling design for PHEV battery pack. The battery are: an optimum operating temperature
cells used in the proposed case are LFP cylinders range and small temperature variations. Some
suitable for an hybrid application. researchers have analyzed methods to optimize the
geometrical and fluid dynamic parameters in
2 Background research on cooling design [4].
The proposed research aims to define a
battery design methodology suitable to evaluate different kinds of
Every rechargeable lithium ion (Li-ion) battery is cooling systems. The objective is to develop a
constituted by a packing of cells (also known as workflow for battery pack thermo-fluid simulation,
battery pack) so that the final electric power is where the electrochemical heat source is evaluated
the sum of all single elements [3]. by an analytical calculation. The main difference
The process of Li-ion battery design concerns with the state of the art is the development of a
modeling and simulation regarding whole battery spreadsheet for rapidly computing thermal power
pack. In industry, the actual strategies for battery values at different rates of charge and discharge.
pack design are mainly related to the market size. The second difference is the application of the
The market of large-scale production requires analytical thermal power values, previous
reliable batteries for standard applications. calculated, into the VP tools for the cooling fluid
Conversely small productions, with a little dynamic analysis.
volume of customized units, focuses the
satisfaction of particular customer specifications 2.1 Thermal issue
coupled to a reduced time-to-market. On the one
In Li-ion battery cells the temperature influences:
hand, large battery manufacturers can use
electrochemical reactions, round trip efficiency,
expensive design tools and subdivide additional
charge acceptance, power and energy capability,
costs on a large number of sales. On the other
reliability, life cycle and cost [5]. Furthermore, in
hand, small and medium enterprises have to
electric vehicles, battery temperature increases
increase the own knowledge and formalize the
during discharge with a huge peak for high current
rules of expert designers in low-cost customized
rates.
tools, in order to contain excessive costs. The

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 2
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0177

Generally an excessive temperature degrades literature several researchers have investigated

performance and a limited thermal dissipation only the electrochemical behavior. However those,
can even produce burning in Li-ion battery cells. who have analyzed the thermal analysis, have also
Great importance is also given to battery safety investigated the electrical model characterization.
concerning heat accumulation due to the lithium Below an explanation regarding recent works on
flammability and explosiveness. The temperature battery simulation.
raises two very important issues in Li-ion The electrochemical simulations on Li-Ion
batteries: safety risks and material aging. The batteries concern the evaluation of voltage curves
safety risks are lead to high temperature values and state of charge (SOC) at different current rates
during discharge over the permitted level. The in charging and discharging [8]. Often, in
material aging are related to the operative cycles electrochemical simulations, the input current
and to the achieved temperature levels during profile wants really reproduce the effective driving
working conditions. Moreover, in Li-ion battery cycles load for electric vehicles. This type of
pack is also very important the temperature simulation supports the engineer in modeling of
distribution. A non-uniformity in temperature of tailored electrical layout configuration for batteries
a cell pack favors localized deterioration in This type of simulations supports the engineer in
single battery elements, this introduce efficiency modeling of tailored electrical layout for batteries.
losses (as ohmic resistance) and lifespan Generally equivalent circuit describes the electrical
reduction in whole pack [6]. behavior of the cell system as an approximation of
The thermal model characterization for a Li-ion non-linear real phenomena. Therefore
cell is important to predict the average conventional linear circuit [9] allows the engineer
temperature at different current rate in the early to predict a reliable voltage over poles with simple
design phase. The explained thermal analysis is component like an RC filter (resistor-capacitor)
based on a study by Thomas and Newman [7] and an ohmic resistance. The calculation of real
which evaluates the heat produced by the single voltage curves gives an important feedback
cell. In particular, a simplified form of Newman regarding the battery performance in terms of
formula (1) has been considered in an analytical energy and power analysis. In addition, the SOC
approach to estimate the heat source at different valuation is an important feature to predict the
current rate for each lithium battery cell. battery autonomy [5].
E On the other hand, the thermal simulation regards
Q  I (V  E0 )  IT 0 (1) the heat generated by the electrochemical reactions
T and also analyzes the cooling performance. Many
In Eq. (1) the term I indicates the instant current simulation techniques have been involved in
value, the term V indicates the related electrical literature. A development of Genetic Algorithm
potential (voltage measured at the cell poles), has been extended to optimizing the NTU
while E0 is the relative open circuit voltage, and (Number of Transfer Unit) model in lithium cells
for heat exchange study [10]. This approach
T indicates the temperature value.
proposes good results, but the application is often
customized to specific cooling system, as i.e. in
2.2 Battery simulation [10] where the heat exchange is encouraged by
Several research works have analyzed the issue tubes with cool air. Using a Genetic Algorithm
of simulation for lithium-ion battery packs. That approach is possible to achieve good results but it
mainly concerns the electrochemical and thermal is not possible to define a routine common for
cell characterization. Particularly the electrical different battery types and models.
aspect of lithium-ion cells regards the property to There are several types of thermal simulation
generate a potential difference between negative models based on: lamped element model, finite
and positive pole (voltage), while the chemical element analysis (FEA) and computational fluid
aspect concerns the capacity to move quantity of dynamics (CFD) analysis with finite volume
electrons and ions through each battery sub- methods (FVM) [11]. Generally the lamped
layers. The thermal behavior of Li-ion cell is element model approach provides simple thermal
related to the heat dissipated from models in lithium battery cell [12], but this
electrochemical reactions inside single battery requires the definition of several parameters
element. through experimental measurements or technical
Thus electrochemical and thermal analysis are datasheet. These technical specifications on cell
coupled features in simulation virtual analysis. In sub-layers are often not provided by manufacturer,

