[Getdp] Magneto Dynamic

[Christophe Geuzaine]

Alejandro Fernández Sela wrote:
> Dear Christophe:
> 
> I am trying to understand the MagnetoDynamic 3D problem (wiki page) 
> using the h-phi formulation, and I have some questions about it:
>   I don't know why does the program need to call the MagSta formulation, 
> Is it possible to build the code in only one formulation (one file) ?

In this particular formulation, you need a way to compute a "compatible" 
source field. That's why the MagSta formulation is used. Look at the 
reference mentioned in the .pro file for more info.

You can put all the formulations in a single file if you prefer, of course.

>   Does the MagSta give the initial conditions, or is it used for all the 
> resolution process?

It is used to compute the source field.

>   Why do you  use  "Integral" instead of "Galerkin"(MagSta)?

These are actually synonyms. I've changed "Integral" into "Galerkin" in 
the wiki.


>   When I launch the program, using "getdp CoilAboveACore -solve 
> MagDyn_hphi -pos Map_b Section_b Map_j Section_j" the convergence is not 
> very fast. Could it be a problem of my "Solver" file? (I use the 
> standard values).

You might want to increase the number of non-zero elements in the 
incomplete ILU by increasing "Nb_Fill" in solver.par.

If you want to use a direct solver instead of an iterative algorithm, 
you should download the source version and compile it with PETSc support 
(./configure --enable-petsc).

>   Is it really necesary to use "cuts" of the regions (done by the gmsh 
> file) in order to solve that kind of problems (eddy currents)?

With this particular formulation: yes.


Christophe



>   Thank you very much in advance, and thank you for developing that 
> fantastic tool! I am really enjoying my work with it, and I am learning 
> a lot about electromagnetism and FEA.
> 
> Best regards!
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> getdp mailing list
> getdp at geuz.org
> http://www.geuz.org/mailman/listinfo/getdp


-- 
Christophe Geuzaine
Assistant Professor, Case University, Mathematics
http://www.case.edu/artsci/math/geuzaine