Effective thermal management is a cornerstone for the safety, longevity, and performance of electric vehicle (EV) battery packs. Managing the heat generated during charge and discharge cycles is critical. Suboptimal thermal conditions can accelerate capacity fade, increase internal resistance, and in extreme cases, lead to thermal runaway. For large-format lithium-ion batteries commonly used in energy storage and electric vehicles, maintaining an optimal operating temperature range, typically between 20°C and 40°C, with a maximum temperature difference below 5°C within the EV battery pack, is essential. This study investigates the thermal performance of three distinct liquid cooling plate designs for a 280 Ah lithium iron phosphate (LFP) EV battery pack through Computational Fluid Dynamics (CFD) simulation, providing a comparative analysis to guide design selection.
Liquid cooling has emerged as a superior method for battery thermal management due to its high heat transfer coefficient, compactness, and precise temperature control compared to air-cooling or phase-change material systems. The design of the cold plate—its internal flow channel geometry, manufacturing process, and integration method—directly impacts the thermal uniformity and maximum temperature within the EV battery pack. This paper focuses on evaluating three common cold plate manufacturing processes: extruded aluminum profiles, stamped brazed aluminum plates, and extruded “harmonica” tubes. Each offers different trade-offs in terms of thermal performance, weight, structural integrity, and cost.
1. Physical Model of the EV Battery Pack and Cooling Designs
The subject of this analysis is a modular EV battery pack with a configuration of 1P52S (1 parallel string of 52 series-connected cells), subdivided into four modules of 1P13S each. Each cell is a large-format prismatic LFP battery with a nominal capacity of 280 Ah and dimensions of 174.4 mm × 207 mm × 72.4 mm. The primary heat source is the internal resistance of the cells. Under a 0.5C continuous operating condition, the average heat generation per cell is approximately 10.5 W, leading to a total heat load of 546 W for the entire pack. Adjacent cells are separated by aerogel insulating pads to mitigate thermal propagation, and cells are interconnected via busbars.
Three cooling strategies, defined by their cold plate design and integration, are modeled:
- Design-1: Extruded & Friction Stir Welded Aluminum Profile. This design serves as the structural baseplate of the EV battery pack. The cold plate is created by joining two symmetrical extruded aluminum sections via friction stir welding. The internal flow channels are formed by the ribs of the extrusion profile, creating a serpentine path of rectangular channels. A thermal interface material (TIM) is placed between the cell bottoms and the cold plate surface to fill air gaps and enhance heat conduction. Its advantages include high structural strength and good surface flatness, but it tends to be heavy.
- Design-2: Stamped & Brazed Aluminum Plate. This design also functions as a baseplate-cooled system. The flow channels are formed by stamping a complex serpentine-parallel path into an aluminum sheet, which is then brazed to a flat cover plate. This process allows for highly customized and optimized flow channel designs. Cooling liquid enters a manifold, distributes into four parallel flow branches covering the pack footprint, and then collects into an outlet manifold. It offers excellent heat transfer performance and good pressure resistance but requires tooling investment.
- Design-3: Extruded “Harmonica” Tube. This design employs a side-cooling approach. Thin, multi-port extruded aluminum tubes (resembling a harmonica) are sandwiched between adjacent battery modules. Thermally conductive adhesive is applied on both sides of the tube to bond it to the large faces of the prismatic cells. The coolant flows in a U-shaped path through the upper and lower levels of the tube. This design is lightweight and low-cost but offers a smaller effective contact area with the cells.
2. Theoretical Foundation for Heat Transfer Analysis
The thermal analysis of an EV battery pack is governed by fundamental principles of heat transfer. The following equations establish the theoretical framework for the CFD simulation.
2.1 Cell Energy Balance and Heat Diffusion
Assuming steady-state conditions, constant material properties, and neglecting internal convection and radiation, the energy balance for a single cell is given by the heat diffusion equation with an internal heat generation term:
$$ \rho C_p \frac{\partial T}{\partial t} = \lambda_x \frac{\partial^2 T}{\partial x^2} + \lambda_y \frac{\partial^2 T}{\partial y^2} + \lambda_z \frac{\partial^2 T}{\partial z^2} + \dot{q} $$
where \( \rho \) is density, \( C_p \) is specific heat capacity, \( T \) is temperature, \( \lambda_x, \lambda_y, \lambda_z \) are thermal conductivities along the principal axes, and \( \dot{q} \) is the volumetric heat generation rate. The heat generation from a cell under operating current \( I \) and internal resistance \( R \) is \( P_{cell} = I^2R \).
