The growing global demand for energy and increasing scarcity of resources have positioned electric vehicles (EVs) as a critical pathway to alleviate energy pressures. Lithium-ion batteries, favored for their high voltage, low self-discharge rate, high energy density, and good stability, have become the primary power source for modern EVs. However, the performance and safety of these batteries are intrinsically linked to their operating temperature. Excessive heat accumulation can severely degrade battery capacity, power output, and cycle life, with the extreme risk of thermal runaway. Therefore, an effective Battery Thermal Management System (BTMS) is not an accessory but a fundamental necessity for safe and efficient EV operation.
The core function of any battery management system is to maintain the battery pack within an optimal temperature range (typically 15°C to 35°C) and ensure minimal temperature variation (ideally below 5°C) between cells. Among various cooling technologies—including air cooling, phase change material (PCM) cooling, and heat pipe cooling—liquid cooling stands out for its superior heat transfer coefficient and precise thermal control, making it the dominant solution in high-performance automotive battery management systems. Indirect liquid cooling, where a coolant flows through channels embedded in or attached to cold plates that contact the cells, is particularly prevalent due to its effectiveness and reliability.
The performance of such a liquid-cooled battery management system is heavily influenced by the design of the cooling channel geometry. An ideal design must achieve a delicate balance: maximizing heat dissipation to lower the average pack temperature, promoting temperature uniformity to prevent localized hot spots, and minimizing the pumping power required to circulate the coolant. This study addresses the challenge of thermal management under high discharge rates by proposing and optimizing a novel parallel turbulent channel architecture. We systematically analyze the thermal and hydraulic performance of several baseline channel designs before introducing an optimized structure that synthesizes their strengths. Furthermore, we investigate the impact of critical operating parameters—coolant flow rate and inlet temperature—on the system’s performance, providing guidance for the integrated design and control of advanced battery thermal management systems.
1. System Modeling and Methodology
1.1 Geometric Model and Baseline Designs
The battery module under study consists of 18 cylindrical 21700 lithium-ion cells arranged in a 3×6 configuration. The key parameters of the battery cell are summarized in Table 1. A thermal interface material (silicone pad) is placed between adjacent cells to facilitate heat conduction to the cooling plate. The cooling plate is made of aluminum alloy, and the coolant is a 50% by volume ethylene glycol-water solution.
To establish a performance baseline and guide the optimization, five distinct cold plate channel designs (Structures a through e) were initially conceived, as illustrated schematically. Structures a, b, and c are parallel-flow designs with varying manifold layouts. Structure a uses a simple “H” pattern, Structure b employs a more centralized inlet/outlet, and Structure c features a symmetrical layout. Structures d and e are serial-flow designs; Structure d has a standard serpentine channel, while Structure e introduces periodic 90-degree bends acting as turbulence promoters within a serpentine path. All channels have a width of 8 mm, a wall thickness of 2 mm, and a 2 mm gap between the cell wall and the cooling plate surface. The inlet and outlet extensions are 20 mm long.
| Parameter | Value |
|---|---|
| Rated Capacity | 4.9 Ah |
| Rated Voltage | 3.6 V |
| Charge Cut-off Voltage | 4.2 V |
| Discharge Cut-off Voltage | 2.5 V |
| Charge Temperature Range | 0 ~ 45 °C |
| Discharge Temperature Range | -20 ~ 60 °C |
1.2 Electro-Thermal Model
The heat generation rate within a battery cell is a critical input for the thermal simulation of the battery management system. This study employs the widely accepted Bernardi model to calculate the volumetric heat generation ($q$). The model accounts for both irreversible Joule heating and reversible entropic heating:
$$ q = \frac{I}{V} \left[ (U – E) + T \frac{dE}{dT} \right] = \frac{I}{V} \left[ I R + T \frac{dE}{dT} \right] $$
where $I$ is the current (A, positive for discharge), $V$ is the cell volume (m³), $U$ is the terminal voltage (V), $E$ is the open-circuit voltage (V), $T$ is the absolute cell temperature (K), $R$ is the internal resistance (Ω), and $\frac{dE}{dT}$ is the entropic heat coefficient (V/K). For a 2C discharge rate, the current $I$ is 9.8 A.
