Numerical Investigation of the Thermal Performance of a Hybrid Phase Change Material and Forced Air Cooling System for a Three-Cell Lithium-Ion Battery Module

Abstract

The thermal performance of a lithium-ion battery module comprising three cells contained within a casing was investigated at discharge rates of 3C and 5C with three different cooling strategies: forced air, phase-change material (PCM), and a hybrid system using a combination of forced air and the PCM. Three levels of fan speed (5000 rpm; 7000 rpm; and 9000 rpm) for cooling air flow were considered. A numerical simulation of heat transfer was performed using the ANSYS Fluent software. The electrochemical modelling of a battery was developed based on the NTGK approach, and the phase-change phenomenon was treated as an enthesis–porosity problem. The composite PCM, aluminum metal foam embedded in n-octadecane, had better heat dissipation performance than forced air convection. The PCM is significantly more effective at heat dissipation than forced air. Interestingly, when using a hybrid cooling system that combines forced air and a PCM, although it meets the operational requirements for Li-ion batteries in regard to maximum temperature and temperature uniformity at a 3C discharge rate, the airflow appears to have a negligible effect on thermal management and yields an indiscernible change in temperature. This can be attributed to a complex flow pattern that developed in a casing as a result of the suboptimal design of the inlet and outlet. Further studies will be required for the optimal positioning of the inlet and outlet, as well as the effectiveness of combining liquid cooling methods.

1. Introduction

Battery technology is considered to be the most crucial part of electric vehicles (EVs), as the battery costs account for approximately 45% of the cost of EVs. Recently, different types of batteries have been used in both pure and hybrid EVs, such as lead–acid (in the General Motors EV1), Li-ion (in the Tesla Roadstar), and nickel–metal hydride (in the Toyota Prius) [1]. Batteries should have low cost, long life, and environmentally friendly characteristics, and provide the necessary energy for EVs in various operating conditions, especially load-point shifting, regenerative braking, and electric auxiliaries. Li-ion batteries are far superior to the other types of batteries due to their relatively high specific power and energy density.

During charging and discharging processes, chemical reactions inside batteries produce a large amount of heat, resulting in an excessive rise and uneven temperature distribution within a cell. This significantly affects the performance and lifespan of the battery. It has been reported that the optimum operating temperature ranges from 25 °C to 40 °C, with a temperature variation of less than 5 °C [2]. A number of different types of battery thermal management systems (BTMSs) have been studied to address these issues. Active BTMSs, including forced air and liquid cooling, require additional power for fans and pumps. In contrast, the use of PCMs as a passive BTMS consumes no power and offers intriguing cooling performance by utilizing latent heat during phase transition. They also have a low weight and simple structure. However, they have the drawbacks of low thermal conductivity and volumetric expansion over the phase-change process. In addition, when the material completely changes its phase, the temperature control becomes difficult to operate continuously [3].

Composite PCMs have been designed and studied for thermal conductivity enhancement through the use of macroscale and nanoscale additives. Akula et al. [4] investigated the thermal management of fins and expanded graphite (EG) combined with eicosane for a Panasonic NCR18650BD battery. It was reported that a combination of simple heat sinks and PCM-EG produced much better thermal performance than complex heat sinks integrated with pure PCM. A design with 130 fins and 30% EG provided significantly more temperature reduction at various discharge rates compared to 260 fins filled with eicosane. Wu et al. [5] investigated the effects of different EG mass fractions embedded in paraffin on the thermal behavior of prismatic cells under extreme operating conditions, and showed that the EG mass fraction in the range of 15% to 20% had the best thermal performance. In addition, a PCM with a pyrolytic graphite sheet (PCM-PGS), which has a convective heat transfer coefficient of 50 W/m²·K, performed similarly to the 200 W/m²·K PCM module. Zhang et al. [6] showed that the mixture of paraffin with EG, aluminum nitride (AlN), and epoxy resin resulted in a substantial 19.4% reduction in battery temperature, with a temperature variation of less than 1 °C.

