From: Derek G. <fri...@gm...> - 2008-06-02 15:31:24
|
On Jun 2, 2008, at 8:15 AM, John Peterson wrote: > Hey Derek, > > I'm sure you already have it, but I just wanted to say: be sure to > multiply by the test function (in Libmesh "phi_face[i]" or > "phi_side[i]"). I just remembered we had forgotten to do this once > before in the Laplace-Young application :-) Hehe... yes... thanks for the reminder. I had already made sure that was the case here! Didn't want to make that mistake twice.... > If "u-BC" differs appreciably from zero, the residual > contributions from this term will be large. We typically use a value > on the order of 10^10. What value are you using? Right now I've settled on 10^5... but I might revisit that. > I don't *think* the penalty should affect the convergence rate > strongly. But I don't have a lot of experience with the matrix-free > Newton-Krylov techniques so I can't say for sure. It's weird... it's not really affecting the _rate_ as much as it's actually affecting the convergence itself... ie the problem isn't fully converged when SNES says it is. Maybe I just need to play with the convergence tolerances some more... Thanks for your help! As it is now... the code is working, I just have an unsatisfactory constant in there (1e5)... Derek |