CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Which CFD package to use? (

headwing October 28, 2010 23:28

Which CFD package to use?

I am a graduate student in biology in the middle of my thesis (a pretty basic computational fluid dynamic problem), and I'm having trouble with my flow solver :( package (U2NCLE, developed at Mississippi STate University).

At this point, it looks as if this solver may not be useable, so I'm trying to figure out what to do. I already have surface grids and volume meshes that I have generated in another program (SolidMesh, also developed at MSU), but I doubt they will be widely compatible with other CFD packages.

What I would likt to know is: WHat is your opinion regarding what software you would use if you were me (my project is unfunded, so commercial packages are not really an option). But... What would you suggest as a fairly intuitive combination of pre-processr and solver (with the easiest workflow for someone not well-versed in computer science).

Pre-processors I have been considering are discretizer (, SALOME (, and Engrid (

Solvers I have been considering are these:



If any of you have experience with any of these (or other) software packages and can suggest what might be the most intuitive option for me at this point (bearing in mind that I don't have lots of experience editing code and would prefer to use a GUI for anything I can up until I actually invoke the command to run the solver...), then please give me your opinions/guidance.



PS: Also... If you know of any tutorials outlining the entire workflow for something like this, likewise... let me know!

andyj October 31, 2010 08:58

All of the packages you mentioned are good.
The easiest one to use would be Elmer Fem.
Available at sourceforge.
Elmer has been developed for about 15 years and is a complete package with preprocessor, solver, and post processor. Also has a GUI.
There is a large amout of documentation that comes with it and more available on the web. I would look and see if it has the solver you need. It is a Windows based program. Also available in Linux though.
Salome and Openfoam are slightly more complicated. Openfoam would be used for more complex problems.

CAELinux is a complete Linux operating system that has all of the programs you mentioned preinstalled, along with Elmer. It is Open Source.

Openinnovation and CentosFoam have other complete Linux OS, with the software already installed. Openfoam ver 1.4 has a GUI, later versions do not, at least for the time being.
Tutorials come with all of these programs.
You should be able to convert the work you have already done with CAELinux.

My suggestion would be to use Caelinux, since it has all of the software you are looking at, as well as Latex and about 20 other science programs.
You can dual boot Windows and Linux, takes a little studying and internet how-to's and careful planning.

headwing November 1, 2010 13:53


Thanks for the info! I will look into this. From the documentation however, it doesn't appear that Elmer will take my grid format (.b8.ugrid), but maybe I can just create new grids from my nurbs geometries.

In any case, it will be good to have CAE Linux and a whole bunch of extra tools to play with. Maybe I can find a use for some of them.

I was referred by someone else to a solver developed at NASA's Langley Research Center called FUN3D.

Apparently it is the precursor to the code I was using (U2NCLE) and is still supported, so it may be my simplest option at this point (it actually does accept my grid format).

Nevertheless, I'd like to become familiar with other, more mainstream packages so that if I continue on in CFD things will be easier for me.

Thanks again for the help! :D


All times are GMT -4. The time now is 02:54.