Questions - Valparaíso - Chile

Alejandro Angulo aacusm at
Mon Jul 31 21:38:55 CEST 2000


I apologize you to make you so many questions, but my
interest to learn how to use GetDP it is enormous (I
want to use it as learning tool for courses of
Eletromagnetics Fields!!).  With regard to the
questions thats I sent you in a previous e-mail about
How to imlement conditions of Neumman type in the
boundary (not homegeneous) in Magneto-Thermal
formulations, I have not been very successful to the
moment to formulate them.  I made a test program (only
thermal!!) to implement this boundary conditions. 
Initially it only had convection conditions in the
boundary, but whith them the problem oscillated (it
didn't converge).  Then, to vetify some theories thet
I had, I forced the temperature (Drichlet condition)
in part of the boundary and there the program yes it
worked.  Apparently, the problem that is presented is
that the integral in the boundary is not considered
for GetDP, in spite of the fact that these terms
appear in the formulation ( I also saw the generated
matrix - Matrix_Printing = 1 en Solver - and they
didn't modify when I modified the convection
coefficients; even more, they stayed similar with or
without convection conditions in the boundary).
Please, I request you that  revise this test program
(attached) and tell me which their problems are.
With regard to my questions on connections of
conductors I found your information very good, but I
that I really need is to know how to use constraints
of 'NETWORK' type (like 6.5 in GetDP manual) to
analize SOLID conductors (not Stranded). I want to
locate hot points in transformer windings and in this
problems type the current distributions cannot suppose
uniform (eddy current are very important!!!).
With my sincere regards and thanks in advance for your


PD: I attach you the test files in ZIP format

Do You Yahoo!?
Obtenga su dirección de correo-e gratis