[overture] Re: interpolation coefficients for an implicit solver
- From: Bill Henshaw <henshaw@xxxxxxxx>
- To: overture@xxxxxxxxxxxxx
- Date: Wed, 28 May 2008 07:50:22 -0700
Hi Giorgos:
The interpolation coefficients are filled into a coefficient matrix in
the function getInterpolationCoefficients in GenericGridCollectionOperators.C
This is done when you call
coeff.finishBoundaryConditions();
You could try to change the getInterpolationCoefficients function. A safer
way would be to write a similar function to getInterpolationCoefficients
and call it after coeff.finishBoundaryConditions() (so that you do not
break the rest of Overture).
Regards,
Bill.
> Greetings,
>
> I am using the oges solver to solve for a two component grid function. For the
> type of scheme I am using the interpolation coefficients I am getting in the
> coefficient matrix are not compatible with this scheme, in the sense that they
> use interpolee points other than the ones I would like to use and they lead to
> wrong results. (my grids consist of orthogonal grids of different levels of
> refinement) At which part the code sets the interpolation coefficients in the
> coefficient matrix so that I can see how it picks the interpolee and
> interpolation points? If I want to customize the interpolation coefficients,
> you've got a hint of how should I go about in order to zero out the default
> interpolation coefficients at the coefficient matrix and replace them with
> alternative ones at the proper coefficient matrix locations?
>
> Thanks,
>
> Giorgos
Other related posts:
