dolfin team mailing list archive

[Anders Logg <question102501@xxxxxxxxxxxxxxxxxxxxx>]

Question #102501 on DOLFIN changed:
https://answers.launchpad.net/dolfin/+question/102501

    Status: Open => Answered

Anders Logg proposed the following answer:
On Fri, Feb 26, 2010 at 11:45:10AM -0000, Chris Richardson wrote:
> New question #102501 on DOLFIN:
> https://answers.launchpad.net/dolfin/+question/102501
>
> What is the best way to specify coefficient fields, e.g. conductivity, permeability, elastic constants ?
> I have seen the example "tensor-weighted-poisson" which seems very convoluted.
>
> I am interested in solving Darcy flow with variable permeability k (usually a scalar, but may be a tensor with diagonal entries)...
> Also, elastic problems with anisotropic material whose properties vary through space.
> There are 9 independent coefficients for elasticity, (out of 81 entries in Cijkl, which relates stress s_ij and strain e_kl).

It depends on where your data for the field comes from.

The reason for the tensor-weighted-poisson demo being a bit complex is
that in addition to demonstrating tensor-valued coefficients, it also
demonstrates

1. Use of C++ code from Python (for efficiency)
2. Reading cell-based field data from file

--
Anders

-- 
You received this question notification because you are a member of
DOLFIN Team, which is an answer contact for DOLFIN.