open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

What is the best way to set up a river boundary, to avoid velocity-violation errors due to boundary nodes drying up?

I have solved this in the past by enforcing water depths at the boundary in BORD (e.g. HBOR(#)=2), but this time it dosent work.

Is there a better way?

Thanks,
Peter

Hello,

On entrances with prescribed flowrate you can also use the keyword:

VELOCITY PROFILES

(one value per liquid boundary)

The default is 1 but 4 and 5 are two variants that give a profile proportional to sqrt(depth), which naturally gives 0 on dry zones.

Regards,

JMH

Hi,

Thanks a lot for the quick replies.

I changed 'VELOCITY PROFILES' (from 1 to 4) in the cas file and there was no change in error:

ITERATION 0 TIME: 0.0000 S
DEBIMP : PROBLEM ON BOUNDARY NUMBER 3
GIVE A VELOCITY PROFILE
IN THE BOUNDARY CONDITIONS FILE
OR CHECK THE WATER DEPTHS

Riadh, I don't really understand your response (sorry!). I have an off-shore tidal boundary (works fine) and lots of dry areas (TREATMENT OF TIDAL FLATS = 1), but here my problem is a river bouldary further up the catchemnt.

Any more suggestions?

Thanks again,
Peter

Hello,

Actually you need to have water at an entrance where you want to prescribe a flowrate, otherwise the velocity to ensure the discharge will be infinite. If you can give a non zero depth at the entrance in the initial conditions it will probably work. Another possibility is to modify the bottom at the entrance so that there is always water.

JMH

Great,
Working fine now.
Thanks a lot!

Hi ucs803,
I had similar problem at upstream flowrate boundary of my model. Can you tell me what worked for you fine?
Thanking you in advance!

Hi,
There needs to be water at the river boundary at time step 1 (when I forced water depths using HBOR in BORD.f, maybe this hadn't been read eary enough).

So when I forced water depths (e.g. 1 m) at river boundary nodes (i) in CONDIN.f, it worked fine.
e.g. H%R(i) = 1

Hope that helps,
Peter

Hello

What do you mean by "river boundary"?
In the past, you enforce water depth, that's mean that you have a physical boundary condition (hbor=2) or is is purely fictitious condition? in this case, the generated velocity is not physical.
In general, in 2D model the river is located far enough from lateral boundaries which are dry in most cases (with solid boundary condition). Telemac will handle drying cells depending on the option you've introduced in your steering file and depending on the kernel you are using (finite volumes or finite elements)

In finite volumes, the drying is automatically tackled without any user prescription (by the called hydrostatic reconstruction). In Finite elements, several options are possible :
TREATMENT OF THE TIDAL FLATS :1 ==> specific treatments to avoid any division by h,this could generate negative water depth and a loss of mass especially if you use the clipping option.
TREATMENT OF THE TIDAL FLATS :2 ==> drying elements are removed from computation (you need to define a threshold value MINIMUM VALUE OF DEPTH)
TREATMENT OF THE TIDAL FLATS :3 ==> like option 1 but using a method based on porosity.
You can see the user manual that gives you more details connected to the use of option 1 and 3.

I hope that this helps

My best regards
Riadh ATA