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 3
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0178

so it is fundamental the planning of useful obtain the thermal power levels as a curve over
experiments to achieve the necessary data. time. The level of Pack Layout Configuration is
Actually there are reliable commercial tools for connected to the CAD system tool and to a
battery modeling, but cost and complexity of collection of template models. This third level
these tools limits their diffusion. So that several supports the engineer in the geometrical layout
researchers have analyzed customized tools for definition and in the assembly of all battery cell
specific cells and battery pack layout. However, arrays in a 3D model geometry. Generally,
rapid tools are necessary to support engineer for development of KBE applications involves a
evaluating thermal battery behavior inside the reduced time of product configuration phase, aids
European SMEs that are working in automotive in decision-making activities and automates
industry. This tools have to provide a common repetitive procedures. In this paper has been
workflow design for several lithium-ion battery explained the proposed platform architecture, but
cases. the KBE tools has not been very detailed in order
to give more relevance to the thermal analysis and
3 The framework methodology virtual prototyping in battery modeling.
For the proposed research it has been formalized
a modeling workflow to support the engineer
during the early design phase. The research aim
is not only battery modeling, but also the
reduction of lead time into a typical SME. The
target is the design of cooling layout for a battery
pack in an automotive powertrain. As cited
before, a reproducible methodology is very
accepted to support the design of customizable
products in SMEs.
In particular Figure 1 describes the main
framework platform which has been followed for
the test case proposed in next sections. The
scheme presents four different modules: HD
tools, KBE, DB and VP tools. The first module
regards the hardware tests of a single cell at
electrical test bench for data gathering. The
voltage curves and the open circuit voltage Figure 1: The scheme of proposed methodology
behavior are the main data which are analyzed
during the cell electrical test. The proposed VP tools module concerns the
The KBE module represents the knowledge virtual prototyping simulations for battery
based engineering tools which support the modeling, and also the stand alone module of CAD
designer in his tasks. Specifically the KBE system. Particularly the virtual prototyping
application is a special type of knowledge based analysis regards two levels: the Cell FVM thermal
system which has an important focus on product analysis, and the Pack FVM Thermal Analysis
engineering design and downstream activities (where FVM means Finite Volume Methods). The
such as analysis, manufacturing, cost estimation first level investigates the heat distribution in the
and even sales. But in our approach it has been virtual model for a single Li-ion cell. The
analyzed only the phase regarding the product boundary condition is given by the input thermal
design and performance analysis. The KBE tool power profile calculated in the analytical thermal
is composed by the Test Configuration form analysis using the KBE tool. While the second
where the designer defines the electrical test for level concerns the virtual simulation of thermal
each type of cell. This level is connected to a and cooling behavior in complete battery pack
database of models which can be selected by the model. In particular this type of simulation regards
designer through a filtering scheme. Next level in two aspects: the CFD analysis of fluid cooling
KBE tools is represented by the First Analytical flow, and the thermal analysis of each elementary
Thermal Analysis, which is constituted by a cell in order to investigate the temperature
parametrical spreadsheet for evaluating heat distribution. For lithium-ion batteries is
source. Using this tool, the designer can recommended a constant working temperature
introduce an electrical current profile and then with a maximum difference of 2-3°C between

