In recent years, the escalating issues of global environmental pollution, energy crises, and climate change have intensified worldwide focus on innovative energy technologies. The proposal of carbon peak and carbon neutrality targets has accelerated the transition from traditional fossil fuels to renewable energy sources, positioning energy storage as a critical enabling technology. Among various energy storage solutions, lithium-ion batteries stand out due to their advantages in power density, capacity, and durability, making them widely adopted in commercial and residential energy storage systems. However, energy storage cells generate significant heat during charging and discharging cycles, originating from reaction heat, side reaction heat, Joule heating, and polarization heat. This heat accumulation adversely affects the lifespan of energy storage cells and, in large-scale containerized energy storage cabinets where multiple cells are connected in series and parallel, can lead to reduced efficiency or even hazardous events like thermal runaway. Thus, developing effective thermal management systems to dissipate heat and ensure safe operation is paramount.
Current cooling methods for energy storage cells include air cooling, liquid cooling, and phase change material cooling. Liquid cooling, which utilizes a circulating coolant to absorb and remove heat via convection in flow channels, is favored for its high efficiency and cost-effectiveness. Numerous studies have focused on optimizing liquid cooling performance, particularly through flow channel design enhancements to improve flow distribution and reduce resistance. This paper introduces a novel stamped flow channel structure for liquid cooling of energy storage cells and employs computational fluid dynamics (CFD) to establish a fluid-thermal-solid coupling model. The internal flow field and overall temperature distribution are analyzed to evaluate cooling performance and flow resistance. Furthermore, the impact of inlet flow rate on system performance is investigated, and structural optimizations, such as the incorporation of convex hull features, are proposed to enhance fluid disturbance and convective heat transfer. The findings provide valuable insights for thermal management design in lithium-ion energy storage systems.
The physical model comprises a liquid cooling plate assembly and a battery pack of energy storage cells. The assembly includes a base plate, a flow channel plate, and inlet/outlet nozzles, while the battery pack consists of 52 cell units. The energy storage cells are positioned atop the base plate, with a thermal interface material (thermal conductivity: 0.8 W/(m·K)) applied to minimize thermal resistance. Foam spacers (thermal conductivity: 0.022 W/(m·K)) are inserted between cells to reduce cross-heat transfer and prevent localized hotspots. The flow channel plate is fabricated via stamping to form recessed channels, which are then brazed to the base plate to create the liquid cooling module. The flow channel design features a depth of 4.5 mm, with an inlet splitting into three parallel branches—two of which further subdivide—and an outlet combining flows from two main paths. All branches follow a serpentine layout to promote uniform cooling.
For CFD analysis, the three-dimensional model is discretized using unstructured polyhedral grids. To balance computational accuracy and cost, local mesh refinement is applied, particularly in critical regions. Boundary layer grids are implemented near channel walls to resolve near-wall flow and heat transfer phenomena, with a first-layer thickness of 0.2 mm to ensure y+ values below 5, three boundary layers, and a growth rate of 1.1. Grid independence is verified by comparing results from three mesh configurations, as summarized in Table 1. The simulations indicate negligible deviations in outlet pressure and average battery temperature (below 1%), confirming that mesh #2 achieves sufficient accuracy for subsequent analyses.
| Mesh ID | Max Element Size (mm) | Min Element Size (mm) | Element Count | Outlet Pressure (Pa) | Battery Avg Temp (°C) |
|---|---|---|---|---|---|
| #1 | 17 | 1.5 | 2,817,518 | 705.07 | 34.17 |
| #2 | 16 | 1.2 | 4,055,112 | 701.46 | 34.51 |
| #3 | 15 | 1.0 | 5,547,631 | 700.81 | 34.77 |
The governing equations for fluid flow and heat transfer are solved using the Reynolds-averaged Navier-Stokes (RANS) approach in the Fluent solver. The conservation equations for mass, momentum, and energy are expressed as:
$$ \frac{\partial \rho}{\partial t} + \nabla \cdot (\rho \vec{u}) = 0 $$
$$ \frac{\partial (\rho \vec{u})}{\partial t} + \nabla \cdot (\rho \vec{u} \vec{u}) = -\nabla p + \nabla \cdot (\mu \nabla \vec{u}) + \nabla \cdot (-\rho \vec{u}’\vec{u}’) $$
$$ \frac{\partial (\rho e)}{\partial t} + \nabla \cdot (\rho \vec{u} e) = \nabla \cdot (k_{\text{eff}} \nabla T) + \nabla \cdot (\vec{u} \cdot \tau) $$
where $\rho$ is density, $p$ is pressure, $\mu$ is dynamic viscosity, $\vec{u}$ is the mean velocity vector, $-\rho \vec{u}’\vec{u}’$ represents Reynolds stresses, $e$ is total energy, and $k_{\text{eff}}$ is the effective thermal conductivity. The SST k-ω turbulence model is employed for closure, with its transport equations given by:
$$ \frac{\partial (\rho k)}{\partial t} + \nabla \cdot (\rho \vec{u} k) = \nabla \cdot (\Gamma_k \nabla k) + G_k – Y_k $$
$$ \frac{\partial (\rho \omega)}{\partial t} + \nabla \cdot (\rho \vec{u} \omega) = \nabla \cdot (\Gamma_\omega \nabla \omega) + G_\omega – Y_\omega + D_\omega $$
Here, $k$ is turbulent kinetic energy, $\omega$ is specific dissipation rate, $G_k$ and $G_\omega$ are production terms, $\Gamma_k$ and $\Gamma_\omega$ are effective diffusivities, $Y_k$ and $Y_\omega$ are dissipation terms, and $D_\omega$ is the cross-diffusion term.
