[jutta]

Hi everybody,

I have a problem with the surface interpolation, in particular with the weighting factors of owner and neighbour cell values, for they are always scalars. But I need tensors as weighting factors for my interpolation of the velocity U.

OpenFOAM doesn't provide such an option or am I missing something? I'm quite new here.

Does someone already use such a scheme? And how can I use these weighting tensors for constructing the fvVectorMatrix of the convection term?

By the way, I'm working with OpenFOAM-1.3.
Thanks for your help.

Jutta

[hjasak]

As I said in the other postings, you have to be very very careful with this sort of thing:

- is the thing you are interpolating actually a vector? If so, having different interpolation factors for each component would violate Galilean invariance. Think about rotating your reference coordinate system by 30 deg and doing the same thing again: will you get the same result?

- do you wish to make ANYTHING related to your interpolation implicit? If so, this implies a block-coupled matrix and this is something I am still working on.

I assume you are on the wrong track if what you do has any relation with fluid flows, heat transfer or any other "standard" physics. If not, drop me an E-mail and let's talk.

Hrv