As the global shift toward sustainable transportation accelerates, electric vehicles (EVs) have emerged as a pivotal technology, particularly in markets like China EV, where government policies and consumer demand drive rapid adoption. Unlike traditional internal combustion engine vehicles, EVs rely on battery packs for power, making battery performance critical to vehicle range, efficiency, and safety. Among various battery types, lithium-ion batteries are widely used in electric vehicles due to their high energy density, low self-discharge rate, and long cycle life. However, these batteries generate significant heat during charging and discharging cycles, which, if not properly managed, can lead to reduced lifespan, thermal runaway, and safety hazards. Thus, effective battery thermal management systems (BTMS) are essential for maintaining optimal operating temperatures and ensuring the reliability of electric vehicles.
Common BTMS approaches include air cooling, liquid cooling, and phase change material (PCM) cooling. Air cooling systems are simple and cost-effective but suffer from low thermal conductivity and poor temperature uniformity, limiting their effectiveness in high-demand applications. Liquid cooling, often using water or glycol-based coolants, offers superior heat transfer but adds complexity and weight due to piping or cold plates. PCM cooling leverages latent heat absorption but faces challenges like material degradation and leakage over time. To address these limitations, I propose a novel air-liquid hybrid cooling BTMS that combines the strengths of both methods, enhancing heat dissipation and temperature uniformity for electric vehicle battery packs. This study focuses on a module composed of 16 cylindrical 18650 lithium-ion batteries, arranged in a 4×4 configuration, integrated with a herringbone fin microchannel cold plate. Using computational fluid dynamics (CFD) simulations, I develop a three-dimensional fluid-solid coupling thermal model to evaluate the system’s performance under various operating conditions.
The geometric model of the BTMS is designed with dimensions of 109 mm × 109 mm × 70 mm (length × width × height), featuring an aluminum shell with a wall thickness of 2 mm. The herringbone fin microchannel cold plate, made of aluminum, has a thickness of 3 mm, with microchannels measuring 4 mm × 1.5 mm in cross-section. Batteries are connected to the cold plate via sleeves, with horizontal and vertical spacing set to 25 mm. This configuration aims to maximize heat exchange area and improve cooling efficiency. The battery parameters, based on standard 18650 cells, include a length of 65 mm, diameter of 18 mm, nominal voltage of 3.2 V, and capacity of 1.35 Ah. Key thermal properties are summarized in Table 1.
| Parameter | Value |
|---|---|
| Length (mm) | 65 |
| Diameter (mm) | 18 |
| Nominal Voltage (V) | 3.2 |
| Capacity (Ah) | 1.35 |
| Equivalent Density (kg/m³) | 2018 |
| Equivalent Specific Heat Capacity (J/(kg·K)) | 1282 |
| Radial Thermal Conductivity (W/(m·K)) | 0.9 |
| Axial Thermal Conductivity (W/(m·K)) | 2.7 |
To model battery heat generation, I employ the Bernardi equation, which accounts for irreversible and reversible heat effects. The equation is given by:
$$Q = I^2 r + I T \frac{\partial U_{OCV}}{\partial T} / V_{battery}$$
where \( Q \) is the heat generation rate per unit volume (W/m³), \( I \) is the discharge current (A), \( r \) is the internal resistance (Ω), \( T \) is the battery temperature (K), \( U_{OCV} \) is the open-circuit voltage (V), and \( V_{battery} \) is the battery volume (m³). Assuming uniform heat generation within the battery and constant material properties, the heat generation rates at different discharge rates are calculated and listed in Table 2. These values are incorporated as heat source terms in the CFD simulations.
