Learning Forum

[andy.zhou.nuaa]

Hi, Everyone.
Ｉ used maxwell to do an power inductor simulation with eddy curent solver. i found there is an interesting phenomina: There is an huge winding loss difference with different mesh settings, while all of other conditions are keep same.
Does anyone met this phenomina before ?
Here is some simulation picture :

Does anyone know the reason? which simulation result is more realistic or both of them is incorrect?

[Navya C]

Hi @andyzhou Maxwell uses a proprietary adaptive mesh technique for both magnetostatic and Eddy current solvers.
Adaptive mesh refinement algorithms are designed to automatically detect and refine the mesh in regions with the largest field error.
This process helps to most accurately and efficiently reduce the required simulation mesh the end goal is minimal energy error criteria and increased simulation confidence.
Adaptive refinement repeats until either of the termination criteria(defined during the Analysis setup) are met either the number of passes has reached the maximum or the energy error and Delta energy fall below the percentage target after the adaptive field solution is done.
So, you need not define any additional mesh settings in Eddy current solver, unless you are very sure of what you are doing. You can check the simulation convergence after the analysis is done.
If you think the convergence is not achieved increase the number of steps in the simulation setup.
Also make sure you have defined the solution region sufficient enough.

Regards Navya

[andy.zhou.nuaa]

Hi, Navya.
Did u mean the simulation result based on adaptive mesh technique is the most reasonable result?