CWIEME Berlin 2016 : Daily pickup
Multi-objective Optimization Based on Multiphysics Analysis

Last Updated 2016-05-10

Multi-objective Optimization Based on Multiphysics Analysis

The problems we must consider in the design phase are things such as thermal design, mechanical strength design, and occasional multiple physical phenomena. Because mechanical strength declines due to the flux barrier in an IPM motor design, the intensity of the bridge part will be one of the constraint conditions.
We obtained the Pareto optimum solution by including stress on the bridge part at maximum rotation speed as the constraint condition for multi-objective optimization that set average torque maximization and magnet volume minimization as the evaluation function. Torque obtained from magnetic field analysis and stress obtained from structural analysis were evaluated at the same time.


Fig. 1 Designs with stress distribution for a Pareto optimum solution.

The gray figure on the left shows the geometry information of the magnet and flux barrier used as design variables of the multi-objective optimization.
The three contour plots on the right show the Mises stress distribution when maximum rotation speed of 5500 rpm is applied to each design proposal. The upper right is the design proposal for maximum torque; the bottom left is the design proposal for when the magnet is minimized. The center design proposal shows a balance of both.


Fig. 2 Changes in the Pareto optimum solution due to stress.
The black dotted line shows a Pareto optimum solution with average torque and magnet volume as the objective function. The red dotted line shows a Pareto optimum solution with stress added as a constraint condition.
(Hiroyuki Sano)
2016-04-27

Archive

Archives