Hi all
I'm trying to implement an adaptive strategy for a problem in non-linear solid
mechanics where internal variables are stored at the level of the quadrature
points. It's in the spirit of the extension of step-18 as proposed in the
documentation. The objective is to project data (scalar, vector, tensor)
stored at quadrature points between quadrature points on successively refined /
coarsened meshes.
The first snag I have hit is that I have vector valued shape function for the
displacement (in R^3) and the concentration (in R). The linear system is
furthermore split into blocks. The underlying shape functions are primitive
however.
I wish to use the function "
FETools::compute_projection_from_quadrature_points_matrix" to perform the first
step of projecting the quadrature point data on the old mesh to the nodes of
the old mesh. The snag is that "fe" has two components (displacement and
concentration resp.) and hence the method fails.
i.e. if i use
FETools::compute_projection_from_quadrature_points_matrix(fe,
quadrature_formula, quadrature_formula, projection_matrix);
i get the following error:
An error occurred in line <1924> of file
</Users/andrewmcbride/lib/real_deal_svn/deal.II/deal.II/source/fe/fe_tools.cc>
in function
static void
dealii::FETools::compute_projection_from_quadrature_points_matrix(const
dealii::FiniteElement<dim, spacedim>&, const dealii::Quadrature<dim>&, const
dealii::Quadrature<dim>&, dealii::FullMatrix<double>&) [with int dim = 3, int
spacedim = 3]
The violated condition was:
fe.n_components() == 1
The name and call sequence of the exception was:
ExcNotImplemented()
Additional Information:
(none)
Is there a simple way around this? Can I send a restriction of "fe" into the
function " FETools::compute_projection_from_quadrature_points_matrix"?
Has anyone else implemented similar functionality into deal.II? It seems to be
essential for adaptive problems in (non-linear) solid mechanics. If so, any
general suggestions would be appreciated. Perhaps we can formalise this
functionality into deal.ii if there is interest.
Thanks again
Andrew
_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
