In the pursuit of enhancing energy efficiency and safety in electric vehicles, the thermal management of the EV battery pack has emerged as a critical area of research. A well-designed thermal control system can significantly mitigate risks such as uncontrolled thermal runaway and combustion, thereby extending the lifespan and reliability of the EV battery pack. Among various cooling methods, liquid cooling remains a predominant choice due to its superior heat transfer coefficients compared to air cooling, while phase change material cooling is still in developmental stages. This study focuses on investigating the effect of liquid cooling initiated mid-discharge in an EV battery pack, utilizing a comprehensive three-dimensional transient analysis to evaluate cooling performance under different operational conditions.
The core of this research involves a complete EV battery pack composed of three battery modules, each with multiple lithium-ion cells. To monitor temperature rise during discharge, conductive busbars and NTC temperature sensors are placed on the pole surfaces. The approach involves numerical simulation using computational fluid dynamics (CFD) in Fluent, where the EV battery pack is discharged at a 1C rate under ambient driving conditions until the temperature reaches 38°C, at which point the liquid cooling system is activated. This strategy aims to optimize energy usage by delaying cooling until necessary, thus reducing parasitic losses in the EV battery pack. The results indicate that upon activation, the temperature of the EV battery pack initially rises due to thermal inertia before declining, with cells near the outlet experiencing higher temperature increases than those near the inlet. However, the overall temperature range remains within safe operational limits for electric vehicles.
To improve temperature uniformity and cooling efficiency, a novel flow channel structure is proposed based on existing serpentine designs. This new channel incorporates a combination of series and parallel paths with progressively increasing widths, aiming to enhance coolant distribution and heat extraction. Further analyses are conducted by varying inlet coolant velocity, temperature, and thermal pad thickness to assess their impact on the cooling performance of the EV battery pack. The findings demonstrate that the novel channel design reduces maximum temperature and temperature differentials, while parameters like coolant flow rate and thermal pad thickness must be optimized holistically for effective thermal management in the EV battery pack.
The foundation of this analysis lies in the physical and mathematical modeling of lithium-ion batteries. A simplified model is developed for a 50Ah prismatic cell with nickel-cobalt-manganese (NCM) cathode and graphite anode. Key parameters are summarized in Table 1, which outlines the basic performance characteristics of the cell used in the EV battery pack.
| Technical Parameter | Value |
|---|---|
| Rated Capacity | 50 Ah |
| Rated Voltage | 3.7 V |
| Charge Cut-off Voltage | 4.2 V |
| Discharge Cut-off Voltage | 2.8 V |
| Maximum Charge Current | 2C |
| Maximum Discharge Current | 3C |
| Battery Dimensions | 148 × 27 × 91 mm (Length × Width × Height) |
The heat generation model for the EV battery pack is derived from Newman’s theory, which simplifies the complex electrochemical reactions into a manageable form. The total heat generation rate per volume, \( q_v \), is expressed as:
$$ q_v = I V \left( (E – U) – T \frac{dE}{dT} \right) $$
where \( I \) is the current, \( V \) is the volume, \( E \) is the open-circuit voltage, \( U \) is the terminal voltage, and \( \frac{dE}{dT} \) is the entropy coefficient. This equation accounts for reversible and irreversible heat sources, providing a basis for simulating thermal behavior in the EV battery pack during discharge.
Thermophysical properties of the materials within the EV battery pack are crucial for accurate simulation. These include density, specific heat, and anisotropic thermal conductivity of the battery cells, as well as properties for components like thermal pads, cooling plates, and coolant. The effective properties for the cell are computed using weighted averages. For instance, the density \( \rho \) is given by:
$$ \rho = \frac{\sum \rho_i \cdot V_i}{\sum V_i} $$
where \( \rho_i \) and \( V_i \) are the density and volume of each component layer. Similarly, specific heat \( c_b \) is calculated as:
$$ c_b = \frac{\sum c_i \cdot m_i}{\sum m_i} $$
with \( c_i \) and \( m_i \) being the specific heat and mass of each layer. The thermal conductivity varies with direction due to the layered structure; for the in-plane directions (x and y), it is:
$$ \lambda_x = \lambda_y = \frac{\sum \lambda_i \cdot l_i}{\sum l_i} $$
and for the through-plane direction (z):
$$ \lambda_z = \frac{\sum l_i}{\sum \frac{l_i}{\lambda_i}} $$
where \( \lambda_i \) and \( l_i \) are the thermal conductivity and thickness of each layer. Table 2 compiles the thermophysical parameters used for all materials in the EV battery pack model.
| Material | Specific Heat (J/kg·K) | Density (kg/m³) | Thermal Conductivity (W/(m·K)) |
|---|---|---|---|
| Battery Cell | 1033 | 2218 | 17.4 (x,y), 5.3 (z), 23 (avg) |
| Thermal Pad | 1800 | 2000 | 1.8 |
| Epoxy Board | 1581 | 1800 | 0.2 |
| Insulation Board | 1260 | 1150 | 0.2 |
| Aerogel | 1180 | 230 | 0.025 |
| Aluminum (Cooling Plate) | 903 | 2707 | 237 |
| Coolant (50% Ethylene Glycol) | 3300 | 1071 | 0.384 |
| Insulation Layer | 1700 | 65 | 0.034 |
| Heating Film | 1130 | 1840 | 1.2 |
Heat generation data for the EV battery pack under 1C discharge is obtained experimentally, fitting a curve that shows power density variations over time. This data is imported as a user-defined function (UDF) in Fluent to simulate transient thermal effects. The heat generation rate peaks towards the end of discharge due to increased polarization resistance, which is critical for modeling the EV battery pack’s thermal response.
