The rapid development of electric vehicles (EVs) is pivotal in addressing global energy and environmental challenges. The EV battery pack, as the core energy storage component, directly determines vehicle performance, particularly its charging efficiency and safety in extreme climates. Lithium-ion batteries, favored for their high energy density, face significant performance degradation in low-temperature environments. Reduced ionic activity within the electrolyte increases internal resistance, severely limiting allowable charging currents and prolonging charging times. Furthermore, aggressive charging at low temperatures accelerates capacity fade. Therefore, effective thermal management during charging—specifically, preheating in cold conditions and cooling during high-rate charging—is essential for unlocking fast-charging potential, ensuring safety, and prolonging pack life.
Current research focuses on external heating methods like Positive Temperature Coefficient (PTC) heaters and liquid cooling systems. While studies have shown that PTC-based heating can effectively raise battery temperature from sub-zero conditions, and liquid cooling can mitigate thermal runaway risks, an integrated strategy governing the transient interplay between heating, charging, and cooling remains a critical challenge. The primary issues involve optimizing energy consumption during preheating, determining the optimal temperature threshold for switching between heating and cooling modes, and maintaining temperature uniformity within the EV battery pack throughout the entire charging process.
This study addresses the transient temperature control and energy consumption problem for an EV battery pack during low-temperature charging. A prismatic lithium-ion battery pack is investigated using a co-simulation approach integrating three-dimensional CFD analysis in Star-CCM+ with one-dimensional system modeling in Amesim. A novel serpentine-channel liquid cooling plate design is implemented. A comprehensive charging transient thermal management strategy is proposed and analyzed. The effects of key operational parameters—preheating power, target preheating temperature, and cooling activation temperature—on charging time, temperature uniformity, and system energy consumption are systematically evaluated to identify an optimal control strategy.
EV Battery Pack Model Development
1.1 Geometric Model of the EV Battery Pack
The studied EV battery pack adopts a modular architecture, consisting of six battery modules connected in parallel. Each module integrates 20 prismatic cells in a series configuration. Thermal pads are placed between adjacent cells for physical isolation and thermal insulation. The module structure is multi-layered, comprising, from top to bottom: foam padding, a top cover made of Sheet Molding Compound (SMC), side plates, end plates, and thermally conductive gap filler. The liquid cooling plate is positioned at the bottom of the pack, welded to the aluminum lower enclosure. The cooling plate features a serpentine flow channel design with a circular cross-section to enhance heat exchange efficiency. For simulation efficiency, minor electrical connectors and mounting hardware were simplified.
1.2 Meshing and Independence Verification
The geometry was meshed in Star-CCM+ using a Poly-Hexcore mesh type. A mesh independence study was conducted to ensure solution accuracy without excessive computational cost. The simulation was run under a 1C constant current charge with an inlet flow rate of 15 L/min and ambient temperature of 25°C. The maximum temperature of the battery module was monitored across different mesh counts. The results, presented in the table below, show that the maximum temperature stabilizes as the cell count exceeds approximately 4.2 million. Therefore, a mesh with about 4.19 million cells was selected for all subsequent simulations.
| Mesh Cell Count | Maximum Module Temperature (°C) |
|---|---|
| 3,657,151 | 34.05 |
| 3,902,681 | 33.95 |
| 4,192,514 | 33.85 |
| 4,475,576 | 33.83 |
| 4,808,898 | 33.82 |
| 5,151,173 | 33.81 |
| 5,451,173 | 33.80 |
1.3 Co-simulation Modeling Framework
A coupled 1D-3D simulation model was established. The 1D model in Amesim calculates the battery’s heat generation rate and the coolant temperature after passing through the PTC heater, based on inputs of battery temperature and coolant inlet temperature. The 3D CFD model in Star-CCM+ simulates the conjugate heat transfer between the EV battery pack and the coolant, outputting the reduced coolant outlet temperature. These two models exchange data (battery temperatures, coolant temperatures) in real-time via a TCP/IP interface, forming a closed-loop system that dynamically captures the transient thermal interactions during the combined heating, charging, and cooling processes.
