Re: Using the dw_diffusion_coupling term with inhomogenous materials

Hi Vladimir,

On 03/12/2013 09:47 PM, Vladimír Lukeš wrote:

Hi Ben,

I apologize, I was wrong, in dw_diffusion_coupling term a vector, not a matrix, must be used as the material parameter.

I see the problem in the material function, try this:

def diffusion_coupling_coefficient(ts, coors, mode=None, **kwargs): if mode == 'qp': return {'f': np.zeros((coors.shape[0], 2, 1))}

 else:
     return None

The values are returned only in 'quadrature points' mode.

Just a small note - the "return None" branch does not need to be there explicitly. Usually, if only 'qp' mode is used in a material function, I use the pattern:

if mode != 'qp': return

...

Also, Ben, the call modes are described in [1], as well as the syntax of other function hooks.

Cheers, r.

[1] http://sfepy.org/doc-devel/users_guide.html#functions