Liquid cooling has been known to have a high specific heat and fast heat transfer rate. For this reason, many researchers have developed hybrid systems with liquid and PCM cooling to obtain higher thermal performance than a single method. Cao et al. [7] carried out experiments to explore the influences of water temperature, flow rate, and kinds of PCMs on the temperature of 20 cylindrical cells. This thermal management system achieved the greatest performance if the water temperature was kept under 40 °C, as close as possible to ambient temperature, if the flow rate was maintained as low as 100 mL/min, or if an EG-based composite PCM with 67 wt% RT44HC was applied. Kong et al. [8] proposed a potential design that could reduce the temperature of 21,700 batteries to 41.1 °C, and the temperature uniformity remained under 4 °C at the end of 3C discharge. Plus, the efficiency of heat dissipation would be inconsiderable if the cell-to-cell distance exceeded 5 mm. An et al. [9] realized that a flow velocity of 0.08 m/s was ideal to achieve the best thermal performance in temperature rise and temperature difference. Over 0.08 m/s, a higher liquid velocity resulted in lower heat improvement. Moreover, the maximum and initial temperature remained unchanged at a velocity of 0.04 m/s and an EG of 6 wt%, which met the requirements for battery operation.

Both heat pipes and PCMs have benefits and drawbacks; BTMSs can operate more productively based on the combination of these two techniques. In line with this idea, Jiang et al. [10] considered essential factors that influence the temperature of 3.2 V-8 Ah prismatic batteries during the charging and discharging processes. They reported that the melting point of the PCM is roughly 3 °C higher than ambient temperature, the phase-change ratio is nearly 0.55, the thickness ratio is 0.17, and the heat transfer coefficient should range from 30 W/m²·K to 60 W/m²·K, which kept batteries safe in operating conditions. Putra et al. [11] proved that the maximum battery temperature rapidly dropped by 33.42 °C in the case of heat pipes associated with RT44HC. Compared with beeswax, RT44HC was more outstanding because its melting point is in the battery working temperature. Chen et al. [12] indicated that there were several methods to decrease the maximum battery temperature; however, they expanded the temperature uniformity, including a rising convective heat transfer coefficient, a declining environmental temperature, or increasing the latent heat and thickness of the PCM. There was a remarkable decrease of 30% in the temperature difference after optimization.

Much attention has been paid to a novel BTMS that integrates air cooling and PCM. The thermal behavior of six cylindrical batteries in a 1S6P array at different ambient temperatures was evaluated by Jilte et al. [13]. The maximum battery temperature was less than 5 °C, even under the most severe conditions. The temperature uniformity was less than 0.05 °C and 0.12 °C at 2C and 4C discharge rates, respectively. Singh et al. [14] investigated the effect of airflow velocity and PCM thickness on SONY 18,650 cells. The battery temperature decreased significantly by 30 K and 45 K under 1C and 5C discharge rates in the case of a 1 mm thick PCM. In addition, the diamond-shaped battery organization produced better thermal performance than the square shape, which can achieve a temperature reduction of 3 K to 4 K in the absence of a PCM. Akkurt et al. [15] carried out transient simulations of a BTMS for three cylindrical batteries in a rectangular pack, considering different battery spacing and airflow velocity. Khan et al. [16] found that the cylindrical Li-ion batteries in a triangular chamber had the lowest temperature and largest heat transfer coefficient at airflow velocities of 0.005 m/s and 0.02 m/s among various PCM shapes, such as circular, lozenge, hexagonal, and square chambers. The heat transfer coefficient of the cooling system was enhanced by 65% by incorporating graphene nanoparticles at a 4% volume fraction into an organic PCM. Moreover, an increase in airflow speed resulted in decreased battery and outlet temperatures and an improved chamber heat transfer. Gu et al. [17] proposed an optimal BTMS of paraffin and expanded graphite with an axial thickness of 45 mm and a radial thickness of 8 mm. Based on the design, airflow at 1.23 m/s and fins were employed to keep the battery temperature below 60 °C under 2C, 3C, and 4C discharge rates. The heat transfer was considerably improved, and the flow dead zone was narrowed by utilizing c-type fins or o-type fins. Recently, Ahmad et al. [18] utilized metal fins and air cooling to enhance the thermal conductivity and secondary heat dissipation of PCMs, respectively. The best design with a 1 mm thick PCM, 162 fins, and a 3 mm diameter fin kept the battery temperature below 40 °C with less power consumption. Chen et al. [19] investigated the effect of PCM thickness, fin type, air velocity, and the width of the support frame on the thermal performance. The bifurcated fin design with an angle of 75° and a length of 40 mm resulted in decreases of 0.4 °C, 0.4 °C, and 13% in terms of the maximum temperature, temperature difference, and energy consumption, respectively, compared with the beak fin design. E et al. [20] investigated the effects of the airflow inlet and outlet on power consumption and cooling performance. At an inlet temperature of 301.15 K, the lowest energy consumption and temperature difference were obtained using the optimal BTMS with five holes and a 7 mm thick PCM.

