On Mon, 17 Sep 2012, John Peterson wrote: > In geometric multigrid, anyway, I think the constrained DOF can get > coarsened out in the V-cycle, and then you are indefinite again?
Depends how you implement the constraint. "Loop over all nodes and then constrain when you hit the pinned one" is definitely playing with fire. What we do in fem_system_ex1 is "Loop over all elements and then constrain when one contains the pinned point", which ought to be safe as long as the coarsest grid doesn't decimate the geometry enough to put that point outside the domain. But likewise I don't know what to "expect" from AMG on either this case or the Lagrange multiplier case. Maybe having a non-positive-definite matrix is the trouble, irrespective of whether the matrix is actually indefinite or not? --- Roy ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
