hi,
Can fenics deal with the multi-point constraints? Any hint is appreciated. Thanks!
Can you describe in more detail what you want to do?
Thank you, Kent-Andre.
The simplest one is to enforce the displacements at two nodes to be identical during the deformation process. For instance, an initially vertical edge remains vertical (i.e., the horizontal displacements along that vertical edge being equal).
Are you constraining points on the boundary or in the interior of the domain. The former can be done by DirichletbBC, the latter requires some work on the user side.
DirichletbBC
One option is to use the "R" space and a Lagrange multiplier approach. This would be similar to the "neumann-poisson" demo. But you would need to localize the "R" functions (which are global). One way of localizing them is to multiply them with functions that are zero away from the points you are interested in.
hi, MireK,
would you please give more details? I am going to constrain some nodes on the boundaries, with theirs displacements coincident but unknown (to be solved).
I know that in numerical implementation of FEM, the linear constraint can be done by modifying the stiffness and load vector. But for fenics, this might be rather cumbersome.
Thanks.
Is there anyone who can help?