1.3 Material Properties and Boundary Conditions
The thermophysical properties of all materials used in the model are listed in Table 2. The battery is modeled with orthotropic thermal conductivity, reflecting its layered internal structure. The initial temperature for the battery, cooling plate, and coolant is set to the ambient temperature of 25°C. Natural convection with a heat transfer coefficient of 5 W/(m²·K) is applied to all external surfaces. For the coolant domain, a velocity inlet boundary condition is specified at the inlet, and a pressure outlet condition (atmospheric pressure) is set at the outlet. The base coolant inlet temperature is 25°C, and the base inlet velocity is 0.05 m/s. These parameters form the foundation for evaluating the thermal management system’s performance.
| Material | Density (kg/m³) | Specific Heat (J/(kg·K)) | Thermal Conductivity (W/(m·K)) | Dynamic Viscosity (Pa·s) |
|---|---|---|---|---|
| Battery Cell | 2673 | 831 | Kx=Ky=1.2, Kz=20.9 | – |
| Aluminum Plate | 2690 | 900 | 218 | – |
| Coolant (50% EG) | 1082 | 3300 | 0.4 | 0.0034 |
| Silicone Pad | 1700 | 1360 | 2 | – |
1.4 Numerical Setup and Validation
Computational Fluid Dynamics (CFD) simulations were performed using a commercial finite volume solver. The energy equation and the Reynolds-Averaged Navier-Stokes (RANS) equations with a standard k-ε turbulence model were solved. A mesh independence study was conducted for the final optimized design to ensure the accuracy of the results. The key monitoring parameters were the battery pack’s maximum temperature ($T_{max}$), average temperature ($T_{avg}$), and maximum temperature difference ($ΔT_{max}$). The results stabilized when the mesh size was refined to 1.25 mm, which was subsequently adopted for all simulations.
Experimental validation was carried out by constructing a prototype of the optimized cooling plate (Structure f) and testing it on a battery module under a 2C discharge cycle. The simulated surface temperatures showed excellent agreement with the thermocouple measurements, with a maximum error of approximately 3% (0.6 °C). This close correlation validates the fidelity of the numerical model and confirms its suitability for analyzing and optimizing the battery thermal management system.
2. Analysis of Baseline Cooling Channel Structures
2.1 Thermal Performance Assessment
The thermal performance of the five baseline structures (a-e) under a 2C discharge is quantified by three metrics: the maximum cell temperature in the pack ($T_{max}$), the maximum temperature difference within the pack ($ΔT_{max}$), and the volume-weighted average pack temperature ($T_{avg}$). A superior battery management system aims to minimize all three values simultaneously.
The analysis reveals a clear hierarchy in cooling effectiveness. Structure a, with its simple “H” pattern, performs the worst, exhibiting the highest $T_{max}$ and $T_{avg}$. Structure c, featuring a symmetrical parallel layout, shows significant improvement over Structures a and b. This is attributed to more balanced flow distribution and shorter, more uniform flow paths to the central cells, reducing thermal accumulation. Among serial designs, Structure e outperforms Structure d. The periodic bends in Structure e disrupt the hydraulic boundary layer, enhance local fluid mixing, and increase the effective heat transfer surface area, leading to better overall cooling. Consequently, Structure e achieves the best thermal performance among the baseline set, with the lowest $T_{max}$ (32.032°C), $T_{avg}$ (29.160°C), and a $ΔT_{max}$ of 4.197°C. These results are summarized in Table 3, highlighting the advantage of induced flow disturbance for heat transfer enhancement in a battery thermal management system.
