Where does r^2 go? Inside or outside of inner product? Doesn't seem to matter?

3D can be rescaled by maximum. 1D seems perfect up until material conservation kicks in at boundary.

Probably need proper dependence on phi_tot instead of direct dependence on x.

But only up to a factor between different partitionings/diffusion coefficients outside.
Problem: Maybe need different concentration scaling for radial problem?

Take care of volume elements everywhere! Needs to be in all integrals (also t) and mass conservation check.