In this study, a large-format (52.3 Ah) Li-ion pouch battery provided by SM Bexel Co., Ltd. (Gumi, Republic of Korea) was utilized to investigate the thermal performance for operating requirements. On the same type of battery, Ho et al. [21] applied forced air cooling to dissipate the thermal energy generated by the ten-cell battery module. The maximum temperature of the Li-ion battery decreased from 74 °C to 36.4 °C and from 114 °C to 49.7 °C at the end of the 3C and 5C discharge rates, respectively. Additionally, Huynh et al. [22] analyzed the thermal behavior of a single cell cooled by an n-octadecane PCM integrated with aluminum foam. At the end stage of the discharging process at 3C and 5C rates, the maximum battery temperatures declined to 34.3 °C and 50.7 °C, respectively.

Furthermore, the current study establishes a battery model in a case for practical orientation using the exact category of Li-ion battery mentioned. A fan, which in previous studies has been assumed to operate as a plane with uniform velocity, has been modeled in 3D in ANSYS Fluent to provide an accurate description of the airflow. The NTGK semi-empirical model was adopted to predict the heat generation inside the Li-ion battery, relying on input parameters from the discharge experiment. The use of the NTGK approach with a specific type of Li-ion battery collectively contributed to a more comprehensive understanding of a new technology in the electric vehicle industry. The thermal behavior of a Li-ion battery model was investigated at 3C and 5C discharge rates with three strategies, including forced air, PCM, and a hybrid system using a combination of forced air and PCM. In the presence of airflow, three different fan speed levels were selected to study the effect of the fan speed on the maximum temperature and temperature uniformity within a cell.

2. Methodology

2.1. Battery Mathematical Model

A Li-ion battery model, composed of large-format pouch cells connected in series in a case, was considered. Each cell consists of two current collectors, two electrodes, and a separator, as shown in Figure 1 and

Table 1. The thickness of each layer in a cell.

Layer	Thickness (µm)
Positive current collector, ${\delta }_c^p$	20
Positive electrode, ${\delta }_e^p$	80
Separator, ${\delta }_s$	20
Negative current collector, ${\delta }_c^p$	114
Negative electrode, ${\delta }_e^p$	15

. Cells are connected in series via copper busbars, the heat generation of which is neglected.

The effective physical properties of the battery, such as its thermal conductivity, specific heat, and density, are calculated using the following formulas:

$$x_{eff}=\frac{{0.5x}_c^p{\delta }_c^p+x_e^p{\delta }_e^p+x_s{\delta }_s+x_e^n{\delta }_e^n+{0.5x}_c^n{\delta }_c^n}{{\delta }_{total}}$$

(1)

$${\delta }_{total}={0.5\delta }_c^p+{\delta }_e^p+{\delta }_s+{\delta }_e^n+{0.5\delta }_c^n$$

(2)

Meanwhile, the electrical conductivity is evaluated in Equations (3) and (4).

$${\sigma }_p=\frac{{0.5\sigma }_c^p{\delta }_c^p+{\sigma }_e^p{\delta }_e^p}{{\delta }_{total}}$$

(3)

$${\sigma }_n=\frac{{0.5\sigma }_c^n{\delta }_c^n+{\sigma }_e^n{\delta }_e^n}{{\delta }_{total}}$$

(4)

The specifications of the cells are summarized in

Table 2. The physical, thermal, and electrical properties of Li-ion battery module.