| Structure | $T_{max}$ (°C) | $ΔT_{max}$ (°C) | $T_{avg}$ (°C) | Flow Type |
|---|---|---|---|---|
| a | 34.171 | 6.336 | 30.326 | Parallel |
| b | 33.285 | 5.450 | 29.440 | Parallel |
| c | 32.648 | 4.813 | 28.979 | Parallel |
| d | 33.221 | 5.386 | 29.262 | Serial |
| e | 32.032 | 4.197 | 29.160 | Serial |
2.2 Flow Field and Hydraulic Analysis
While thermal performance is paramount, the hydraulic performance, characterized by pressure drop ($ΔP$), is equally critical for the energy efficiency of the battery management system. A high pressure drop requires a more powerful and energy-consuming pump. Analysis of the flow fields provides insights into this trade-off.
The parallel structures (a-c) exhibit lower average flow velocities in the channels compared to the serial structures (d-e). This is due to flow division at the manifolds, which reduces the volumetric flow rate per channel. While this leads to a lower pressure drop, it also limits convective heat removal. In contrast, the serial structures maintain a high, single-stream flow velocity, enhancing heat transfer but at the cost of significantly higher pressure drop, especially in Structure e where bends induce additional flow resistance. Structure e, despite its excellent cooling, suffers from the highest pressure drop. This analysis underscores a key design conflict in liquid-cooled BMS: serial designs favor heat transfer but penalize pumping power, while parallel designs offer lower pressure drop but can suffer from flow maldistribution and reduced local cooling.
3. Development of an Optimized Parallel Turbulent Structure
3.1 Design Rationale for Structure f
Based on the lessons learned from the baseline analysis, an optimized channel geometry, designated as Structure f, was proposed. The design synthesizes the advantageous features of the previous structures to create a hybrid solution that balances thermal and hydraulic performance. The core design principles are:
- Symmetrical Parallel Manifold: Adopting the symmetrical inlet/outlet layout from Structure c to ensure even flow distribution and balanced cooling paths to all cells, addressing the uniformity weakness of simple parallel designs.
- Integrated Turbulence Promoters: Incorporating the periodic bend concept from Structure e directly into the parallel channels. These bends are placed strategically in the gaps between battery cells to disrupt the thermal boundary layer and enhance local heat transfer.
- Streamlined Transitions: Replacing sharp, right-angled junctions in the manifold with smooth, arcuate transitions. This minimizes flow separation and local pressure losses at分流 points, improving the hydraulic efficiency of the parallel architecture.
This hybrid approach aims to harness the low system pressure drop of parallel flow while actively boosting the local heat transfer coefficient via targeted flow disturbance, a strategy tailored for an efficient battery thermal management system.
3.2 Comprehensive Performance of Optimized Structure
The performance of the optimized Structure f was evaluated under the same 2C discharge conditions. The results, compared directly with the best-performing baseline (Structure e), demonstrate a comprehensive improvement, as detailed in Table 4.
| Performance Metric | Structure e | Structure f | Improvement |
|---|---|---|---|
| $T_{max}$ (°C) | 32.032 | 31.858 | ↓ 0.174 °C |
| $ΔT_{max}$ (°C) | 4.197 | 4.169 | ↓ 0.028 °C |
| $T_{avg}$ (°C) | 29.160 | 28.026 | ↓ 1.134 °C |
| Pressure Drop, $ΔP$ (Pa) | 147.210 | 20.466 | ↓ 86.1% |
The thermal management performance shows decisive gains. The maximum and average pack temperatures are lower, indicating more effective heat removal. Crucially, the maximum temperature difference is also reduced, signifying improved temperature uniformity—a vital goal for any battery management system to ensure balanced aging and safety. The most dramatic improvement is in hydraulic performance. The pressure drop of Structure f is only 20.466 Pa, which is less than one-seventh of that of Structure e. This massive reduction in pumping power requirement translates directly into higher overall system efficiency for the EV. Structure f successfully demonstrates that through intelligent geometric synthesis, it is possible to break the traditional trade-off, achieving superior cooling with drastically lower flow resistance in a liquid-cooled battery thermal management system.
