Background: The increasing power density requirements of traction electric machines, particularly Permanent Magnet Synchronous Motors in electric vehicle applications, necessitate advanced thermal management strategies to ensure performance, reliability, and longevity. Accurate thermal modeling is crucial during the design phase to predict temperature distribution and identify potential hot spots. While numerical methods like Computational Fluid Dynamics (CFD) offer high fidelity, they are often computationally intensive for system-level analysis and transient simulations. The Lumped Parameter Thermal Network (LPTN) method provides a computationally efficient alternative, offering a balance between accuracy and speed. However, the accuracy of an LPTN model is contingent upon the appropriate definition of the network topology (number of nodes) and the accurate estimation of thermal resistances and heat sources. This research addresses the need for a flexible and validated LPTN modeling tool capable of capturing the thermal complexities of modern traction PMSMs, from simplified representations to high-resolution models suitable for design optimization and cooling system analysis.
Aim: The primary objective of this research is to implement and validate the Lumped Parameter Thermal Network (LPTN) method for the thermal analysis of traction Permanent Magnet Synchronous Motors. The specific aims are:
To develop a computational tool in the MATLAB environment capable of constructing and solving LPTN models with varying degrees of complexity.
To validate the tool's accuracy by replicating reference simple LPTN models (4-node and 7-node) from existing literature under transient conditions.
To construct a comprehensive 64-node LPTN model for the Nissan Leaf 2012 traction motor, incorporating empirical correlations for thermal resistance calculation and electromagnetic loss data.
To validate the developed 64-node model against a commercial tool (MotorCAD) under both steady-stat