The charging current is not constant but is determined by a pre-defined look-up table based on the real-time battery temperature (T) and State of Charge (SOC). This table, provided by the cell manufacturer and shown below, defines the maximum allowable charging rate (in C-rate) to prevent lithium plating and other degradation mechanisms.
| T (°C) | SOC (%) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 10 | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | 95 | 100 | |
| -10 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| -5 | 0.1 | 0.1 | 0.1 | 0.1 | 0.07 | 0.07 | 0.05 | 0.05 | 0.05 | 0.05 | 0.05 | 0.05 |
| 0 | 0.2 | 0.2 | 0.2 | 0.2 | 0.12 | 0.12 | 0.08 | 0.08 | 0.08 | 0.08 | 0.08 | 0.08 |
| 5 | 0.35 | 0.35 | 0.35 | 0.35 | 0.25 | 0.25 | 0.19 | 0.16 | 0.13 | 0.12 | 0.12 | 0.12 |
| 10 | 0.68 | 0.68 | 0.68 | 0.56 | 0.5 | 0.5 | 0.37 | 0.37 | 0.37 | 0.37 | 0.37 | 0.2 |
| 15 | 0.9 | 0.9 | 0.9 | 0.9 | 0.7 | 0.6 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.2 |
| 20 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0.5 | 0.5 | 0.2 |
| 25 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0.5 | 0.5 | 0.2 |
| 30 | 0.9 | 0.9 | 0.9 | 0.9 | 0.9 | 0.9 | 0.9 | 0.9 | 0.9 | 0.5 | 0.5 | 0.2 |
| 35 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.5 | 0.5 | 0.2 |
| 40 | 0.7 | 0.7 | 0.7 | 0.7 | 0.7 | 0.7 | 0.7 | 0.7 | 0.7 | 0.5 | 0.5 | 0.2 |
| 50 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.2 |
| 60 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.28 | 0.2 |
| 65 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
1.4 Governing Equations for Thermal and Fluid Dynamics
The thermal behavior of the EV battery pack is governed by several fundamental equations. The heat generation rate within a battery cell is calculated using a simplified Bernardi model, which accounts for irreversible Joule heating and reversible entropic heat:
$$q = \frac{I_L^2 R_T + I_L T \frac{\partial U_{oc}}{\partial T}}{V_b}$$
where \(q\) is the volumetric heat generation rate (W/m³), \(I_L\) is the current (A), \(R_T\) is the total internal resistance (Ω), \(T\) is the cell temperature (K), \(\frac{\partial U_{oc}}{\partial T}\) is the entropy coefficient (V/K), and \(V_b\) is the cell volume (m³).
Convective heat transfer between the battery surface and the ambient air is given by:
$$\phi = h (T – T_a)$$
where \(\phi\) is the heat flux (W/m²), \(h\) is the convective heat transfer coefficient (set to 3 W/(m²·K) for natural convection in a still environment), and \(T_a\) is the ambient temperature (K).
The flow and heat transfer of the 50% ethylene glycol coolant within the liquid cooling plate are described by the conservation laws of mass, momentum, and energy for an incompressible fluid:
Mass Conservation:
$$\frac{\partial \rho_l}{\partial t} + \nabla \cdot (\rho_l \vec{\nu}) = 0$$
Momentum Conservation (Navier-Stokes):
$$\frac{\partial (\rho_l \vec{\nu})}{\partial t} + \nabla \cdot (\rho_l \vec{\nu} \vec{\nu}) = -\nabla P$$
Energy Conservation:
$$\frac{\partial (\rho_l c_l T_l)}{\partial t} + \nabla \cdot (\rho_l c_l T_l \vec{\nu}) = \nabla \cdot (k_l \nabla T_l)$$
Here, \(\rho_l\) is density (kg/m³), \(c_l\) is specific heat capacity (J/(kg·K)), \(\vec{\nu}\) is the velocity vector (m/s), \(T_l\) is coolant temperature (K), \(k_l\) is thermal conductivity (W/(m·K)), and \(P\) is pressure (Pa). The standard k-ε turbulence model was employed in Star-CCM+.
The energy consumption of the liquid cooling system, \(W_c\), is calculated by integrating the pump power over time:
$$W_c = \int P_{pump}(t) \, dt \approx \int \Delta P(t) \cdot Q(t) \, dt$$
where \(Q\) is the volumetric flow rate (m³/s).
Battery Characterization and Model Validation
2.1 Experimental Setup
Experiments were conducted on a single prismatic LiFePO₄ cell (150 Ah, 3.2 V nominal) to characterize its internal resistance and validate the thermal model. The test platform included a battery cycler (CE-6008n), a thermal chamber (SC-80-CB-2), and a data acquisition system (Neware CA-4008) with K-type thermocouples attached to the cell surface.
