As the global energy transition accelerates and carbon neutrality goals gain traction, electric vehicles have emerged as a pivotal solution in the transportation sector due to their low-carbon footprint. The lithium-ion battery, with its high energy density and long cycle life, serves as the core power source for these vehicles. However, the performance, safety, and longevity of lithium-ion batteries are intrinsically linked to their operating temperature. Studies indicate that when battery temperature deviates from the optimal range of 20–40°C, capacity degradation accelerates, and the risk of thermal runaway increases significantly, especially under fast charging or extreme operating conditions. In electric vehicles and energy storage systems, an efficient Battery Thermal Management System (BTMS) is crucial to ensure batteries operate within an ideal thermal environment. Among various cooling techniques, liquid cooling stands out for its high thermal conductivity, superior heat dissipation efficiency, and precise temperature control capabilities. This has made it a focal point of recent research. This article presents a novel liquid cooling channel design aimed at enhancing the thermal management of square lithium-ion batteries. Through numerical simulation, we analyze the impact of coolant flow velocity, inlet temperature, and channel thickness on cooling performance, seeking to balance heat dissipation efficiency with pump power consumption for optimized system design.
The thermal behavior of a lithium-ion battery during operation is governed by heat generation principles, which can be modeled using the Bernardi equation. This model assumes uniform internal heat generation and constant material properties, neglecting radiation effects. The total heat generation includes reversible reaction heat and irreversible Joule heat. The reversible heat component is given by:
$$\phi = \theta_d \frac{dU_{oc}}{d\theta_d}$$
where $\phi$ is the reversible heat (negative during charging, positive during discharging), $\theta_d$ is the battery temperature in °C, and $U_{oc}$ is the open-circuit voltage in V. The overall volumetric heat generation rate $Q$ (in W/m³) is expressed as:
$$Q = \frac{I(U_{oc} – U) + \phi}{V_{cell}}$$
Here, $I$ is the charge-discharge current in A, $U$ is the operating voltage in V, and $V_{cell}$ is the battery cell volume in m³. For a square ternary lithium-ion battery with a nominal capacity of 37 Ah and nominal voltage of 3.65 V, the reversible heat factor $\phi$ is calculated as 0.0514 V. At discharge rates of 1C and 2C (corresponding to currents of 37 A and 74 A, respectively), the volumetric heat generation rates are 7,780 W/m³ and 20,773 W/m³. This heat generation model forms the basis for simulating thermal dynamics in lithium-ion batteries.
In liquid cooling systems, the coolant (typically a 50% ethylene glycol-water mixture) flows through channels to remove heat from battery surfaces. The flow regime is characterized by the Reynolds number, which determines whether the flow is laminar or turbulent. For rectangular channels, the Reynolds number is defined as:
$$Re = \frac{\rho_l v_s \cdot 2ab}{\nu(a + b)}$$
where $\rho_l$ is the coolant density in kg/m³, $v_s$ is the coolant velocity in m/s, $\nu$ is the dynamic viscosity in Pa·s, and $a$ and $b$ are the cross-sectional length and width of the channel in m, respectively. In our simulations, the maximum coolant velocity is 0.6 m/s, yielding a Reynolds number below 2,300, confirming laminar flow conditions. The governing equations for the fluid include conservation of mass, momentum, and energy. For incompressible, steady-state laminar flow, these are expressed as:
Mass conservation: $$\nabla \cdot \mathbf{v_s} = 0$$
Momentum conservation: $$\rho_l \frac{\partial \mathbf{v_s}}{\partial t} = -\nabla p + \mu \nabla^2 \mathbf{v_s}$$
Energy conservation: $$\rho_l c_l \frac{\partial \theta_l}{\partial t} + \nabla \cdot (\rho_l c_l \mathbf{v_s} \theta_l) = \nabla \cdot (\lambda_l \nabla \theta_l)$$
Here, $p$ is pressure in Pa, $\mu$ is the dynamic viscosity in Pa·s, $c_l$ is the specific heat capacity in J/(kg·K), $\theta_l$ is the coolant temperature in K, and $\lambda_l$ is the thermal conductivity in W/(m·K). These equations are solved numerically to simulate heat transfer between the battery and coolant.