Property	Value
Thermal conductivity, $k_b$ (W/m·K)	25.5
Specific heat, $c_b$ (J/kg·K)	566
Density, ${\rho }_b$ (kg/m3)	2695
Electrical conductivity, ${\sigma }_{+}$ (S/m)	3.77 × 107
Electrical conductivity, ${\sigma }_{-}$ (S/m)	5.96 × 107
Width × height × thickness (m)	0.249 × 0.227 × 0.008
Nominal voltage, $V_n$ (V)	3.75
Nominal capacity, $Q_n$ (Ah)	52.3
Internal resistance, $R$ (Ω)	6.1 × 10−4

, along with their thermo-physical and chemical properties.

Kim et al. [23] established a general multi-scale, multi-physics Li-ion battery model framework to solve the coupled electrochemical, electrical, and thermal physics within a cell. This approach was validated in the analysis of physical properties at various solution domains, such as the particle, electrode, and cell. During the charging and discharging process, the current flux is governed as follows:

$$𝛻·\left({\sigma }_{+}𝛻{\phi }_{+}\right)=-j$$

(5)

$$𝛻·\left({\sigma }_{-}𝛻{\phi }_{-}\right)=j$$

(6)

In ANSYS Fluent, the volumetric heat source comprises irreversible heat, $j[E-({\phi }_{+}-{\phi }_{-}\left)\right]$; reversible heat, $jT(dE/d{T}_b)$; and ohmic heat, ${(\sigma }_{+}·{𝛻}^2{\phi }_{+}+{\sigma }_{-}·{𝛻}^2{\phi }_{-}$), as follows:

$$q_b=j\left[E-\left({\phi }_{+}-{\phi }_{-}\right)-T\frac{dE}{d{T}_b}\right]+{\sigma }_{+}·{𝛻}^2{\phi }_{+}+{\sigma }_{-}·{𝛻}^2{\phi }_{-}$$

(7)

2.2. PCM Mathematical Model

There are several considerations for PCM applications. First, the melting point should be in the range between the ambient temperature and the heat source temperature. Second, the melting latency characterizes how much energy can be stored during the phase change. A longer melting latency means a higher energy efficiency of the system. Another criterion is that PCMs should have high thermal conductivity to avoid thermal bottlenecks at the source. Finally, the material should be chemically and physically stable over repeatable fusion cycles [24]. From our literature review, a potential candidate was n-octadecane with a solidus temperature of 301.15 K and a liquidus temperature of 303.15 K [25].

Venkateshwar et al. [26] enhanced heat transfer performance of pure n-octadecane by integrating it with aluminum metal foam. The effective thermal conductivity of the composite PCM in the solid and liquid states are determined by the following equations:

$$k_{s,eff}=A_1\left[\epsilon ·k_{s,PCM}+\left(1-\epsilon \right)k_{MF}\right]+\frac{1-A_1}{\frac{\epsilon }{k_{s,PCM}}+\frac{1-\epsilon }{k_{MF}}}$$

(8)

$$k_{l,eff}=A_1\left[\epsilon ·k_{l,PCM}+\left(1-\epsilon \right)k_{MF}\right]+\frac{1-A_1}{\frac{\epsilon }{k_{l,PCM}}+\frac{1-\epsilon }{k_{MF}}}$$

(9)

The thermal conductivity of 2.394 W/m·K can be achieved with a porosity of 0.972. The details of the basic thermal properties of the composite PCM are presented in

Table 3. The material properties of the composite PCM.

Property	Solid Phase	Mushy Zone	Liquid Phase
Density, $\rho$ (kg/m3)	814	769	724
Specific heat, $c$ (J/kg·K)	2150	225,000	2180
Thermal conductivity, $k$ (W/m·K)	2.394	2.394	2.394
Pure solvent melting heat, $L$ (J/kg)	225,000	225,000	225,000

[25,26].

