Learning Forum

[jrozsa]

I am trying to simulate the charge induced on two parallel copper plates (20mmx12mm) by a charged, ferrite sphere (60 micron diameter) using electrostatic analysis. However, I haven't been able to get any solutions to converge because of the amount of memory and processing power that the simulation is using. I'm running the simulation on a supercomputing node with 512 GB of RAM and I still haven't been able to get any solution to converge. Is there anything that I can do to get the simulation to run faster or use less memory? It seems to me like this would be a relatively simple phenomenon to simulate, so I'm wondering if there's something I'm doing that's keeping it from converging. I'm attaching a screenshot of the model to give a better idea of what I'm trying to do. 

[mgardner]

You probably need to use less mesh elements.  Here are some suggestions:

 

1) Reduce your simulation region size.  Your region probably doesn't need to extend very far beyond the plates.

2) If you are assigning mesh settings, try making them looser.

3) Meshing perfectly curved surfaces is hard.  You could try modeling the sphere as a 12-hedron or even as a cube.

 

Generally, you are going to have some difficulty because your sphere is so much smaller than the other objects in your geometry.  This means Maxwell will have to use really small elements to accurately model the region near the sphere.  Because your geometry is so large relative to this minimum mesh size, Maxwell will end up using a ton of elements.