Gmsh GetDP question

[Guenter Milde]

On Tue, 06 Mar 2001 17:12:28 +0100 Christophe Geuzaine <Christophe.Geuzaine at ulg.ac.be> wrote:

> Guenter Milde wrote:
> > 
> > For the project I am currently working on (simulation of a pyroelectric IR
> > sensor) I look for a FEM program that can solve a harmonic temperature
> > problem (1st. step) and a coupled pyroelectic/piezoelectric problem (with
> > the temperature field of the first step as input). I wonder whether
> > Gmsh/GetDP could solve this kind of problems or whether it would be possible
> > to extend the program to do so (relatively) easily.
> 
> Hello,
> 
> GetDP is still largely a "beta" software, but it may be of some interest
> for the problems you describe (Gmsh is quite multipurpose, so there is
> no real applicability problem). But what do you exactly mean by a
> _harmonic_ temperature problem? Is it nonlinear, but treated in the
> frequency domain? Or is it linear? Is it possible/suitable to consider a
> time domain analysis followed by some Fourier transform?

The problem is shortly described as follows:

* incident radiation is absorbed and causes temperature change

* heat conduction results in a typical temperature distribution in the sensor

* pyroelectric effect leads to change in spontaneous polarization

* change in polarisation leads to change of charge on electrodes 

  pyro-current or pyro-voltage is proportional to mean temperature in active
  volume

* surface leakage -> only temperature/radiation changes are detected 

-> for measurement of a steady IR-radiation flow: choppering of incident
   radiation. Phi -> Phi_0 * theta(t) (modulation function)

-> We are only interested in the non-steady part: -> Fourier decomposition ->
   Approximation: Phi -> Phi_0 * cos(omega*t)

With this, time dependence can be eliminated by using complex values for Phi
and T. (i.e. analysis in the frequency domain for a set frequency dependence)
This is what ANSYS calls harmonic anlaysis but only offers for
mechanical and electrical problems, not for temperature field analysis.

If you are interested, I could send you an postscript file (550 KB) of a
speech about the project.

> >    INTERNET FINITE ELEMENT RESOURCES:
> >    www.engr.usask.ca/~macphed/finite/fe_resources/fe_resources.html, and

Sorry, IFER has it already (I looked at the wrong place ("with source code")).


sincerely

Guenter

--
G.Milde at physik.tu-dresden.de