In ANSYS Fluent, the enthalpy–porosity is implemented to address phase-change problems. The interface is not explicitly tracked, and the liquid fraction is defined at each iteration based on an enthalpy balance as follows:

$$\begin{array}{ll}\beta =0& \mathrm{for}T<T_{solidus}\\ \beta =1& \mathrm{for}T>T_{liquidus}\\ \beta =\frac{T-T_{solidus}}{T_{liquidus}-T_{solidus}}& \mathrm{for}{T}_{solidus}<T<T_{liquidus}\end{array}$$

(10)

The energy equation for thermal behavior in the PCM domain is considered as follows:

$$\frac{\partial }{\partial t}\left(\rho H\right)+𝛻·\left(\rho \mathit{v}H\right)=𝛻·\left(k𝛻T\right)+S$$

(11)

2.3. Numerical Computation and Validation

A model of a battery module with three cells in a case was considered, as shown in Figure 2. The thermal behavior of two different types of cooling systems were investigated. The first cooling approach is forced air convection (Figure 2a), and the other is a hybrid system combining forced air and PCM, in which a pouch-type battery was sandwiched between two layers of PCM (Figure 2b). The thicknesses of the PCMs were 0.5 mm, 1 mm, 1 mm, and 0.5 mm from left to right. Three different levels of fan speed were applied for airflow, as provided in

Table 4. The specifications of fan were found at SENSDAR (Shenzhen, China).

Speed (rpm)	Mass Flow Rate (kg/s)
5000	0.00140
7000	0.00197
9000	0.00240

.

A hexahedral-based mesh system was generated for the battery, PCM, and air domains, while the fan domain was divided by tetrahedral elements, as shown in Figure 3. The total number of elements was about 4,344,000, with 435,027 for the battery domain, which ensures that y+ is less than 1.

The conjugate heat transfer scheme is employed with adiabatic boundary conditions on all surfaces, except the mass flow rate at the inlet and atmospheric pressure at the outlet. A k-ω SST model was applied for the turbulent model. Based on the finite volume method in ANSYS Fluent, the coupled algorithm was implemented for the pressure–velocity coupling. In particular, the second-order interpolation scheme was used in the momentum, energy, and potential equations, while the PRESTO scheme was used for the interpolation of the pressure field.

The procedure of the battery simulation was validated by Ho et al. [21], where the same type of Li-ion battery was modeled and compared to the experiment. The heat production within the battery mainly includes an irreversible source, a reversible source, and ohmic heating. A proper comparison between the simulating and experimental battery temperatures proved a reliable 3D battery model in terms of heat generation. As shown in Figure 4, the error levels were within 0.5% for the maximum temperature at the end of discharge. In addition, the PCM modeling was also validated with Javani et al. [25] and showed excellent agreement with a small error of 0.8%.

3. Results and Discussion

In this study, the effects of air convection and a PCM on the cooling performance of a three-cell Li-ion battery module were numerically investigated at 3C and 5C discharge rates. In addition, the results were compared by considering different fan speed levels.

3.1. Effect of Cooling Method

Three different BTMSs, including forced air (a speed of 5000 rpm), PCM, and a combination of forced air (a speed of 5000 rpm) and PCM were considered and compared for thermal performance. With forced air convection, there was a decrease of 8 °C and 10.2 °C at the end of the 3C and 5C rate discharge, respectively, but with a similar slope of the temperature rise curve to no cooling, as shown in Figure 5. It is evident that the PCM significantly reduces the battery temperature during the discharge period. In the case of hybrid forced air and PCM cooling, the airflow has a negligible effect on thermal management with an indistinct change in temperature. It is noteworthy that the airflow in the hybrid system can cause the degradation of temperature uniformity, as shown in Figure 5b.

At the end of the discharging process, the maximum battery temperature in the hybrid cooling was slightly higher than in the PCM cooling due to the movement of air particles, as observed in Figure 6. For hybrid cooling, the thermal energy generated by the battery was dissipated through forced convection, in addition to the use of PCM. The inlet did not align with the outlet, which formed a Z type. That led to complex airflow behavior, with the formation of recirculation zones, as well as vortices, as shown in Figure 6b. Moreover, when one particle gains thermal energy from the battery, it vibrates more rapidly and collides with neighboring particles, transferring some of its energy to them. This process continues, creating a chain reaction of energy transfer, resulting in the chaotic motion of air particles. Therefore, the air particles encountered difficulties when exiting through the outlet, resulting in the formation of a superheated air layer that surrounded the battery model. In contrast, the heat transfer mechanism in PCM cooling is natural convection. Based on the inherent buoyancy forces to drive the flow, the air motion occurred easily and orderly through two outlets due to the temperature differences, as seen in Figure 6a. Therefore, the thermal dissipation was more effective than the case of hybrid cooling.