2.2 HPPC Test for Internal Resistance
The Hybrid Pulse Power Characterization (HPPC) test was performed at various temperatures (-30°C to 50°C) and SOC levels (10% to 90%) to map the internal resistance, \(R_T\). A sample of the obtained resistance values is summarized in the table below, highlighting the strong dependence of resistance on both temperature and SOC, especially at low temperatures.
| T (°C) | Internal Resistance, \(R_T\) (mΩ) at SOC | ||||
|---|---|---|---|---|---|
| 10% | 50% | 90% | |||
| -30 | 23.51 | 3.56 | 0.98 | ||
| 0 | 2.08 | 1.14 | 0.88 | ||
| 25 | 1.07 | 0.73 | 0.68 | ||
| 40 | 0.71 | 0.71 | 0.58 | ||
2.3 Charge Temperature Rise Test and Model Validation
The cell was charged at constant currents of 0.33C, 0.5C, and 1C from an initial temperature of 25°C. The temperature rise was recorded. A corresponding 3D CFD model of the single cell was created in Star-CCM+. The simulated temperature profiles showed excellent agreement with experimental data. The maximum temperature error was less than 3.2% across all test cases, validating the accuracy of the thermal model and the adopted material properties.
Analysis of Charging Transient Thermal Management Strategy
3.1 Proposed Multi-Stage Strategy
Based on the charging current limits, a multi-stage transient thermal management strategy was designed for the EV battery pack starting from -30°C:
Stage 1 – Low-Temperature Preheating & Low-Rate Charging: When the pack’s average temperature (\(T_{ave}\)) is below 0°C, a PTC heater warms the coolant circulating through the cooling plate. Simultaneously, a minimal charging current (e.g., 0.2C) is applied. This dual approach gently increases the temperature to enhance ionic conductivity while avoiding damage.
Stage 2 – Normal Charging: Once \(T_{ave}\) reaches 0°C, the PTC heater is deactivated. The charging current is allowed to increase according to the look-up table as temperature and SOC rise, potentially reaching 1C.
Stage 3 – Active Cooling during High-Rate Charging: To prevent excessive temperature rise (>45°C) which accelerates aging, the liquid cooling system is activated when \(T_{ave}\) reaches a predefined threshold (e.g., 40°C). This suppresses the temperature, allowing the charging current to remain at a higher level for a longer duration compared to a scenario without cooling.
3.2 Effect of Preheating Power
The impact of PTC heater power (4000W, 5000W, 6000W) on the preheating phase was analyzed. The goal was to heat the EV battery pack from -30°C to 0°C. The results are summarized below:
| PTC Power (W) | Heating Time (min) | Avg. Heating Rate (°C/min) | Heating Energy (kWh) | Max. Temp. at 0°C (°C) | Max. ΔT at 0°C (°C) |
|---|---|---|---|---|---|
| 4000 | 125.9 | 0.238 | 8.37 | 2.01 | 5.33 |
| 5000 | 100.5 | 0.298 | 8.42 | 2.37 | 5.98 |
| 6000 | 88.0 | 0.341 | 8.77 | 2.72 | 6.62 |
Increasing the PTC power from 4000W to 6000W significantly reduced the preheating time by 37.9 minutes (30.1%) and increased the average heating rate by 43.3%. However, this came at the cost of a slight increase in energy consumption (4.8%) and a noticeable increase in the maximum temperature and temperature difference within the pack at the end of preheating. A 6000W heater offers the fastest warm-up, crucial for user convenience in extreme cold, and was selected for the final strategy despite the higher temperature non-uniformity.
3.3 Effect of Target Preheating Temperature
The influence of the target preheating temperature (\(T_{target}\)) on the overall charging process was studied. The EV battery pack was heated from -30°C to different \(T_{target}\) values (-5°C, 0°C, 10°C) using the 6000W PTC before commencing normal charging.
| \(T_{target}\) (°C) | Total Charging Time (min) | Preheating Energy (kWh) | Max. Temp. during Charge (°C) | Max. ΔT during Charge (°C) |
|---|---|---|---|---|
| -5 | 116.75 | 7.57 | -2.42 | 6.05 |
| 0 | 97.52 | 8.77 | 2.72 | 6.62 |
| 10 | 79.83 | 11.22 | 12.95 | 7.60 |
Preheating to a higher temperature dramatically reduces total charging time—by 31.6% when increasing \(T_{target}\) from -5°C to 0°C, and by an additional 18.1% when heating to 10°C. This is because a warmer battery can accept a higher charging current immediately, as per the current limit table. However, preheating to 10°C requires 28.0% more energy than preheating to 0°C and results in a higher peak temperature and greater thermal gradient within the EV battery pack. Balancing time, energy, and thermal stress, a \(T_{target}\) of 0°C was chosen as the optimal compromise.
