60 Ah Pouch Cell Thermal Model: Fluent liquid-cooled battery simulation

This Ansys Fluent model predicts how a 60 Ah lithium-ion pouch cell heats up and cools during a full charge and discharge cycle. An aluminium T-shaped cold plate with water flowing through it covers the top band and centre of the cell. It reproduces the validation test of Dileep et al. (2026) and compares the cell's average temperature with their measurements and their own CFD results.

What it's good at

• Shows the cell coldest under the top bar and hottest at the uncovered sides

How it works

The NTGK battery model solves two electric potentials in the cell, using fitted curves of depth of discharge for equilibrium voltage (U) and conductance (Y) to get the current and the heat generated. Heat sources are reaction and entropy heat in the cell and Joule heating in the tabs, with conduction through the solids and convection into the water and the air. The laminar coolant flow is solved once and then held fixed, because the water properties are constant. Time stepping is second-order implicit with 1 s steps and 15 iterations per step.

Setup

• Cell: 60 Ah NMC/graphite pouch, 3.7 V nominal, 156 × 245 × 12 mm, with two 50 × 33 mm tabs
• Cold plate: T-shaped, 9 mm thick, with a 5 mm water channel, two inlets and one outlet
• Model extent: half the cell thickness, with a symmetry plane at the centre
• Mesh: 993,157 polyhedral cells made in Fluent Meshing; faces 1 mm overall, 0.5 mm at the tabs, 0.3 mm at the ports; minimum orthogonal quality 0.20
• Cell materials: density 2400 kg/m³, specific heat 750 J/kg·K, thermal conductivity 22 / 5 / 1 W/m·K; aluminium plate and positive tab, copper negative tab, water with constant properties
• Inlet: 0.002 kg/s at 40 °C (half of 0.004 kg/s for the half model); pressure outlet
• Outer surfaces: h = 5 W/m²·K to 45 °C air; everything starts at 45 °C
• Load cycle: starting depth of discharge 0.95; 1C charge for 3600 s, 600 s rest, then 2C discharge to the 3 V cut-off; half cell is 30 Ah, so 30 A at 1C and 60 A at 2C

Validation and results

The average temperature has an RMSE of 0.67 °C against the experiment, with a maximum error of 2.14 °C (5.2%). The paper's CFD gives 0.80 °C RMSE and 4.5% maximum error, from curves digitised from Fig. 4 (the paper states 4.3%). Peak cell temperature is 48.16 °C, against 47.2 °C measured and 48.0 °C in the paper's CFD, and the discharge ends at 5938 s at 3 V. Cutting the solver to 15 iterations per step changed the temperature curve by at most 0.04 °C.

How to use it

The case file (.cas.h5) holds the mesh, materials, boundary conditions and battery model settings, and a PowerPoint summary of the set-up and validation is also provided. Open it in Ansys Fluent 2024 R2 or newer; a full run takes about 26 hours on 14 cores.

Limitations

The model misses the rapid early cooling in the first 600 s, the temperature bump near 1600 s in the charge and the continued heating after 5300 s in the discharge; the paper's CFD misses these too. The flow is frozen and laminar, heat generation is uniform within the cell, there is no contact resistance, and only one half cell is modelled, not the paper's three-cell module. Cell conductivity in y (5 W/m·K) and the cell size (156 × 245 mm, against 178.5 × 286 mm in the paper's Table 2) are not confirmed against the paper, and the U and Y curves come from the original Fluent case and are unchecked against their source.

Reference

Dileep, H., Mahapatra, P. S., & Pattamatta, A. (2026). Lightweight thermal management strategy for Li-ion pouch cells using localised cold plate and graphite sheet. Thermal Science and Engineering Progress, 71, Article 104570. https://doi.org/10.1016/j.tsep.2026.104570