2.2 Cold Plate Hydraulics and Fluid Energy Transport
The flow within the cold plate channels is characterized by the hydraulic diameter \( D_h \), crucial for non-circular ducts:
$$ D_h = \frac{4 A_c}{P_w} $$
where \( A_c \) is the cross-sectional area of the flow channel and \( P_w \) is its wetted perimeter. The flow regime is determined by the Reynolds number:
$$ Re = \frac{\rho_f V_m D_h}{\mu} $$
with \( V_m = \frac{\dot{V}}{A_c} \) being the mean flow velocity, \( \dot{V} \) the volumetric flow rate, \( \rho_f \) the fluid density, and \( \mu \) the dynamic viscosity. For \( Re > 2300 \), the flow is turbulent, which enhances mixing and heat transfer. The energy transport for the turbulent coolant flow is modeled using the Reynolds-Averaged Navier-Stokes (RANS) equations with a standard k-ε turbulence model.
2.3 Cooling System Sizing
The required coolant flow rate \( \dot{V} \) to remove the total heat load \( P_{pack} \) from the EV battery pack can be estimated from a simple energy balance on the coolant:
$$ \dot{V} = \frac{P_{pack}}{\rho_m C_{p,m} \Delta T} $$
where \( \rho_m \) and \( C_{p,m} \) are the density and specific heat of the coolant (50% ethylene glycol solution), and \( \Delta T \) is the desired temperature rise between inlet and outlet. Targeting a \( \Delta T \) of ~2°C leads to a designed flow rate of 5 L/min for this study.
3. CFD Simulation Setup and Material Properties
The three-dimensional models of the EV battery pack assemblies were simplified by omitting small components like wires and connectors. The cells were modeled as homogeneous anisotropic blocks. Steady-state simulations were performed using ANSYS Fluent with the realizable k-ε turbulence model for the fluid domain. The boundary conditions were kept consistent for a fair comparison:
- Coolant: 50% Ethylene Glycol solution, inlet temperature = 20°C, flow rate = 5 L/min.
- Ambient Temperature: 25°C.
- Total Heat Load: 546 W (10.5 W/cell × 52 cells), applied uniformly as a volumetric source within each cell.
- Cell and Component Materials: Properties are assigned as per the table below.
| Material | Density (kg/m³) | Specific Heat (J/kg·K) | Thermal Conductivity (W/m·K) |
|---|---|---|---|
| Lithium-ion Cell | 2152.9 | 1060 | x: 3.56, y: 9.04, z: 11.0 |
| Aluminum (Cold Plate/Busbar) | 2702 | 903 | 237 |
| Thermal Interface Material | 2300 | 1000 | 2.0 |
| Aerogel Insulation Pad | – | – | ≤ 0.04 |
| 50% Ethylene Glycol | 1073.35 | 3281 | 0.38 |
| Air | 1.225 | 1006.43 | 0.0242 |
4. Simulation Results and Comparative Analysis
The temperature contours of the cells within the EV battery pack for the three designs are analyzed to assess maximum temperature (\(T_{max}\)) and temperature uniformity, often represented by the maximum temperature difference (\(\Delta T_{max}\)).
4.1 Pack-Level Temperature Distribution
All three designs successfully maintained the average cell temperature within the safe operating window (20-40°C). However, significant differences in thermal uniformity were observed.
- Design-1 (Extruded Profile): The highest temperature region was located on the inner cells of the module closest to the coolant outlet. The lowest temperature was on the outer cells near the inlet. The pack-level \(\Delta T_{max}\) was approximately 1.59°C.
- Design-2 (Stamped Plate): This design showed the most uniform temperature distribution. The parallel flow path design provided more even cooling across the entire pack base. The pack-level \(\Delta T_{max}\) was the lowest at 1.29°C, and the overall \(T_{max}\) was also the lowest.
