CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- Main CFD Forum (https://www.cfd-online.com/Forums/main/)

- - Open Source (https://www.cfd-online.com/Forums/main/7238-open-source.html)

Folks, Has anyone got any experience running problems with any of the open source tools? Seems that the Gerris solve is the most developed (good job Stephane) but I am looking more to be a user and not a developer (Yes I know that makes me a parasite) and my attempt to get gerris working under msys/mingw failed miserably.

I am taking a peak at opengrid, the first package from opencfd. Does anyone know who this person is?

Any thoughts on any other tools out there?

The problems of interest to me (remember I am a user), involve turbulent external flows below mach .5, internal flows involving heat transfer, and free surface flows around planing water craft.

Any suggestions on whats out there for free.

Any suggestions on free meshers for complex geometries?

Thanks

Andy

Have a look at the following section

http://www.cfd-online.com/Resources/soft.html

Yes I've taken a look at most of that. You will find if you drill down a lot of links that are dead and a lot of free code that is not free.

I am looking for any real world experience anyone has had with the codes I mentioned or any other open source codes.

Again I am interested in interal turbulent flows with heat transfer and free surface flows past planing hulls.

Any experience there?

- Andy R

Free for academic use

I think a bit mode detail may be fair:

Quoted from http://www.nabla.co.uk/main/mainhe6....cademicPricing

Nabla Ltd offer licences of FOAM to academic institutions free of charge.

We provide absolutely no support with this licence, but users will still have the following resources at their disposal: PDF manuals; Doxygen documentation of FOAM; web site; discussion group.

We will offer support packages for specific projects that require it and have the necessary resources to pay for it. Support options include: basic support; development support; and, teaching support.

This licence is subject to terms and conditions which prohibits its use for commercial purposes.

A commercial licence and support is, of course, not free.

Re: Free for academic use

Ok let me reiterate my very simple question in terms that even someone with a PHD can understand.

Does any one have any experience or direct knowledge of any of the open source CFD codes?

Now thats a damn simple question!

Now Foam is an excellent code but it is not open source as far as I know.

So again has anyone on this list used

GERRIS OPENCFD

or any other open source code?

Thanks - Andy

This email proof that one should not post after that second scotch on the rocks ;)

Re: Free for academic use

Andy,

Yes, I have experimented with some of the open-source codes. Gerris is very innovative and interesting, but it is still (i.e. when I last played with it) an incompressible Euler solver, which is rather limiting. The other one that I have spent some time on is DUNS, which has got a turbulence model, can handle compressible flows well and was originally written (I think) for mostly chemical processing. DUNS is quite promising, but like most of these codes, documentation is a bit hit and miss, which makes it a bit tricky to use. Also, it uses multi-block structured grids and has no built-in post-processor, although with the availability of codes like GMV and Paraview that is not a show-stopper. Also, CGX (graphic pre-and post-processor for the open-source FEM code CCX) works with DUNS, but its capability is limited. If you want it to be more useful, do the work and share it with the rest of the world .... To put it in a nutshell, you might get somewhere with an Open-source code, but it is far from free, in the sense that you will have to spend a very serious amount of time learning it (with little documentation and few examples), writing pre- and post-processor interfaces, etc. It is not a "free, easy" option. If you've got a serious job you need to do, rather try to get a short-term license for a commercial code. If you just want to learn, go with open-source. As for free-surface, you're on your own!

Charles,

Thanks, thats the kind of input I was looking for. I have no illusions about free and easy. Just looking at what would be the most interesting and productive tool for me to work with if I wanted to get involved.

Quite frankly the biggest issues I have with all the open source is packaging and interoperability.

A friend is using MAYAVI as a post processor and loves it. However that tool can't read CGNS files. It reads its own format, plot3d and ensight (of all things) data formats.

Gerris uses GTK and GTK uses glib. But the version of glib it uses is not the current one, so that becomes an installation issue.

If I get involved my self appointed task will be to address some of these issues.

The reason that LINUX has really taken off is that it reached a point where someone not involved in the development of that program could still use it. In other words there were people out there who were simply users. As there were more users, more developers saw it as a project worth working on. This amplification has yet to happen with any of the open source engineering tools.

Time will tell.

Thanks - Andy