Thanks Daniel, that is what I wanted. FiPy is really impressive piece of
software.
Respectfully,
Dariusz Kedzierski
> Hi Dariusz, The way to handle this is as follows. Say you have an
> internal boundary condition that says
>
> D \frac{\partial \phi}{\partial z} = \lambda at z = h
>
> then you create a mask at the boundary, say 1 on the boundary and 0
> elsewhere
>
>
> X, Y, Z = mesh.getFaceCenters()
> mask = FaceVariable(mesh=mesh, value=False)
> mask[Z==h] = True
>
> Say the equation is "TransientTerm() == DiffusionTerm(D)" then do this
> instead
>
> eqn = TransientTerm() == DiffusionTerm(D * ~mask) + (lambda * mask
> * (0,1)).getDivergence()
>
> That should apply an internal boundary along the z faces at height h.
> Something like this is what you need. Cheers.
>
> 2010/5/17 Dariusz Kędzierski <[email protected]>:
>>
>> Thanks for fast reply,
>>
>> Do you have more examples of this procedure? Except point 5.5.3 from
the
>> manual? It would be great to have more examples.
>>
>> Thank you in advance.
>>
>> Respectfully,
>> Dariusz Kedzierski
>>
>>>
>>> Hi Dariusz, Thanks for your interest in FiPy. I would say that, in
>>> general, one can incorporate these boundary conditions either as
>>> changes in coefficient across material boundaries or as source terms
>>> localized to cells close to the boundaries or a combination of the
>>> two. You must deal with it in this way as FiPy doesn't have explicit
>>> internal boundary conditions. Hope this helps.
>>>
>>> 2010/5/17 Dariusz Kędzierski<[email protected]>:
>>>>
>>>> Dear fipy users and developers,
>>>>
>>>> I'm beginning my adventure with computer modelling in FiPy (which is
a
>>>> great
>>>> software), and I have some questions considering a modelling of the
>>>> superlattice.
>>>>
>>>> I have a paraelectric/ferroelectric (PE/FE) superlattice, consist of
>>>> ferroelectric and paraelectric layers. The aim of the simulation is
to
>>>> calculate the polarization and electrostatic potential distribution
>>>> versus
>>>> temperature.
>>>>
>>>> Because I have different sets of equations for P and \phi in PE and
FE
>>>> I've
>>>> decided to make two meshes, one with FE layer and second with PE
layer.
>>>>
>>>> The problem I have is to incorporate boundary conditions. I have a
>>>> boundary
>>>> conditions in a form (at the interface between FE and PE layers)
>>>> {latex}:
>>>> \begin{eqnarray}
>>>> \frac{\partial \phi^f}{\partial z} - \epsilon_p \frac{\partial
>>>> \phi^p}{\partial z}& =& 4 \pi P, \label{pefe1}\\
>>>> \phi^f& =& \phi^p, \label{pefe2}\\
>>>> \frac{\partial P}{\partial z}& =& \lambda P \label{pefe3}
>>>> \end{eqnarray}
>>>>
>>>> The question I would like to ask is: how to incorporate those
>>>> conditions
>>>> and
>>>> join these two meshes together.
>>>>
>>>> Any help would be greatly appreciated.
>>>>
>>>> Respectfully yours,
>>>> Dariusz Kedzierski
>>>>
>>>>
>>>
>>>
>>>
>>
>>
>>
