open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everybody,
I got an easy system (straight channel with trapezoidal profile). And the flow is actually only in one direction. Actually the velocity profile is expected to be symmetrical and clearly formed.
Unfortunately, there occur small oscillations at the peak of the velocity profile, no matter where I look at the velocity throughout the channel.





Does anybody have an idea what could be wrong or what's the reason for that? When using other twodimensional software this oscillations do not occur.

I added the used files.

Cheers
Sarah

Hello,

try to use higher eddy viscosity like 0.5 or use the Elder turbulence model. Why this dip in the velocity profile happens... maybe the moderators can answer.
As far as I know programs like Hydroas use default constant eddy viscosity values of 0.5 to 1.0.

Your steering file has some inconsistencies, e.g.
you don't specify a time step or you want to apply friction for your walls in which case you should turn on the rough wall law.
Enclosed you will find an example steering file, a boundary condition file and a liquid boundary file which you can use for your case.

Hope this helps,
Clemens

Hello Sarah,

the attached steering file in the previous post may be misleading because I simply used an old project file but without updating some values and comments etc.
Please find attached a clearer and proper steering file to your case. Most of the values are default values but with this steering file you can check the most important keywords,in my view, with the help of the reference manual and they can be used for a lot of different applications.

Best regards,
Clemens

Thank you very much Clemens! That helped me a lot.
Unfortunately, the oscillations are the same when using the k-epsilon model. With the elder model there are no oscillations but the velocity profile is a little different and does not fit that well to the other models as with a constant or k-epsilon model.

Cheers Sarah

Hi,

I also made some tests with your case. I used constant viscosity, Smagorinsky and k-epsilon. I don't obtain as big oscillations as yours.



I have lowered a bit FREE SURFACE GRADIENT COMPATIBILITY from 0.9 to 0.5 but this is should not affect so much your velocity profile.
I also have IMPLICITATION FOR DEPTH and IMPLICITATION FOR VELOCITY = 1 (OK for steady state, might add some diffusion).

Good luck,
PL

Hi Pilou,
could you please upload the *.cas file you used. I don't get it.

Cheers
Sarah

Hi,

I attach it here. Note that I changed you downstream boundary condition to prescribed elevation. I also attach the .cli file.
Otherwise the parameters are pretty much generic, no significant differences with the examples provided by Clemens.

Regarding your question on Edler model, I can't help since I never used this turbulence model.

Good luck!
PL