[Hoyskipoyski]

Hey,
due to faster mesh generation i am working on a script to let icem do all the meshing in batchmode on windows. I am not running the replay file in the icem gui. I am expierincing some troubles during my process with non manifold vertices, especially some very nasty tent like structures, which by manually smoothing the shells produce negative volumes afterwards.

First i load the tetin, set all the necessary attributes to my parts, then save the tetin and meshing parameters (saved to job.tin).
If i set the octree meshing parameters there is no attribute specifying the "fix non-manifold" checkbox (for fix holes there is one). So I am not sure if it's activated in the tetra built-in cutter command.
Therefore I would like to run the cutter seperatley because there you can specify it by controlling the fix_nonman argument (which doesn't work in the tetra command).
The Programmers guide says to load the tetin before running the cutter. When executing the following:

ic_run_tetra "job.tin" "tri2tet_temp0_uncut.uns" run_cutter 0 fix_holes 1 write_volume 1 n_tri_smoothing_steps 20 in_process 1 log tetra_cmd0.log n_processors $num_proc
ic_save_unstruct "tri2tet_temp0_uncut.uns"
ic_load_tetin "job.tin"
ic_run_cutter "job.tin" "tri2tet_temp0_uncut.uns" "tri2tet_temp0.uns" saveall 0 savefams surface fix_nonman 1
ic_smooth_elements upto 0.4 iterations 30 smooth TRI_3 metric {Aspect ratio} laplace 0 all_elements 1 refine 0 n_processors $num_proc
ic_unload_mesh
ic_run_tri2tet "tri2tet_temp0.uns" "tri2tet_mesh.uns" safety 1 use_tg_tri2tet 1 use_tg_tgrid_af 0 tetexpand 1.1 family FLUID bgmesh 0 errors tri2tet_errors verbose 0

I get the following error at the beginning of the cutter:

reading nodes...
reading surface model...
Usage: periodic {x y z nx ny nz angle}|{translational dx dy dz}
error executing command: periodic
error in job.tin
error reading surface model

It doesn't make any difference wheter i load the tetin before executing the cutter command or don't. I am meshing a turbine blade with fillets and circumferential periodic boundaries.
Does anyone have any experience or hints for me?
Do you know if the fix non manifold is inclued in the tetra cutter command?
I am running Ansys Version 15.0 and the cutter says it is 14.0 although i don't have version 14.0 installed.


Edit: Seems the "fix_nonman 1" argument for the tetra command works just fine. Don't know what went wrong there yesterday. I also increased the edge_criterion back to the 0.2 default.