| Discharge Rate | Heat Generation Rate (W/m³) |
|---|---|
| 1C | 5318 |
| 2C | 19452 |
| 3C | 42400 |
| 4C | 74163 |
The numerical model is built using Ansys Fluent, based on the RNG k-ε turbulence model to capture complex flow and heat transfer phenomena. The governing equations include the continuity, momentum, and energy equations. For an incompressible fluid, the continuity equation is:
$$\frac{\partial \rho}{\partial t} + \frac{\partial (\rho u)}{\partial x} + \frac{\partial (\rho v)}{\partial y} + \frac{\partial (\rho w)}{\partial z} = 0$$
where \( \rho \) is density (kg/m³), \( t \) is time (s), and \( u \), \( v \), and \( w \) are velocity components in the x, y, and z directions (m/s), respectively. The momentum equations in tensor form are:
$$\frac{\partial (\rho u_i)}{\partial t} + \frac{\partial (\rho u_i u_j)}{\partial x_j} = -\frac{\partial p}{\partial x_i} + \frac{\partial \tau_{ij}}{\partial x_j} + F_i$$
where \( p \) is pressure (Pa), \( \tau_{ij} \) is the stress tensor, and \( F_i \) represents body forces. The energy equation is expressed as:
$$\frac{\partial (\rho T)}{\partial t} + \frac{\partial (\rho u_j T)}{\partial x_j} = \frac{\partial}{\partial x_j} \left( \frac{k_T}{c_p} \frac{\partial T}{\partial x_j} \right) + S_T$$
where \( k_T \) is the thermal conductivity (W/(m·K)), \( c_p \) is the specific heat capacity (J/(kg·K)), and \( S_T \) is the heat source term. The RNG k-ε model equations for turbulent kinetic energy \( k \) and dissipation rate \( \varepsilon \) are:
$$\frac{\partial (\rho k)}{\partial t} + \frac{\partial (\rho u_i k)}{\partial x_i} = \frac{\partial}{\partial x_j} \left( \alpha_k \mu_{\text{eff}} \frac{\partial k}{\partial x_j} \right) + G_k + G_b – \rho \varepsilon$$
$$\frac{\partial (\rho \varepsilon)}{\partial t} + \frac{\partial (\rho u_i \varepsilon)}{\partial x_i} = \frac{\partial}{\partial x_j} \left( \alpha_\varepsilon \mu_{\text{eff}} \frac{\partial \varepsilon}{\partial x_j} \right) + C_{1\varepsilon} \frac{\varepsilon}{k} (G_k + C_{3\varepsilon} G_b) – C_{2\varepsilon} \rho \frac{\varepsilon^2}{k} – R_\varepsilon$$
Here, \( G_k \) and \( G_b \) represent turbulence generation due to velocity gradients and buoyancy, \( \mu_{\text{eff}} \) is the effective viscosity, and \( C_{1\varepsilon} \), \( C_{2\varepsilon} \), \( C_{3\varepsilon} \), \( \alpha_k \), and \( \alpha_\varepsilon \) are model constants. The term \( R_\varepsilon \) is a modification factor defined as:
$$R_\varepsilon = \frac{C_\mu \rho \eta^3 (1 – \eta / \eta_0)}{1 + \beta \eta^3} \frac{\varepsilon^2}{k}$$
where \( \eta = S k / \varepsilon \), \( S \) is the strain rate modulus, and \( C_\mu \), \( \eta_0 \), and \( \beta \) are constants. The thermal properties of materials used in the BTMS are provided in Table 3.
| Material | Density (kg/m³) | Specific Heat Capacity (J/(kg·K)) | Thermal Conductivity (W/(m·K)) |
|---|---|---|---|
| Water | 998.21 | 4128 | 0.6 |
| Air | 1.22 | 1006.43 | 0.0242 |
| Aluminum | 2719 | 871 | 202.4 |
Boundary conditions are set to simulate realistic operating scenarios. The initial temperature for the battery module, cold plate, cooling water, air, and environment is 298.15 K. The water and air inlets are defined as velocity inlets (\( v_{\text{water}} \) and \( v_{\text{air}} \)), while outlets are set as pressure outlets with zero gauge pressure. Convective heat transfer with the environment is applied to external surfaces, with a heat transfer coefficient of 10 W/(m²·K). The interfaces between batteries, cold plate, and air are treated as fluid-solid coupling walls, with batteries modeled as heat sources. The SIMPLE algorithm is used for pressure-velocity coupling, and second-order upwind schemes discretize momentum and energy equations. Convergence criteria are set to residuals of 10⁻⁴ for continuity and velocity, and 10⁻⁶ for energy.
