Re: Octave PDE initial value problem solvers

[Carlo de Falco]

On 28/ott/05, at 06:31, <address@hidden> <address@hidden> wrote:

Hello,

I need to put together a model that describes how moisture and salts

move through a layer of unsaturated mud as it dries (by evaporation off

the top, and by drainage out the bottom).  I'll start by trying to get a

moisture transport model working, then will worry about coupling the

salt equation.  At some point I'll also need to concern myself with

volume changes as the mud dries out.  I'm going to start by trying to

numerically solve the "Richards Equation", which is a nonlinear

parabolic equation.  The usual way to solve this seems to be to use a

Galerkin-type finite element scheme - something I need to read up on.

Anyway, I'm wondering if anybody knows of publicly available Octave code

that might be useful to me?  I'll only be working on a 1D model.

As somebody who used to use Matlab (though in signal processing and

control, not for solving equations) at uni, I've decided that using

Octave should get me started faster then trying to use C++ or Delphi...

Thanks,

Neil

Neil,

You might find it usefull to have a look at my scripts for solving

advection/diffusion/reaction PDEs in 2D at

http://www1.mate.polimi.it/CN/MNME/MFV.zip

and 

http://www1.mate.polimi.it/CN/MNME/FVSG.zip

They use two different FEM discretization schemes and 

are used for a lab course on finite element methods for 

microelectrinic applications.

They are written for matlab but I tested them under

octave and everything (except visualization) works ok.

good luck,

carlo

--

Carlo de Falco

PhD student,

MaSSC, Dipartimento di Matematica "F. Enriques"

Universita' degli Studi di Milano

via Saldini, 50 - 20133 MILANO ITALY

phone: +39 02503.16166

or

MOX - Modeling and Scientific Computing

Politecnico di Milano,

via Bonardi,32 - 20133 MILANO ITALY

email: carlo <DOT> defalco <AT> mate <DOT> polimi <DOT> it

phone: +39 02 2399 4604