Figure 7 displays the maximum temperature at the end of discharge for the 3C and 5C discharge rates. At the 3C discharge rate, the implementation of different cooling strategies, i.e., forced air convection, PCM, and a hybrid system with forced air and PCM, led to temperature reductions of 8.0 °C, 35.8 °C, and 34.5 °C, respectively. Similarly, the temperature reductions at the 5C discharge rate for the previously mentioned cooling strategies are 10.2 °C, 47.1 °C, and 44.0 °C.

The thermal performance of hybrid cooling in this study was compared with those of Singh et al. [14] to identify heat improvements. Singh et al. [14] examined the effectiveness of forced air cooling integrated with PCM for a group of 25 cylindrical cells. The presence of a 1 mm thick PCM layer and an air inflow velocity of 0.1 m/s reduced the battery temperature by 30 °C at the 1C rate and 45 °C at the 5C rate. Similarly, our analysis yielded a decrease of 34.5 °C at the 3C rate and 43.9 °C at the 5C rate in the battery temperature.

Moreover, the battery temperature in the case of PCM cooling was similar to the use of hybrid cooling, as presented in Figure 7 and Figure 8 in the study of Singh et al. [14]. A similar thing can be observed in our results with a small difference in those two cooling methods. From 5000 rpm to 9000 rpm of fan speed, the maximum battery temperature with hybrid cooling decreased. If the fan speed is high enough (over 9000 rpm), the battery temperature can decline lower than 32.12 °C at the 3C rate and 47.01 °C at the 5C rate, which is obtained with PCM cooling.

The temperature difference across the module was higher in the case of forced air cooling as compared to no cooling, as shown in Figure 8. Despite this, forced air cooling was able to lower the average temperature. The PCM was successful in sustaining substantial temperature uniformity at both the 3C and 5C discharge rates. This demonstrates that the PCM and the hybrid cooling approaches meet the requirement for an optimal operating temperature below 5 °C, especially at the 3C discharge rate.

Figure 9 shows a comparison of the average temperatures of each cell using different cooling methods. It is evident the cell in the middle has a higher temperature among the three cells because of the narrowed airflow passage.

Heat generation was typically concentrated near the current-collecting tabs, as illustrated in Figure 10 and Figure 11. Forced air convection effectively dissipated more thermal energy in the region on the left, which is close to the air inlet in both the 3C and 5C rates. Meanwhile, the cooling performance improved significantly at the bottom region when PCM cooling was employed.

3.2. Effect of Fan Speed

3.2.1. Forced Air Cooling

Although a forced air convection-based cooling system using a fan contributes to battery heat dissipation, the absolute amount of removed heat is insufficient due to the low specific heat and thermal conductivity coefficient of air. Moreover, the presence of dust and pollutants in airflow can further decrease the thermal performance. At a fan speed of 5000 rpm, the maximum temperatures were approximately 83 °C and 84 °C, as shown in Figure 12a. While the temperature decreased only slightly with each 2000 rpm interval, a higher fan speed led to a lower maximum temperature of the battery module at the end of the discharging process.

The temperature distribution became more uniform as the speed of the fan was decreased. Based on Figure 12b, six cases did not meet the criterion of maintaining a temperature difference of less than 5 °C to balance the battery life cycle and efficiency.

Table 5. The maximum temperature uniformity during discharging process with forced air cooling (°C).

C-Rate	5000 rpm	7000 rpm	9000 rpm
3C	15.43	16.17	16.36
5C	26.52	28.69	29.85

shows the temperature differences for discharge rates of 3C and 5C. At the 5C discharge rate, the temperature change per 2000 rpm was highest, specifically around 2.17 °C from 5000 rpm to 7000 rpm. In comparison to the 3C discharge rate, the temperature difference increased rapidly during the 5C discharge at a shorter time period.