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 4
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0179

minus and maximum values. Thus the cooling presents a single 25 kW drive electric motor with a
validation concerns also the verification of 7.55 kW lithium-ion battery pack; while the
temperature distribution, in order to guarantee an internal combustion engine has 4 cylinders and a
uniformity in heat exchange. Using CFD tools it 1.2 L in capacity. The battery size focuses the
is possible to set parameters in virtual model hybrid aspect of the prototype vehicle. This vehicle
such as cooling temperature, and fluid flow can work in electric mode during the short paths
velocity, while the geometrical parametric around buildings, and use the internal combustion
variation is related to the CAD model. engine crossing the city districts. The max vehicle
The connection between each platform module velocity is 70 km/h if gasoline fueled, while 50
are symbolized in Figure 1, where dashed line km/h if electric powered.
means that the design workflow can turn back in In details, the prototypal battery model provides
presence of failure settings. The platform 212 LFP 10 Ah cylindrical cells (Table 1),
provides the First Analytical Thermal Analysis separated in 4 module of 59 cell each one. The use
tool to early valuate the thermal behavior of one of a low capacity cell is explainable by the
cell. This analytical tool carries out a thermal technical characteristics suitable for a prototype
characterization based on previous experimental hybrid vehicle. The selected cell has a different
tests (HD tools). The heat source values are application range due to its cylindrical geometry,
calculated by an analytical formula and depend which is suitable for packaging, and due to the
on the cell’s physical parameters, current rates high discharge current rate of 3C in continuous and
and time period. This early analysis mainly 10C in peak. Thus, this Li-ion cell type is able to
considers a one battery cell working in natural be adopted in hybrid powertrain application, where
convection in order to reproduce the test bench battery provides support in starting, parking and
conditions in virtual environment. But it is also driving for short paths. It is not required an
possible to calculate the cell heat exchange in important fixed kilometers range for hybrid
forced convection by analytical formulation. vehicle. Farther, the proposed application requires
Input data for the whole proposed workflow is slow velocity in main working condition for
constitute by the cell details such as type, curbside collection. Besides, the chosen cells can
electrical capacity, energy and power density, be rapidly charged at least 2C rate, instead of the
heat capacity, thermal conductivity. Other input usual limit of 0.5C or 1C.
is the electric layout of battery pack: cell count, Table 1 Technical data sheet of LFP cell 10 Ah
electrical connection, etc.
Output data are the geometrical battery layout, a Chemistry LFP (LiFePO4)
simplified battery model, a maps of a Nominal Voltage 3.2 V
temperature distribution, and the definition of Geometry Cylinder
cooling parameters. Nominal Capacity 10 Ah
Max Discharge 3C (30 Ah)
4 Battery modeling for PHEV Max Charge 2C (20 Ah)
An hybrid application has been chosen to explain Weight 330 g
the proposed methodological approach. In
particular a prototype Li-ion battery module for a Each battery module is air-cooled by two compact
PHEV has been analyzed through virtual tools. fan wheels, and the geometry layout has been
The related vehicle, which is designed for the defined using virtual tools and physical
European market, is a prototype van for curbside experiments on the selected cells.
collection. This special vehicle is configurable on Next section explains the thermal model
the customer’s requirement and has an hybrid characterization for one cell using the analytical
propulsion with gasoline engine and electric analysis of heat source. The design methodology
motor, which are coupled in mild-parallel mode has been applied to the modeling and simulation of
with regenerative braking. The internal proposed Li-ion battery pack. Then, the virtual
combustion engine is suitable for extra-urban temperature distribution has been analyzed at FVM
distances at constant velocity, while the electric tool. Afterwards, the research work proposes the
powertrain is recommendable for the urban start- geometrical parameterization of battery pack and
and-stop paths at slow velocity. Furthermore, the CFD simulation for a defined model of 59 LFP
using the energy recovery brake is possible to cells.
recharge a part of Li-ion battery during the
journey routes. This prototypal hybrid vehicle

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 5
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0180