4. Parametric Study on the Optimized Battery Management System
Having established an optimized static structure, the dynamic response of the BTMS to key operating parameters was investigated. This study is essential for developing control strategies for the BMS.
4.1 Effect of Coolant Flow Velocity
The coolant inlet velocity was varied from 0.01 m/s to 0.09 m/s while keeping the inlet temperature at 25°C. The impact on $T_{max}$, $ΔT_{max}$, and $ΔP$ is shown in Figure 1. As expected, increasing the flow velocity enhances convective cooling, reducing $T_{max}$. However, the relationship is non-linear. The cooling benefit is very pronounced when velocity increases from 0.01 m/s to 0.05 m/s ($T_{max}$ drops by ~3.4°C), but diminishes significantly beyond 0.05 m/s. This indicates a point of diminishing returns for heat transfer enhancement. Conversely, the pressure drop increases quadratically with velocity ($ΔP \propto v^2$), leading to exponentially higher pumping power. Furthermore, $ΔT_{max}$ initially decreases with velocity but then begins to increase at higher flows. This is because higher velocity coolant gets less heated as it traverses the channel, creating a larger temperature gradient between the inlet and outlet regions of the pack. For the optimized structure, a velocity of 0.05 m/s appears to be a sensible operating point, offering a good compromise between effective cooling, manageable temperature spread, and low pumping power consumption.
4.2 Effect of Coolant Inlet Temperature
The coolant inlet temperature was varied from 15°C to 35°C while maintaining a constant flow velocity of 0.05 m/s. The results, summarized in Figure 2, reveal another critical trade-off. Lowering the coolant temperature ($T_{in}$) is highly effective at reducing the peak battery temperature ($T_{max}$), as it increases the driving temperature difference for heat transfer. However, this comes at a significant cost to temperature uniformity. With very cold coolant (e.g., 15°C), the cells near the coolant inlet are overcooled while cells downstream receive pre-warmed coolant, resulting in a large $ΔT_{max}$. As $T_{in}$ increases, $T_{max}$ rises, but $ΔT_{max}$ improves because the temperature gradient along the channel lessens. This analysis is crucial for BMS control logic. While using very cold coolant might seem advantageous for peak temperature suppression, it can be detrimental to pack longevity and safety due to high internal stresses from large temperature gradients. An inlet temperature around 25°C (close to the optimal operating temperature of the cells) provides a balanced solution for the battery thermal management system, maintaining acceptable peak temperatures while ensuring excellent uniformity.
5. Conclusion
This study focused on the structural and operational optimization of a liquid-cooled battery thermal management system (BTMS) for high-discharge-rate applications. Through a systematic analysis of baseline channel designs, we identified that symmetrical layouts improve temperature uniformity, while periodic flow disturbances enhance heat transfer at the cost of increased pressure drop in serial configurations.
The core contribution is the development of an optimized parallel turbulent channel design (Structure f). This design ingeniously combines a symmetrical parallel manifold for low flow resistance and even distribution with integrated arcuate turbulence promoters for enhanced local heat transfer. The result is a battery management system that breaks the conventional trade-off: compared to the best serial baseline, it achieved lower maximum temperature (31.858°C), better temperature uniformity (ΔTmax = 4.169°C), and a dramatic 86% reduction in pressure drop (20.466 Pa). This signifies a major step towards energy-efficient and high-performance thermal management.
Furthermore, the parametric study provided essential insights for BMS control strategy. It demonstrated that simply increasing coolant flow or decreasing its temperature is not optimal. A flow velocity of 0.05 m/s and a coolant temperature of 25°C were identified as balanced operating points for the optimized structure, effectively managing peak temperatures and gradients while minimizing pumping energy. These findings offer a valuable framework for the design and real-time control of advanced, multi-objective battery thermal management systems, contributing directly to the safety, performance, and longevity of electric vehicle battery packs.