The geometric model focuses on a square lithium-ion battery module. Each battery cell has dimensions of 148 mm in length, 27 mm in width, and 98 mm in height. The novel liquid cooling channel is positioned between two battery cells. Traditional channel designs often feature simple parallel or serpentine layouts, which may lead to uneven cooling and high pressure drops. Our innovative design starts with a rectangular contact area between the channel and battery group. This area is divided into multiple grid regions using diagonal lines. The midpoint of one side is designated as the inlet, and the midpoint of the opposite side as the outlet. By removing lines perpendicular to the flow direction, we create a channel centerline sketch, which is then used to construct the three-dimensional channel geometry. This approach increases the contact area between the coolant and battery surface, promoting more uniform heat dissipation while potentially reducing flow resistance. The materials involved include the square lithium iron phosphate battery, the coolant (50% ethylene glycol-water mixture), and aluminum for the channel walls. Their thermophysical properties are summarized in the table below.
| Material | Density (kg/m³) | Specific Heat Capacity (J/(kg·K)) | Thermal Conductivity (W/(m·K)) | Dynamic Viscosity (Pa·s) |
|---|---|---|---|---|
| Square Lithium Iron Phosphate Battery | 2,056 | 812 | 4.500, 0.780, 4.500* | — |
| 50% Ethylene Glycol-Water Solution | 1,069 | 3,494 | 0.419 | 0.00315 |
| Aluminum | 2,700 | 2,719 | 202.4 | — |
*Thermal conductivities along the length, width, and height directions, respectively.
Numerical simulations were conducted using ANSYS Fluent software. The battery heat generation model was implemented as a volumetric source term. The initial ambient temperature was set to 25°C, with a convective heat transfer coefficient of 5 W/(m²·K) on external walls. The discharge rate was fixed at 2C, corresponding to a volumetric heat generation of 20,773 W/m³. The coolant flow was modeled as laminar, and conjugate heat transfer was enabled between the battery and channel. Boundary conditions included velocity inlet and pressure outlet for the coolant. A mesh independence study was performed to ensure result accuracy. Different mesh sizes were tested, and the battery maximum temperature and channel pressure difference were monitored. The table below shows the results for varying mesh counts.
| Mesh Count | Battery Max Temperature (°C) | Pressure Difference (Pa) |
|---|---|---|
| 1.2×10⁵ | 35.88 | 295.19 |
| 2.6×10⁵ | 34.91 | 283.00 |
| 6.3×10⁵ | 34.70 | 244.35 |
| 8.5×10⁵ | 34.63 | 222.32 |
| 1.2×10⁶ | 34.32 | 221.86 |
As the mesh count increased, changes in maximum temperature and pressure difference diminished. At 8.5×10⁵ elements, both parameters varied by less than 1% upon further refinement, indicating mesh independence. This mesh size was adopted for all subsequent simulations to balance computational accuracy and efficiency. Model validation was performed by comparing simulation results with experimental data from literature under 2C discharge. The maximum error was approximately 4.2% at around 1,300 seconds, which is within an acceptable range (below 5%). Discrepancies may arise from simplifications in the numerical model, such as uniform heat generation and constant material properties, whereas real-world conditions involve variable internal resistances and environmental factors.
The performance of the novel liquid cooling channel was first verified by comparing it with a traditional baseline channel design. Under identical conditions (ambient temperature 25°C, coolant velocity 0.5 m/s, channel thickness 3 mm), the novel channel reduced the maximum battery pack temperature by approximately 8.26°C. This significant improvement underscores the effectiveness of the enhanced contact area and flow path design in boosting heat dissipation for lithium-ion battery thermal management.
