Discussion Forum

Before you begin to prepare a model, you should make sure you understand the physics. For example, if all your fields are static, there will be no induced currents. (In that case, your problem is already done, so you don't need to build a model.) So, let's assume non-static fields. But let us bear in mind that a "charged spherical shell" would be a very odd way to produce a time dependent field. Hmmm. But perhaps you can explain your plan for that. So, moving on, if your subject conductor is a perfect conductor, it will contain no bulk (i.e., interior) current density, nor will it have any internal electric field, but it may have a surface current density and a surface charge density. Regardless, I believe there are many examples of possibly-relevant models of applied EM fields inducing currents, whether in and on imperfect conductors or on the surfaces of perfect conductors, provided in the Comsol-supplied application library. If everything in your region of interest is small compared to a free space wavelength, then begin with the AC/DC module. If not, then you may need to use the RF module. If your fields are purely-sinusoidal in time, then select a frequency domain model. If your fields have some other kind of time dependence, then use a time domain model. (And if your fields are static, then you are not actually inducing any currents, as already noted earlier.) If you can post a more complete picture and description of your physics problem, the expected time or frequency behavior of the field from your "charged spherical shell", and more about what exactly, you are trying to compute, then you will likely receive a more useful answer. Good luck.

Yes, this seems to be a case for some textbook reading. It is inevitable to understand the physics before using modeling tools. My preference is the Alonso Finn series.