Author |
Message |
Joe Pierce (jpierce)
New member Username: jpierce
Post Number: 1 Registered: 04-2008
| Posted on Monday, April 14, 2008 - 05:00 pm: | |
I have a stress problem that has a symmetry plane that is a 45deg angle to the x axis. I think the boundary condition should be value(Up)=Vp and value(Vp)=Up. This does not work with FlexPDE. How do you create a BC where Up=Vp in FlexPDE? Thanks
|
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 1099 Registered: 06-2003
| Posted on Monday, April 14, 2008 - 11:19 pm: | |
value(Up)=Vp and value(Vp)=Up do not impose two boundary conditions. It is simply one condition stated twice. This creates two identical rows in the matrix and makes the system singular. I do not know what the proper combination of boundary conditions is on the diagonal boundary. I tried using Natural(Up)=normal(Sx,Txy), but this is a tautological equation that adds no new information to the system. The solution is similar to the Quarter_Sym problem, but not exactly the same. You need some condition that actually imposes the same kind of constraint that the upper leg frozen at the left imposes. I assume this will be conditions on the normal and tangential stresses. The Quarter_Sym system solves very quickly, so I'm not sure what the motivation is for imposing the diagonal symmetry plane.
|
Joe Pierce (jpierce)
New member Username: jpierce
Post Number: 2 Registered: 04-2008
| Posted on Tuesday, April 15, 2008 - 08:59 am: | |
Thanks for the fast response. The models I posted are very simple examples of the diagonal symmetry plane. The real problem is much more complicated and has small relevant features that slow the solution down quite a bit. Will the addition of a 3rd variable like w=Up-Vp with value(w)=0 on the diagonal sym plane? I tried this and the solution diverged. I may have applied it wrong. |