Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE TWELFTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY
Equivalent Orthotropic Material Properties for Stators Of Electric Cars
P. Millithaler1,2, E. Sadoulet-Reboul2, M. Ouisse2, J.-B. Dupont1 and N. Bouhaddi2
1Vibratec, Écully, France
P. Millithaler, E. Sadoulet-Reboul, M. Ouisse, J.-B. Dupont, N. Bouhaddi, "Equivalent Orthotropic Material Properties for Stators Of Electric Cars", in , (Editors), "Proceedings of the Twelfth International Conference on Computational Structures Technology", Civil-Comp Press, Stirlingshire, UK, Paper 201, 2014. doi:10.4203/ccp.106.201
Keywords: homogenization, laminated structures, equivalent homogeneous material, stator of electric machine, super-elements..
In this paper, a novel method of equivalent material identification is proposed for orthotropic and multi-layered laminated structures. It gathers a set of simple static simulations on finite element models (i.e. virtual testing), inspired from experimental testing processes, in order to identify the equivalent material's nine elastic coefficients directly: Young's moduli, shear moduli and Poisson's coefficients. Unlike the other homogenization techniques, the method may be applied to superelements in addition to fully-defined finite element models. The method is applied to the models of two stators of electric cars. As it is usually the case, a stator consists of a stack of several hundreds of steel sheets, less than a half-millimetre-thick and separated from each other by an insulating resin in order to prevent eddy currents from taking place. Taking into account the additional heterogeneity arising from the weld beads in the stator's lateral face, the model is divided into several zones, each of which stands for a specific equivalent material. The simulations performed at low frequencies enable prediction of both the real stators' dynamic behaviour with good precision, according to experimental data. In particular, the cylinder modes, most critical for industrial studies as a result of their high probability of coincidence with electromagnetic excitations, are well predicted. These results offer interesting perspectives in the prediction of the dynamic behaviour of stators of electric machines during the design stages, without the need for time-consuming and delicate model updating procedures with experimental data from expensive prototypes.
purchase the full-text of this paper (price £20)