OBJECTIVE

The main objective of this project is to perform the “CFD & thermal analysis on a battery pack”, while working on three different cooling strategies like natural air cooling, forced air cooling & liquid cooling.

And to post-process the simulation results to compare these cooling strategies with their advantages & disadvantages based on aspects like cost, feasibility & complexity in these three cooling strategies.

THEORY

INTRODUCTION

As part of the energy of electric vehicles, power battery has a significant impact on vehicle performance. Because the merits of small size, high power density, high voltage, low self-discharge and low resistance, lithium-ion batteries are widely used in the power system and obtain a lot of notice from many automobile manufacturers.

In the general case, there is a small space available for power battery while designing the car. In addition, the lithium-ion battery will be accompanied by a lot of heat during the charging and discharging process.

It is easy to lead to the accumulation of heat inside the battery and then affect

battery performance and safety if the heat is not dissipated timely. If the heat dissipates unevenly, it will cause a large temperature difference inside the battery pack.

The unevenness of the battery pack temperature field will cause unbalance of the battery modules and each cell performance, and finally,

affect the performance of the entire battery pack and security. Therefore, in order to make the battery power system available to play a better performance while ensuring security, it is necessary to optimize the design of the battery cooling solution.

That’s why in order to assist this issue the concept of Battery Thermal Management System (BTMS) comes into effect where we use different modelling & simulation techniques to analyse the best cooling criteria & battery model to maximize the power output & increase the battery life.

BACKGROUND RESEARCH

Every rechargeable lithium-ion (Li-ion) battery is constituted by a packing of cells (also known as a battery pack) so that the final electric power is the sum of all single elements. The process of Li-ion battery design concerns modelling and simulation regarding the whole battery pack. In industry, the actual strategies for battery pack design are mainly related to the market size.

The market of large-scale production requires reliable batteries for standard applications. Conversely, small productions, with a little volume of customized units, focuses the satisfaction of particular customer specifications coupled to a reduced time-to-market. On the one hand, large battery manufacturers can use expensive design tools and subdivide additional costs on a large number of sales.

On the other hand, small and medium enterprises have to increase their own knowledge and formalize the rules of expert designers in low-cost customized tools, in order to contain excessive costs.

The traditional design approach, commonly used in many SMEs, is mostly based on the experience of engineers. So that designers use basic calculation tools to analyse and elaborate new solutions. The engineer calculates the electrical layout configuration, determines the geometrical shapes and defines the cooling system size. Then, the corresponding physical prototype is

manufactured and physical tests give him/her the performance quality. Possible design errors introduce expensive iterations, thus time to market gets longer. The main activities are repetitive tasks and designers are limited to use own experience on past project data (such as empirical tables, drawings, reports and feedbacks).

The main parameters in battery designing are: the operative conditions, the single cell type, the cell number, the battery layout, and finally the evaluation of heat released by the electrochemical reactions. According to the temperature problems in Li-ion cells, a battery pack needs successful thermal energy management and a cooling system to favour heat dissipation and thermal runaway.

The two required aspects for an automotive lithium-ion battery are:

an optimum operating temperature range and small temperature variations. Some researchers have analysed methods to optimize the geometrical and fluid dynamic parameters in cooling design.

THERMAL ISSUES RELATED TO BATTERY

In Li-ion battery cells the temperature influences: electrochemical reactions, round trip efficiency, charge acceptance, power and energy capability, reliability, life cycle and cost. Furthermore, in electric vehicles, battery temperature increases during discharge with a huge peak for high current rates. Generally, an excessive temperature degrades performance and a limited thermal dissipation can even produce burning in Li-ion battery cells.

Great importance is also given to battery safety concerning heat accumulation due to the lithium flammability and explosiveness. The temperature raises two very important issues in Li-ion batteries: safety risks and material ageing. The safety risks are leading to high-temperature values during discharge over the permitted level.

The material ageing is related to the operative cycles and to the achieved temperature levels during working conditions. Moreover, in Li-ion battery pack is also very important the temperature distribution. A non-uniformity in temperature of a cell pack favours localized deterioration in single battery elements, this introduces efficiency losses (as ohmic resistance) and lifespan reduction in the whole pack. The thermal model characterization for a Li-ion cell is important to predict the average temperature at the different current rate in the early design phase. The explained thermal analysis is based on a study by Thomas and Newman which evaluates the heat produced by the single cell. In particular, a simplified form of Newman formula has been considered in an analytical approach to estimate the heat source at the different current rate for each lithium battery cell.

