‘-clface’ seems to be abnormal in some cases

[smallrain1996]

Hello, I use -clface to control the mesh size of the specified surface, but it seems that in some cases the mesh size of some areas has not been adjusted (y==0 everything is normal, but there are some problems with y==10).
Version 4.2.1-14
Thank you very much.

neper -T -n 200 -domain "cube(10,10,10)" -statcell size,diameq,group,sphericity -o testy0
neper -M testy0.tess -nset faces -faset none \
-clface "(y==0)?0.3:1" \
-format inp,msh -o testy0
neper -V testy0.msh -print testy0 

neper -T -n 200 -domain "cube(10,10,10)" -statcell size,diameq,group,sphericity -o testy10
neper -M testy10.tess -nset faces -faset none \
-clface "(y==10)?0.3:1" \
-format inp,msh -o testy10
neper -V testy10.msh -cameracoo 40:60:50 -print testy10

[rquey]

You are dealing with real numbers, so, use a tolerance: -clface "(y>9.99)?0.3:1" instead of -clface "(y==10)?0.3:1".

I would also recommend to use -reg 1 in -T (or at least / instead -pl 100 in -M) to avoid the local mesh refinements around small edges.

neper -T -n 200 -domain "cube(10,10,10)" -statcell size,diameq,group,sphericity -o testy10 -reg 1
neper -M testy10.tess -nset faces -faset none \
-clface "(y>9.99)?0.3:1" \
-format inp,msh -o testy10
neper -V testy10.msh -cameracoo 40:60:50 -print testy10

[smallrain1996]

Thank you very much for your reply, which not only solved my problem, but also gave me a very important suggestion.