[ngj]

Good evening,

I suppose that you are using a turbulence model, correct? Try to disable it, and I would suspect that the problem will disappear.

Kind regards

Niels

[Ya_Squall2010]

I am now running the case with turbulence switched off. Will keep you updated.

BTW, what if the turbulence must be on?

[kumar2]

Dear Niels,

I have a question regarding the "Validation case for breaking waves" ( section 3.3) in your journal publication ( Wave Generation toolbox ... ; International Journal for Num. Methods in Fluids , 2011)

In figure 8, what is the length of the section which has a depth of 0.01 m ? And in your experiments did you use a wave outlet relaxation zone?

Thanks a lot

Best regards

Kumar

[ngj]

Hi Kumar,

Thanks for the questions. Unfortunately, I cannot remember how long the horizontal distance was, and I will not be able to check within the first couple of weeks.

With respect to the "outlet" boundary, then I do not use any, as I rather use a fully reflecting wall with no-slip and Neumann on the pressure; as indicated in the sketch of the numerical wave flume.

Having an outlet at the beach would remove very important nearshore processes.

Kind regards,

Niels

[kumar2]

Hi Niels,

Thanks so much for the important information regarding the outlet boundary.

As far as the pressure is concerned will a zeroGradient work ? ( If I need to go with a fixedGradient, I have no idea what the value of the gradient will be). In the case of the velocities, I do not see a boundary available in OpenFOAM that will reflect the normal velocity. (Please point me to the correct one if I am wrong). I was wondering if you custom made your own boundary condition.

Best regards

Kumar

[reynoldsStress]

Just a quick note:

waves2Foam compiles beautifully out of the box for 1.6-ext but fails to compile for 2.2.x. See attached log of compilation.

[ngj]

@Kumar: No, it is straight out of the box, with (0 0 0) for the velocity and zeroGradient for the pressure.

@reynoldsStress: Yes, I do know, however, I have not had the time to correct this. Reason is an interface change for the sampling or functionObject classes from version 2.2.0 to 2.2.x. Until further notice please correct the interface in the appropriate files yourself or remove them from the compilation. The latter will result in a failure to compile the surfaceElevation tool.

Kind regards

Niels

[kumar2]

Hi Niels,

Thank you very much for the detailed information on the outlet BC. So I went and modeled the breaking wave case. In my experiments the first 3 breakers are created, however once the first breaker dissipates near the top of the slope, the wave activity in the whole domain seems to be dying out.
Please take a look at the video. The movie starts at t=8 and goes on upto t=14 with a time step of 0.2 sec.

The link to the free surface is here:
http://www.flickr.com/photos/98137989@N05/9179801190/

The link to the TKE field is here:
http://www.flickr.com/photos/98137989@N05/9177652759/

Briefly, i used the stream function waves as given in your publication. I used the generateStreamFile and used N=2, to calculate the coefficients A,B and Ubar. These parameters were also input in waveProperties. I also removed the relaxation at the outlet. In these simulations the length of the section where h=0.01m ( Figure 8 of journal publication) is about 1 m. The aspect ratio of my cells are 2. The delX and delY of the cell near the water line, say in the relaxation zone near the inlet is 0.02m and 0.01m respectively. Also the slope starts immediately after the relaxation zone at the inlet. I used the modified k omega turbulence model and the wall functions given in your paper. Near the bottom the 'y' spacing in about 0.00046m in the relaxation zone near the inlet. I was wondering if you can give me some pointers as to why the waves are dying out in the domain.

Thanks a lot

Best regards

Kumar

[Ya_Squall2010]

Just an update by switching off the turbulence. As can be seen from the attached pictures, things are improved but waves are obviously lower than the original state. Anyone got any suggestions?

w1.jpg
w2.jpg

Just found mesh has significant effects on the results, see attached below. What is happening in the relaxation zone?

mesh.jpg
mesh2.jpg

[Ed R]

Hi,

I've been playing with the bichromaticSecond wave type and can't seem to get the Tsoft parameter in work despite having this message in the log.setWaveFields file:

Setting the wave field ...

--> FOAM Warning :
From function setWaveField::setWaveField( const fvMesh & mesh, volVectorField & U, volScalarField & alpha, volScalarField & p)
in file setWaveField/setWaveField.C at line 73

The specified value of Tsoft is non-zero in the waveType: `bichromaticSecond'
specified in the sub-dictionary waveProperties::inletCoeffs

Consequently, the initialised `wave field' is set to a horizontal free surface with zero velocity.

It still sets the wave fields as if Tsoft equals zero. Is this property functional for this wave type?

Thanks,

Ed

[Ed R]

Regarding my previous post (above) it would appear that the Tsoft parameter doesn't work for the bichromaticSecond wave type as the exact same case with the wave type changed to bichromaticFirst works as expected. Could this be due to my recent update which included the introduction of the bichromaticSecond wave type?

Thanks,

Ed

[Ed R]

Hi all,

I'm trying to create a new wave type and struggling.
I've copied one of the other waveTypes and just changed to name of the 3 files <name>.C <name>.dep <name>.H and every instance of the original waveType in these files and then recompiled waves2Foam by typing ./Allwmake in the root directory. Is this the right way to do this? Is there some list I need to add the name of my new waveType so that it compiles? At the moment my simulations says that the waveType doesn't exist.

Thanks,

Ed

[rye]

Hi,


I am running openFoam 2.11 on Ubuntu 12.04. I am trying to run waves2Foam, but something, somewhere, goes wrong.

I followed the tutorial found on the wiki to modify interDyMFoam into wavesDyMFoam, but,
-I am not sure where I have to put the directory waves2Foam
-I got various errors during compilation (exemple : could not open file relaxationZone.H for source file waveFoam.C), even if the files are present at the (what I think) right place.

