| The High-speed permanent magnet machine(HSPMM)is widely used as the core driving mechanism for centrifugal compressors due to their outstanding performance,including high-efficiency and high-power density,making them suitable for refrigeration,energy storage,and other fields with vast development prospects.However,the high-speed and high-frequency characteristics of HSPMMs pose certain design challenges.Due to the small volume of the rotor in HSPMMs,the rotor losses increase significantly under high frequency and high-speed conditions,causing problems such as demagnetization of the permanent magnet due to excessive rotor temperature rise.Therefore,it is crucial to study the mechanism and influencing factors of rotor losses,optimize rotor temperature rise and improve machine efficiency from the perspective of machine design for safe and efficient operation of HSPMMs.In this study,a 160 k W compressor with a HSPMM was selected as the research object,and the initial design scheme of HSPMM was determined based on both electromagnetic performance and rotor strength.A two-dimensional electromagnetic finite element model of the machine was constructed using the finite element method(FEM)to calculate the electromagnetic performance of HSPMM under both no-load and load conditions.Meanwhile,the influence of air gap length,sleeve thickness,and conductivity on rotor eddy current losses were studied based on FEM,combining the mechanism of eddy current losses,revealing the distribution of rotor eddy current losses under different sleeve materials.Finally,the sleeve design scheme was determined based on both rotor strength and eddy current losses.A three-dimensional fluid model of the HSPMM air gap and the R134 a state equation were established,and the numerical simulation method was used to study the effects of machine speed,inlet flow rate,and gas species on the fluid flow state in the air gap and the rotor friction loss,revealing the law of Taylor vortices generation and transition on the rotor friction loss under different Reynolds numbers in the air gap.Based on the numerical results of rotor friction losses,the accuracy of different friction loss analytical models was compared,and the research indicated that the Bilgen model considering the axial flow torque has the smallest relative deviation from the numerical results.Meanwhile,due to the strong effect of air viscosity,the rotor friction loss generated by R134 a under the same machine speed is much larger than that of the rotor air friction loss.To address the problem of excessive rotor losses,a HSPMM multi-objective optimization method based on a multidisciplinary analysis model and a multi-objective genetic algorithm was proposed.By using the Kriging surrogate model and rotor friction loss analytical model to replace the numerical method for analysis and calculation,the time cost of the optimization process was greatly reduced.Meanwhile,by taking rotor losses as the optimization objective,the operation performance of the optimized machine was improved while ensuring the multi-physics field performance of HSPMM,providing method support for the multi-objective optimization design of HSPMMs. |