I have read here that gradient calculations do not work properly for cylindrical 2D grids, because of discretization erros. Does that mean that DiffusionTerm and/or other terms in FiPy are also broken for cylindrical 2D grids? Or is it just the gradient method alone?
Just the .grad
method. DiffusionTerm
doesn't use it (except with anisotropic diffusion coefficients).