On Apr 29, 2013, at 5:45 PM, Frederic Durodie <[email protected]> 
wrote:

> In a simple 2D heat conduction problem I try to impose a power flux on
> part of the outer boundary and a fixed temperature on the remaining
> part.
> 
> However the resulting temperature distribution is not what I expect and
> furthermore depends on how the problem geometry is rotated (alphadeg : I
> did this to try and understand what is happening).
> 
> To impose the powerflux I use the method described in
> http://matforge.org/fipy/wiki/BoundaryConditions.

That wiki page is a bit misleading, as it says that you are specifying 
$\hat{n}\cdot\nabla\phi$ (a scalar), when in fact you need to specify 
$\nabla\phi$ (a vector).

The official documentation:

http://www.ctcms.nist.gov/fipy/documentation/USAGE.html#applying-fixed-gradient-boundary-conditions-neumann

is more accurate, in that it makes it clear that the constraint must be rank-1, 
but still not as helpful as it could be. I believe what you want is

T.faceGrad.constrain(PowerFlux/K * mesh._faceNormals, 
where=mesh.physicalFaces["semi circle"])


_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
  [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]

Reply via email to