[ICEM] parallel mesher error

beguxa · January 28, 2011, 11:44

I had cleaned a geometry in ICEM and ran octree mesher (global scaling factor =1 and max global size = 64) the mesh that was created was ~ 35 million tet cells.

But for the same geometry and without changing any other meshing parameters, when I run the same octree mesher in ICEM, it shows that mesh count is ~180 Million. I only changed the no of processors while I created that mesh. Has any one faced this issue before? How can it corrupt the geometry? How to trouble shoot it?

Thanks in advance.

PSYMN · January 29, 2011, 21:30

Are you sure you don't have some other sizes set or something (maybe curvature and proximity based refinement?)... If not, then this is a parallelizing bug we have not seen before. You should expect very small differences when the mesh generation is parallelized, but nothing like this.

Can you email the tetin file and description of the problem to techsupp@ansys.com? You can tell them your are reporting a defect...

Best regards,

Simon

beguxa · February 1, 2011, 02:56

thanks for the reply.
1) I am not setting any other sizes. Infact the way I faced this problem is following:
when the first mesh I created (~25 million cells) was not saved in absence of a disk space, the mesher reported error.
After I fired the meshing run second time (with required disk space and, I did not close ICEM) the same geometry gave me a mesh of ~ 194 million (80% element sizes were ~ 7 to 15 mm where as my global max size was 64 and scaling factor of 1).

2) Moreover, I completely discarded the geometry and reconstructed it from 'mesh to facets' options. This is also giving me a very large mesh now (similar size distribution).

3) I need to tune the scaling factor (> 1) to get appropriate mesh. but this kinda of bug is quite unsettling. I am not sure if I will be able to give this geometry to Ansys due to proprietery issues. In order to start afresh, what are the files I need to discard if their corruption is causing this issue? I am no longer using the same geometry file, and created a new geometry file from mesh.

Thanks for the patience.
-aditya

January 28, 2011, 11:44	parallel mesher error	#1
beguxa New Member Annon Join Date: Apr 2010 Posts: 18 Rep Power: 16	I had cleaned a geometry in ICEM and ran octree mesher (global scaling factor =1 and max global size = 64) the mesh that was created was ~ 35 million tet cells. But for the same geometry and without changing any other meshing parameters, when I run the same octree mesher in ICEM, it shows that mesh count is ~180 Million. I only changed the no of processors while I created that mesh. Has any one faced this issue before? How can it corrupt the geometry? How to trouble shoot it? Thanks in advance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Compile calcMassFlowC	aurore	OpenFOAM Programming & Development	13	March 23, 2018 08:43
Accessing phi from a fvPatchField at same patch	johndeas	OpenFOAM	1	September 13, 2010 21:23
Installation OF1.5-dev	ttdtud	OpenFOAM Installation	46	May 5, 2009 03:32
How to get the max value of the whole field	waynezw0618	OpenFOAM Running, Solving & CFD	4	June 17, 2008 06:07
Compiling problems with hello worldC	fw407	OpenFOAM Installation	21	January 6, 2008 18:38

January 29, 2011, 21:30		#2
PSYMN Senior Member Simon Pereira Join Date: Mar 2009 Location: Ann Arbor, MI Posts: 2,663 Blog Entries: 1 Rep Power: 47	Are you sure you don't have some other sizes set or something (maybe curvature and proximity based refinement?)... If not, then this is a parallelizing bug we have not seen before. You should expect very small differences when the mesh generation is parallelized, but nothing like this. Can you email the tetin file and description of the problem to techsupp@ansys.com? You can tell them your are reporting a defect... Best regards, Simon

February 1, 2011, 02:56		#3
beguxa New Member Annon Join Date: Apr 2010 Posts: 18 Rep Power: 16	thanks for the reply. 1) I am not setting any other sizes. Infact the way I faced this problem is following: when the first mesh I created (~25 million cells) was not saved in absence of a disk space, the mesher reported error. After I fired the meshing run second time (with required disk space and, I did not close ICEM) the same geometry gave me a mesh of ~ 194 million (80% element sizes were ~ 7 to 15 mm where as my global max size was 64 and scaling factor of 1). 2) Moreover, I completely discarded the geometry and reconstructed it from 'mesh to facets' options. This is also giving me a very large mesh now (similar size distribution). 3) I need to tune the scaling factor (> 1) to get appropriate mesh. but this kinda of bug is quite unsettling. I am not sure if I will be able to give this geometry to Ansys due to proprietery issues. In order to start afresh, what are the files I need to discard if their corruption is causing this issue? I am no longer using the same geometry file, and created a new geometry file from mesh. Thanks for the patience. -aditya