> i have following problem. For implementation of moving interface by use
> of level set method i need to compute boundary integrals on the zero
> level of level set function. What i need is determination of quadrature
> points which are nearest to the zero level of level set function. I know
> values of level set function from previous time step in quadrature
> points. Is it possible to set this quadrature points as part of
> boundary, i.e. set their, something like, boundary indicator to some
> number and then during assemble procedure use standard fe_face_values
> procedures? Does there exist any such functionality or should i
> implement it by myself? Or generally, is it possible to set boundary
> indicators "dynamically", conditionaly during assembly, or is it
> possible only during the setup phase? Thank you

I think you'll have to implement something yourself. Boundary indicators 
can only be assigned to actual faces and they need to be on the boundary 
of the domain. You may be able to set either material_id or use the 
user_index/user_pointer if you want to first determine which cells you 
have to do extra work on, and do the actual work later.

As for FEFaceValues: this class is really meant to only do the integration 
on faces of cells, not on arbitrary manifolds of dimension dim-1. It won't 
help you in this case. What you probably have to do is determine 
quadrature points on your interface yourself, initialize a quadrature 
formula with these points, and then use a FEValues on this set of 
quadrature points. Since you do an integration over a dim-1 manifold, you 
have to adjust the quadrature weights as well when summing over quadrature 
points.

Does this give a first idea of what to do?

Best
 W.

-------------------------------------------------------------------------
Wolfgang Bangerth                email:            [email protected]
                                 www: http://www.math.tamu.edu/~bangerth/

_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii

Reply via email to