`Q ̇=I(V-Eo)-IT(∂E_o/∂t)`

In the above equation, the term I indicates the instant current value, the term V indicates the related electrical potential (voltage measured at the cell poles), while E_o is the relative open-circuit voltage, and T indicates the temperature value.

BATTERY SIMULATION TECHNIQUES

Several research works have analysed the issue of simulation for lithium-ion battery packs. That mainly concerns the electrochemical and thermal cell characterization. Particularly the electrical aspect of lithium-ion cells regards the property to generate a potential difference between negative and positive pole (voltage), while the chemical aspect concerns the capacity to move a quantity of electrons and ions through each battery sublayers. The thermal behaviour of the Li-ion cell is related to the heat dissipated from electrochemical reactions inside a single battery element. Thus, electrochemical and thermal analysis are coupled with features in simulation virtual analysis. In literature, several researchers have investigated only the electrochemical behaviour. However, those, who have analysed the thermal analysis, have also investigated the electrical model characterization. Below an explanation regarding recent works on battery simulation. The electrochemical simulations on Li-Ion batteries concern the evaluation of voltage curves and state of charge (SOC) at different current rates in charging and discharging. Often, in electrochemical simulations, the input current profile wants really reproduce the effective driving cycles load for electric vehicles. This type of simulation supports the engineer in the modelling of tailored electrical layout configuration for batteries This type of simulations supports the engineer in the modelling of tailored electrical layout for batteries. Generally, the equivalent circuit describes the electrical behaviour of the cell system as an approximation of non-linear real phenomena. Therefore, the conventional linear circuit allows the engineer to predict a reliable voltage over poles with a simple component like an RC filter (resistor-capacitor) and an ohmic resistance. The calculation of real voltage curves gives an important feedback regarding battery performance in terms of energy and power analysis. In addition, the SOC valuation is an important feature to predict battery autonomy.

On the other hand, the thermal simulation regards the heat generated by the electrochemical reactions and also analyses the cooling performance. Many simulation techniques have been involved in literature.

Development of Genetic Algorithm has been extended to optimizing the NTU (Number of Transfer Unit) model in lithium cells for heat exchange study. This approach proposes good results, but the application is often customized to the specific cooling system, as i.e. in where the heat exchange is encouraged by tubes with cool air. Using a Genetic Algorithm approach is possible to achieve good results but it is not possible to define a routine common for different battery types and models.

There are several types of thermal simulation models based on: lamped element model, finite element analysis (FEA) and computational fluid dynamics (CFD) analysis with finite volume methods (FVM). Generally, the lamped element model approach provides simple thermal models in a lithium battery cell, but this requires the definition of several parameters through experimental measurements or technical datasheet. These technical specifications on cell sub-layers are often not provided by the manufacturer, so it is fundamental to the planning of useful experiments to achieve the necessary data. Actually, there are reliable commercial tools for battery modelling, but the cost and complexity of these tools limit their diffusion. So that several

researchers have analysed customized tools for specific cells and battery pack layout. However, rapid tools are necessary to support engineer for evaluating thermal battery behaviour inside the European SMEs that are working in the automotive industry. These tools have to provide a common workflow design for several lithium-ion battery cases.

FRAMEWORK METHODOLOGY

The basic methodology adopted in the research & development activities related to the optimisation & modelling of batteries have a common process flow, and one most common process flow diagram is stated below:

Li-iON BATTERY

The operating principle of Li-ion battery:

Lithium-ion batteries mostly contain anode, cathode, electrolyte and separator. The cathode is made of carbon materials such as graphite, while the anode is mainly made of a lithium-containing transition metal-oxides and its doped compounds such as LiCoO2 and LiMnO2. The chemical equation of battery charging and discharging procedure is proposed to be

where M is Co, Mn and Fe, etc. The lithium-ion battery will produce or absorb heat by a variety of ways during its working process. However, this will cause a changing of the temperature inside the lithium-ion battery itself. If the rate of heat dissipation is less than the rate of heat generation without good control, then it will lead to thermal runaway. Each heat generated by the various processes within the cell includes polarization heat and reaction heat. Compared with Polarization heat, the reaction heat of the lithium-ion battery is relatively smaller.

One uses a microcalorimeter to show the thermal behaviour of LiXMn2O4. When the charge rate is less than 1C, the thermal process is reversible; while when the charge rate is greater than 1C, the thermal behaviour is controlled by an irreversible process (ohmic resistance and overpotential). Battery thermal process keeps to the following equation,

where Q_T is the total heat generated within the battery; Q_e is the heat passed to the outside of the battery; Q_a is the heat absorbed by the battery itself, which directly affect the battery temperature.

Some common reference names of the Li-ion Batteries:

Equations governing thermal analysis of batteries

The governing equations which were used to solve the time-dependent three-dimensional flow problems include the continuity equation, momentum equation, and the energy equation.

Continuity Equation:

x-momentum:

y-momentum:

z-momentum:

Energy equation:

OPERATING REQUIREMENTS OF BATTERY

The battery temperature should be controlled within temperature limits to avoid the thermal issues and improve the performance. The temperature range affects the battery power and battery cycle life, the figure below shows the same. At the same time, the temperature distribution should be even to guarantee the battery performance and lifetime. That is also the reason why the battery thermal management system is necessary to the battery system.

When temperature ranges from 20°C to 40°C, battery power reaches to the maximum.

The cycle life goes down slowly below 10°C because of anode plating and drops off quickly above 60°C due to the breakdown of electrode materials, as seen in the figure above. Generally, the temperature must be controlled between 20°C and 40°C to ensure the performance and cycle life.

IMPORTANCE OF FUEL CELL MODELLING

The need of fuel cell vehicles (hybrid electric vehicles) & EV is increasing & so is the market trend changing, as much as the demand of EV increases so does the need of fuel cell modelling.

HOW DOES FUEL CELL MODELLING ADDRESSES THE CHALLENGES?

HOW FUEL CELL MODELLING IS ASSISTED IN STAR CCM+

1. PARAMETRIC MODELS:
   
   
   
   
   
2. EASY PHYSICS SETUP:
   
   
   
   
   
   
3. FLEXIBLE & ROBUST MESHING:
   
   
   
   
   
4. COUPLING WITH MULTIPHYSICS:
   
   
   
   
   
5. ELECTROCHEMISTRY:
   
   
   
   
6. INTELLIGENT DESIGN EXPLORATION:
   
   
   

MODELLING FUEL CELLS IN STAR CCM+:

1. MULTIPHYSICS: Tight coupling between physics models
   
   Simcenter STAR CCM+ is a powerful multiphysics environment coupling flow, energy & species:
   a.) FLOW:
   1. A comprehensive range of turbulence models.
   2. Single or multiphase models can be combined with single species or multi-components fluids.
   
   b.) ENERGY:
   1. Energy is computed both in fluid and solid regions, accurately capturing heat transfer at fluid-solid interfaces.
   
   c.) SPECIES:
   1. Gas, liquid, or multiphase flows can include electrochemical reactions between species.
   
   
   
   
   
   
2. ELECTROCHEMISTRY MODELS - with single-phase flow:
   

   Single-phase flow modelling provides an economical and reliable prediction of reaction rates at the interface between the two electrodes.

   a.) Any number of reactions can be specified, involving multiple fluid species & ions.

   b.) Electrochemical species transport equations are accounted for in both solids or fluids.

   c.) Both electrochemical and Ohmic heating are accounted for.
   
   
   
   
   
   These charts & above simulation provides us with the best picture to where there are flaws in the model or is it optimized completely or not.
   
   
   

3. ELECTROCHEMISTRY MODELS - with the multiphase flow:
   By modelling a multiphase system, more detailed and realistic predictions of fuel cell behaviour can be made:
   
   a.) Modelling liquid water as a separate phase provides a more accurate reaction rate and current density/voltage curve.
   
   Critical behaviours can only be predicted using multiphase modelling:
   
   a.) Flooding and membrane dry out.
   

   b.) Parasitic reactions for degradation and lifetime.

   c.) Site blockage in PEM (Proton Exchange Membrane) fuel cells.

   d.) Sorption.

   
   

PROCEDURE OF BATTERY PACK MODELLING & THERMAL SIMULATION IN STAR CCM+

1.) BATTERY MODELLING:

The basic battery pack information we have from the problem statement is:

Based on the above-mentioned data we start modelling of our battery pack & follow the steps mentioned ahead.

Procedure for battery pack modelling is as follows:

1. The first step is to enter into the STAR CCM+ GUI & click on new to start creating a new simulation file.
2. The next step is to open the simulation tree & under the geometry click on 3D CAD model & select edit option to start modelling the battery packs.
   
