[I sent this reply to Julien, but neglected to send it to list. The new list server defaults to reply-to-poster, which I prefer, but it's easy to forget after so long]
On Oct 19, 2011, at 12:46 PM, Julien Derr wrote: > the boundary conditions are defined by a source term (value c20 )in a > circular region : > and a null value on the outside, > > > X, Y = mesh.getFaceCenters() > XX = FaceVariable(mesh=mesh, value=X) > YY = FaceVariable(mesh=mesh, value=Y) > sourcec2 = ((XX-CX)**2+(YY-CY)**2 <= Rstart**2 ) Are these faces interior to the mesh? .constrain() is not supported for internal boundary conditions. > c2.constrain(c20, where=sourcec2) > BCsforc2 = FixedValue(faces=mesh.getExteriorFaces(), value=0) I don't recommend combining .constrain() and old-style `FixedValue` boundary conditions. Choose one or the other. ----------------------------------- On Oct 19, 2011, at 1:39 PM, Julien Derr replied: > Thanks Jonathan, > I put all the conditions via constrain, but the result is the same. OK, not surprising. As I said, I don't recommend mixing .constrain() and FixedValue(), but I don't know for a fact that they don't work together. > what do you mean by faces interior to the mesh? I have a 2D mesh, and I > wanted to constrain a full circle in the middle of the mesh, so yes in a way > it is not the outside boundary of the mesh... Yes, exactly. > what is the alternative ? I guess using an implicit plus an explicit source > directly in the equation? Yes, that's the recommended procedure. > I am still confused with the syntax fipy is using, where can I find > documentation about the syntax ? Syntax with respect to what? _______________________________________________ fipy mailing list [email protected] http://www.ctcms.nist.gov/fipy [NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy]
