Discussion Forum

Hi,

I am trying to solve a non-steady state problem in spherical coordinates using a square as a geometry for possible latter extrusions to other linear geometry.

Thus the real equation I want to solve is (du/dt = grad(-c*r^2 grad(u)))

In the generic equation, (du/dt =grad (-c grad (u))), I introduce the diffusion term as follows c= D*y^2 being y^2 the height of the square (effectively the radial direction). I then introduce y^2 as well in the mass coefficient (as if all the original equation had been multiplied by y^2).

Is this approach correct?

thanks