|
From: Moore, S. <st...@sa...> - 2021-10-04 18:37:11
|
You can try the changes in https://github.com/sparta/sparta/pull/244, there is a high likelihood the issue will go away with that branch, but two (obscure) regression tests are still failing with those change, which is why that branch hasn't been merged yet. These other issues on Github are probably all related: https://github.com/sparta/sparta/issues/218 https://github.com/sparta/sparta/issues/203 https://github.com/sparta/sparta/issues/137 https://github.com/sparta/sparta/issues/7 HTH, Stan ________________________________ From: Natalia Hladczuk <N.A...@tu...> Sent: Monday, October 4, 2021 12:10 PM To: spa...@li... <spa...@li...> Subject: [EXTERNAL] [sparta-users] SPARTA issue Dear all, I am struggling with the exact same error as Zuheyr (see the post below): Created orthogonal box = (-3 -3 -3) to (3 3 3) Created 1000 child grid cells parent cells = 1 CPU time = 0.00268237 secs create/ghost percent = 59.656 40.344 Reading surface file ... 11689 points 23402 triangles -1.5996 1.56995 xlo xhi -0.972 0.972 ylo yhi -0.489699 0.993697 zlo zhi 1e-06 min triangle edge length 1.065e-10 min triangle area 50 = cells with surfs 26245 = total surfs in all grid cells 2473 = max surfs in one grid cell 1.66667e-06 = min surf-size/cell-size ratio 0 0 = number of pushed cells 50 0 = cells overlapping surfs, overlap cells with unmarked corner pts ERROR: Surface check failed with 3 points on triangles (../surf.cpp:1559) Interestingly, while changing the grid resolution to much denser, the error changes to the following one: Created orthogonal box = (-3 -3 -3) to (3 3 3) Created 1000000 child grid cells (…) Cut3d failed in cell ID: 534956 parent 0 level 0: NxNyNz 100 100 100: child 534956 55 49 53 lo corner 0.3 -0.06 0.18 hi corner 0.36 0 0.24 # of surfs = 150 out of 23402 surfs: (<surf listed>) ERROR on proc 15: LP: More than one positive area with a negative area (../cut3d.cpp:658) Could you provide me some more information on what exactly those errors mean and how to tackle this problem? Many thanks in advance, Natalia |