Learning Forum

[Magnet_man]

I've previously used other multiphysics software for magnetostatic FEA, and have now switched to Ansys.

One thing that surprises me is the adaptive meshing feature - and if it's really as amazing as it's advertised.

I vividly remember from my University classes that pouring more elements into a model is not always a good solution, as it rapidly increases computation time, especially if it's done without careful consideration.

I'm currently looking at a 3D magnetostatic model, which have gone from an initial 375k elements to now over 3 million in 9 adaptive passes. I would have thought that reduction of step size would be the better solution, as i wonder how much extra computation time is added by having to mesh, solve, re-mesh and solve again.

Anybody can give some insights on this?

Cheers!

[Ashish Khemka]

If you see that adaptive mesh is taking too much time then define the mesh controls manually. It may be less time consuming.

Regards Ashish Khemka

[Rob]

It's a very useful too, but needs to be used with care. Remeshing may be more efficient with compute resources, but adaption may save you time. Which is in shortest supply, cpu or person time?