   
   
   
   
   
3. Now, on features, we right-click & select the solid primitives under which we select the block.
   
   
   
4. Now, in order to create the block we need to enter the battery pack specification i.e., w*d*h = 200*100*400 in x, y & z respectively while keeping the base centre as x=0.0mm, y=55.0 mm, z=0.0mm as each block represents a battery pack.
   
   
   
   
5. Similarly, we create the next block for the second battery pack parallel to the first one, by moving the centre.
   Assumption: A 10mm air gap for a thin film of fluid to move and cool is left in between both the battery packs.
   Hence, in order to assist this, we displace the centre of the next battery pack.
   

   Here also the dimensions would be similar i.e.,

   Length x= 200 mm;
   Length y= 100 mm;

   Length z= 400 mm;

   Base centre:
   x=0.0 mm;

   Y=-55 mm;

   Z= 0 mm;

   
   
   
6. Now, the next step is to make the outer jacket to basically define the fluid dimension needs to be created outside both the battery packs.
   Dimensions: Length x= 220 mm; Length y= 230 mm; Length z= 420 mm;
   Base centre: x=0.0 mm; Y=0.0 mm; Z=-10 mm;
   
   
   

   And under restore views option above, we click on projection mode as parallel.

   And rename the body parts as Cell_01, Cell_02 & Battery_Pack.
   
   
   
   

7. Extracting the volume:
   Select all three of the body parts & then right-click --> Boolean --> Subtract
   
   
   
   
   
   
   
   
   
   

   As we can see in the images & clicking on the transparent view helps us getting the view of the extracted volumes.
   
   
   

   Now, click on close 3D CAD & the model gets saved.

8. Creating Geometry Scene:
   

   Now, under create or open scenes click on the geometry --> Right-click on the 3D CADE MODEL 1--> Click on “new geometry part” --> under pop-up for parts creation option--> ensure solid body is selectedà click OK.
   
   
   
   

   Now, in the tree under the scenes, open geometry scene 1 --> open displayers --> open geometry 1 --> click on parts --> open surface & select default surface -->OK.
   
   
   
   
   
   

   Now, right-click on geometry scene 1 --> under apply representation --> select geometry --> a block appears in the geometry scene view --> this is basically the fluid domain.
   
   

9. Creating Groups by using split by patch to define fluid domain:
   
   Under geometry tree --> parts --> surface --> default -->right click on default -->select split by patch.
   
   
   
   Now, combine 7,8,9,10,11,12àName Cell_Surface_01
   
   
   
   Next combine 13,14,15,16,17,18àName them as Cell_Surface_02
   
   
   
   
   Next combine 2,3,4,6 --> Name them as Ambient_Surface
   
   
   
   Now, we are left with 1 & 5 out of which one is inlet & other is an outlet.
   
   
   
   
   
   
10. Creating a region for the fluid domain:
    
    In simulation tree inside parts --> right-click on Battery_Pack --> Select assign region to parts --> in pop-up select Battery_Pack and also confirm the selecting the three things-->(1) Create one region for all parts, (2) Create a boundary for each part surface (this is necessary to input boundary conditions for each part surface), (3) Create one feature curve for all part curves --> Click on apply.
    
    
    
    Under regions --> open region --> open boundaries --> Here, all boundaries based on each surface is shown. So basically, we have created regions.
    

    Hence, the basic part of battery pack modelling ends here. And next, we need to setup the boundary conditions for different cases.

2.) DECIDING BOUNDARY CONDITIONS

Under regions selectàregion --> under boundariesàclick on the ambient surface.

Now, in the menu bar --> click on the window tab --> click on propertiesàthen we see under the properties window we can select various properties.

So, now we will click on the respective boundaries from here & enter the desired properties for each respective boundary based on the problem statement.

Now, we select various properties for each boundary region in a general way here & will later on decide each of them for a specific case setup.

Inlet --> Velocity Inlet

Outlet --> Pressure outlet & for the remaining three boundaries we select wall.

3.) CREATING PHYSICS CONTINUA

Right-click on the continua under geometry tree --> Click on New --> Select physics continuum --> Right-click on Physics1 --> Click on Select Modelsà Pop-up appears.

Under Space select 3D.

Under time select steady --> select material as liquid (for one particular case) --> flow is selected to be segregated flow --> equation of density --> constant density -->Viscous regime --> TurbulentàReynolds Averaged Turbulence --> K-Epsilon Turbulence --> Click on close.

Note: In the simulation under “continua” --> liquid --> H₂O --> Clicking on material properties we can see & change the material properties most of them are standard properties.