Mesh independence is critical for accurate CFD results. I conducted a grid sensitivity analysis by varying the number of elements and monitoring the maximum temperature of the BTMS, pressure drop across microchannels, average mesh quality, and average orthogonal quality. As shown in Table 4, when the grid count exceeds 3,001,362, these parameters stabilize, indicating mesh independence. Thus, I selected a grid with approximately 3 million elements for all simulations to balance computational efficiency and accuracy.
| Grid Count | Maximum Temperature (K) | Microchannel Pressure Drop (Pa) | Average Mesh Quality | Average Orthogonal Quality |
|---|---|---|---|---|
| 1,500,000 | 302.5 | 210 | 0.85 | 0.82 |
| 2,500,000 | 301.8 | 205 | 0.88 | 0.85 |
| 3,001,362 | 301.2 | 200 | 0.90 | 0.88 |
| 3,500,000 | 301.1 | 199 | 0.91 | 0.89 |
I compared the performance of three cooling strategies: air cooling alone, liquid cooling alone, and the air-liquid hybrid cooling. Under a discharge rate of 2C, with inlet water velocity \( v_{\text{water}} = 0.05 \, \text{m/s} \) and inlet air velocity \( v_{\text{air}} = 0.30 \, \text{m/s} \), the hybrid system demonstrated superior temperature uniformity and lower maximum temperature. For air cooling, the temperature gradient was significant, with hotspots near the outlet reaching up to 310 K. Liquid cooling reduced the maximum temperature to approximately 300 K but still showed variations along the flow direction. The hybrid approach achieved the best results, with a maximum temperature of 299.5 K and improved uniformity due to combined convective heat transfer. The pressure distribution in the air side and microchannels was analyzed; the air side pressure drop was less than 10 Pa, while the microchannels exhibited a drop of around 200 Pa, indicating efficient flow management.
To optimize the hybrid system, I investigated the effect of inlet water velocity (\( v_{\text{water}} \)) while keeping \( v_{\text{air}} = 0.30 \, \text{m/s} \). As \( v_{\text{water}} \) increased from 0.01 m/s to 0.10 m/s, the maximum battery temperature decreased from 301.84 K to 299.97 K, a reduction of 1.87 K. Temperature uniformity also improved, as higher flow rates enhanced heat removal. The relationship between inlet water velocity and maximum temperature can be approximated by:
$$T_{\text{max}} = 302.0 – 20.5 \cdot v_{\text{water}} + 105.0 \cdot v_{\text{water}}^2$$
This quadratic fit highlights the diminishing returns at higher velocities. For instance, increasing \( v_{\text{water}} \) from 0.05 m/s to 0.10 m/s only reduced \( T_{\text{max}} \) by 0.27 K, suggesting an optimal range for practical applications in electric vehicles.
Next, I examined the impact of inlet air velocity (\( v_{\text{air}} \)) with fixed \( v_{\text{water}} = 0.10 \, \text{m/s} \). Raising \( v_{\text{air}} \) from 0.10 m/s to 3.00 m/s lowered the maximum temperature from 302.30 K to 300.34 K, a decrease of 1.96 K. However, the effect was less pronounced compared to water velocity changes. The temperature reduction followed a logarithmic trend:
$$T_{\text{max}} = 302.5 – 0.65 \cdot \ln(v_{\text{air}} + 0.1)$$
This indicates that beyond 1.00 m/s, further increases in air velocity yield minimal benefits. For example, boosting \( v_{\text{air}} \) from 1.00 m/s to 3.00 m/s only reduced \( T_{\text{max}} \) by 0.62 K. Thus, in designing BTMS for electric vehicles, prioritizing water flow rate optimization is more effective for heat dissipation.
The results underscore the advantages of the air-liquid hybrid cooling BTMS for electric vehicles, especially in the context of China EV markets where high-performance thermal management is crucial for battery longevity and safety. The hybrid system mitigates the weaknesses of individual cooling methods, offering a balanced solution for temperature control. Future work could explore multi-objective optimization involving factors like energy consumption, cost, and weight, or integrate advanced materials to further enhance performance. Additionally, real-world testing in varying climatic conditions would validate these findings and support widespread adoption in the electric vehicle industry.
In conclusion, this study demonstrates that the air-liquid hybrid cooling BTMS significantly improves heat dissipation and temperature uniformity compared to standalone air or liquid cooling systems. Inlet water velocity has a more substantial impact on maximum temperature reduction than inlet air velocity, guiding optimization efforts. For electric vehicle applications, particularly in growing markets like China EV, this approach offers a reliable path toward safer and more efficient battery thermal management, contributing to the broader goals of sustainable transportation.