I had a look at the Stokes example. It uses model bricks to form the system. From the documentation, it seems using model bricks implicitly chooses the bilinear variational form. I would like to define my own bilinear form to form the matrix, in which I can define terms like div(q) + c u (q = vector, c = constant, u = another variable besides q). How would I go about this?
Thanks, -Jehanzeb On Sun, Jul 5, 2009 at 11:05 AM, Iago Barbeiro<[email protected]> wrote: > Dear Jehanzeb, > > I think you may start by looking the Stokes example. > It deals with two variables (û=uî+v^j and p) and has the term div(û).dp. > Bon courage! > > Iago > > On Sun, Jul 5, 2009 at 6:07 AM, Jehanzeb Hameed <[email protected]> > wrote: >> >> Hello, >> >> Is there a simple example in getfem where assembly is for two >> variables? E.g. we may want to solve for "u" and "q" (with an equation >> defining the relationship between u and q) . Such a case arises in >> mixed methods. Is there an example for mixed-poisson problem >> somewhere? I know mixed-elasticity problem is given with getfem, but I >> am not familiar with that particular problem. >> >> In particular, I am not sure how to refer to "q" and "u" in >> generic_assembly routines. Say my weak form involves div(q) . v ? How >> will I write this in "assem.set" routine? (I am guessing thats what I >> am supposed to do). >> >> Thanks, >> -Jehanzeb >> >> _______________________________________________ >> 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
