Hey guys, I really can't figure it out what's going on with that program. I've printed the matrix but actually this doesn't helped me at all. I still think the problem is most likely to be in the assembling cause I've not completely understand how the shape functions are handled and I've used the same test shape functions for both the "u" and the "v" variables as they are modelled with equal order and FE type but I would not be able to actually use different models for "u" and "v" because I would not know how to treat the indexes. If someone catch an incongruence between the weak formulation found in the file attached (equation 9) and the assembling elasticity function let me know!
Thanks, Rodrigo Broggi 2014-08-20 23:23 GMT+02:00 John Peterson <[email protected]>: > > > > On Wed, Aug 20, 2014 at 2:39 PM, Rodrigo Broggi <[email protected]> > wrote: > >> >> Articolo.pdf >> <https://docs.google.com/file/d/0B49xyYU5aZowazV5TmNyeS1GYXc/edit?usp=drive_web> >> Thanks for the attention guys. Here it goes the file. About the program, >> I don't get any errors and actually there is also a "std::cout" that >> garantees that the PressurexR works fine. It's really an odd result, I can >> even postprocess on Paraview but the solution is just constant and zero. >> I've also changed the elastic constants to see if there was a problem with >> magnitudes but still got the same problem. >> > > Are the signs on the two boundary terms in (9) really supposed to be the > same? > > For an annulus, I'd expect the outward normal on the inner surface to be > in the -x direction, the outward normal on the outer surface to be in the > +x direction. > > -- > John > > ------------------------------------------------------------------------------ Slashdot TV. Video for Nerds. Stuff that matters. http://tv.slashdot.org/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
