Hi Daniel, Thank you for your reply. You are right. Sorry I made a bad example.
Regards, Zhiwen On Mon, Nov 10, 2008 at 9:35 AM, Daniel Wheeler <[EMAIL PROTECTED]>wrote: > > On Fri, Nov 7, 2008 at 3:45 PM, Zhiwen Liang <[EMAIL PROTECTED]> wrote: > > Hi all, > > This might not be relevant, but I am working on boundary conditions that > not > > only depend on spatial positions but also a couple of variables. For > > example, > > mesh=... > > u=CellVariable(mesh=mesh,...) > > v=CellVariable(mesh=mesh,...) > > > bcs=FixedValue(faces=mesh.getFacesLeft(),value=u.getFaceGrad()*v.getFaceValue()) > > Do you think this will cause convergence difficulties? > > In general if you increase the interdependence and non-linearity then, > yes, I guess that systems have slower or worse convergence properties > mostly. As for specifics of this boundary condition, I don't really > know. > > FYI In the above you are passing in a vector (getFaceGrad()) field > into an argument that expects a scalar field. Maybe this works in 1D > but in general I don't think it makes sense. > > -- > Daniel Wheeler > >
