A study on the design optimization of a switched reluctance motor for improving sound quality

Abstract

A Switched Reluctance Motor (SRM) has high efficiency and high torque density without rare earth material. In addition, the structure is simple, and operation is possible over a wide speed range. Thus, compared to brushless DC motors and interior permanent magnet synchronous motors, which use rare earth materials, SRM can be competitive in terms of price and environment. Since the SRM has a doubly salient structure, however, it generates large vibration and noise. In order to commercialize and popularize SRMs, vibration and noise problems of SRMs should be improved. Particularly, the demand related to the noise of the consumer of the product continues to increase, and the noise of the product greatly affects the purchase of the consumer. Therefore, the evaluation of SRM's noise felt by humans and the study for its control are essential procedures in design stage of SRMs. There have been many researches related to the noise reduction of SRMs. A study on the primary noise source of a SRM, and a vibration and noise analysis with experiments of SRMs have been performed. Also, design optimization studies have been carried out to reduce the noise of SRMs. Most of these design studies have focused on indirect causes of noise, such as torque ripple, rather than directly handling the sound pressure radiated from the stator. However, the results of such optimization studies cannot easily predict exactly how much the radiated sound pressure has been reduced and the result design may differ from a design in which the radiated sound pressure is directly calculated and reduced. For this reason, it is necessary to obtain the radiated sound pressure in the time or frequency domain of the SRM and to study the design optimization considering the calculated sound pressure. The degree to which humans feel about noise is difficult to be expressed simply by the physical quantity of pressure. Sound quality metrics were developed by psychoacoustic researchers in Germany as an objective indicator of human’s subjective feelings of sound. There are four main metrics of sound quality, which are loudness, sharpness, fluctuation strength and roughness. Some studies have been carried out to consider the sound quality metrics. However, no studies on SRM’s sound quality have been performed before. The main cause of the noise emitted from the SRM is the radial attractive force generated between the stator and the rotor. This force becomes an exciting force that causes an oscillation of the stator and eventually generates a sound field in the outward direction of the stator. In this study, we evaluate both average torque and sound quality metrics which are loudness, sharpness, fluctuation strength and roughness from the radiated sound pressure of a SRM and consider them as an objective function in the design optimization process. On the other hand, when designing to reduce the noise of the SRM, the average torque should be evaluated to prevent the decrease in average torque, and the average torque should be considered in the design optimization process. In order to evaluate the average torque, we perform a magnetostatic Finite Element Analysis (FEA). Next, based on the input currents during actual operation, other magnetostatic FEAs are performed to obtain a local force distribution acting radially on the inner surface of the stator at each angle of the rotor position. A transient structural FEA of the stator is carried out by using these radial forces as excitation forces to obtain the acceleration in the radial direction of the outer surface of the stator. Next, regarding these radial accelerations as input sources, the transient acoustic FEA is conducted to evaluate the radiated sound pressure at the receiving point 1 m away from outside the stator. Then, by using LMS Test.Lab, sound quality metrics are calculated. Finally, we conduct design optimization of a SRM to increase the average torque and to reduce the annoyance defined by correlation between sound quality metrics and jury test results.