[maille]

Hi,

I'm quite new at using OpenFoam. I've been looking for tutorials or examples applied to natural environments such as lakes/pounds without success.

I'm trying to model seiches in a water basin as a result of wind forcing.
I have started my project with a very simple basin, using (RAS)interFoam. My first results were not making much sense, so I set U to 0m/s and it appears that despite this set up significant velocities appear in the run, provoking some oscillation of the water body.

Would you have some advices or directions for me?

Thank you for your attention.

[gwierink]

Hi maille,

I think the problem are parasitic currents, which are more pronounced for greater viscosity and density ratios for the two fluids. These are usually quite difficult to deal with in a general way, although there are some tricks for specific situations. Have a look at this discussion for instance. More info on parasitic currents you can find here.

If you run the damBreak case in zero gravity (so that there is no driving force), you'll see parasitic currents as well. Do you have any pics of your lake? Have you tried interDyMFoam (dynamic meshing)?

One thing that may be of your interest is a report by Jan Potac. It's about snow drift, but there is some explanation on what BCs were used etc.

[gwierink]

Right, I forgot to mention some other things ... Perhaps you know, but in any case there is an interface compression factor called cAlpha in system/fvSolution. Increasing cAlpha above 1 compresses the interface, which may help. Also, it may help to set momentumPredictor to yes in system/fvSolution.
This is all for inter(DyM)Foam, using the VoF method. You could also have a look at an interface tracking method like interTrackFoam.

P.S. I became curious and tried a simple 2D water basin with 10 m/s wind. The pics are after 60 s simulation in a 300 m domain. The water seems to start moving a bit, but I think the simulation time should be longer and/or higher wind velocity. The case lives here.

[maille]

Hi,
thank you for all your advices and help. I had found the second link on parasitic currents and tried to use some of the advices there but without sucess. The gravity seemed to be part of the trouble.
What would be the advantage of using interDyMFoam?
I have played around with cAlpha and momentumPredictor as well but without getting much improvements.
I have to put this project on break for a while but I will run your example with 10m/s (that should be more than sufficient to produce basin oscillations) and 0m/s to check the results, I will compare the configuration. I have already seen that you are using OpenFoam 1.6 while I'm on OpenFoam (p=>p_rgh), and you are also using the refineMeshDict that I have not used yet but seems extremely convenient here.
I will post a bit more details and results as soon as I can get back on this.
Thank you again.

[gwierink]

Hi,

gravity seemed to be part of the trouble

In terms of parasitic currents you mean? Actually, if you're talking about parasitic currents, the problem is that if surface tension forces become dominant the VoF method freaks out and gives weird velocities at the interface.

What would be the advantage of using interDyMFoam?

interDyMFoam is basically interFoam with dynamic meshing (this is what the "DyM" part stands for). The solver finds areas with string gradient in the colour function (alpha) and starts refining the mesh locally to resolve this strong gradient. Your interface will thus be sharper and better resolved.

I used OF-1.6 to set up this case, that's right. But it's not such a job to convert it to 1.7. Have a look and a try, I'm happy to help if I can