On Thu, 9 May 2013, Paul T. Bauman wrote: > What David said is correct (and how I currently deal with cylindrical > coordinates). Nevertheless, while there are no formal plans, I've thought > it would be nice to try and deal with alternative (to Cartesian) coordinate > systems at the libMesh level. E.g. JxW comes premultiplied by r, > curl/Laplacian/div/etc formulae have the right terms so that the same code > could be used regardless of coordinate system, etc. Alas, it hasn't been > high enough priority for me to spend any time thinking about it and > proposing how to do it, e.g. whether it should be an FE type, etc.
I looked into this too; it's been a long while, IIRC I ran into issues with covariant vs. contravariant vectors in some formulation that made it non-obvious how to enable coordinate-system-independent code. > That said, Peter, if you wanted to take a crack at it, I (and probably > others) would be happy to give guidance. Likewise. --- Roy ------------------------------------------------------------------------------ Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. This 200-page book is written by three acclaimed leaders in the field. The early access version is available now. Download your free book today! http://p.sf.net/sfu/neotech_d2d_may _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
