Improvement of Thermal CFD EM-Models by Introduction of FMU Sub-Models
The thermal simulation of electric traction motors is very sensitive to a variety of physical processes that cannot be calculated within one solver. There are two solutions for this situation. One is to guess expected values using experience, and the other is co-simulation. Classic co-simulation offers high resolution but very low performance. The performance can be significantly improved by introducing metamodels to the solver. The time to achieve results is only marginally increased, but the quality of the model can be significantly improved.
In this webinar we demonstrate how, by applying co-simulation, the system design can be evaluated more precisely, making the safety factors more tangible and opening space for further design improvement. The generation of metamodels is performed by introducing simplified sub-models into the optiSLang sensitivity study and extracting an FMU directly from optiSLang. This FMU can then be implemented directly into the thermal simulation, be it 1D-analytical or 3D-CFD. The transferability of this approach enables validated models to be implemented in a wide range of simulation tools, improving the thermal simulation from acquisition to application.