In the rapidly evolving field of electric vehicles (EVs), the thermal management of the EV battery pack is a critical factor influencing performance, safety, and longevity. As a researcher focused on enhancing EV battery pack systems, I embarked on a study to optimize the thermal management of a basic battery module using Computational Fluid Dynamics (CFD) thermal analysis. The goal was to design a reliable foundation for a full-scale EV battery pack that meets specific operational requirements, particularly under demanding driving conditions. This work emphasizes forced air cooling with parallel ventilation, a cost-effective and manageable approach for EV battery pack thermal control. Through a combination of experimental validation and iterative CFD simulations, I aimed to refine the design of a basic parallel module, ensuring that its thermal behavior adheres to strict temperature limits, thereby contributing to the overall stability and efficiency of the EV battery pack.
The design process began by defining the parameters for the target EV battery pack. Based on the requirement for a 250 km range under comprehensive driving conditions, the total energy needed was calculated. Using 18650 lithium-ion cells with a nominal voltage of 3.7 V and a capacity of 2200 mAh, I determined the series and parallel configuration. The nominal voltage for the EV battery pack was set at 120 V, leading to a series count a:
$$a = \frac{U}{U_0} = \frac{120}{3.7} \approx 33$$
However, to ensure a margin and account for voltage drops, 34 cells in series were selected. The total energy requirement of 7.54 kWh, considering an effective capacity coefficient η of 0.8, dictated the parallel count b. The system total capacity cz is:
$$c_z = \frac{W}{U \cdot \eta} = \frac{7.54}{120 \times 0.8} \approx 79 \text{ Ah}$$
Thus, the parallel number is:
$$b = \frac{c_z}{c_0} = \frac{79}{2.2} \approx 36$$
Therefore, the EV battery pack configuration was established as 34 series and 36 parallel, totaling 1224 cells. This “parallel-then-series” grouping simplifies Battery Management System (BMS) monitoring and reduces costs for the EV battery pack. The key parameters for the EV battery pack are summarized in Table 1.
| Parameter | Requirement |
|---|---|
| Total Energy | 7.54 kWh |
| Usable Energy | ≥6.7 kWh |
| SOC Operating Window | ≥80% |
| Nominal Voltage | 120 V |
| Operating Voltage Range | 96 V to 144 V |
| Continuous Discharge Power | ≥9.5 kW (for 30 min at 25°C) |
| Operating Temperature Range | -30°C to 60°C |
| Service Life | 8 years or 200,000 km |
The selection of a cooling method is pivotal for the EV battery pack. Forced air cooling with parallel ventilation was chosen due to its structural simplicity, low cost, and ease of maintenance. Compared to serial ventilation, parallel ventilation allows air to flow simultaneously through channels between cells, promoting more uniform heat dissipation and better temperature homogeneity within the EV battery pack module. The overall design strategy involved first optimizing a basic module comprising 36 cells in parallel, then integrating 34 such modules in series to form the complete EV battery pack. This modular approach facilitates targeted thermal management improvements.
The basic module was modeled with a rectangular enclosure. Cells were arranged with a 2 mm gap between them to allow airflow. The enclosure featured an inlet and an outlet on opposite sides, creating a Z-type flow path for forced air cooling. The initial dimensions were set, but as will be shown, this initial design required optimization to meet thermal criteria. The specifications of the 18650 cell used are detailed in Table 2.
| Parameter | Value |
|---|---|
| Model | 18650 |
| Cathode Material | Lithium Cobalt Oxide (LCO) |
| Nominal Capacity | 2200 mAh |
| Nominal Voltage | 3.7 V |
| Voltage Range | 2.75 V to 4.20 V |
| Internal Resistance | Approx. 40.85 mΩ |
Accurate thermal simulation of the EV battery pack module hinges on reliable heat source data from the single cell. I conducted experiments to characterize the internal resistance and heat generation of the 18650 cell. The internal resistance was measured at different states of charge (SOC) under constant current discharge rates of 2 A, 3 A, 4 A, and 5 A. The results indicated that internal resistance increases with higher discharge currents and rises sharply when SOC falls below 0.3. This behavior is crucial for predicting heat generation during operation. The heat generation rate q for the cell was calculated using a simplified Bernardi model, which is commonly applied for lithium-ion batteries in EV battery pack studies:
$$q = I \left( V_{oc} – V \right) – I T \frac{dV_{oc}}{dT}$$
Where I is the current, Voc is the open-circuit voltage, V is the terminal voltage, and T is temperature. For simulation purposes, the volumetric heat generation rate was derived based on experimental data at different discharge rates (C-rates). The thermal properties of the cell, essential for CFD analysis, are listed in Table 3.
