ICEM meshing problem
 

Hi all,

I am facing a strange problem using ICEM 5.1 for tetra meshing. I made a fairly complicated building geometry using workbench and import the ParaSolid into ICEM, however it ends up a huge number of cells and ICEM crashes. First, it gave me error message:

=======================

Assertion failed!

Program: …iles\ANSYS Inc\v90\icemcfd\5.1-win\bin\med.exe File: cell_base.C Line: 1065

Expression: cells!=0

For information on how your program can cause an assertion failure, see the Visual C++ documentation on asserts

(Press Retry to debug the application – JIT must be enabled)

==================================

I didn't try to 'debug' since I don't understand this at all. I tried to simplify my geometry but the same problem repeated. I was thinking that the potential cell number of my geometry might be too big so I tried some simple geometries but it seems the problem is still there. From the log information I saw something unusual:

========================

Can't rename unnamed.jrf to … (the same location the project file saved): error renaming "unnamed.jrf': no such file or directory.

Project saved.

Warning: Running tetra with automatic sizing

1 processes about to subdivide …

============================

I am not able to generate any mesh under this situation. Does anyone have any idea about this problem?

Thanks a lot

Forrest

Re: ICEM meshing problem
 

Hi,

The following posting discusses options of meshing large meshes in ICEM: http://www.cfd-online.com/Forum/cfx.cgi?read=10786

Not sure it that is the problem here, your problem may be different. Unfortunately ICEM has quite a lot of bugs and unhelpful error messages.

Regards, Glenn

Re: ICEM meshing problem
 

Hi Glann,

Thank you very much. I am now trying to split the geometry.

Thanks a lot

Regards, Forrest

Re: ICEM meshing problem
 

Hi Glenn,

I tried to reduce half of people I should model for the theatre (roughly 300) and tidy up the geometry. It seems I still have the same problem. I used thin surface feature to model the stairs at different level and the smallest edge in the domain is 0.3 meter. I didn't expect it would end up with a huge number of cells (causing the icem to crash). Here the link below is part of the geometry (complicated part of the theatre). Any suggestion or clue about the reason why icem crashes even when I mesh it using coarsening mesh? Thanks a lot.

http://www.iesd.dmu.ac.uk/~yingchun/...eatingrake.png

The full dimension for this figure is about 11 by 12 by 3.6 meter (just the seating area and air reservoir behind the seating rake)

Thanks a lot

Regards, Forrest

Re: ICEM meshing problem
 

Hi,

Some questions: 1) It appears you created the model in ANSYS Workbench, so why not mesh it in Workbench? Workbench also has limitations on the size mesh it can generate, but it is a lot simpler to use so you are likely to get a useable mesh faster. 2) The geometry appears to be symmetrical. If the flow is symmetrical you can halve the size of the mesh and that will help greatly.

Regards, Glenn