HI All, In my code I need to construct a residual analytical function (Navier-Stokes) using FVM. The solution of any code based on SIMPLE algorithm on a colocated grid I introduce in my code does not satisfy the mass conservation, since I don't know how to correct the velocity on the face explicitely from a given (U,V,P) on the cell center. I tried the Rhie and Chow interpolation but it seems implicit.

In the staggered arrangement case, I don't have this pb since the velocities given are on the cell face.

Any idea?

Thanks a lot

You could try to implement the staggered arrangement. It is harder to implement since you need different locations for each velocity and pressure but it makes life easier since you don't need to use interpolation and get the velocities in the CV face. In Patankar's book you can find an explanation on how to correct the velocities and specify boundary conditions for the pressure equation on staggered grids.

Thank you Peter for your suggestion. The staggered case works very well for the reason you explained. Actually I need to work with colocated arrangement, since they are easier to implement for non orthogonal grids or 3D cases. For this reason I need an explict way to recompute the velocities on the faces from the field on the cell center so that the mass conservation is satisfied.

thanks again