Due to the complex energy conversion processes and limited avenues for heat dissipation in the giant magnetostrictive transducer (GMT). Large losses and elevated internal temperature rise are inevitable, this would bring into serious reliability problem for GMT. Accurately analyzing its electromagnetic losses and hot-spots temperature are crucial for designing and optimizing the transducer. High-power GMT typically use permanent magnet (PM) to provide a bias magnetic field for enhanced power density. Although, this structure could significantly reduce the volume of driving part, it also caused issues such as non-uniform magnetic field distribution and bias magnetic field, which have great influence on the magnetic energy loss and hot-spots temperature rise. The previous researches seldomly investigate these influences on total loss at GMT system level. Additionally, the differential excitation coil loss caused by AC resistance among multi-layers can’t be neglected, which can lead to variations in coil loss and temperature rise. Therefore, this paper proposed a detailed electromagnetic losses and hot-spots temperature rise calculation model from material level to GMT system level. This method can obtain the distribution of electromagnetic losses within multiple factors, and the hot-spots temperature rise considering the temperature non-uniform distribution. Firstly, we have developed a precise loss calculation model, which takes the influence caused by PM bias and non-uniform magnetic field on magnetic energy loss into account. Additionally, the proposed model could conclude the impact of inter-coil resistance variations on the eddy current loss in excitation coil. Secondly, based on of the non-uniform distribution of losses in driving part, using the divisional thermal modeling approach and a 2-D T-shaped equivalent thermal network, we further constructed a transient temperature calculation model for GMT. This approach can effectively improve the energy conversion efficiency of GMT, so as to enhance the thermal reliability of the transducer. Finally, a testing system for the loss-thermal of GMT was established to validate the effectiveness of proposed model. Results indicate that the average error between our loss calculation model and experimental results is 2.93%, while the corresponding error for without considering the above factors is 15.97%. This confirms the necessity of conducting a detailed electromagnetic losses calculation model. In addition, the average error of the temperature rise calculation model is less than 5%, while the error for without considering the losses distribution and heat dissipation paths difference is exceed 10%. In conclusion, our proposed method accords well with the measured results, and can provide significant fundamental for design and optimization of GMT.
Read full abstract