Hi,

I recently came across a open source code called FreeCFD. It seems to have very nice features, although I haven't tested it yet.

http://www.freecfd.com/about/

Yay, horray, another open source cfd code, just what the world was waiting for

(you will have to forgive my complete and utter lack of enthusiasm)

I still remember about 8 years ago you could only find ancient 1D/2D CFD codes written in Fortran with no interfaces, gui or documentation and with many limitations such as having imposed simple geometries.

Yes, just look how far we've come: now you can download 3D CFD codes with no interfaces, gui, documentation and many limitations such as having imposed simple geometries (unless you buy a third party mesher).

GV

... and some open-source codes with superior capabilities than commercial codes.

This looks pretty well written. Its not too complicated if someone wants to start using it and even modifying it.

Hi, colleagues!

I haven't found description of turbulence models used in the FreeCFD code. I am afraid, if the code involves no models, then it is only appropriate for flow simulation at small Reynolds numbers.

The navier-stokes equations describe flow at any regime as long as they are well discretized. Turbulence models are only approximations.

Oh well thats ok then, here am I using turbulence models when all the time I could have been running DNS on complex industrial turbulent geometries. What a dope am I!!

"Oh well that's ok then, here am I using turbulence models when all the time I could have been running DNS on complex industrial turbulent geometries. What a dope am I!! "

You're missing the point dude, it's like free, and like, better than the free 2D codes from 8 years ago. It's also better than than evil commercial CFD. Apparently. I always thought that turbulence was over-rated.

Sigh

GV

Father, I see the error of my ways.

Anybody got a million CPU cluster I can borrow for the next decade or two?

Use the force*

GV

*However, you'll have to provide your own mesher and post-processing tools. Support is not free. There also isn't any documentation. Or turbulence models. Don't try and move the mesh. Your statutory rights are not effected.

You are free not to use it.

Here is some insider comments from Free CFD.

Almost no open-source/free project is born with all the functionality/documentation and features. Nor they are bug free. But the point is, those projects have a potential of attracting wide interest (if they are lucky) and if they are responsive enough, user base help pick the bugs, generate tons of documentation through the discussion boards and even contribute codes. I am sure several of you are enjoying your Firefox.

Now let's break a myth here before it forms. We don't have a support-ticket system but we have a discussion board and we do our best to help there. We don't charge for that. In fact, we don't charge for anything.

Addressing a few of the concerns mentioned here:

1. GUI: You are right. A graphical user interface is not there. And you are absolutely right that it should be there. While we plan on designing one, we believe that the current interface is not hard at all to use for now.

2. Turbulence Model: A turbulence model is at the top of our list. There is already some preparatory work in the code. But as you will agree, it is not a straightforward task if you want it to be robust.

3. Moving/deforming mesh: This one is also high up in our list of to-do's. I think we can easily say that we have quite a bit of expertise in this area.

4. Documentation: There is a user's manual available and theory/tutorial/coding manuals will be prepared too.

There are also a number of other features we will be adding once we tackle the list above, including a reaction model, multi-phase...

Let's focus on the fact that this is not a code that we used for some time and now throwing as open source. This is a project started from scratch with the sole intention of becoming a capable, but easy to use/develop code for the CFD community. It will be in continuous, active development. We intentionally released it early in the game so that it is easier to get familiar with the fundamentals.

So there. We hope you enjoy it and we are looking forward to your feedback.

I wish to help once I get some more knowledge about the code and I think it is a good idea to release the code as open source early in the development stage so that the learning curve is not too steep.

I'm curious, where do you get the fund to do this project? Are there still people out there who fund such project? How is such project crucial? Compared to FEA (structural) codes, there are more projects on CFD free codes. Why is this choice made I often wonder? Is it because to surplus CFD codes in the world so that the giant CFD companies make theirs cheaper, I assume? If yes, relatively speaking, the giant CFD codes are not too expensive compared to giant FEA codes so I keep on wondering why CFD, not FEA.. (business-wise, there are more demands in structural analyses codes).

Yours wondering, autobot

It is simple really. We don't have any funding or any source of income related to Free CFD. Not a dime. Not that we couldn't find funding, we just didn't seek any.

"How is such a project crucial?"

This part is pretty important. Commercial CFD packages are nice, feature packed and proven. If you are willing to pay for them, that is a relatively painless way of getting some simulation results for your problem. But it mostly doesn't go any farther than that. We are not competing with any CFD package, nor we are trying to pull their prices down. That just simply wouldn't work.

Free CFD is trying to fill in another niche. The whole source code is available. This fact is actually more important than distributing the code for free. Hopefully Free CFD will provide the community with an easily accessible platform to try out new things / learn / teach. We really believe that it will play some role in the advancement of CFD field.

Business-wise, it may be better to work on an FEA code maybe but then again, we are not a business. We are CFD researchers, doing what we do best and taking joy in that.

There are many open source FEA packages also: Elmer, Code-aster, Calculix, nemesis, etc.

The list is long.

I looked a bit on your site and code and I'm impressed. It looks very interesting and I will make sure to link to it properly from CFD Online in the future. As you mention it is a very nice code to base research and education on. Keep up the good work!

To Harold Skywalker and Bob(a fett) who also posted in this thread: Please don't make that kind of derogatory comments, especially not when posting from the same anonymous relay. It looks very much like you have a hidden objective with your posts.

Thanks a lot for your support.