Shtyrkn/b2ag#23
Conversation
hrdoktorx
left a comment
hrdoktorx
left a comment
There was a problem hiding this comment.
The change in b2us_prep is acceptable. The other is dangerous. Instead consider searching for the place where the physical discontinuity is, i.e. where the cells are no longer neighbours of each other, or where the guard cells are, and use that to determine nxtl and nxtr.
| nxtr=nxtr+1 | ||
| enddo | ||
| endif | ||
|
|
There was a problem hiding this comment.
I strongly dislike this hack. If there is a fundamental problem that some faces are mislabeled in b2fgmtry, then there should be hard error and the user is meant to fix the problem. Maybe then provide information that also for a manual fix inside the b2fgmtry file.
There was a problem hiding this comment.
The problem is that it is impossible to create b2fgmtry file without this fix - the example is attached here.
Maybe the user should manually edit *.geo file, however, Nikita haven't found any errors in *.geo, and suggested the fix to b2ag instead.
There was a problem hiding this comment.
As I said in the general review. You need to propose a better algorithm that this one that will give the wrong values most of the time. b2ag knows the vertex coordinates. Find where they are no longer continuous and assign nxtr and nxtl to that gap.
3 participants
Nikita and myself have occasionally found that in some EAST DDN meshes the cuts near the inactive X-point are badly determined, which prevents from a successful mesh generation in b2ag: the error message looked like 'iCv.gt.nCv'.
Here is the Nikita's suggestion to fix it.