电工技术学报2025，Vol.40Issue(22)：7193-7203,11.DOI:10.19595/j.cnki.1000-6753.tces.241836
高速永磁电机的三维集总参数热网络模型研究
Research of the Three-Dimensional Lumped Parameter Thermal Network Model for High-Speed Permanent Magnet Motor
摘要
Abstract
High-speed permanent magnet motors(HSPMMs)are increasingly applied in industrial fields,particularly in environments requiring high power density and rotational speeds.The issue of rising temperatures has become prominent.Excessive temperature rise can lead to a decline in the insulation performance of the motor,demagnetization of the magnetic materials,and even cause motor failure.Therefore,accurately predicting and assessing the rise in temperature during the design phase is crucial.Traditional methods for temperature rise calculation mostly rely on numerical analysis techniques.However,these methods face significant challenges in designing high-speed motors due to their low computational efficiency and high computational load.This paper proposes an efficient computational model based on a three-dimensional lumped parameter thermal network(LPTN)to improve the thermal network approach,ensuring accurate temperature rise predictions while reducing computational complexity. The proposed LPTN model simplifies the complex heat transfer processes within the motor by establishing a rational topological structure.This model employs hollow cylindrical T-shaped and+-shaped topologies,solid cylindrical topologies,and tile-shaped topologies.Specifically,hollow cylinders represent the stator yoke,titanium alloy sheath,permanent magnets,motor housing,end caps,and bearings.At the same time,tile-shaped bodies are used to model the stator teeth and slot winding.The rotation shaft is treated as a combination of hollow and solid cylindrical bodies.The assembly gap is also modeled using a heat resistance representation based on flat-plate conduction.To better account for the impact of temperature on electromagnetic losses,the study also incorporates an electromagnetic-thermal network coupling solution method.This method achieves bidirectional coupling between magnetic and thermal fields through iterative calculations,enabling the consideration of the feedback effect of temperature variations on electromagnetic losses.Consequently,the accuracy of the temperature rise prediction is enhanced. The finite volume method(FVM)calculates the motor's temperature field.It is found that the LPTN offers significant advantages in terms of setup and computational time.The time for a single iteration calculation is only 0.89 seconds,significantly improving computational efficiency,while the temperatures at all nodes in the thermal network are within the range of the FVM results.Additionally,a temperature rise experiment is conducted.Due to limitations imposed by the experimental platform and bearings,tests are carried out under operating conditions with an output power of 7.5 kW and a rotational speed of 15 000 r/min.The results show that the temperature rise errors at the end-winding and bearing locations are within 7%.关键词
高速永磁电机/三维集总参数热网络/磁热双向耦合/热分析Key words
High-speed permanent magnet machines/three-dimensional lumped parameter thermal network/electromagnetic-thermal bi-directional coupling/thermal analysis分类
动力与电气工程引用本文复制引用
张超,司马秉奇,王凯东,孙宁,王晓宇..高速永磁电机的三维集总参数热网络模型研究[J].电工技术学报,2025,40(22):7193-7203,11.基金项目
辽宁省科技厅联合基金项目面上资助计划项目(2023-MSLH-247)和辽宁省教育厅基本科研项目(JYTMS20231232)资助. （2023-MSLH-247）