A three-dimensional model of the complete EV battery pack is constructed, including the battery modules, air domain, end plates, busbars, NTC sensors, insulation layers, and the liquid cooling plate. The EV battery pack consists of 39 cells (3p13s) arranged within a protective enclosure. The air domain inside the pack allows for natural convection modeling, while the cooling plate with a serpentine flow channel is placed beneath the modules to facilitate heat transfer. Thermal pads are inserted between the cells and cooling plate to enhance contact and reduce thermal resistance. Mesh generation is performed using Fluent Meshing, with local refinements at key areas like inlets, outlets, and epoxy boards. The final mesh comprises approximately 2 million cells, ensuring adequate resolution for accurate simulation of the EV battery pack’s thermal behavior.
Boundary conditions are set to mimic real-world operating conditions. The initial ambient temperature is 25°C, with convective heat transfer coefficients of 3 W/m²·K applied to external surfaces. The k-ε turbulence model is employed for coolant flow, and coupled interfaces are defined between the battery cells, thermal pads, and cooling plate. The simulation runs in two phases: first, discharge without cooling until the NTC sensors detect 38°C; second, activation of liquid cooling with an inlet velocity of 0.4 m/s until discharge completion. Results show that without cooling, the EV battery pack temperature soars to 47°C, exceeding safe limits. With cooling activated at 38°C, the temperature initially rises to 40.1°C due to thermal inertia before decreasing, and the maximum temperature differential remains below 5°C, indicating effective management of the EV battery pack’s thermal state.
To address temperature non-uniformity observed in serpentine channels, a novel flow channel design is proposed. This design integrates series and parallel segments with widths increasing in an arithmetic sequence: 50 mm, 50 mm, 50 mm, 60 mm, 60 mm, and 80 mm from inlet to outlet. The narrower inlet channels increase coolant velocity for rapid heat absorption, while wider outlet channels expand the contact area to dissipate heat even as coolant temperature rises. Additionally, turbulence promoters are added at the inlet to improve flow distribution. The advantages and disadvantages of serial and parallel channels are summarized in Table 3, guiding the design rationale for the EV battery pack cooling system.
| Channel Type | Advantages | Disadvantages |
|---|---|---|
| Serial Channel | Better cooling performance, lower maximum temperature | Higher temperature differentials, increased pressure drop |
| Parallel Channel | Improved temperature uniformity, lower pressure drop | Reduced cooling efficiency, higher maximum temperature |
Simulation of the novel channel in the EV battery pack shows a maximum temperature of 38.8°C, a reduction of 1.3°C compared to the serpentine design. The temperature differential decreases to 4°C, an improvement of 11.1%, and the pressure drop across the channel reduces by 32.5% to 1013.5 Pa, lowering energy consumption. These enhancements demonstrate the efficacy of the novel design in maintaining thermal homogeneity and efficiency in the EV battery pack.
Further parametric studies investigate the influence of coolant flow rate, inlet temperature, and thermal pad thickness on the cooling performance of the EV battery pack. For flow rate, velocities of 0.4, 0.6, 0.8, and 1.0 m/s are tested. The results, plotted in Figure 1, indicate that increasing flow rate reduces both maximum temperature and temperature differential, but with diminishing returns. For instance, at 0.4 m/s, the maximum temperature is 38.8°C with a 4.0°C differential; at 1.0 m/s, these values drop to 38.4°C and 3.8°C, respectively. However, the pressure drop escalates nonlinearly, suggesting an optimal range of 0.4–0.8 m/s for the EV battery pack to balance cooling and energy costs.
$$ \text{Pressure Drop} \propto v^n \quad \text{where } n > 1 \text{ for turbulent flow} $$
Coolant inlet temperature variations (15°C, 25°C, 35°C) reveal trade-offs between cooling capacity and temperature uniformity. At 15°C, the EV battery pack’s maximum temperature is lowest (38.4°C), but the differential peaks at 4.4°C due to excessive heat absorption. At 35°C, the maximum temperature rises to 39.2°C, but the differential minimizes to 3.8°C as heat transfer stabilizes. Thus, 25°C is selected as a compromise for effective and uniform cooling in the EV battery pack.
Thermal pad thickness, often overlooked, significantly affects heat transfer in the EV battery pack. Thicknesses of 0 mm (no pad), 0.5 mm, 1.5 mm, 2.5 mm, and 4 mm are analyzed. Without pads, direct contact yields the lowest maximum temperature (38.4°C) but the highest differential (4.2°C) due to localized hotspots. As thickness increases, thermal resistance rises, leading to higher maximum temperatures but lower differentials; at 4 mm, the maximum temperature reaches 39.5°C with a differential of 3.5°C. The relationship can be expressed as:
$$ R_{\text{th}} = \frac{t}{\lambda A} $$
where \( R_{\text{th}} \) is thermal resistance, \( t \) is thickness, \( \lambda \) is thermal conductivity, and \( A \) is area. Based on this, a thickness of 1–3 mm is recommended for the EV battery pack to optimize both cooling and uniformity.
In conclusion, this study demonstrates that delayed activation of liquid cooling based on temperature monitoring can effectively manage the thermal state of an EV battery pack while conserving energy. The novel differential flow channel design enhances temperature uniformity and reduces pressure drop, contributing to more efficient thermal management in the EV battery pack. Parametric analyses underscore the importance of optimizing coolant flow rate, temperature, and thermal pad thickness to achieve balanced cooling performance. Future work could explore adaptive control strategies or hybrid cooling systems to further improve the reliability and longevity of the EV battery pack. The insights gained here provide valuable guidelines for engineers designing thermal management systems for electric vehicles, ensuring that the EV battery pack operates safely and efficiently across diverse driving conditions.