Thermal design optimization of traction battery modules using meta-model-based simulations

Fast simulation models that predict the temperature difference in a battery cell depending on the geometric cell design are needed for the evaluation, comparison and thermal optimization of battery modules with tab cooling and other cooling approaches. Here, a new simulation model based on a neural network is proposed which rapidly predicts the maximum temperature difference in a single battery pouch cell after a charging process. The model considers all relevant geometric and thermal parameters. Using this model, a charging process of a battery module can be computed in less than 18 s. This enables the thermal optimization of pouch cell designs and the comparison of different cooling approaches on module level, which is demonstrated on two exemplary test cases. Furthermore, a tab-cooled battery module prototype using thermally conductive plastic materials is developed and analyzed. The simulation model is partly validated with the prototype and its cooling performance is evaluated on.