[Gmsh] Changing the angle threshold for recombining to degenerate quadrangles

jan . janneggers at outlook.com
Thu Jul 16 10:13:54 CEST 2015


Update 2:

I’ve read the blossom paper [1], which is amazing stuff, very nice work. So now I understand that the algorithm doesn’t really work like I imagined and 
I probably had a mesh which did not have any Q4 match (e.g. like the Tutte’s example in figure 6). I tried changing the swap angle:
Mesh.AllowSwapAngle = 90; // default = 10
But that didn’t really make a difference. Then I tried changing the T3 meshing algorithm and this worked for me. Apparently for my geometry 
algorithm 8 (delquad) leaves many triangles after recombination, while algorithm 5 (Delaunay) converts to 100% quads.

Anyway, I just wanted to put my experience here for future reference. 

It would have been even better if there was a bit in the documentation where Recombine Surface { expression-list } | "*" < = expression >; is explained,
detailing that it only applies to RecombinationAlgorithm = 0.

[1] J. Remacle, J. Lambrechts, Blossom‐Quad: A non‐uniform quadrilateral mesh generator using a minimum‐cost perfect‐matching algorithm, Int. J. …. (2010) 1–6.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20150716/3eb7d3e6/attachment.html>