Electromagnetic-thermal coupled multi-objective optimization on HPC systems for rotating electrical machines used in phev applications

Sign in to download the data

Download PDF

David Philipp Morisco, GS-EH/EDM1, Robert Bosch GmbH

Abstract

During the design process of electrical machines for hybrid- and electrical cars, multi domain considerations are required.
Most conventional design algorithms include thermal influences on the material behavior only partial, by considering them separately after the electromagnetical design evaluation.
Although the individual review of electromagnetical and thermal domain increases the effort within the designing process, a coupled analysis between the two domains allows an more realistic design iteration process.
In case of a multiobjective optimization process the usage of multicore high performance computing systems significantly increases the speed and the amount of calculated designs.
For the developed optimization algorithm the electromagnetical-thermal coupled analysis in JMAG designer is used to optimize an expert design using multicore high performance computing systems.

Search Filter

  • All Categories