CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Pointwise & Gridgen (http://www.cfd-online.com/Forums/pointwise/)
-   -   Pointwise/Gridgen (http://www.cfd-online.com/Forums/pointwise/70833-pointwise-gridgen.html)

Amiga500 December 7, 2009 13:58

Pointwise/Gridgen
 
Any users here?

How does it stack up compared to ICEM?


I am getting somewhat annoyed at the inconsistencies of ICEM when performing geometric operations.


Or does anyone else have any other meshing software they'd recommend?

CapSizer December 7, 2009 17:28

I've used both, although much more Icem than Gridgen. The underlying philosophies are very different indeed. Coming from Icem, I had much difficulty getting Gridgen or Pointwise to do what I wanted, but I suspect that an experienced Gridgen user going the other way would have similar problems. I would say though that neither is brilliant when it comes to geometric operations. Geometric operations in Gridgen seemed to be much more tedious, but Icem has other geometric frustrations. The grid generator that I quite liked for its ease of geometry handling was ESI/CFDRC's CFD-Geom. Getting the actual grid exported in a useable format may be a problem though. I have to say though, that current (and also future) thinking is biased towards doing most or all geometry in CAD, before going into the mesh generator.

Amiga500 December 7, 2009 19:00

Sorry, I wasn't clear when I said "geometric operations".

I mean CAD cleanup, and perhaps subsequent addition of features and details.


I find ICEMs inability to correctly project lines onto surfaces, or subsequently obtain accurate surface-to-surface intersections very frustrating.

CapSizer December 8, 2009 02:48

That doesn't actually change my comment at all. I've come to the reluctant conclusion that if you don't do nearly all the CAD cleanup in the CAD, you will be in for a world of trouble. I had the novel experience in Icem 11 of creating a simple geometry entirely within Icem, passing all the topology checks, and the tetra mesher blithely ignoring my internal geometry. I complained to Ansys, and there reply was "Use Design Modeler". I did, and it worked, but really, if you are simply doing a sphere inside a cube it shouldn't be necessary to resort to CAD!


I share your frustrations exactly, but I found creating extra geometry in Gridgen quite difficult, certainly less easy than in Icem. What I did like about Gridgen was the ability to "quilt" patches together with a tolerance.


It seems to me that the best answer to the "BAD CAD" problem is a mesh generator with a good surface wrap capability, and perhaps Adapco have the answer there.

jchawner December 8, 2009 11:41

Brendan & Charles:

I hope you don't mind if I offer my biased opinions! While I cannot offer any comparison to ICEM (never used it myself) I can address some of the issues you cite in isolation.

Dealing with geometry from a CAD system continues to be a challenge for everyone. ANSYS' relationship with SpaceClaim is evidence of the work going on in this area. I will admit that Gridgen's geometry modeling capability cannot compete with a CAD system - it simply wasn't designed to. By geometry modeling I mean the ability to create geometry from scratch. The software does support creation of many curve and surface types but not a lot and not in the way that a CAD system would do it. The new software, Pointwise, has improved on this somewhat (especially the workflow) and will continue to do so in the future.

Both Gridgen and Pointwise were designed to bring in a CAD model and make it analysis ready. So they both have tools for intersections and projections and trimming. From the standpoint of dealing with sloppy CAD, both offer two approaches: fault tolerance (in which the grid is merged over gaps an junk in the CAD) and solid modeling (where the geometry is formed into a watertight solid). (Note: solid modeling will be released in Pointwise shortly after the first of the year.)

Geometry modeling (including import, export, creation, assembly, etc.) will continue to be an area of focus for us (and I suspect everyone else in CAE) for the foreseeable future. To put it all in perspective, it is reported that the annual cost of CAD interoperability problems exceeds the market capitalization of all CAD companies combined. So I suspect there's plenty of work to go around for all of us.

If you'd like to discuss this offline in more detail, feel free to email me directly.


All times are GMT -4. The time now is 21:21.