Ansys Meshing & ICEMCFD - A comparison
 

Please find attached the comparison of Ansys meshing and ICEMCFD that I've drawn from my user experience.

I'd like to have suggestions on correcting something (or) including more points into this comparison.

I'd appreciate any help.

In version 14.5 (released during November) ICEM CFD is integrated in workbench too.

Thanks for the correction ill update this into the comparison

I've modified the comparison a bit by correcting some of the points now

I do not agree with this Analysis.

Could you please correct me if Im wrong. I'd appreciate the help because my intention to post it on the forum is to recieve such suggestions.

I also disagree, but have not had time to comment... This is not an official ANSYS repsonse, but rather a top of my head commentary at the end of a long day...

Going point by point, I would start by saying that ICEM CFD is much LESS geometry dependent than ANSYS Meshing...

Octree is totally different than Delaunay, so comparing speed between the tools is pointless. If you compare the Octree in ICEM CFD with Patch Independent in ANSYS meshing, you will find ICEM CFD is faster and uses less memory.

ICEM CFD smoothing is very good and very fast... ANSYS Meshing doesn't really have an equivalent smoother, so it is not a surprise that smoothing takes "longer" in ICEM CFD. You should also compare the quality of the results. Maybe reducing the ICEM CFD smoothing to 1 iteration would give a more comparable timing and quality.

You say pre prism is better than post, but you have another post on CFD Online where you had a Post prism that worked, but you can't get the delaunay tetra to generate... http://www.cfd-online.com/Forums/ans...n-icemcfd.html I see that quite a lot... Don't get me wrong... Pre-prism is pretty great (it is based on TGrid tech), especially for external aero, and it is getting better every release... But they are tough to compare and both options are available in ANSYS Meshing (the actual ICEM CFD post prism is in ANSYS Meshing)... Both should also be in ICEM CFD for the next release.

You say that the prism mesher is stringent, but you mean the quality criteria is harsh... that is true. ICEM Prism quality looks bad until you take it to another tool that is less harsh... Then suddenly it looks great.

You say ICEM is better at handling STL, I would say that ANSYS Meshing can't handle it at all.


ANSYS Meshing is totally geometry dependent. For Volume Meshing, it requires closed solids and does not support faceted data or imported mesh at all. Yes, its patch conforming algorithm is faster than octree, but it compares more with the ICEM CFD patch conforming algorithm.

I will come back another time and post the offical list of how to choose between these two ANSYS products...