dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

VMTK, the Vascular Modeling Toolkit, can now export meshes directly to
DOLFIN XML format, which shortens the step from medical imaging data
to simulation considerably.

For more about VMTK, visit the VMTK web page:

  http://villacamozzi.marionegri.it/~luca/vmtk/

Thanks to Luca Antiga (author of VMTK) and Kent for implementing the
new file format in VMTK!

Since a few days, it is possible to attach "arbitrary" user-defined
data to meshes, handled by the class MeshData. This in combination
with a new optional <data> section in the mesh XML format enables
VMTK to communicate boundary indicators to DOLFIN.

(The possibility of dynamically attaching data to meshes may simplify
the implementation of other things, like for example storing data for
parallel communication without needing to modify the mesh class.)

Segments of the boundary are labeled 0, 1, 2 etc so defining a
boundary condition u = g on segment i is just a matter of writing

  bc = DirichletBC(g, mesh, i)

No need to handle separate MeshFunctions. Here's the Python code
for the newly added demo for illustration:

  # Create mesh and finite element
  mesh = Mesh("../../../../data/meshes/aneurysm.xml.gz")
  element = FiniteElement("Lagrange", "tetrahedron", 1)

  # Define variational problem
  v = TestFunction(element)
  u = TrialFunction(element)
  f = Function(element, mesh, 0.0)
  a = dot(grad(v), grad(u))*dx
  L = v*f*dx

  # Define boundary condition values
  u0 = Function(mesh, 0.0)
  u1 = Function(mesh, 1.0)
  u2 = Function(mesh, 2.0)
  u3 = Function(mesh, 3.0)

  # Define boundary conditions
  bc0 = DirichletBC(u0, mesh, 0)
  bc1 = DirichletBC(u1, mesh, 1)
  bc2 = DirichletBC(u2, mesh, 2)
  bc3 = DirichletBC(u3, mesh, 3)

  # Solve PDE and plot solution
  pde = LinearPDE(a, L, mesh, [bc0, bc1, bc2, bc3])
  u = pde.solve()
  plot(u, interactive=True)

-- 
Anders