hi everybody! thanks for the answer, but sorry for my bad questioning. i think the vector identity mentioned will work. unfortunatly i forgot to say that i also have terms like integral curl(A) to deal with. So my main question is, if there exists a function like vRot(#i), like the other comp commands, or way to deal directly with curls.
regards and thanks in advance, alex Richard George wrote: > I guess you get an authorative answer from someone like Yves Reynard > who wrote getfem, but > > could you express curl(curl(A)) in terms of available functions using > an identity such as > > curl(curl(A)) = grad(div(A)) - vector_laplacian(A) > > would that make sense / allow enough numerical accuracy? > > Richard George > > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > > > On Sun, 9 Sep 2007, Alex Magnes wrote: > >> Hi everybody! >> >> >> Is there any possibility to assemble a term like integral curl(curl A) >> where A is a vector field? I just found examples dealing with gradients >> and divergences. >> >> regards and thanks in advance, alex >> >> >> >> >> _______________________________________________ >> Getfem-users mailing list >> [email protected] >> https://mail.gna.org/listinfo/getfem-users >> _______________________________________________ Getfem-users mailing list [email protected] https://mail.gna.org/listinfo/getfem-users