I runned the 3Dwaves tutorial, and it went fine.

When I tried to run the "moving box" case, found on this topic, I got the following error : "/opt/openfoam211/bin/tools/RunFunctions: ligne 47: waveDyMFoam : commande introuvable" in the log.waveDyMFoam file, which makes me think of an error in the compilation : nobody is able to found the waveDyMFoam command.

Of course, the only thing I am interested in is the 6degrees of freedom model, the one which don't run. That's why I am writing here.
I can give you more details if needed on the errors I encountered.

[Abdel]

Hi everyone,

currently, I'm running waveFlume with 10sec of creation of wave.
However, I observe that my waves (StokesFirst model) disappear progressively.
I think it's because of energy dissipation.

waveProperties :

timeShift 0.0;
seaLevel 0.00;
relaxationNames ();
initializationName init;
pName p_rgh;
inletCoeffs
{
waveType stokesFirst;
Tsoft 2;
depth 0.700000;
omega 7.853981634;
phi 0.000000;
waveNumber (6.283185307 0.0 0.0);
height 0.05;
initCoeffs
{
waveType potentialCurrent;
U (0 0 0);
Tsoft 2;
};


I want to know if it is possible to suppress this phenomena.
(I tryed to modify my equations in the src/waveTheory/regular/stokesFirst but without results .. maybe my waves are too short)
Thank You in advance

Kind regards,

Abdel

[rye]

To add to my post of yesterday, I have two problems :

A-I can't run the tutorials in any other directory than in the one I installed waves2Foam. and I don't want to run the calculations in the opt/ directory.
B-I can't run at all waveDyMFoam I installed using the wiki


A-

-it is only possible for me to run cases with waveFoam put in the directory I installed openFoam and waves2Foam. I presume this is linked to a call for files, but I am not able to find which ones :

I modified all "./../../" in the Allmake files of the tutorials to go to the correct position, but, I think I made a mistake since It says :
sh: 0: Can't open opt/openfoam211/applications/utilities/waves2Foam/bin/prepareCase.sh
But the case runs for a few second (8 iterations, instead of the 200 asked), then crashed. The only error I found is "/opt/openfoam211/bin/tools/RunFunctions: ligne 47: faceSet : commande introuvable" in the "log.faceSet" file.

B-

-still unable ti run waveDyMFoam, I don't understand what I missed in the tutorial (well, if I knew the source of my problem, i would'nt be there). Could someone help me understanding what to do to install waveDyMFoam

[ngj]

Hi Rye,

Only the tutorials rely on a bash script, which makes the relative location to the waves2Foam installation directory very important. Simulations in general can be conducted in any directory. The bash-script ensures compatibility of the tutorials across a multiple of OF-versions.

With respect to waveDyMFoam I hope that other users will help you, since I have not even written that particular part of the wiki.

Kind regards

Niels

[ngj]

@Kumar: Sorry for the long wait, though I have not been able to see your movies until know (lack of internet). The only thing I can assume is that you are using a turbulence model without the density terms included. This leads to excessive diffusion of turbulence across the interface.

Furthermore, as you know from my article, I get poor results for the wave breaking, when I am using an aspect ratio of anything but 1.

Good luck

Niels

[Rotto]

Hi Niels and others,

I am trying to install Waves2FOAM on a new machine which runs OpenFOAM 2.2.0 on an Ubuntu 13.04 OS machine (I've been enjoying using Waves2FOAM toolbox on my current system, which runs OpenFOAM 2.1.0 on Ubuntu 12.04 OS). OpenFOAM 2.2.0 is installed and runs with no problem. I am, however, running to issues with installing Waves2FOAM.
After downloading the toolbox, and when I try ./Allwmake I receive the following error:

FATAL ERROR.
The directory path /applications/utilities/waves2Foam does not exist.
Correct the path in bin/bashrc

This error makes sense, as there is actually no such directory (there is no 'waves2Foam' under 'utilities')! I created an empty directory to see whether that would resolve anything, no luck! I thought it may matter where waves2Foam is being downloaded and installed. So I tried to install it under 'root' directory, and this time I receive the following similar error:

FATAL ERROR.
The directory path /root/OpenFOAM/root-2.2.0/applications/utilities/waves2Foam does not exist.
Correct the path in bin/bashrc

which, the error makes sense, again! As there is no directory called 'root-2.2.0', but instead 'OpenFOAM-2.2.0'! I do not understand why bashrc would look for 'root-2.2.0'. I am afraid whether the problem is how 'WAVES_DIR'=$WM_PROJECT_USER_DIR' in the bashrc file defined. I could not figure out how this variable is defined and/or how to modify it, if that is the problem.

Any chance you would have an idea on what the problem is, and how can I fix it to install waves2Foam properly? Has anyone run to this issue?
Any help is highly appreciated. Please let me know if more information is needed.

Thank you,
Masoud

[ngj]

Hi Massoud,

All you need to do is to type the absolute path to the waves2Foam directory, so simply modify to

export WAVES_DIR=<absolute path to waves2Foam directory>

Kind regards

Niels

[kumar2]

Hi Niels,

Thank you very much for your reply and checking out the animation.

I had modified the default incompressible k-omega model in OpenFOAM. In the default k-omega model, the density is not 'visible' since the coding is with viscosity/density. From you reply, it looks like this will not work and I need to bring density into the formulation. Am I correct ? And would modifying the 'compressible default kOmegaSST' be a good option?

I believe that the wall functions that you used are not present by default in OpenFOAM. Did you make your own custom made wall function or did you implement them inside the turbulence model? ( perhaps by getting rid of omega_.boundaryField().updateCoeffs(); and setting the patches to the k, omega and nut values)

Sorry I am asking you so many questions.

Best regards

Kumar