CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Velocity spikes at interfase (interFoam) (http://www.cfd-online.com/Forums/openfoam-solving/69573-velocity-spikes-interfase-interfoam.html)

santiagomarquezd October 27, 2009 15:27

Velocity spikes at interfase (interFoam)
 
Hello all, I'm running interFoam under a OpenFoam-1.5 distro. Actually I'm facing problems described by Henry in #27 (http://www.cfd-online.com/Forums/ope...-solver-6.html)

Quote:

Look carfully at where the velocity is causing the maximum Courant number and why. I guess there is some level of numerical instability causing spikes in the velocity field which need to be delt with. Are these spikes at boundaries? near the interface? near corners? Are they reduced or removed by using upwind? Also how good is the mesh? Try running checkMesh on it if you are unsure about the quality
namely velocity spikes, they are near the boundaries and persists even using upwind schemes for U and gamma, by the way mesh quality is excellent. Ratio between Co_max and Co_mean is 250, in a quasi uniform mesh, it implies near a ratio of 250 in velocities too.
This spikes causes spurious deformations in the free surface that ought to be sinusoidal (is a mass pendulum, sloshing problem).

Any clues in solving?, I've been working in this for several weeks, without success...

Thanks in advance.

chiven October 27, 2009 20:34

Hi, dear Santiago, you can try to use setFields/funkySetFields to set different initialized fields to calculate your case. Hope it works.
Best regards,
Chiven

santiagomarquezd October 27, 2009 21:36

Chiven, thanks for your reply. If I understand your answer I think actually my problem is not the initialization. I initialize the problem by hand, setting a sinusoidal wave as the free surface. The idea is that this form remains in time but attenuating, oscillating as a mass pendulum. Problems arises few steps after the beginning, velocity spikes starts to show up at the intefase zone, then the sinusoidal is completely smeared out (see another post from me: http://www.cfd-online.com/Forums/ope...-sloshing.html). I know the problem is the indicator function advection or the compressive scheme, but I can't figure what parameter must I change.

Thanks in advance.

jploz October 28, 2009 05:01

Hi,

what value of cGamma are you using? I prefer values between 0.5 and 1.0, not higher. The cGamma coefficient determines the strength of the 'artificial compressive velocity' in the interface region. Too high values can distort the free surface.

HTH,
Jean-Peer

simt June 19, 2013 09:05

Anything new on this topic?

I also get very low courant no. sometimes when using interFoam (OF 2.1.x) for free surface ship flow, similar to the wigley hull case. Oddly, no extreme velocities are observed in the postprocessing though.

The low courant-"spikes" happens independent of :
  • numerical scheme (I've tried using upwind for all convection terms)
  • for cAlpha = 0 (no compession)
  • nNonOrthogonalCorrectors
  • even for low time step equivalent to maxAlphaCo = 0.5 & maxCo= 0.5 with nAlphaSubCycles = 5
So my questions are,
  1. Anyone know how to deal with this?
  2. Is there any alternative which is more robust (navalFoam to 1.6-ext, shipFoam, using MULES::Implicit instead of explicit as in interFoam etc) ?
Best regards, simt.

simt June 24, 2013 06:59

Got rid of the spikes by limiting grad scheme, cellLimited leastSquares 1.0;.

santiagomarquezd July 26, 2013 12:14

It's been a long time since I started this thread. I'm solving the sloshing again, an slightly different case and the spikes are also present. I'm doing a parametric analysis, going up in maxCo. Things were good for small Co numbers, then, when I increased the Co number the spikes appeared, they seem to be related with a deficient p-U coupling, so that the problem desapeared increasing the PISO corrections and activating the mometum predictor.

Regards.

jasonchen September 26, 2014 18:04

Quote:

Originally Posted by santiagomarquezd (Post 442215)
It's been a long time since I started this thread. I'm solving the sloshing again, an slightly different case and the spikes are also present. I'm doing a parametric analysis, going up in maxCo. Things were good for small Co numbers, then, when I increased the Co number the spikes appeared, they seem to be related with a deficient p-U coupling, so that the problem desapeared increasing the PISO corrections and activating the mometum predictor.

Hi Santiago,

Thanks a lot for your work on description of interFoam solver. Did you get good comparison now for the sloshing case, at the end of the tutorial? In another thread, you mentioned: "This spikes causes spurious deformations in the free surface that ought to be sinusoidal (is a mass pendulum, sloshing problem)."

I also found that the velocity in the air may be so large that the wave surface is distorted. Do you solve your problem by reducing maxCo and increasing PISO correctors?

Thanks,
Jason


All times are GMT -4. The time now is 09:11.