4.1 Thermal model Then an experimental test has been conducted,

The proposed thermal model solves Eq. (1) to applying this current profile on one cylinder cell,
evaluate the electrochemical heat source during in order to acquire the voltage curve related to the
charge and discharge. The geometry domain defined driving cycle.
analyzed is the cell cylindrical volume. The aim
of the analytical thermal model is to provide a
tool to aid the designer in the estimation of the
max heat generated during the working
condition. The input is a current profile which
reproduces the electrical load during the standard
operative phases. Instead the required physical
parameters, related to cell type and chemistry,
are recovered from database or gathered from
experimental tests. These physical parameters
are: the OCV (open circuit voltage), indicated as Figure 3 ECE R15 Urban Driving Cycle repeated 4
the term Eo in Eq. (1), and the variation of OCV times (780 s)
in function of temperature (∂Eo/∂T). Test for
evaluating OCV curves concerns discharge steps
of 10% SOC at 1/3 C current rate. While the
study of temperature effect over OCV are
analyzed by the maintaining of cell at controlled
temperature in a climatic chamber. In Figure 2 is
shown the trend of OCV acquired through
experimental tests at several SOC steps.

Figure 4 Graphic of current profile for one cell under the

driving cycle analysed.

Figure 2 OCV trend for selected cell type

The operative condition, chosen for the proposed
Figure 5 Graphic of the analytical calculation of heat
test case, is the thermal load related to one cell source generated by electrochemical reaction for one
during a driving path powered by electrical cell under load condition of four repetitive ECE R15
motor. In particular, as a standard run, it has been driving cycles.
chosen a path constituted by four repeated ECE
R15 Urban Driving Cycles (Figure 3). The The graphic in Figure 5 shows the variation of heat
standardized ECE cycle reproduces the typical generated during the driving cycle, in particular
driving conditions of busy European cities, this profile has been calculated resolving the
characterized by low engine load, and a analytical formula to evaluate the heat source
maximum speed of 50 km/h. This kind of driving released by main electrochemical reactions.
cycle is suitable to reproduce the real operating The follow section reports how the cell thermal
condition introduced for this test case, even if in profile can be reproduced through the numerical
the last years several researches have promoted formulation of the virtual simulation, in order to
alternative cycles more realistic. analyze the maximum average temperature
Figure 4 describes the electrical load on one cell achievable under operative condition. While the
into the prototype battery pack. The current analytical profile of heat generated becomes an
profile related to the selected driving path has input for the simulation analysis.
been calculated considering the electrical pack
configuration and the vehicle dynamic behavior.

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 6
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0181

4.2 Cell simulation compared with the experimental values. This result
The methodological approach includes two levels is acceptable, however it should be considered that
of virtual simulation: the one cell thermal the simulation is limited to a transitory of only 13
analysis, and the CFD simulation of a battery minutes and the heat outlet is no so high during
pack. This section explains and describes the standard ECER15 driving cycles. The max
thermal simulation approach for one cylindrical temperature estimated is almost 21,6°C, while the
LFP cell of 10 Ah in capacity. The target of these real value is 21,3°C and the starting value is 20°C.
simulations is to rapidly valuate the cell During this operative condition a limited low
temperature level using virtual tools. Particularly current discharge does not cause significant
the application case concerns a battery air cooled. increases in temperature, however the problem
The main input, in this valuation, is the heat appears after a continuous loop of several cycles.
generation rate which is time dependent. The The characterization of a virtual model requires
values of thermal power are calculated by the also data such as heat capacity, cell density, and
analytical tool as cited before. the superficial emissivity values. Even if the
The virtual boundary condition reproduces the lithium-ion cells are constituted by several
thermal generation in operative condition for one repetitive sub-layers, the material properties have
cell. At this step it has been analyzed the natural been considered as average values over volume.
convective heat exchange, so that it has been Thus the simulation target is limited to analyze the
possible to reproduce the temperature profile average temperature distribution, in order to
during test condition inside a climatic chamber. simplify the virtual model and to reduce the project
The convective heat transfer coefficient has been lead time. While all the simulation parameters
evaluated by empirical formulations for natural (such as those regarding the turbulence model),
convection. have been validated comparing the experimental
The profile of heat generation, during time, has data with the simulation results.
been set according to the analytical calculation The implementation of virtual prototype methods
described in previous phase. As cited before, it allows the engineer to simulate different load
has been reproduced the thermal load due to a profiles. In this test case the analysis related to the
path of four repeated ECE R15 Urban Driving repetitive ECE R15 cycles has been described, but
Cycles. In Figure 6 the temperature profile these methods can be extended to any customized
elaborated by an thermal FVM analysis in driving cycles.
comparison with the real temperature trend Figure 7 shows the cell temperature profile
during test, at the defined working condition. monitored by IR camera during the load related to
defined driving cycle. While Figure 8 shows a
report of virtual temperature distribution calculated
by FVM solver, as a comparison with previous
image (Figure 7Figure 8).

