I need a commercial mesh generator. Actually I've been using in-house tools which is quite sophisticated for specific application compared with the ones that come with cfd packages. But mine is not good any more for my new project. I got enough information from each vendors. Before I go for it, I hope to hear from cfd-online. My application requires shock wave capturing, structure analysis, 1000 IGS input etc.

-Thanks.

usually CFD packages will have there own mesh generation codes. But i feel ICEM CFD mesh generation is very good and it is a popular choice for commercial companies.

Where I work, most of the meshes are also generated with ECEM CFD. It is very powerful, and, when used with the module MOM3D it can perform grid adaptation based on a solution (such as the MACH number.)

Check out Tiger mesh generator. You can download a demo from the web

I know there are some famous but not smart enough.

You may want to find an application very similar to yours. Just check www.truegrid.com or one of its contribution ( www.llnl.gov/str/Raboin.html ) to a project which is going to use the fastest computer in the world, ASCI. It may tell the quality and its IQ. Thanks.

Scott

Let me start by saying that I currently work in aerospace, so some of my comments may not be applicable. I have used a variety of grid generation packages over the last 12 years. I am currently using several packages, depending on the application:

- CFD-GEOM from CFDRC for most work

- Hypgen - Hyperbolic generation of structured grids (NASA)

- Gridgen from Pointwise for complicated blocked grids where smoothing is required.

A summary of the commercial/government packages I have used and the relative strengths and weaknesses.

1) ICEM CFD

+ Strengths - Excellent handling of CAD geometries. CAD manipulation is very strong since it is based on a CAD package.

-Weaknesses - Difficulty in using. The learning curve is very steep and using the code with proficiency requires a fair amount of work. It is not the code for you if you are an occasional grid generator. The CAD package is surface based rather than the newer solids based packages.

2) Gridgen

+ Strengths - Excellent grid generation technology. Particularly for structured grids. Elliptic smoothing algorithms are without rival. Ability to import and tweak external grids.

- Weaknesses - Weak CAD import. Does not handle trimmed surfaces well.

3) Hypgen

+ Strengths - Smooth structured volumes, fairly robust. High quality cells in boundary layer.

- Weaknesses - No multi-block capability. No geometry capability (this is only a single block structured GG package)

4) CFD-GEOM

+ Strengths - Good CAD import. Fairly robust CAD manipulation and cleanup tools. Fast tetrahedral grid generation. Good undo/redo option. Latest version has journaling/scripting. Fairly easy to use. Good projection of structured grids to underlying CAD geometry.

- Weaknesses - Smoothing algorithms for structured grids could use improvement. Needs to have hyperbolic marching for structured grids added.

My advice if you are heavily dependent on CAD import (IGES, STEP, etc):

- Try importing and working with your sample files from your mechanical design department. Don't depend on just the samples from the grid generation vendor.

- Make sure the package has sufficient CAD capabilities to repair geometries.

- Don't expect the GG package to be a full-up CAD package. The CAD vendors do CAD much better than CFD vendors.

- Learn to be reasonably proficient with your companies standard CAD package so you can work with your mechanical designers and do some manipulations within the CAD package yourself.

Hope this information helps.

Take a look at Gridgen (http://www.Pointwise.com).

Dear john,

If i'm not mistaking, you graduated from UTA under the supervision of Dr. Dale Anderson.

If so, I am currently working on my Ph.D in the CFD lab at UTA and recently, while i was looking for a mesh generator, i found your master thesis at the library.

I asked Dr. Anderson if he'd still have a copy of the TFI2D code, but unfortunatly, he does not. If on the contrary you still do have a copy of this program, i would be extremely delighted if you could send me a copy of the source code!!

Thank you for your time and consideration,

Sincerely,

Frederic Felten CFD Laboratory, MAE Dept. Woolf Hall, University of Texas at Arlington, Arlington, Texas, USA.

email: felten@uta.edu