open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear All,

I am using T2D for simulation through the attached stretch of the river. In the attached picture the variation of bed elevation in the stretch is shown. The domain covers the main channel and the flood plain. I want to patch initial depth of 50 cm from the bed all along the stretch only in the main channel. Any suggestions how can I go about that.




Thanks in advance,

Best regards,
Sumit

Picture of the bed

Hi

This could be done in fudaa (editor) if you activate initial conditions.
You could define (or import) a line which follow your river with an initial value for each node.
From this line, fudaa automaticaly define cross-sections and interpolate the value to all the nodes

I suppose this could also be done in BK if you define the free surface variable...

Regards

Hello,

Another possibility that will take only computer time is to do a first run, starting from a dry bed, to fill the river.

JMH

Hi,

But if I start with completely dry bed than it gives me an error, or am I doing some thing wrong ? So with complete dry bed the problem will be having one upstream flow and downstream bc will be changed to the code 4 4 4 right ??

Yes, you will have to put some water at least at the entrance, so that the program can compute velocities that will give the correct discharge (otherwise without water there is a division by 0). At the exit during the flooding you can try 4 4 4 (it may crash, in this case put Thompson bundary conditions at the exit).

Ragards,

JM

Hello JMH,

This is again in regard with the stretch of the river that I am trying to model. At the upstream side I have a constant flow of 5.7 m3/s and have also prescribed a stage discharge curve.

At the downstream side I have the case of known depth and again I am constraining the downstream boundary with a stage discharge curve.

I just want to run the simulation for 24 hrs and fill the domain with realistic water depth. With the aforementioned simulation set up, although the inflow is only 5.7 m3/s the outflow becomes more than 40 m3/s after 12 hrs.

I have simulated the exactly same case with HEC-RAS and I did not have this problem. Do you have any suggestion for me ?I

Hello,

You should not have a stage-discharge curve while you fill the river, just a free exit (or with Thompson as we said), until you have a reasonably steady state flow.

Regards,

JMH

Hello JMH,

Once again the way that I have set up this problem is as follows

a) Upstream boundary is inflow boundary with prescribed discharge and I am also specifying stage-discharge table for this boundary

b) Downstream boundary is coded as 4 4 4 and I have included thomps.f subroutine in the user subroutine.

c) Bit of initial water depth is specified through condin subroutine.

But I have still the same problem, for the specified inflow of 5.7 m3/s the outflow becomes 50 m3/s after 12 hrs. Do you have any more suggestions ?

Thanks a lot for all the help.

Best regards,
Sumit

Hello,

The program is exactly mass-conservative, so if you wait for a steady state you will get the same discharge at the entrance and at the exit. However make sure that you have the keyword :

CONTINUITY CORRECTION : YES

it is useful when you have a prescribed depth at the exit (otherwise the continuity equation is not solved at the exit because it is replaced by equation depth = prescribed depth).

Another thing: Thompson is not obtained by adding the subroutine thomps.f in your Fortran, but with the keyword :

OPTION FOR LIQUID BOUNDARIES : 1;2

here it is an example with 2 boundaries and you want Thompson on the second.

With best regards,

Jean-Michel Hervouet