An Improved Dual Iterative Transient Thermal Network Model for PMSM With Natural Air Cooling

Abstract

Thermal analysis is vital in the design stage of permanent magnet synchronous motors (PMSMs). Though recent studies have developed the thermal equivalent circuit models for internal temperature calculation, they cannot consider transient modification of thermal conditions, and have a notable drawback of low accuracy. Therefore, this study addresses the neglected aspect of transient thermal analysis. A thermal network model (TNM) is proposed with inclusive coupling mechanisms, covering the increase of loss power and change of thermal parameters. Proposed TNM is combined with a dual iteration technique in order to achieve accurate prediction of thermal condition during time-stepping calculation. Also, the solving process is improved without calculating the inverse matrix after Laplace transform of the TNM system. Targeting the 10/12 PMSM that is designed and manufactured, temperature rises on motor parts are evaluated by analytical TNM, finite element analysis, and thermal tests. The study identifies the well-agreed temperature results and the necessity of transient analysis for accurate prediction of temperature rise.