4.) CREATING MESH CONTINUA

Right-click on the continua --> New --> Mesh continuum --> simulation tree gets updated --> right-click on models -->select meshing models -->

In pop-up --> select surface remesher --> volume mesh --> polyhedral mesher (for creating conformal volumetric meshes) --> prism layer mesher (to record the temperature gradient on walls as very small elements are required) --> close.

Now under continua in simulation tree --> mesh1 appears --> under which we can see different values for each of the boundary regions.

Like defining the mesh base size, number of prism layers, surface size etc.

Also, if we see under regions --> properties --> Select type as a fluid region.

Now since we don’t need prism mesh at inlet & outlet so we will disable them in both these regions.

For this, click on battery pack inlet (under boundaries) --> customize prism mesh --> under properties section --> select disable.

Similarly, click on battery pack outlet --> customize prism mesh --> under properties section à select disable.

5.) CHECKING VARIOUS INPUT VALUES

Now, we need to input heat flux & keeping it a constant value for both Cell_Surface_01 & Cell_Surface_02.

And for Ambient_Surface we need to enter the static temperature which is a constant.

Under shear stress condition we take the case as no slip.

6.) SOLVING, REPORT PLOTTING, AND POST PROCESSING

Now, before generating the mesh & running the thermal simulation for a particular case we need to generate the report.

• Generating the reports:

Right-click on the reports --> select new report --> element count…surface uniformity --> surface average --> right-click on surface average --> select temperature as field function --> in properties section --> Under representation select volume mesh --> And under parts in this property section itselfàselect Cell_Surface_01 --> Surface Average 1 gets created;

Similarly, we create this for Cell_Surface_02 à it will be named as Surface Average 2.

• Generating Plots:

Now in reports generated above selected both of them together & right-click, then select create monitor & plot from report --> Click on a single plot.

Now, as you see under the plots there will be reports plot & residuals.

This residual plot is important to see if our simulation reached convergence or not.

• Generating Surface Mesh:

Save the file now to completely ensure that all inputs we have given are registered.

Now, click on generate surface mesh option in the menu bar at the top. This will generate the surface mesh.

• Updating Geometry Scene:

Now, under the scenes right-click on geometry scene --> Select apply representation --> Select remeshed surface --> In geometry scene the model gets updated.

• Generating Volume Mesh:

Now, after generating the surface mesh we need to generate the volume mesh. So in the menu bar next to the surface mesh option we have volume mesh generation option, select this option & wait till the mesh generates.

Here, we can see the prism layers on the walls along with the conformal polyhedral mesh types.

• Generating Scalar View:

Now, before initializing the simulation we need to create a scalar view to see if the cooling trend & analyse the same throughout the battery.

Under Derived Parts --> New Part --> Section --> Plane --> Choose the plane.

Now under the scenes --> scene1 --> displayers --> Scalar scene1 -->Scalar1 -->parts --> Select Cell_Surface_01 & Cell_Surface_02 as region --> OK --> Under GUI the scene gets updated.

This helps in better visualization & effect of various physical and thermal parameters while the simulation is running.

• Initializing the Simulation:

Now, we can save the model & click on initialization option in the menu bar at the top.

• Running the solution:

Now, as the initialization gets complete, we click on the run & let the simulation run for the defined time steps which here we have taken to be 1200.

COMPARATIVE ANALYSIS OF THREE COOLING STRATEGIES BASED ON THERMAL SIMULATION RESULTS, NATURAL CONVECTION, FORCED CONVECTION, LIQUID COOLING

1.) COMPARISON OF ALL THREE CRITERIA at same Mesh sizes while only the COOLING material varies:

10mm Mesh Size:

Other input parameters are as follows:

Since the battery capacity is 20Ah each so the power output obtainable from the complete battery pack = 20*48*2= 1920Wh and this is mentioned in the table above where we used this as a heat source power.Calculations:

Comparison of Scalar Scenes & Residual Plots indicating temperature values:

Natural Air Convection Scalar Scene & Residual Plot:

II.) Forced Air Convection Scalar Scene & Residual Plot:

III.) Water Cooling Scalar Scene & Residual Plot:

Based on the above plots we can see that the minimum difference in the max value & min. value of the temperature across the battery is in the case of water cooling & also the rate of cooling is very fast compared to the other 2 cases, this is when the mesh size is kept same in all the modes & solver is steady state.

