Harmonic mean values of positive definite tensors: Motivation for developing a MPFA-scheme
User of UG and DUMUX-Codes often apply a generalized formula of harmonic means for permeablilities. This formula, originally derived for 1D-Codes, has some serious drawbacks in its simple generalization. I want to discuss this topic and present a different method for calculating a mean tensor.
Starting with these thoughts on interpretation of weighted permeability tensors in the application of finite volume discretizations, I want to go on with a motivation for a certain multipoint flux approximation (MPFA) method.