[victorz]

Hi all,

I'm relatively new to CFD, and have been using Fluent for the past few months. I'd like to consult you guys whether switching to OpenFoam as the general purpose solver is a good idea.

The geometries I will be using may be relatively complicated, with a large number of cells.
In general, I'll be dealing with incompressible flows in the following applications:
- two phase flows/condensing flows
- flows in and around heat exchangers
- heat transfer
- porous media
- flows through fans
- pipe flows

Please, write down any ideas that might be relevant, and that will help me decide whether it is a good idea (cons and pros, what questions should I ask myself before the change, what steps should I take, etc.)

Thank you very much!

[msbealo]

Victorz,

You should probably also post this message in the fluent forum as it might be slightly biased here.

I used fluent for a short time. I found it very frustrating to use, especially the CAD type interface when you're setting up geometries and meshes.

OpenFOAM is free, however if you had a service contract with Ansys then it might make it worth paying for, it all depends on your computer skill and patience. I found OpenFOAM a lot more satisfying than Fluent, but it's no picnic.

As for performance, memory usage and accuracy I don't know. I've not compared the two, however my gut feeling is that OpenFoam would be quite memory efficient, but I've heard people say that it's slower at solving, probably because most solvers are transient rather than steady state.

I suggest you look at the documentation and play with the tutorials and see if you like it.

Mark

[alberto]

Originally Posted by victorz

Hi all,

I'm relatively new to CFD, and have been using Fluent for the past few months. I'd like to consult you guys whether switching to OpenFoam as the general purpose solver is a good idea.

It largely depends on your expertise and on the applications you have to deal with. I agree with Mark about asking also on FLUENT's forum and decide independently: you will find bias on both sides.

The geometries I will be using may be relatively complicated, with a large number of cells.

This should not be a problem. You can rely on external meshers, if you prefer, or on the snappyHexMesh tool if it works for you.

Please, write down any ideas that might be relevant, and that will help me decide whether it is a good idea (cons and pros, what questions should I ask myself before the change, what steps should I take, etc.)

A few questions to ask yourself:

- Do you know C++?
- Do you have time to eventually add the sub-models you might need, if they are not available (it might be the case for condensing flows, for example, depending on what you need)?
- Are you comfortable with the command-line interface?

To conclude, if you decide for OpenFOAM, you might consider to attend one of the training courses to speed things up, if needed.

Good luck!

[victorz]

Thank you msbealo and Alberto for your response!

To answer Alberto's questions:
- I have some experience with C, and wrote some relatively complex programs, in a somewhat "engineering programming" style, rather than in a more efficient and professional manner.
- My time is not unlimited, but on the other hand, I do have some "grace period" to learn new things.
- Similarly to the first answer, I have some programming experience, so it doesn't scare me.

From my point of view, the main reason to switch to OpenFOAM is to spare the tens (or hundreds) thousands of $ needed for a good commercial package. The price I pay is that I have to spend much more time to get the same experience.
Another benefit of OpenFOAM is, as I read in several similar posts, that I get to REALLY understand what's going on "behind the scenes", get experience in programming and be a better engineer. The challenge is without a doubt huge.

Anyway, thanks for the reply, I'll post a similar thread to see what the Fluent biased guys can tell

Victor.