[Sideshore]

Hello everybody!

Currently I'm working on resistance calculations of a IOM sailing boat (http://www.sailing.org/classesandequipment/IOM.php). As these are remote control boats of 1 meter long the calculations are done at "full" scale at 2 m/s.

I'm using OpenFOAM-v1812, the vessel is free to sink and trim using dynamicMotionSolverFvMesh and the sixDoFRigidBodyMotion, and a sail force is added as a sixDofRigidBodyMotionRestraint linearSpring.

Most of the settings I used are taken from the DTC_Hull tutorial.
I've attached the following files:
dynamicMeshDict
controlDict
fvSchemes
fvSolution

Unfortunately I noticed I have spurious waves in my freesurface and also the water interface along the hull doesn't look very good, see the attached pictures. I traced these problems to the icAlpha option of the alpha.water term in the fvSolution. Setting this to 0 eliminates the spurious waves, as you can see in the other attached pictures where I restarted the calculation and let it run for some timesteps.

The only information I can find about icAlpha is the following:
https://openfoamwiki.net/index.php/InterFoam

and this thread onfortunetely without any answers.
What is scAlpha vs. icAlpha vs interfaceCompression

Can anybody give me any advice which value for icAlpha (and scAlpha) I should use? And what is the range for icAlpha?
Any description would be much appreciated!

Thanks in advance for your comments!

regards,
Jan Willem

[JM27]

Hi,


I am also interested in the functionality of icAlpha and scAlpha. It seems that documentation on this is very limited.



Unfortunately, spurious waves are a common problem in VOF simulations. A multitude of things could be triggering the presence of such spurious waves, so I think you need to do some detective work...



Some artificial compression like icAlpha and scAlpha can help but only one you are sure your setup is right



I would try different settings and compare the magnitude of these parasitic currents before trying to implement icAlpha or scAlpha. In any case I think their behaviour is similar to cAlpha, 0 for no compression, 1 for compression and >1 for enhanced compression. This should be explained in the source code.



Have you tried different Co number setting in your simulation?



Goodluck!

[Sideshore]

Hi,

Thanks for you reply!
For me it's not clear from the source code that the values for icAlpha & scAlpha should be 0 for no compression, 1 for compression and > 1 for enhanced compression. It seems obvious but for me that was/is just an assumption.

So I basically narrowed it down to the icAlpha coefficient. A value of 1 gives me spurious waves and 0 not. However for other cases with other mesh settings I did not have any problems with icAlpha.

This is inline with what I read somewhere (sorry cannot find it anymore) where using an icAlpha value > 0 can give spurious waves if the mesh has high aspect cells around the freesurface.

I tried lowering the maximum courant number in the freesurface interface which did not change the results. I tried lowering the tolerances and changing some numerical schemes without any luck.

Changing icAlpha to 0 was the only solution I found without going to first order methods.

Thanks again,
Jan Willem

[JM27]

Originally Posted by Sideshore

Hi,

Thanks for you reply!
For me it's not clear from the source code that the values for icAlpha & scAlpha should be 0 for no compression, 1 for compression and > 1 for enhanced compression. It seems obvious but for me that was/is just an assumption.

So I basically narrowed it down to the icAlpha coefficient. A value of 1 gives me spurious waves and 0 not. However for other cases with other mesh settings I did not have any problems with icAlpha.

This is inline with what I read somewhere (sorry cannot find it anymore) where using an icAlpha value > 0 can give spurious waves if the mesh has high aspect cells around the freesurface.

I tried lowering the maximum courant number in the freesurface interface which did not change the results. I tried lowering the tolerances and changing some numerical schemes without any luck.

Changing icAlpha to 0 was the only solution I found without going to first order methods.

Thanks again,
Jan Willem

Hi Jan,

If you have a look at alphaControls.H you will find some info on icAlpha and scAlpha.

Have you tried changing the cells in the free surface region so that they are close to 1? Maybe that would solve your issue with icAlpha working for some cases and not for others...

What value are you using for cAlpha?