dolfin team mailing list archive

[Kristian Oelgaard <k.b.oelgaard@xxxxxxxxxx>]

Quoting Peter Brune <prbrune@xxxxxxxxx>:

> As a quick solution, you could, knowing the coordinates of a node, constrain
> everything within some epsilon smaller than the mesh feature size of the
> node (such as DOLFIN_EPS).

Yes, and then make sure to use the "pointwise" method when creating the
DirichletBC. See comment in dolfin/fem/DirichletBC.h.
You can also look at demo/pde/dg/advection-diffusion/main.cpp for an example on
how to use the "geometric" approach. The majority of DOLFIN demos uses the
"topological" approach which is the default.

Kristian

> - Peter
>
> On Sat, Aug 8, 2009 at 11:27 PM, Catherine Micek <mice0012@xxxxxxx> wrote:
>
> > Hi,
> > I'm working with a group that's using Fenics to study the equations of
> > linear elasticity (with a pure displacement formulation), and I have a
> > question about boundary conditions.  Our goal is to study the pure traction
> > problem, and this means we have to be careful in formulating the boundary
> > conditions.  In order to prevent a singularity in our equations, we need to
> > constrain certain nodes so as to prevent translations and rigid rotations.
> >  There are many examples in the fenics demos about how to prescribe mixed
> > conditions, but I have only seen examples where the Dirichlet condition
> > occurs along an edge (as opposed to occurring at a few individual nodes).
> >  So my question is, given a mesh, how does one prescribe Dirichlet
> > conditions for individual nodes?  Are there any demos that address this?
> >
> > Any help would be greatly appreciated -- thanks!
> >
> > Katy
> >
> > ---
> > Catherine (Katy) A. Micek
> > Graduate Assistant, U of MN Mathematics Department
> > http://www.math.umn.edu/~mice0012 <http://www.math.umn.edu/%7Emice0012>
> >
> >
> >
> >
> >
> > _______________________________________________
> > DOLFIN-dev mailing list
> > DOLFIN-dev@xxxxxxxxxx
> > http://www.fenics.org/mailman/listinfo/dolfin-dev
> >
> >
>