Hi Franz, I can indeed reproduce the issue on Windows (the Mac and Linux versions work fine). We have seen problems like this (accuracy problems leading to meshing failure with large element size gradients) for a while on Windows, but have not yet found a fix. They might be related to the behavior of the robust geometrical predicates, but we haven't been able to pinpoint the issue.
Christophe > On 12 May 2017, at 10:16, Bormann, F. <[email protected]> wrote: > > Hi, > I run currently into problems when I want to mesh my model. > > I have two large equally sized connected 2D domains of total size LxH. Due to > my problem I need a high resolution at following points: > > 1. Small area around Joint line of both domains between a<=x<=L with > a>0 > 2. Small circle around point at x=L and y=H/2 > 3. Small area around line x=L and H/2-b<y<H/2+b, where b is small in > comparison with H > > For the refinements I use the Attractor and Threshold fields with a combined > Field = Min. The minimum mesh size is 0.125 for refinement 2 and 0.25 for > refinements 1 and 3. > > I have run the meshing procedure with 5 variations. 4 of those failed. The > only parameters I varied are the: Length, height, maximum element size and > (for the successful simulation) as well the parameter a. The model size > varies between 4000x1600 and 6000x2000. And the maximum element size between > 50 and 100. > > Depending on the different variations I get different failure modes. > 1. Warning: Skipping bad triangle; Error: Unable to recover an edge. – > only one of the domains is meshed > 2. GMSH crashes > 3. For Meshing of both domains – Error: Identical points in > triangulation; Error: Unable to recover an edge. Increasing Mesh.RandomFactor > does not help > 4. Error: Unable to recover an edge. – only one of the domains is meshed > > The .geo files are attached with the suffix corresponding to the failure > mode. Suffix 5 corresponds to the working mesh generation. > > Here I used GMSH version 2.12.0 and 3.0.1. The operating system is Windows 7, > 64-bit. > > What is the reason behind the failure and is there a way to remedy the > problem? It would really help me a lot. > > Thanks already in advance. > > Cheers, > Franz > > <Bug_Meshing_4.geo><Bug_Meshing_1.geo><Bug_Meshing_2.geo><Bug_Meshing_3.geo><Bug_Meshing_5.geo>_______________________________________________ > gmsh mailing list > [email protected] > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info _______________________________________________ gmsh mailing list [email protected] http://onelab.info/mailman/listinfo/gmsh
