Discussion Forum

We do a lot of eddy current calculations and are often interested in response to a crack. Currently, the only way we have to do this is to put in a very thin region and declare it to be air. This, of course results in a very fine mesh or a mesh with very distorted elements. It would be nicer to just declare that Jnormal to a boundary was zero. I've tried using a transition boundary condition but that doesn't produce the right result. Any ideas on how the current boundary condition could be imposed? Typically we work in mf as there are no external current sources.