Date of Award
6-2026
Degree Name
Master of Science in Engineering
Department
Mechanical and Aerospace Engineering
First Advisor
Richard T. Meyer Ph.D.
Second Advisor
Sandun S. Kuruppu Ph.D.
Third Advisor
Shrabanti Roy, Ph.D.
Keywords
ANSYS, CFD, EV thermal system, ICI system, thermal analysis, thermal management
Access Setting
Masters Thesis-Open Access
Abstract
This work presents a computational and experimental investigation into the thermal management of power electronic modules for an electric vehicle (EV) integrated charger-inverter (ICI) system, with a focus on the Inverter module, Dual Active Bridge (DAB) converter with a High- Frequency Transformer (HFT) with an external leakage inductor. Analytical loss models are used to calculate conduction and switching losses for the DAB and inverter SiC MOSFET modules, and the Improved Generalised Steinmetz Equation is applied to determine HFT core and winding losses. The adequacy of passive natural convection as a thermal management strategy is first assessed, and its limitations at elevated power levels are demonstrated.
A serpentine liquid-cooled aluminium cold plate solution is subsequently designed and evaluated using conjugate heat transfer Computational Fluid Dynamics (CFD) simulations in AN-SYS Fluent, employing a laminar flow model with SIMPLEC pressure-velocity coupling through-out. Simulations were conducted at operating power levels of 1.2kW, 5kW, and 15kW. With a coolant inlet temperature of 22◦C and a flow velocity of 0.5ms−1, the cold plate maintained DAB MOSFET temperatures of 24.59◦C and 25.87◦C at 1.2kW and 5kW respectively, and inverter MOSFET temperatures of 24.04◦C and 29.22◦C at the same power levels. At the worst-case US drive technical team road map (USDTTRM) condition of an 85◦C coolant inlet at 15kW, the DAB and inverter MOSFET temperatures reached 86.72◦C and 100.43◦C respectively, with calculated junction temperatures of 92.91◦C and 121.22◦C, all within the 180◦C device limit. For the HFT at 15kW, five of the seven interface gap configurations evaluated - comprising combinations of direct contact, air-gap, and thermal interface material fill at the cold plate–winding, primary-secondary windings, and between core leg windings - maintained maximum core and winding temperatures below the 100◦C USDTTRM limit at the 85◦C worst-case inlet condition, with the two non-compliant designs identified as those in which the inter-winding gaps are modelled as air layers.
Experimental validation was carried out for all three primary modules using thermistor measurements. Thermistor-to-CFD discrepancies for the DAB were 1.63% and 0.86% at 1.2kW and 5kW, respectively, and 1.05% and 0.65% for the inverter at the same conditions. At 15kW, the inverter cold plate transient analysis final CFD prediction result agreed with experimental measurements final reading to within 4.32%, with the transient discrepancy attributed to the finite thermal time constant of the assembly and the absence of external coolant loop thermal capacitance in the model. HFT validation at 1.2kW yielded agreement within 1.77% across all measurement locations. In all DAB and HFT cases, maximum discrepancies remained below 5%, confirming the appropriateness of the CFD methodology for these two modules. However, the leakage inductor CFD model substantially over-predicted experimental temperatures by approximately 62◦C, at-tributed to an overestimation of the applied heat source and the neglect of forced convection from laboratory air movement.
Recommended Citation
Wijethunga, Pihillawala, "Finite Element Analysis and Computational Fluid Dynamics to design and analyze a liquid cooling system for an Integrated Charger Inverter System" (2026). Masters Theses. 5564.
https://scholarworks.wmich.edu/masters_theses/5564