3.2.2. Hybrid System of Forced Air and PCM Cooling

The temperature elevation curve of the battery is affected by the PCM and undergoes three stages: solid sensible heating, the latent heat of fusion, and liquid sensible heating. Figure 13a illustrates the limited impact of the fan speed on the temperature reduction of battery module. At the 3C discharge rate, the battery module obtained a peak temperature of under 40 °C, which is within the optimal temperature range. However, when discharged at a rate of 5C, the maximum temperature is above the optimal range, but still remains within safe range below 60 °C.

The combination of forced air and the PCM in the cooling system, depicted in Figure 13b, was capable of maintaining temperature uniformity below 5 °C while discharging at 3C. However, a marked increase in temperature difference was noticed for a 5C discharge rate, especially after the initial 300 s.

Table 6. The maximum temperature uniformity during discharging process with forced air and PCM cooling (°C).

C-Rate	5000 rpm	7000 rpm	9000 rpm
3C	3.53	3.07	2.83
5C	14.65	13.97	13.62

reports the maximum values at various fan speeds.

3.3. Effect of Fan Speed on Thermal Performance in Forced Air and Hybrid Cooling System

Figure 14 illustrates the maximum temperature of the battery module at the end of the discharging process at the 3C and 5C rates. Using the PCM substantially improved the thermal performance in the BTMS for all fan speeds. In the hybrid cooling system, the battery temperature was reduced by 26.5 °C, 23.8 °C, and 22.5 °C at fan speeds of 5000 rpm, 7000 rpm, and 9000 rpm, respectively, for the 3C discharge rate, when compared to forced air cooling. At the 5C discharge rate, these values correspond to 33.7 °C, 32.1 °C, and 30.9 °C.

Figure 15 and Figure 16 present comparisons of the average temperatures of each cell using forced air cooling and hybrid cooling under the 3C and 5C discharge rates. The use of the PCM helped to narrow the ranges of the temperature differences among the three cells. The cell located in the middle of the module always obtained the highest temperature in both cases, including the forced air and hybrid methods, under the 3C rate as well as the 5C rate. Within the cells, the fraction below the average temperature accounted for the larger volume at the end of the 3C discharge with forced air cooling. This was opposite for the hybrid air and PCM cooling.

At the end of the 5C discharging process, the fraction below the average temperature occupied a wider volume within the cells. However, the use of the PCM almost maintained a balance between the higher and lower temperature volumes.

The temperature distribution on the battery surfaces depends on the discharge rate, the cooling method, and the fan speed. Generally, the high temperature near the current-collecting tabs was efficiently dissipated by the fan, as shown in Figure 17 and Figure 18. Although the heat on the left tabs was mostly eliminated, it still persisted on the right tabs because of their distance from the fan. Thus, the thermal energy was primarily concentrated on the right half of the battery module. The use of the PCM created an even heat dissipation, resulting in a uniform temperature distribution. The surface of a battery is divided into a higher temperature region at the top and a lower temperature region at the bottom. At the 3C discharge rate, the temperature within the lower half fluctuated between 30 °C and 30.5 °C, and at the 5C discharge rate, it ranged from around 42 °C to 44 °C. With forced air cooling, the temperature tended to rise from left to right, while with the hybrid air and PCM cooling, the increase was from bottom to top. In addition, the thermal energy was highly concentrated in the right half when using forced air. Meanwhile, it was concentrated in the upper half with the use of hybrid cooling.

Forced air cooling systems enable the direct interaction of air particles with all surfaces of the battery, including the space between the cells. For high fan speeds, thermal energy is effectively directed towards the outlet via airflow. At 3C discharge, with fan speeds of 5000 rpm, 7000 rpm, and 9000 rpm, respectively, the average outlet airflow temperature was 47.1 °C, 44.2 °C, and 42.0 °C. At the end of the 5C discharging process, the corresponding temperatures were measured as 61.8 °C, 56.9 °C, and 54.7 °C. It is worth noting that the air temperature at the inlet was 25 °C. Based on the temperature distribution, battery surfaces may be divided into two regions: a left region with a lower temperature and a right region with a higher temperature.