Figure 6 Comparison between simulated temperature Figure 7 Temperature distribution monitored by IR

(solid red line) and real values monitored by IR- camera at the end of test cycle
Camera (dark dotted line) for one single cell in natural
convection during the defined driving cycles.
The red solid line (Figure 6) describes the
simulated average temperature profile for the
analyzed battery cell (cylindrical LFP 10 Ah)
under the operating condition related to the
defined driving cycle. While the dotted line
represents the real temperature values for one
cell monitored by IR camera at test bench inside
the climatic chamber in natural convection. The
average error for simulation values is almost 5%
Figure 8 Temperature report of cell thermal simulation

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 7
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0182

4.3 Battery pack parameterization an offset arrangement has been chosen with a cross
According to the explained methodology, the air flow due to the difference between inlet and
Battery Pack Layout Configuration concerns the outlet into battery pack (Figure 10 and Figure 11).
definition of geometrical layout. This level
provides 2D sections which are related to a
library of 3D models. Some datasheets have been
implemented to configure typical battery plant
configurations. The engineer can reuse existing
template models during design process or define
a new model structure and add it in the related
database. Each battery template model has to
include the main dimensions such as cell size,
elements distances, battery sizes and cells
patterning. As example Figure 9 reports a
parametric section of a small battery array with
cylindrical elements. In this figure each
dimension is represented by a parameter. The list
of parameters becomes an input for the
generation of simplified 3D model useful for Figure 10 A frontal isometric view of analysed battery
virtual simulation. The aim of the geometrical module with the two inlet sections
modeling is the definition of a battery layout for
cooling optimization. This method gives
flexibility to the workflow platform in order to
satisfy the SME needs.

Figure 11 A rearward isometric view of analysed battery

module with the two outlet sections
Figure 9: Example of parameterization for a
small array with cylindrical cells Following the proposed workflow, at this level the
Generally a Li-ion battery is constituted by engineer can select battery layout and define main
repetitive modules, and the virtual analysis only dimensions. In Figure 10 is shown the geometrical
focus on the elementary cells array. CAD model which has been configured for the test
The layout shape concerns the cell arrangement case using the Pack Layout Configuration tool.
into a battery pack or one repetitive module. A However, the geometrical details are manually
common used layout is that linear, where cells modeled by the engineer which has to review the
are ordered in parallel rows. While the offset generated model. So, the configuration tool is an
layout regards a crossed arrangement of aid support, and this aspect lets to give flexibility
staggered rows (like a zigzag form) as shown in to the virtual analysis process.
Figure 9. This offset structure is mostly used for Next section discusses how the defined
cylindrical elements, while the parallel shape is geometrical model becomes a virtual domain for
common for the soft pouch type cells. the CFD simulation phase.
The cooling flow may be aspirated or
pressurized, but generally is used the aspirated 4.4 Battery module simulation
mode for electronic equipment like battery. The second level of virtual prototyping regards the
The cylindrical cells, used in the proposed test fluid dynamic simulation of a simplified battery
case, are suitable for an offset packaging. Thus model. Thus, at this section the CFD simulation

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 8
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0183