| Property | Value |
|---|---|
| Density (ρ) | 2722 kg/m³ |
| Specific Heat Capacity (cp) | 970 J·kg⁻¹·K⁻¹ |
| Radial Thermal Conductivity (λr) | 2.6 W·m⁻¹·K⁻¹ |
| Azimuthal Thermal Conductivity (λθ) | 2.6 W·m⁻¹·K⁻¹ |
| Axial Thermal Conductivity (λz) | 28.0 W·m⁻¹·K⁻¹ |
The volumetric heat generation rates at different discharge C-rates, calculated from the model and experiments, are summarized in Table 4. These values serve as the heat source inputs for subsequent simulations of the EV battery pack module.
| Discharge C-rate | Heat Generation Rate (W/m³) |
|---|---|
| 1.0C | 27,815.75 |
| 1.5C | 50,688.63 |
| 2.0C | 79,538.17 |
| 2.5C | 114,364.39 |
To validate the simulation model, I performed temperature rise experiments on a single cell and compared the results with CFD simulations. The cell was discharged at 1.0C, 1.5C, 2.0C, and 2.5C rates in a temperature-controlled chamber set at 25°C. The CFD model of the single cell was set up with the same thermal properties and boundary conditions: a convection heat transfer coefficient of 5 W·m⁻²·K⁻¹ and an ambient temperature of 25°C. The simulation results showed good agreement with experimental data, with temperature curves closely matching across all discharge rates. For instance, at a 2.5C discharge, both simulation and experiment recorded a final temperature above 70°C. This validation confirmed the accuracy of the heat generation data and the CFD model, providing a solid foundation for analyzing the more complex EV battery pack module.
With confidence in the single-cell model, I proceeded to the thermal simulation of the basic 36-parallel module, which is the core building block of the EV battery pack. The 3D model of the module enclosure was meshed, and a coupled fluid-thermal analysis was performed in CFD software. The cells were defined as volumetric heat sources with a rate corresponding to a 1.0C discharge (27,815.75 W/m³). The inlet boundary condition was set as a velocity inlet with air at 25°C flowing at 5 m/s, while the outlet was defined as a pressure outlet at 0 Pa gauge. The initial simulation revealed that the temperature distribution was unsatisfactory. The maximum temperature (Tmax) reached 32.38°C, and the maximum temperature difference (ΔTmax) within the module was 5.64°C. For reliable operation of an EV battery pack, it is generally required that ΔTmax be kept below 5°C to prevent localized hotspots and ensure cell balance. Therefore, the initial design required optimization.
The first optimization parameter was the inclination angle of the converging chamber (the section guiding air toward the outlet). In the original design, this chamber was horizontal. I modified the model so that the converging chamber formed an angle α with the horizontal, aiming to accelerate airflow near the outlet and improve heat removal. The angle was varied from 1° to 7° (the maximum possible without contacting the cells). Simulations were run for each angle, and Tmax and ΔTmax were recorded. The results are presented in Table 5.
| Inclination Angle α (°) | Maximum Temperature Tmax (°C) | Maximum Temperature Difference ΔTmax (°C) |
|---|---|---|
| 0 (Original) | 32.38 | 5.64 |
| 1 | 32.15 | 5.47 |
| 2 | 31.94 | 5.31 |
| 3 | 31.75 | 5.16 |
| 4 | 31.58 | 5.02 |
| 5 | 31.51 | 4.89 |
| 6 | 31.45 | 4.77 |
| 7 | 31.41 | 4.66 |
The data shows a clear trend: as α increases, both Tmax and ΔTmax decrease. At α = 7°, Tmax is 31.41°C and ΔTmax is 4.66°C. The reduction in ΔTmax below 5°C is promising, but further improvement was sought to enhance robustness. The velocity field analysis indicated that regions near the middle of the module, symmetrically around two central cells, still experienced lower airflow velocities, contributing to higher local temperatures.
