Numerical experimental investigation of flow and heat transfer characteristics in micro-channel cold plates for electric vehicle power electronics

This paper presents a numerical experimental investigation of fluid flow and heat transfer characteristics in micro-channel cold plates for EV power electronics cooling. We conduct a systematic parametric study of geometric and operational parameters including channel width (0.3-2.0 mm), AR (1:1 to 5:1), flow rate (0.5-3.0 L/min), and coolant properties, covering an R e range of 100–2000 representative of laminar-to-transition flow regimes. The numerical experiments employ a validated CFD model with experimental data from literature, achieving less than 8% deviation in temperature predictions. Results reveal that the optimal thermal–hydraulic performance occurs at channel widths of 0.8-1.2 mm and AR values of 3:1-4:1, where the trade-off between heat transfer enhancement and pressure drop is balanced. A new dimensionless correlation for Nu as a function of channel geometry and flow parameters is developed from 256 numerical experimental cases. The correlation predicts experimental data from three independent studies with MAE of 9.7%. This work demonstrates how numerical experimentation can efficiently explore design spaces that are challenging to investigate physically, providing fundamental insights and practical guidelines for micro-channel cold plate design.