CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Running, Solving & CFD

waves2foam 2d deepwater case boundary reflections

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   August 27, 2012, 10:26
Default waves2foam 2d deepwater case boundary reflections
  #1
Senior Member
 
Kevin Smith
Join Date: Mar 2009
Posts: 104
Rep Power: 17
kev4573 is on a distinguished road
Hi I am having some difficulties using the waves2foam toolbox for a simple 2D deepwater case. It appears that the waves are breaking at the beginning of the wave outlet relaxation zone and causing reflections. I've tried a number of things to fix this, but I've been unsuccessful so far in getting rid of reflections.

Here's my openfoam case, setup for 2.1.x
https://www.dropbox.com/s/qht3ajefrr....025_mesh1.tgz

Thanks,
Kevin
kev4573 is offline   Reply With Quote

Old   August 27, 2012, 10:57
Default
  #2
ngj
Senior Member
 
Niels Gjoel Jacobsen
Join Date: Mar 2009
Location: Copenhagen, Denmark
Posts: 1,900
Rep Power: 37
ngj will become famous soon enoughngj will become famous soon enough
Hi Kevin

With the setup you have posted, the wave will never reach the outlet relaxation zone, simply because the domain is 7 m long, whereas the outlet relaxation zone starts at x = 15.

Kind regards,

Niels
ngj is offline   Reply With Quote

Old   August 27, 2012, 11:11
Default
  #3
Senior Member
 
Kevin Smith
Join Date: Mar 2009
Posts: 104
Rep Power: 17
kev4573 is on a distinguished road
The waveProperties file uses an include statement to merge in the inputs for this case, so if you run expandDictionary on the the file, the output dictionary section looks like this -

Code:
outletCoeffs
{
    waveType        potentialCurrent;
    U               ( 0 0 0 );
    Tsoft           2;
    relaxationZone
    {
        relaxationScheme Spatial;
        relaxationShape Rectangular;
        beachType       Empty;
        relaxType       OUTLET;
        startX          ( 6 0 0 );
        endX            ( 7 0 0.05 );
        orientation     ( 1 0 0 );
    }
}
kev4573 is offline   Reply With Quote

Old   August 27, 2012, 11:31
Default
  #4
ngj
Senior Member
 
Niels Gjoel Jacobsen
Join Date: Mar 2009
Location: Copenhagen, Denmark
Posts: 1,900
Rep Power: 37
ngj will become famous soon enoughngj will become famous soon enough
Hi Kevin

I see, then please disregard my answer, but I would recommend that you make your test cases more eligible in the future, as confusion will definitely occur otherwise.

Looking at the right values, your setup seems sane, so I will not be able to offer you any suggestions beside whether you have tried using a cell aspect ratio of 1 around the free surface. In paraview it looks more like 1:1.5. I have found tremendous improvements in the results using 1:1 around the free surface.

/ Niels
ngj is offline   Reply With Quote

Old   August 27, 2012, 12:41
Default
  #5
Senior Member
 
Kevin Smith
Join Date: Mar 2009
Posts: 104
Rep Power: 17
kev4573 is on a distinguished road
Niels,

Thanks for taking a look at my case. Setting the cell AR~1 does improve the results, although the wave is still abruptly stopped near the beginning of the outlet relaxation zone, causing it to break and ultimately reflect waves back into the domain.

As is, this case uses relaxation zone lengths (input and output) twice the input wave length (lambda=0.5m). I have seen improvement by doubling the zone size. Maybe I need to extend the outlet zone further, and possibly use a cell AR~1.

Best regards,
Kevin
kev4573 is offline   Reply With Quote

Old   August 27, 2012, 13:02
Default
  #6
ngj
Senior Member
 
Niels Gjoel Jacobsen
Join Date: Mar 2009
Location: Copenhagen, Denmark
Posts: 1,900
Rep Power: 37
ngj will become famous soon enoughngj will become famous soon enough
Hi Kevin

To me it sounds like your time steps are really, really small due to some large velocities. This will mean that relaxation is applied many times in the very first cell of the relaxation zone, e.g. building a shock in alpha. This, obviously, cannot be sustained and it looks like wave breaking.

I would like you to try the Courant-dependent relaxation flag. Its implementation in waves2Foam is to my knowledge still untested, but it might be exactly what you are looking for. Merely add the following line to your relaxationZone definition:
Code:
courantCorrection on;
/ Niels
ngj is offline   Reply With Quote

Old   August 27, 2012, 13:40
Default
  #7
Senior Member
 
Kevin Smith
Join Date: Mar 2009
Posts: 104
Rep Power: 17
kev4573 is on a distinguished road
Niels,

I have tried this option and it did improve the results - see attached. These are plots of surface elevation vs time in the middle of the domain. The runs were done with a cells ~0.75 (same as previously uploaded), though I've run the same case with the courant correction on a mesh with cell AR~1 and I get the same result.

When the courantCorrection is on, it looks like a different frequency wave is reflected. The reflected harmonics are not apparent in the surface elevation time history until a later time.

Kevin
kev4573 is offline   Reply With Quote

Old   August 27, 2012, 13:52
Default
  #8
ngj
Senior Member
 
Niels Gjoel Jacobsen
Join Date: Mar 2009
Location: Copenhagen, Denmark
Posts: 1,900
Rep Power: 37
ngj will become famous soon enoughngj will become famous soon enough
Hi Kevin

Interesting. If the problem is indeed too large velocities in the air, then could you try scaling your problem to say 10 s wave, very large depths and 0.5 m wave height. This will hopefully make the wave kinematics more dominant relative to the air velocities.

/ Niels
ngj is offline   Reply With Quote

Old   August 28, 2012, 15:00
Default
  #9
Senior Member
 
Kevin Smith
Join Date: Mar 2009
Posts: 104
Rep Power: 17
kev4573 is on a distinguished road
Niels,

Thanks for the suggestion, I may try that out.

After post-processing the velocity field, it looks like there are some non-physical ripples being generated in the relaxation zone, which seems to be where the high frequency content originates. I cranked up some of the solver settings, which did give better results. There is still some reflection, but not as bad as before. I may also try using an upwind interface compression to see if that helps.

By default I think the below settings use 1 outer corrector and 1 nalphacorr.
Code:
PIMPLE {     nOuterCorrectors 2;     nCorrectors      2;     nNonOrthogonalCorrectors 0;      momentumPredictor no;     cAlpha 1;     nAlphaCorr 2;     nAlphaSubCycles 4; }
Attached Images
File Type: png alldata_stokesSecond_0.5_0.025_mesh4_int.png (92.4 KB, 58 views)
kev4573 is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Domain Imbalance HMR CFX 5 October 10, 2016 06:57
CFX does not continue Shafiul CFX 10 February 17, 2011 08:57
Axisymmetric Case with No Physical Boundary for Rotation kamkari OpenFOAM 0 December 24, 2010 05:41
Boundary conditions? Tom Main CFD Forum 0 November 5, 2002 02:54
Boundary Conditions Jan Ramboer Main CFD Forum 11 August 16, 1999 09:59


All times are GMT -4. The time now is 20:56.