- Design-3 (Harmonica Tube): The side-cooling approach led to a different gradient. The cells in direct contact with the cooling plates were cooler, while the inner cells of the central modules, farther from the cooling surface, were warmer. This resulted in the largest pack-level \(\Delta T_{max}\) of 2.22°C among the three designs.
The transient simulation over a full charge-discharge cycle showed that temperatures peaked towards the end of the discharge due to increasing internal resistance. The peak temperatures from the simulation were: \(T_{max-1}\) = 32.1°C, \(T_{max-2}\) = 31.5°C, and \(T_{max-3}\) = 32.7°C.
4.2 Single-Cell Temperature Gradient
A critical metric is the temperature difference within a single cell (\(\Delta T_{cell}\)), as large gradients induce mechanical stress. The results were:
- Design-1 & Design-2 (Bottom Cooling): Both showed a vertical gradient with the hottest point at the top of the cell and the coolest at the bottom in contact with the cold plate. \(\Delta T_{cell}\) was 4.7°C for Design-1 and 4.6°C for Design-2.
- Design-3 (Side Cooling): Cooling from the large face reduced the through-plane gradient. \(\Delta T_{cell}\) was the smallest at 4.3°C, demonstrating an advantage of side-cooling for individual cell uniformity.
| Design | Cooling Method | Max Pack Temp, \(T_{max}\) (°C) | Max Pack \(\Delta T\) (°C) | Max Single-Cell \(\Delta T\) (°C) | Key Characteristics |
|---|---|---|---|---|---|
| Design-1 | Extruded Profile (Bottom) | 32.1 | 1.59 | 4.7 | Good performance, structurally robust, heavier. |
| Design-2 | Stamped Plate (Bottom) | 31.5 | 1.29 | 4.6 | Best overall pack uniformity and cooling. Flexible design. |
| Design-3 | Harmonica Tube (Side) | 32.7 | 2.22 | 4.3 | Lightweight, lower cost, poorer pack uniformity, best single-cell gradient. |
5. Experimental Validation
Based on the superior simulation performance, the stamped brazed aluminum plate (Design-2) was selected for experimental validation. A thermal test was conducted on the EV battery pack assembly under conditions matching the simulation: coolant (50% EG) inlet at 20°C, flow rate of 5 L/min, and a controlled heat load of 546 W. Temperature sensors were placed at strategic locations corresponding to the simulated points.
The experimental temperature profile over state-of-charge (SOC) aligned well with the simulation trends. Both showed temperature rise during charge/discharge and a drop during rest periods. The peak temperature during the test reached approximately 33°C, which is within a reasonable margin of the simulated 31.7°C, considering real-world losses and environmental factors. Crucially, the measured temperature difference within the pack remained below 2°C, confirming the design’s ability to maintain the required thermal uniformity for the EV battery pack.
6. Conclusion
This study conducted a comprehensive CFD-based thermal analysis of three liquid-cooling plate designs for a large-format EV battery pack. The key findings are:
- All three manufacturing processes—extruded profiles, stamped brazed plates, and harmonica tubes—can form the basis of a functional thermal management system that keeps the battery within a safe temperature range.
- The stamped brazed aluminum plate (Design-2) demonstrated the best overall thermal performance at the pack level, achieving the lowest maximum temperature (31.5°C) and the best temperature uniformity (\(\Delta T_{pack}\) = 1.29°C). Its design flexibility allows for optimized flow distribution.
- The harmonica tube design (Design-3), while resulting in poorer pack-level uniformity, produced the smallest temperature gradient within a single cell (4.3°C), highlighting a potential benefit of side-cooling approaches for reducing cell-internal stress.
- The experimental validation of the stamped plate design confirmed the accuracy of the CFD model, with measured data closely following the simulated trends and meeting all critical thermal specifications for the EV battery pack.
The selection of a cooling strategy for an EV battery pack involves a multi-objective optimization considering thermal performance, weight, structural requirements, manufacturability, and cost. This analysis provides a clear quantitative comparison of thermal performance to inform that decision. For applications where pack-level temperature uniformity is the paramount concern, the stamped brazed plate design is the superior choice. Future work could explore hybrid approaches or advanced channel geometries to further minimize the single-cell temperature gradient while maintaining excellent pack uniformity.