[Getdp] Thermal coupling A-V formulation inside coils

[Gilles Vogt]

Hi,

I am using getdp to solve a 2D magneto-thermic problem based on the wiki
example (http://onelab.info/wiki/Magnetodynamics_with_cohomology_conditions
).

At the moment, I am using the A-V formulation, which is in my opinion
easier than the T-Ω formulation. However, I have some trouble about the
heating power inside my coil domain.

In post-pro, my evaluation of the ohmic losses is right in every domain :
sigma[]*((Re[{ur}]-Im[w*{a}])*(Re[{ur}]-Im[w*{a}]) +
(Im[{ur}]+Re[w*{a}])*(Im[{ur}]+Re[w*{a}]))
where w=2*pi*f, {ur} the electric potential and {a} the magnetic vector
potential
(which is the same as sigma[]*SquNorm[Dt[{a}]+{ur}]).


However, the following formulation does not produce good results when the
coil is included in Omega_c2 :
      Galerkin { [ -sigma[]*((Re[{ur}]-Im[w*{a}])*(Re[{ur}]-Im[w*{a}]) +
(Im[{ur}]+Re[w*{a}])*(Im[{ur}]+Re[w*{a}]))  , {t} ];
      In Omega_c2; Integration CurlCurl; Jacobian Vol;  }
For some reason, the temperature increases far too much in the coil even if
the ohmics losses are still OK when post-processed.

I have been able to work around this behavior with :
Galerkin { [ -qVol[]  , {t} ];
      In ***COIL DOMAIN***; Integration CurlCurl; Jacobian Vol;  }
Galerkin { [ -sigma[]*((Re[{ur}]-Im[w*{a}])*(Re[{ur}]-Im[w*{a}]) +
(Im[{ur}]+Re[w*{a}])*(Im[{ur}]+Re[w*{a}]))  , {t} ];
      In ***OTHER DOMAINS***; Integration CurlCurl; Jacobian Vol;  }
where qVol=(I/section)^2/sigma...

My results seem right, but I am a bit vexed because I cannot understand
where is my mistake... Can anyone pinpoint my error ?

PS : obviously
Galerkin { [ -1/sigma[]* (Re[{js}]*Re[{js}]+Im[{js}]*Im[{js}]), {t} ];
In ***COIL DOMAIN***; Integration CurlCurl; Jacobian Vol;  }
doesn't work either...

Best regards,
--
Gilles VOGT
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/getdp/attachments/20150317/7486fb7c/attachment.html>