regarding one of four identical battery module

provided for the prototype hybrid vehicle, where
one module contains 59 cells. Different thermo
fluid dynamic parameters have been simulated,
and a final configuration has been defined to
provide a suitable temperature, and also a
thermal uniformity during the working condition.
The chassis of the battery pack is an important
constraint in this test case. The space dimension
is fixed and also the inflow and outflow sections
have already been defined in order to apply this
battery inside a specific vehicle frame.
The simulation domain analyzed concerns the
symmetric part of virtual model. The use of
symmetric feature allows to reduce the number
of tetragonal element, generated during the mesh
Figure 13 The temperature map on the middle section
processing, and to limit also the computational with air cooling at 20°C and 50 m3/h volumetric flow
load of un-steady simulations.
The CFD simulation reproduces the thermal
behavior of a battery module (59 cells) under the
operative condition of driving cycle, in order to
evaluate the performance of air cooling. All the
calculation is based on the analytical evaluation
of the electrochemical heat source. This
condition is an input for the FVM analysis.
Summary, the CFD simulations provide a means
to estimate the temperature distribution in a
battery pack module reproducing the operative
condition in a virtual environmental.
Different geometrical layouts have been
Figure 14 Isometric view of battery module including
analyzed, but only one configuration has been temperature map and velocity streamline for the cooling
reported (Figure 11). The air cooling is case of 50 m3/h
constituted of two compact fans (nominal
diameter 100 mm) that suck air from the battery Figure 12 shows the temperature profile related to
interior and push it out through circular sections. the simulation with a volumetric flow of 100 m3/h
The inlet section is arranged in the opposite side sucked by two compact fans. This simulation
respect to the fan wheels. reproduces the thermal load due to the cycle
described above, constituted of 4 repeated ECE
R15 driving cycles. In particular, the temperature
distribution in Figure 12 regards the middle section
of selected battery module. The starting
temperature analyzed is 20°C and the air cooling
flow is at the same temperature during the
simulation. Using this setup, the maximum cell
temperature, simulated after a driving cycle of 780
sec, is 22.05 °C (Figure 12), while the minimum
value is 20.74 °C (with a gap of 1,31 °C).
On the other hand, in Figure 13 is shown the
simulation report with a different air cooling
condition. The intake air elaborated by compact
fan wheels has a volumetric flow of 50 m3/h . This
second analysis aims to investigate the influence of
air flow rate on the cooling performance, in order
to evaluate the battery cooling with virtual
Figure 12 The temperature map on the middle section
with air cooling at 20°C and 100 m3/h volumetric flow
prototyping tools. The maximum cell temperature,

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 9
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0184

simulated for this second condition, is 22.08 °C prototyping tools, which supports the engineer
(Figure 13), while the minimum value is 21.08 during the thermal analysis and the cooling design
°C (with a gap of 1 °C). Then in Figure 14 is phase. The focus is on the cooling simulation of a
shown an isometric view of the virtual model PHEV Li-ion battery. As a test case, a battery pack
including both the temperature distribution over with 10 Ah LiFePO4 cylindrical cells has been
the cell skins and the velocity streamline. This proposed.
representation is related to the volumetric flow of The battery thermal analysis is based on the
50 m3/h. analytical calculation of heat generated by
Comparing Figure 12 and Figure 13 the electrochemical reactions. This values have been
maximum values of temperature are almost the used to reproduce the behavior of one cell in a
same, but the thermal distribution is different. In FVM analysis, and to compare the temperature
first case, with 100 m3/h volumetric airflow, it is profile between real and virtual model. So, the first
shown a less average temperature. However this level of virtual analysis has regarded the
thermal difference is not really clear for several simulation of one cell, but the second level has
causes: the air velocity increasing introduces introduced the CFD computation of a complete
significant losses through the interstices between battery module which contains 59 cells. The
the cell elements, the cooling temperature is the simulated thermal loads are related to an urban
same for both simulation, and in addition the light vehicle. In particular the heat generated has
analysis refers to a short test period of 13 been calculated by analytical formulation using the
minutes with low current rates during the current profile related to the driving paths
simulated driving cycle. The final temperature constituted by four repeated ECE R15 Urban
achieved during simulation is higher than that Driving Cycles. Thus the FVM analysis have been
one of the real test related to one cell (Figure 6). conducted in unsteady condition.
This difference is less than 1° C, but the VP A geometry configuration has been proposed for
analysis presents a forced convection condition one battery module to obtain a suitable
for air cooling, while the test on one cell has temperature and a cell thermal uniformity during
been conducted under natural convection. So the the working condition. Then two simulation
real temperature achieved by a single cell is less reports have been described with different air
than the value reached inside a battery pack. This cooling flow rate. The results show a good
difference can be explained because in a battery thermal uniformity using the analyzed cooling
the presence of several hot elements, packaged in settings.
a close way, limits the heat dissipation and favors In accordance to the needs of small and medium
the temperature increasing. size companies, the research approach leads to a
Anyway using virtual prototyping analysis the reduction in the cost and the project development
designer is able to valuate several settings and to lead-time. In this way the real pilot prototypes are
investigate the causes of different thermal replaced in part by virtual models.
behaviors on parametric geometrical models. For what concerns the research approach, a future
In this case, the compact fan wheels selected for development can be the introduction of BMS
the described PHEV have been chosen (battery management system) behavior during
comparing the thermo fluid dynamics behavior CFD simulation in order to consider different
at different volumetric rate. The final choice has current profiles for each cell. In this case it will be
been to use a fan wheel type with variable also possible to simulate different thermal
volumetric flow, with 50 m3/h produced in the management algorithms during the design phase.
analyzed conditions.
Generally, the geometrical layout analyzed Acknowledgments
favors the heat exchange between the superficial
cell skins and the air flow under the forced The authors wish to thank FAAM Group S.p.A. for
convection condition. An advantage related to their precious contribution in the development of
this arrangement is the possibility to achieve a this research program; particular acknowledgement
uniform temperature between each battery to Eng. Roberto Isidori.
element.
References
5 Conclusion [1] S. Li, Automatic generation of assembly
system configuration with equipment
In conclusion, the proposed approach shows a selection for automotive battery
design methodology, based on virtual

