help with GetDP needed

[Christophe Geuzaine]

Samuel Kvasnica wrote:
> 
> This is a multi-part message in MIME format.
> --------------82D70A3304E631D20FC21837
> Content-Type: text/plain; charset=iso-8859-2
> Content-Transfer-Encoding: 7bit
> 
> > That's strange. Maybe a bug -> send me the .pro and .msh files.
> 
> o.k., I'm sending you .geo and .pro files. If you really want, I'll send you also .msh
> but it has 1.9MB gzipped. Hope you'll be able to generate the mesh with gmsh. I'm
> having strange segfault problems with it at higher node densities...
> 


OK, there are two little mistakes in the FunctionSpace. Here is the
correct version :

  // Magnetic vector potential a (b = curl a)
  { Name Hcurl_a_Mag_2D ; Type Form1P ;
    BasisFunction {
      // a = a  s
      //      e  e
      { Name se ; NameOfCoef ae ; Function BF_PerpendicularEdge ;
        Support Domain_Mag ; Entity NodesOf[ All ] ; }
        {Name se2; NameOfCoef ae2; Function HBF_PerpendicularEdge_2E;
        Support Domain_Mag ; Entity EdgesOf[ All ]; } // CHANGED
    }
    Constraint {
      { NameOfCoef ae ;
        Entity NodesOf UsingConstraint MagneticVectorPotential_2D ; }
      { NameOfCoef ae2 ;
        Entity EdgesOf UsingConstraint MagneticVectorPotential_2D ; } //
CHANGED
    }
  }


Your mistake was that you associated the HBF basis functions to the
nodes, not to the edges...

With these 2nd order functions, your mesh does not need to be so
refined. I used this in your .geo file, and the results were already
quite nice looking :

mult =  3;
/* 0.4,5,12,1 */
pCore = 0.5e-3 * mult;
/*pInt = 8e-3;*/
pInt=6e-3 * mult;
pExt = 10e-3 * mult ;
pAxis = 0.8e-3 * mult;


E-mail me if it works (or if it doesn't!)


PS : is your problem truly 2D? Or is it axisymmetric? In the latter,
your jacobian methods are not correct...

-- 
Christophe Geuzaine

Tel: +32-(0)4-366.37.10    mailto:Christophe.Geuzaine at ulg.ac.be
Fax: +32-(0)4-366.29.10    http://www.montefiore.ulg.ac.be/~geuzaine/