B.) 5mm Mesh Size:

Other input parameters are as follows:

I.) Natural Air Convection Scalar Scene & Residual Plot:As we can see these parameters mentioned above are same as the one while performing a thermal simulation of a battery having 10mm base size, so the plots which will be obtained here can help us identify the changes in the respective categories themselves when their mesh size is reduced from 10mm to 5mm.

Above image reflects the changes happening in the grid dependency test. The energy being delivered is more in a constant manner as the grid size is reduced.

II.) Forced Air Convection Scalar Scene & Residual Plot:

Above image reflects the changes happening in the grid dependency test. The energy being delivered is more in a constant manner as the grid size is reduced.

III.) Water Cooling Scalar Scene & Residual Plot:

Above image reflects the changes happening in the grid dependency test. The energy being delivered is more in a constant manner as the grid size is reduced and the results are most smoothened in case of water cooling.

NOTE:

In case of natural convection through the air, the inlet velocity of air is kept 0 m/s, in rest of all cases it is 5 m/s.

2.) Inference from the CFD & THERMAL SIMULATIONS

Based on the above-mentioned plots we can infer the following things:

• Natural Air convection is disastrous to the battery as temperatures can exceed far above 60 degrees with variation within the cell from inlet to outlet as high as 1000 degrees which is not feasible, neither the battery can withstand so long.
• Forced Air convection is somewhat effective in cooling with 98% of the battery retained at a suitable operating condition less than 60 degree Celsius, but still the rate of cooling is not accurate & areas near outlet still remain at high temperature, which will again deteriorate the battery.
• Water Cooling is found to be most effective of all the above mentioned 3 cases, where the average battery temperature is 34 degree Celsius & the max. the temperature at convergence is 41 degrees which is well within the operating temperature range. Hence, life Is the maximum in this case.
• As we move from the mesh size of 10 mm to 5mm the complexity of the problem increases & the solver takes a longer duration of time to iterate to the final results. However, the mesh becomes refined & more optimised results are obtained for smaller mesh size.
• However, I tried meshing this for 1mm mesh size, but the computational time increased so much that it exceeded my system capacity & finally I had to abort the process which suggests that going for too much-refined mesh also increases cost & decreases the feasibility of the case.
• The max. temperature is observed in all cases near the output of the battery pack & that too at the core of the cells. Which if maintained to a suitable degree will maximise the battery performance.

SOME OF THE IMPORTANT OUTCOMES OF WATER-COOLING SIMULATION

1.) Total Energy Distribution:

2.) The Pressure Distribution:

3.) The Vorticity Variation:

4.) Variation in Heat Transfer Coefficient:

5.) Surface Average Temperature Variation of the Battery Pack:

We can see that without even changing any condition and just the mesh size, the temperature variation across the surface of the battery vary hence, going for a 5mm mesh is more preferable than 10mm mesh size.

CONCLUSION

Based on my study & aspects taken into consideration for running these simulations I come to the following conclusion:

• With the increase of inlet flow speed, the cooling performance improves gradually. After exceeding a certain speed, increasing wind speed cannot improve battery pack cooling performance obviously.
• The change in the ambient temperature also impacts battery performance.

• Comparison of advantages & disadvantages of three cases:
  i.) Natural Air Convection:

Advantage: This mode is suitable when the vehicle or battery pack is in idle condition for a prolonged duration.

Disadvantage: Since, this type of convection occurs only due to the impact of gravity hence, the rate of cooling is very poor & leads to complete deterioration of battery pack if any other mode is not followed.

ii.) Forced Air Convection:

Advantage: This type of cooling is effective when the fluid i.e. water or 50-50 ethylene glycol or some other coolant either is not available.

Disadvantage: This type of cooling is somehow effective if we increase air flow rate but on further increasing the airflow rate there is only loss of energy & power while the cooling is not more effective in that case.

iii.) Water Cooling:

Advantage: This type of cooling is the most effective type of cooling where both the max temperature is well within specified limits & also the average battery pack temperature is very moderate. Hence, this cooling increases battery life & helps in delivering the best power outputs for a longer time & range.

Disadvantage: This requires a proper arrangement to maintain the flow of coolant & also in very cold weather there are chances of coolant getting frozen up thereby choking the flow & hence, becoming dangerous for the battery life as it can produce huge temperature difference on various parts of the battery.

Hence, I conclude with the fact that based on my study under the given constraints & considering the limitations of computational processing the method of water cooling is by far most suitable keeping in mind the ambient conditions & vehicle requirements for battery output.