3.4 Effect of Cooling Activation Temperature
After preheating to 0°C, the cooling system’s activation temperature (\(T_{cool,on}\)) was varied to analyze its effect on managing temperature rise during the high-SOC charging phase. The cooling inlet temperature was set to 10°C with a flow rate of 15 L/min.
| Cooling Strategy | Total Charging Time (min) | Cooling Energy, \(W_c\) (kWh) | Max. Pack Temp., \(T_{max}\) (°C) | Max. Pack ΔT (°C) |
|---|---|---|---|---|
| No Cooling | 97.52 | 0.0 | 51.32 | 9.65 |
| \(T_{cool,on}\) = 30°C | 84.52 | 3.15 | 38.64 | 6.28 |
| \(T_{cool,on}\) = 40°C | 87.58 | 2.58 | 42.95 | 6.55 |
| \(T_{cool,on}\) = 45°C | 90.11 | 1.69 | 46.88 | 7.32 |
Activating cooling is crucial for preventing excessive temperatures. Without cooling, the EV battery pack reached 51.3°C with a large 9.65°C gradient. Activating cooling at 40°C reduced the peak temperature by 8.4°C, narrowed the gradient by 3.1°C, and paradoxically shortened the total charging time by 10.2%. This is because cooling suppresses temperature, preventing it from entering the high-temperature, low-current region of the look-up table (e.g., >50°C at high SOC). Activating cooling earlier (at 30°C) provides even better temperature suppression but consumes more pumping energy and slightly increases charging time as the pack is held at a lower, potentially less electrochemically favorable temperature. Activating later (at 45°C) saves cooling energy but allows higher thermal stress. Activating at \(T_{cool,on}\) = 40°C provides an optimal trade-off between thermal management, charging speed, and ancillary energy use for this EV battery pack.
3.5 Temperature Uniformity Under Optimal Strategy
The final proposed strategy employs a 6000W PTC to preheat to 0°C, charges normally, and activates liquid cooling at a \(T_{ave}\) of 40°C. Under this strategy, the temperature uniformity across the six modules of the EV battery pack was evaluated at key stages.
| Thermal Stage | Average Pack Temp., \(T_{ave}\) (°C) | Max. Module \(T_{max}\) (°C) | Max. Module ΔT (°C) | Inter-Module \(T_{max}\) Spread (°C) |
|---|---|---|---|---|
| End of Preheating | 0.0 | 2.72 (Module 1) | 1.55 – 2.20 | 2.39 |
| Cooling Activation | 40.0 | 40.27 (Module 1) | 2.47 – 3.47 | 1.00 |
| End of Charging | ~36.5 | 36.54 (Module 5) | 2.01 – 2.59 | 0.53 |
The temperature distribution is non-uniform due to the coolant flow path; Module 1 near the coolant inlet warms/cools more effectively than Module 5 near the outlet. However, the designed serpentine cooling plate and the control strategy successfully maintain the maximum temperature difference within any single module below 3.5°C and the overall pack gradient below 5°C throughout the entire charging process, which is critical for the longevity and safety of the EV battery pack.
Conclusion
This research developed and analyzed a transient thermal management strategy for an EV battery pack during low-temperature charging using a validated Star-CCM+/Amesim co-simulation model. The key findings are:
1. Increasing preheating power from 4000W to 6000W raises the average heating rate by 43.2% with only a 4.8% increase in energy consumption, making high-power preheating effective for reducing wait time in cold climates.
2. The target preheating temperature has a profound impact on total charging time. Preheating the EV battery pack from -30°C to 0°C instead of -5°C reduces charging time by 31.6%, while preheating to 10°C saves an additional 18.1% but at the cost of significantly higher energy use and thermal gradients.
3. Activating the liquid cooling system is essential for enabling fast charging. For the studied EV battery pack, initiating cooling when the average temperature reaches 40°C optimally balances performance and energy use. Compared to a no-cooling scenario, it reduces charging time by 10.2%, lowers the peak temperature by 8.4°C, and shrinks the maximum temperature spread by 3.1°C.
4. The proposed integrated strategy—6000W preheating to 0°C, followed by managed charging with cooling activation at 40°C—successfully maintains the maximum temperature difference within the EV battery pack below 5°C. The internal temperature spread within individual modules remains between 2.01°C and 3.47°C, ensuring good temperature uniformity and safe operation.
This work demonstrates that a well-designed co-simulation framework can effectively optimize complex thermal management strategies for EV battery packs, leading to faster charging in cold weather without compromising safety or longevity. The methodology and findings provide a valuable theoretical foundation for the development of advanced battery thermal management systems in electric vehicles.