open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello!
I've a new question for you
I'm imposing at the offshore boundary (OB) velocities from a subregional model.
At each layer of the OB, velocities are read from a file, and vary in time and space.
The external file are read properly.
I'm attaching the results. The first line are the velocity imposed at the boundary at the first time step.
The second line reports velocities at the second output time step (i.e., 30' after, but also in the following) in a node nearby the OB.
Why is the profile always log type? Maybe I miss something in the theory or in the cas file keywords.
I'm also attaching the cas file.

Any advices?
Thanks.
Regards

Gabriella

Hello,

Look at the 5 different options of the keyword : VELOCITY PROFILES. There is also a different behaviour if you put 5 (KENT) on the velocities or 6 (KENTU). 5 is for prescribed discharges and 6 is for prescribed velocities, in the latter case the function VIT3 is called. Then it depends on how you have implemented subroutine bord3d.f.

With best regards,

Jean-Michel Hervouet

Which is the meaning of VELOCITY PROFILES keyword? I mean where and why it is called..
And which is its default value?
I don't have a log or constant profile at all.
I used 4 6 6 as BC, i.e. prescribed velocity and free H at the offshore boundary.
In bord3d.f, i commented the call to VIT3 and imposed UBORL%R(IBORD) equal to Udata, and similar for V component. The data are read correctly, I can also check it during the run at the OB.
But it seems that the velocity inside the domain is force to have a logarithmic profile....

Any suggestion?

G

Hello,

VELOCITY PROFILES corresponds to array PROFVEL in Fortran (1 value per liquid boundary), and the different options are explained in the user manual (if not clear just tell us...). However this is for option 5 of velocities (prescribed discharge). If you look at subroutine bord3d.f around line 362, it could be that you did not add the keyword PRESCRIBED VELOCITIES in you parameter file. The number of precribed velocities given by this keyword is NVIT, so NVIT = 0 skips the call to VIT3. In this case the values read in the boundary conditions file are taken. It is the presence of this keyword that allows to choose between : boundary conditions given by the boundary conditions file or boundary conditions given by keyword through a call to function VIT3. If you have programmed in the section where VIT3 is called, you might just add PRESCRIBED VELOCITIES : 0.;0.;0. ... with a sufficient number of numbers (as many as liquid boundaries).

With best regards,

Jean-Michel Hervouet

Thanks Jean-Micheal!
You were really clear, and I set exactly 0,0,0 for prescribed velocities, while I don't need to set the velocity profile, since I didn't prescribe discharge.
I probably solved my problem set the keyword TREATMENT OF FLUXES AT THE BOUNDARIES equal to 1. I'll keep this post updated..

Thanks.
Regards,

G

Hello,

Yes, this keyword TREATMENT OF FLUXES AT THE BOUNDARIES = 2;2... has the effect that the prescribed values are not immediately taken into account, to ensure the conservation of fluxes, but after some time you must see however the prescribed values.

With best regards,

Jean-Michel Hervouet

I see now.
Actually I don't run my simulation for a long period (just 2 days).
So I'll try to check the results with this option (TREATMENT OF FLUXES AT THE BOUNDARIES)=1, otherwise I'll need more time for the model "spin-up" period...
I'll update you. Hope to fix this problem.

Thanks.
Regards

G

Hello,
I still have this problem, and I need help to fix it.
I tried different options of solvers, turbulence and boundary types, but tel3d still gave me log vertical profiles of velocities in depth in nodes really closed to my Offshore Boundary, where data are read correctly.
What am I missing??
Getting crazy...

Thanks for your support.

Gabriella