Using a hybrid cooling system, the battery surfaces were in direct contact with the PCM rather than air. The majority of airflow passes over the front and back faces of the PCMs, which have a large contact area. To ensure the effective cooling of the battery, it is crucial to remove the heat stored in the PCM layers. The measurement of the liquid fraction indicated that the PCM fully transitions to a liquid state at a 5C rate, while it preserves a partially solid phase in the lower region at a 3C rate. At 3C and 5C discharge rates, the positive and negative tabs reached the highest temperatures of approximately 33 °C and 48 °C, respectively. The application of the PCM to those tabs was not feasible due to complex electrical connections and a high risk of electrical failure. Therefore, the forced air cooling method was the primary solution in these regions. The temperature change remained stable in the horizontal direction.

Two monitoring planes were selected (referred to as Plane 1 and Plane 2 in Figure 2) to examine the temperature and velocity distributions near the inlet and outlet of the case, as detailed in Figure 19, Figure 20, Figure 21 and Figure 22. The forced air cooling approach facilitated a smooth airflow through the channel between two cells as well as the space between the cell and the case. However, the airflow was considerably weakened in the rear half around the outlet, resulting in a higher temperature, as shown in Figure 19a and Figure 20a. The higher fan speed produced higher airflow, leading to a broader cooling region. The cell positioned in the middle experienced a lower cooling effect due to the narrow flow channel between cells.

The hybrid cooling method restricted the airflow passage through the channels due to the placement of the PCM between the cells. Plane 2 has a lower temperature than Plane 1, which differs from the forced air cooling system. At the end of 3C discharge, the outlet recorded average airflow temperatures of 25.8 °C, 26.0 °C, and 26.2 °C at 5000 rpm, 7000 rpm, and 9000 rpm, respectively. The recorded airflow temperatures were 27.9 °C, 28.7 °C, and 29.2 °C at the end of 5C discharge. Moreover, on Plane 1, the PCM layers melted completely and became liquidus, which had a low specific heat. On Plane 2, the PCM layers partly transformed and turned mushy, which reflected a huge energy storage. Consequently, there was a considerable difference between the front half and the rear half, as described in Figure 21a and Figure 22a. The velocity distribution on Plane 2 was almost the same at every fan speed under the 3C and 5C discharge rates. The air velocity near the outlet would be high if the fan speed increased, which played a crucial role in the airflow path.

4. Conclusions

This study investigated the thermal behavior of a Li-ion battery module consisting of three pouch cells using a variety of cooling strategies such as forced air convection, PCM, and hybrid forced air and PCM cooling. In particular, to focus on practical application, a specific Li-ion battery model enclosed within a casing was taken into consideration, accompanied by a three-dimensional fan model and the NTGK semi-empirical model for the battery. The main findings of this study can be summarized as follows:

• The PCM is significantly more effective at heat dissipation than forced air. At the end stage of the discharging process, the maximum temperatures were 24.3 °C and 32.2 °C for the 3C and 5C rates, respectively. In contrast, the air cooling method resulted in temperatures of 59.9 °C and 84 °C for the same rates. Interestingly, when using a hybrid cooling system that combines forced air and PCM, the airflow appears to have a negligible effect on thermal management and yields an indiscernible change in temperature.
  This can be attributed to a complex flow pattern that developed in the case as a result of the suboptimal design of the inlet and outlet.
• The hybrid cooling system meets the operational requirements for Li-ion batteries in regard to maximum temperature and temperature uniformity at a discharge rate of 3C across all three fan speeds. However, at a discharge rate of 5C, the maximum temperature remains within a safe range of below 60 °C, but the temperature difference exceeds 5 °C.

On battery surfaces, the thermal energy accumulated in the region near the current-collecting tabs. Specifically, the heat dissipation on the left tabs was improved effectively because the fan was close compared with the right tabs. In addition, the highest temperature was always found in the middle cell. Further studies will be required for the positioning of the inlet and outlet, as well as the effectiveness of combining liquid cooling methods.