An improved temperature prediction model of motor considering air flow field for in-wheel electric vehicle applications

Abstract

Accurate real-time temperature prediction is critical for optimizing the efficiency, reliability, and operational safety of vehicles equipped with in-wheel motors (IWMs). The thermal behavior of IWMs constitutes a complex multi-physics problem involving tightly coupled electromagnetic, thermal, and fluid flow fields. A significant challenge arises from the IWM’s integration within the rotating wheel assembly: vehicle motion induces highly transient, nonlinear airflow around the motor, profoundly influencing its convective heat dissipation. To address this challenge, an integrated thermal prediction framework is developed for IWMs. Electromagnetic finite element models, core loss models, and a lumped-parameter thermal network (LPTN) are established, explicitly accounting for temperature-dependent material properties and performance parameters. Crucially, computational fluid dynamics (CFD) simulations are conducted from a full-vehicle perspective to characterize the aerodynamic flow and convective heat transfer coefficient (CHTC) distributions on the IWM surfaces within the rotating wheel environment across varying vehicle speeds (VSs). Through integration of these components, a bidirectional electromagnetic-thermal coupled prediction model is formulated, dynamically incorporating spatially resolved airflow velocity and CHTCs on each motor surface. Experimental validation confirms the model’s effectiveness. Results demonstrate that high accuracy in predicting component temperatures under diverse operating conditions is achieved by the proposed framework.