Measurement of the magnetic characteristics of soft magnetic materials with the use of an iterative learning control algorithm

The quality of finite elements simulations is depending on the used magnetic characteristic curves and material coefficients on a high degree. Especially in the automotive industry the power density of electric machines are tending to increase. Thereby the quality of a whole design optimization process for highly saturated magnetic circuits is depending on the accuracy of the material data. To measure the magnetic characteristics of materials at high field strengths and high frequencies, the losses in the coil and the material has to be reduced because of thermal influences. On the one hand the influence of the heating itself is unwanted and on the other hand, especially when small test specimen are used, the maximum measuring time is limited to the temperature and sometimes leads to bad control accuracies. To overcome this problem an iterative measurement cycle is introduced in this paper.