We then investigated the influence of coolant flow velocity on thermal performance. The channel thickness was fixed at 3 mm, and the coolant inlet temperature was set to 25°C. Velocity was varied from 0.1 m/s to 0.6 m/s. The results are summarized in the following table.
| Coolant Velocity (m/s) | Battery Max Temperature (°C) | Temperature Difference (°C) | Channel Pressure Difference (Pa) |
|---|---|---|---|
| 0.1 | 35.02 | 2.08 | 21.56 |
| 0.2 | 34.78 | 2.08 | 60.33 |
| 0.3 | 34.79 | 2.16 | 116.25 |
| 0.4 | 34.72 | 2.08 | 190.15 |
| 0.5 | 34.63 | 2.06 | 222.32 |
| 0.6 | 34.68 | 2.09 | 390.09 |
As velocity increased, the maximum battery temperature decreased slightly, from 35.02°C at 0.1 m/s to 34.68°C at 0.6 m/s. The temperature difference (difference between maximum and minimum battery temperatures) remained relatively stable, around 2.08°C ± 0.10°C, indicating good temperature uniformity provided by the novel channel design. However, the pressure difference across the channel rose dramatically, from 21.56 Pa to 390.09 Pa, following a roughly quadratic relationship with velocity. This highlights a trade-off: higher velocities enhance convective heat transfer but require more pump power to overcome increased flow resistance. For lithium-ion battery thermal management systems, optimizing velocity is crucial to achieve efficient cooling without excessive energy consumption.
Next, we examined the effect of coolant inlet temperature. The channel thickness was kept at 3 mm, and coolant velocity was fixed at 0.5 m/s. Inlet temperature was varied from 25°C down to 10°C in decrements of 5°C. The results are shown in the table below.
| Coolant Inlet Temperature (°C) | Battery Max Temperature (°C) | Temperature Difference (°C) |
|---|---|---|
| 25.00 | 34.63 | 2.06 |
| 20.00 | 31.25 | 2.51 |
| 15.00 | 28.02 | 3.04 |
| 10.00 | 24.87 | 3.67 |
Lowering the inlet temperature significantly reduced the maximum battery temperature, from 34.63°C at 25°C to 24.87°C at 10°C, a drop of 9.76°C. However, the temperature difference increased from 2.06°C to 3.67°C, suggesting slightly less uniform cooling at lower inlet temperatures. Nevertheless, all temperature differences remained within 4°C, which is generally acceptable for lithium-ion battery operation. Compared to flow velocity, inlet temperature has a more pronounced impact on the absolute battery temperature. This underscores the importance of pre-cooling or chiller systems in liquid cooling loops for lithium-ion batteries, especially in high-demand scenarios.
To provide a comprehensive view, we combined variations in both coolant velocity and inlet temperature. The following table presents battery maximum temperatures and temperature differences for different combinations.
| Coolant Inlet Temp (°C) | Battery Max Temperature (°C) at Velocity (m/s) | Temperature Difference (°C) at Velocity (m/s) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.1 | 0.2 | 0.3 | 0.4 | 0.5 | 0.6 | 0.1 | 0.2 | 0.3 | 0.4 | 0.5 | 0.6 | |
| 25 | 35.02 | 34.78 | 34.79 | 34.72 | 34.63 | 34.68 | 2.08 | 2.08 | 2.16 | 2.08 | 2.06 | 2.09 |
| 20 | 31.80 | 31.44 | 31.33 | 31.39 | 31.25 | 31.35 | 2.54 | 2.51 | 2.49 | 2.54 | 2.51 | 2.56 |
| 15 | 28.65 | 28.21 | 28.06 | 28.18 | 28.02 | 28.13 | 3.08 | 3.03 | 3.00 | 3.10 | 3.04 | 3.13 |
| 10 | 25.60 | 25.09 | 24.92 | 25.06 | 24.87 | 25.01 | 3.71 | 3.66 | 3.62 | 3.74 | 3.67 | 3.73 |
This data reinforces that inlet temperature is a dominant factor for absolute temperature control, while velocity adjustments have a moderate effect. For all cases, temperature differences stayed below 4°C, demonstrating the robustness of the novel channel design across various operating conditions for lithium-ion battery cooling.
