Pre-processor (Gambit Alternatives)
 

I was just wondering if anyone out there could recommend a third party pre-processor as a Gambit replacement. I am tired of spending hours to create a half decent mesh in gambit, as I'm sure most people are. I am looking for a program that will make meshing easier and quicker, yet still allow a similar level of control over mesh density and scheme to that offered by gambit.

Thanks

Re: Pre-processor (Gambit Alternatives)
 

ICEM-Cfd works for me. It was a strange program to learn, but it does the job.

Re: Pre-processor (Gambit Alternatives)
 

Hey Charles,

Is this a free software? How do we get and if ot free how much it costs???

Thanks Ravi

Re: Pre-processor (Gambit Alternatives)
 

Ansys ICEM CFD is far from free. The cost is of the same magnitude as a Fluent license. But it's a very nice program to work with, and superior to Gambit.

More info here: http://www.ansys.com/products/icemcfd.asp

Re: Pre-processor (Gambit Alternatives)
 

Thanks for the information, does anyone know what the difference (if any) between ICEM and AI Enviroment?

Re: Pre-processor (Gambit Alternatives)
 

AI Enviroment is ICEM CFD with extra features for export to structural analysis code like Ansys and Nastran. You get a couple of extra tabs in the GUI, otherwise it's the same.

Re: Pre-processor (Gambit Alternatives)
 

I used GridGen, which is better than Gambit, although I had some trouble with grids that did not connect to the geometry. Currently I am working with ANSA. ANSA is more user friendly and has a very clear GUI. Several meshing algorithms are available. The CAD funcions are easy to use even after meshing the geometry.

Re: Pre-processor (Gambit Alternatives)
 

The best gridgenerating software I know till this time is the Gridgen www.poinwise.com

The disadvantage of Gridgen - it can't be used for Numeca grid generation.

Re: Pre-processor (Gambit Alternatives)
 

Does anyone have any information on Harpoon by CEI: http://www.ensight.com/products/harpoon.html

I am considering benchmarking ANSA, Gridgen, ICEM and Harpoon. I am currently working on the criteria for the benchmark along with the tests I will use but my initial musings on this include:

Cost, Ease of Use, Quality of mesh, Control over mesh, Computational Efficiency

Any thoughts on any other criteria or how best to investigate these areas would be very much appreciated.

Re: Pre-processor (Gambit Alternatives)
 

I have installed an evaluation version of Harpoon, I will try it in a few weeks. Harpoon is an automatic mesh generator. It generates a cartesian non-conformal mesh. I didn't have the time to try it, but according to some Fluent users, the quality of simulations with non-conformal interfaces in Fluent is quite low. When comparing the mesh generators, the type of simulation you want to run is an important aspect. For potential flow, grid quality is of no importance. But for viscous calculations, low skewness is a major requirement, especially when using Reynolds stress. When comparing Gridgen and ANSA, I prefer the last. Please let me know when you have any results of your benchmark!

Re: GAMBIT AND FLUENT
 

I have 02 queires. I would be very greatful if someone can help me.

1. What could be BCs for the simple Viscous Fluid falling (-Y direction ) under gravity on a moving surface in (+X direction). 2. Can VOF be used for the above case and if YES how can i Patch in FLUENT.

Kind Regards

Re: Pre-processor (Gambit Alternatives)
 

I've experimented a bit with Harpoon. It's really quick, which is a good advantage. It takes a bit of work to get the mesh resolution right in the areas that you are interested in. Where it comes into its own is where you have lots of computing resources (parallel licences), but not much time. So you just "shotgun" the mesh with Harpoon, i.e. you get a grid that captures the geometry, but you don't try to develop the most efficient grid. To get the necessary detail, the grid will be quite large. In contrast to what some have said, it projects and cuts the Cartesian grid to the geometry surfaces, and there is even the option of building prism layers ("boundary layers" in Gambit-speak). When you run it, convergence is generally very good, because the grid is extremely orthogonal, and there are few high aspect ratio cells. It's a good "rapid prototype" tool, you really don't want to spend much more than an hour or so on meshing, and then you throw the grid at a nice powerful cluster, and you have preliminary answers a few hours later.

If you have lots of time but not lots of computer resources, the best option is to really build and tweak the best hexahedral grid you can, typically with either Icem, Gridgen or even Gambit. Such a grid could easily end up with only 1/10 as many cells as the Harpoon grid, for the same fidelity, but it will take you a couple of days (or weeks!) to build.

If you have neither time nor computing resources you are up the creek without a paddle.

Re: Pre-processor (Gambit Alternatives)
 

Does anyone of you have experience with GridPro?