To address this, I introduced a flow deflector (baffle) placed vertically in the airflow path, directly behind the two hottest cells. The deflector’s purpose is to redirect airflow toward the problematic zones. The height of the deflector, defined as the distance H from its bottom edge to the floor of the module enclosure, was varied. Simulations were conducted for H = 0 mm, 3 mm, 6 mm, 9 mm, 12 mm, 15 mm, and 18 mm, using the optimized model with α = 7°. The thermal performance metrics for each deflector height are compiled in Table 6.
| Deflector Height H (mm) | Maximum Temperature Tmax (°C) | Maximum Temperature Difference ΔTmax (°C) |
|---|---|---|
| No Deflector | 31.41 | 4.66 |
| 0 | 40.96 | 14.50 |
| 3 | 33.15 | 6.48 |
| 6 | 31.25 | 4.22 |
| 9 | 30.37 | 3.56 |
| 12 | 30.85 | 4.01 |
| 15 | 31.12 | 4.35 |
| 18 | 31.33 | 4.68 |
The results demonstrate a non-linear relationship. At H = 0 mm (deflector touching the floor), the cooling performance severely deteriorates because the deflector blocks most airflow from reaching the rear cells, creating a large temperature disparity. As H increases to 9 mm, Tmax and ΔTmax reach their minimum values of 30.37°C and 3.56°C, respectively. This represents a significant improvement: compared to the model with only the 7° inclination, Tmax decreased by approximately 3.3%, and ΔTmax decreased by about 23.6%. The optimized design effectively balances the airflow, allowing a portion to pass over the deflector and another portion to flow beneath it, ensuring that all cells are adequately cooled. For H values greater than 9 mm, the deflector’s influence diminishes, and performance gradually reverts toward the no-deflector case. Therefore, the optimal configuration for the basic module of the EV battery pack is a converging chamber inclination of 7° combined with a flow deflector height of 9 mm.
The final CFD simulation of this optimized module confirmed excellent thermal performance. The temperature field was uniform, with a Tmax of 30.37°C and a ΔTmax of 3.56°C, well within the desired range of 10°C to 40°C and the ΔTmax < 5°C criterion. The velocity field showed a well-distributed airflow pattern through all channels. This optimized basic module serves as a validated building block. For the full EV battery pack assembly, some modules require extended inlet/outlet sections to align all openings in the same plane when modules are staggered in the pack. A simulation of a module with these extended ports yielded a Tmax of 30.55°C and a ΔTmax of 3.82°C, still meeting the design requirements. This consistency underscores the robustness of the optimization for the EV battery pack application.
The success of this optimization can be analyzed through the principles of fluid dynamics and heat transfer. The forced convection heat removal rate from the cells can be expressed as:
$$q_{conv} = h A_s (T_s – T_f)$$
where h is the convective heat transfer coefficient, As is the cell surface area, Ts is the cell surface temperature, and Tf is the coolant air temperature. The modifications to the converging chamber and the addition of the deflector effectively increase the local air velocity around the hottest cells, thereby enhancing the convective coefficient h. The parallel ventilation scheme ensures that the airflow distribution factor is more uniform across the EV battery pack module. The temperature uniformity, critical for cell longevity in an EV battery pack, is quantified by the maximum temperature difference ΔTmax. The optimization reduced ΔTmax from an initial 5.64°C to 3.56°C, a 37% improvement, significantly mitigating thermal stress gradients.
In discussing the implications for a complete EV battery pack, this optimized module design provides a thermal management solution that is scalable and modular. By ensuring each basic module operates within safe thermal limits, the overall EV battery pack’s reliability and safety are enhanced. The forced air cooling approach, while less capable than liquid cooling for extremely high loads, remains a viable and economical choice for many EV applications, especially when optimized through such CFD-guided design. Future work could explore the integration of these modules into the full pack, investigating inter-module airflow and the performance under dynamic driving cycles. Additionally, the impact of higher discharge rates, such as during fast charging or aggressive acceleration, on the EV battery pack’s thermal behavior warrants further study, possibly necessitating adaptive cooling controls.
In conclusion, this study successfully demonstrates the optimization of a basic module for an EV battery pack using CFD thermal analysis. Starting from the specification of a 34-series, 36-parallel EV battery pack, I focused on the thermal management of a single 36-parallel module employing forced air cooling with parallel ventilation. Experimental validation of single-cell heat generation ensured accurate simulation inputs. Initial CFD simulations revealed inadequate cooling, prompting a two-stage optimization: first, adjusting the converging chamber inclination to 7°, and second, incorporating a flow deflector with an optimal height of 9 mm. The final design achieved a maximum temperature of 30.37°C and a maximum temperature difference of 3.56°C under a 1.0C discharge condition, satisfying the stringent thermal criteria for EV battery pack operation. This optimized basic module forms a reliable foundation for constructing a full-scale, thermally managed EV battery pack, contributing to the advancement of safer and more efficient electric vehicles. The methodologies and results presented here underscore the value of iterative CFD simulation in the design and refinement of thermal management systems for modern EV battery packs.