Another critical structural parameter is the channel thickness. We analyzed thicknesses of 2 mm, 3 mm, and 4 mm under different coolant velocities, with inlet temperature fixed at 25°C. The results are compiled in the table below.
| Coolant Velocity (m/s) | Battery Max Temperature (°C) at Thickness | Temperature Difference (°C) at Thickness | Pressure Difference (Pa) at Thickness | ||||||
|---|---|---|---|---|---|---|---|---|---|
| 2 mm | 3 mm | 4 mm | 2 mm | 3 mm | 4 mm | 2 mm | 3 mm | 4 mm | |
| 0.1 | 34.90 | 35.02 | 35.13 | 2.17 | 2.08 | 2.08 | 29.90 | 21.56 | 20.12 |
| 0.2 | 34.51 | 34.78 | 34.90 | 2.15 | 2.08 | 2.09 | 86.57 | 60.33 | 55.06 |
| 0.3 | 34.38 | 34.79 | 34.85 | 2.14 | 2.16 | 2.10 | 156.02 | 116.25 | 105.31 |
| 0.4 | 34.30 | 34.72 | 34.84 | 2.14 | 2.08 | 2.12 | 242.66 | 190.15 | 170.76 |
| 0.5 | 34.27 | 34.63 | 34.79 | 2.14 | 2.06 | 2.13 | 347.49 | 222.32 | 212.85 |
| 0.6 | 34.33 | 34.68 | 34.82 | 2.16 | 2.09 | 2.15 | 469.29 | 390.09 | 380.18 |
Increasing channel thickness from 2 mm to 4 mm led to a rise in maximum battery temperature, from around 34.30°C to 34.84°C at 0.4 m/s, for instance. This is because thicker channels reduce flow resistance (lower pressure difference) but also decrease the convective heat transfer coefficient due to larger cross-sectional area and potentially lower flow velocity gradients near the walls. Pressure difference decreased with increasing thickness, as expected from fluid dynamics principles. Temperature differences remained relatively constant across thickness variations, averaging about 2.10°C ± 0.05°C, indicating that channel thickness has minimal impact on temperature uniformity. However, thickness significantly influences the trade-off between cooling performance and pump power. Thinner channels offer better heat transfer but higher pressure drops, while thicker channels reduce pumping costs at the expense of slightly elevated battery temperatures. For lithium-ion battery thermal management, selecting an optimal thickness is essential to balance these factors.
Based on the comprehensive analysis, we can derive optimal operating and design parameters. The primary goal is to minimize battery maximum temperature and temperature difference while keeping pump power consumption low. From the data, a coolant velocity of 0.4 m/s appears favorable: it reduces the maximum temperature effectively (34.72°C at 3 mm thickness) without incurring an excessively high pressure difference (190.15 Pa). Higher velocities like 0.6 m/s yield marginal further temperature reduction but nearly double the pressure drop, increasing energy costs. For channel thickness, 3 mm strikes a balance: compared to 2 mm, it offers a moderate pressure drop reduction (from 242.66 Pa to 190.15 Pa at 0.4 m/s) with only a slight temperature increase (from 34.30°C to 34.72°C). A 4 mm thickness lowers pressure drop further but raises battery temperature more noticeably. Coolant inlet temperature should be as low as feasible, considering system constraints, to directly lower battery temperatures. However, very low temperatures may increase temperature differences slightly, so active control might be needed. In summary, for the novel liquid cooling channel applied to square lithium-ion batteries, recommended settings are a coolant velocity of 0.4 m/s and a channel thickness of 3 mm, with inlet temperature adjusted based on ambient conditions and cooling capacity.
In conclusion, this study presents a novel liquid cooling channel design for thermal management of square lithium-ion batteries. Through numerical simulation using coupled heat generation and fluid flow models, we demonstrated that the new channel reduces the maximum battery pack temperature by approximately 8.26°C compared to a baseline design. Parametric studies on coolant velocity, inlet temperature, and channel thickness reveal key insights: velocity increases enhance cooling but raise pressure drops quadratically; inlet temperature has a strong effect on absolute temperature control; channel thickness affects the trade-off between heat transfer and flow resistance. Temperature differences remained within 4°C across all tested conditions, indicating good uniformity. For optimal performance balancing efficiency and energy consumption, a coolant velocity of 0.4 m/s and channel thickness of 3 mm are recommended. This work provides a theoretical foundation for optimizing liquid cooling channel structures in lithium-ion battery thermal management systems, contributing to safer and more efficient electric vehicle operations. Future research could explore multi-objective optimization algorithms, transient thermal responses, and experimental validation to further refine the design for practical applications in lithium-ion battery packs.