Steady-state, pressure-based simulations are conducted with an incompressible coolant. Material properties are listed in Table 2. The energy storage cells operate at a 0.5 C discharge rate with a heat generation of 14 W per cell and an initial temperature of 35°C. The inlet boundary condition is set as mass flow inlet with a flow rate of 5 L/min and coolant temperature of 18°C. The Reynolds number is calculated as 2,703.4, indicating a transitional flow regime. The outlet is defined as pressure outlet with zero gauge pressure. Adiabatic conditions are applied to external surfaces, neglecting natural convection and radiation. The SIMPLEC algorithm couples pressure and velocity, with second-order discretization schemes for pressure, momentum, and energy.
| Material | Density (kg/m³) | Viscosity (kg/(m·s)) | Specific Heat (J/(kg·K)) | Thermal Conductivity (W/(m·K)) |
|---|---|---|---|---|
| 50% Ethylene Glycol | 1,074.194 | 4.216e-03 | 3,273 | 0.382 |
| Aluminum (Base/Channel Plate) | 2,719 | – | 871 | 202.4 |
| LFP Energy Storage Cell | 2,284 | – | 1,060 | kx=9.04, ky=11, kz=3.56 |
Model validation is performed by comparing simulation results with experimental data from literature under 1 C, 2 C, and 3 C discharge rates. The close agreement in cell surface temperature trends and minimal errors confirm the accuracy of the Fluent solver for thermal analysis of energy storage cells.
Under a 0.5 C discharge and 5 L/min inlet flow, the flow field and temperature distribution are analyzed. The pressure distribution within the channels shows a maximum total pressure at the inlet, decreasing gradually along the flow path, with a pressure drop of 9.828 kPa between inlet and outlet. Flow streamlines and velocity contours reveal uniform flow distribution without significant recirculation or vortices at bends, indicating high channel utilization. The coolant temperature rises from 18°C at the inlet to 20.49°C at the outlet, with a temperature difference of 2.49°C due to convective heat exchange with the base plate. The temperature distribution across the energy storage cell pack shows a minimum of 33.77°C in the first row and a maximum of 35.17°C in the fourth row, resulting in a maximum temperature difference of 1.40°C. This non-uniformity arises from positional variations: rows above inlet branches benefit from cooler coolant, while the fourth row, above outlet branches, experiences warmer coolant and reduced heat transfer.
The influence of inlet flow rate on cooling performance and flow resistance is examined for rates of 5, 8, 10, 12, and 15 L/min. As flow rate increases, the average temperature of the energy storage cell pack decreases, but the rate of improvement diminishes. For instance, the average temperature drop per unit flow rate decreases from 0.30°C/(L/min) between 5 and 8 L/min to 0.08°C/(L/min) between 12 and 15 L/min. The maximum temperature difference initially decreases from 1.40°C at 5 L/min to 1.22°C at 8 L/min, then gradually increases to 1.39°C at 15 L/min, indicating optimal uniformity at 8 L/min. Flow resistance, however, rises sharply from 9.828 kPa at 5 L/min to 59.441 kPa at 15 L/min, while system thermal resistance decreases modestly from 23.58°C/kW to 21.79°C/kW. This highlights the need to balance cooling performance and flow resistance to minimize pump power consumption in energy storage systems.
To enhance temperature uniformity, the flow channel structure is optimized by introducing convex hulls in the outlet branches. These features disrupt the thermal boundary layer and intensify fluid disturbance, thereby improving convective heat transfer. The original four outlet branches are consolidated into two, with convex hulls evenly spaced along each branch. Vortex identification using the swirling strength method ($\lambda_{ci}$) quantifies fluid disturbance, where the velocity gradient tensor is decomposed to isolate rotational components. Simulations with 3, 4, and 10 convex hull pairs are compared to the baseline. Results show that convex hulls generate distinct vortex structures, increasing average swirling strength and reducing the temperature of the fourth row of energy storage cells. The relationship between convex hull count, average swirling strength, and maximum temperature difference is summarized in Table 3. While more convex hulls improve uniformity, the effect saturates, necessitating a trade-off with manufacturability.
| Convex Hull Pairs | Average Swirling Strength (s⁻¹) | Max Temp Difference (°C) |
|---|---|---|
| 0 (Baseline) | 12.5 | 1.40 |
| 3 | 18.3 | 1.25 |
| 4 | 21.7 | 1.18 |
| 10 | 28.4 | 1.12 |
In conclusion, this study presents a stamped flow channel design for liquid cooling of energy storage cells and utilizes CFD simulations to analyze thermal and flow characteristics. At an inlet flow rate of 5 L/min, the channel exhibits a flow resistance of 9.828 kPa and a maximum temperature difference of 1.40°C, demonstrating good temperature uniformity. However, increasing the flow rate significantly elevates flow resistance with diminishing cooling returns, underscoring the importance of optimizing flow rate to balance performance and energy consumption. Structural optimization via convex hulls enhances fluid disturbance and reduces temperature non-uniformity, with the degree of improvement dependent on hull count. These findings offer practical guidance for thermal management design in lithium-ion energy storage systems, emphasizing the role of flow channel geometry in achieving efficient cooling for energy storage cells.
The development of advanced cooling strategies is crucial for the reliability and longevity of energy storage cells. Future work could explore multi-objective optimization combining flow channel design, coolant properties, and operational parameters to further enhance thermal performance. Additionally, experimental validation under real-world conditions would strengthen the applicability of these CFD-based insights. As energy storage technology evolves, continuous improvements in thermal management will be essential to support the global transition to sustainable energy systems.