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 10
World Electric Vehicle Journal Vol. 6 - ISSN 2032-6653 - © 2013 WEVA Page Page 0185

manufacturing, Journal of Manufacturing Authors

System, 30 (2011), 188-195
Dr. Paolo Cicconi,
[2] R. Raffaeli et al., Modular Product PhD. in Mechanical and
Configuration: An Automatic Tool for Management Engineering at PhD
Eliciting Design Knowledge From School of Università Politecnica
Parametric CAD Models, Proceedings of delle Marche; doctor thesis: Tools
the ASME IDETC/CIE 2010, Montreal, and Methods Based on Knowledge
Quebec, Canada, DETC2010-28242 Elicitation to Engineering Design.
[3] C. Sen, Battery Pack Modeling for the He graduated in Mechanical
Analysis of Battery Management System of Engineering and currently works
a Hybrid Electric Vehicle, Vehicle Power research activities regarding
and Propulsion Conference VPPC '09, Knowledge Management at
ISBN 78-1-4244-2600-3, 2009 Industrial Engineering and
Mathematical Sciences Department
[4] S. Li, H. Wang, Y.T. Lin, J. Abell, S.J. of Polytechnic University of
Hu, Benchmarking of high capacity Marche.
battery module/pack design for automatic Email: p.cicconi@univpm.it
assembly system. Proceedings of the 2010
ASME International Manufacturing PhD. Student Daniele Landi
Science and Engineering Conference, is two years into his PhD studies at
2010, p. 34114 the Università Politecnica delle
[5] AA. Pesaran, Battery thermal models for Marche. Currently he is working on
hybrid vehicle simulations, Journal of methods and tools for virtual
Power Sources, 110(2), 377–82, 2002 prototyping in the Industrial
Engineering and Mathematical
[6] Z. Rao, S. Wang, A review of power Sciences Department of the
battery thermal energy management, Polytechnic University of the
Renewable and Sustainable Energy Marche. He graduated in
Reviews, 15 (2011), 4554– 4571 Mechanical Engineering with a
thesis entitled: “Study of methods
[7] K.E. Thomas, J. Newman, Journal of the
for virtual prototyping”.
Electrochemical Society, 150 (2), A176–
Email: d.landi@univpm.it
A192, 2003
[8] Y. He, W. Liu, B.J. Koch, Battery Prof. Michele Germani.
algorithm verification and development is Associate Professor at the
using hardware-in-the-loop testing, Università Politecnica delle Marche,
Journal of Power Sources, 195 (2010), Department of Industrial
2969–2974 Engineering and Mathematical
Sciences. His main research topics
[9] J. Randles, Kinetics of rapid electrode
are: eco-design methods and tools,
reactions, Discuss Faraday Soc., 1, 11-19,
advanced virtual prototyping tools,
1947
energy efficiency and human-
[10] M. Mousavi, S. Hoque, S. Rahnamayan, I machine interaction. He has written
Dincer, G.F. Naterer, Optimal Design of 150 scientific papers published in
an Air-Cooling System for a Li-Ion international journals and
Battery Pack in Electric Vehicles with a conference proceedings and he is
Genetic Algorithm, IEEE Congress on responsible for several European and
Evolutionary Computation (CEC), 2011 National research projects.
Email: m.germani@univpm.it
[11] D. Ghosh et al., Full Hybrid Electrical
Vehicle Battery Pack System Design,
CFD Simulation and Testing, Advanced
Battery Technology 2010, ISBN 978-0-
7680-3414-1, SAE International, 2010
[12] C. Forgez et al., Thermal modeling of a
cylindrical LiFePO4/graphite lithium-ion
battery, Journal of Power Sources, 195
(2), 2961–2968, 2010

EVS27 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 11