define test cases for release 3.0.alpha2

[tlmquintino]

[barche]

For UFEM, I'd propose a 3D backward-facing step. http://onlinelibrary.wiley.com/doi/10.1002/(SICI)1097-0363(19970615)24:11%3C1159::AID-FLD534%3E3.0.CO;2-R looks promising for comparison, though I've only seen the abstract. The advantage of this case is that it can be meshed using the blockmesher and doesn't need any periodicity or turbulence models.

[barche]

And maybe a square cylinder for an unsteady sample http://www.sciencedirect.com/science/article/pii/S0142727X99000818 but again I need to visit the library :)

[tlmquintino]

I can give you some input on the square cylinder. My wife Anna finished a thesis on the subject.

[tlmquintino]

I think the acoustics entropy pulse that is already committed could be a very good test case for both SFDM and RDM