Multi-objective CFD-based optimization of cold plates with non-uniform inlet and channel geometries

With the increasing heat generation of batteries in electric vehicles, efficient thermal management has become critical. Cooling plates, as primary heat sinks, play a vital role in maintaining battery safety and performance. This study develops a fully coupled three-dimensional multi-objective optimization integrated with CFD to minimize both pressure drop (Δ p ) and thermal resistance ( θ ) in cooling plates featuring irregular inlet profiles and variable channel widths. Two models are proposed: the CO-CP model optimizes the widths of fins and channels, while the CIO-CP model further adjusts the inlet profile. Results reveal that varying channel widths affect both Δ p and θ , whereas the inlet profile predominantly influences Δ p , achieving over 20 % reduction. Correlation analysis identifies the first channel width as the most influential geometric parameter. Through multi-objective decision analysis, the optimal CIO-CP configuration demonstrates an outward-expanded inlet that effectively weakens the recirculation zone and enhances coolant distribution to the hottest region. Compared to the initial design, the optimized CIO-CP achieves simultaneous reductions of 18.1 % in Δ p and 16.7 % in θ . This work provides a comprehensive optimization framework and key design insights for improving the performance of battery cooling plates.