open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello everybody,
I am working on my first telemac2d project and I would like to know how to check the reason why my simulation doesn't reach the steady state.
My objectif is to simulate the extent of flood around a coastal and multi-tributary lake. My domain contain a delta formed by two rivers that flow into the lake. This lake is open to the ocean, is also fed by a stream and is opened onto another lagoon.
Initially I run a steady state using only prescribed elevations for the four boundaries. The steady state has been reached in 2 hours but the result was not good in term of overflow (flood) extent. When I use the prescribed flowrates upstream and prescribed elevations downstream, the steady state is not reached even after 1 day.
I need help to know what is going on and how to fix the problem.

Thank you and Best Regards,
Marleine

Hello Marleine,

As you seem to be new, have you read the rules of this forum:
opentelemac.org/index.php/kunena/help
in particular #7.

Please upload your steering files, potential Fortran files + output files.

How can you say that the steady state is not reached or is reached?

Chi-Tuan

Hello Chi-Tuan,
Thank you for responding.

About the steady state, I chose YES for the key word STOP IF A STEADY STATE IS REACHED. Then the simulation stops when steady state is reached ; when the simulation continues till the end of time loop I consider that the steady is not reached.

Please find attached the geometry file, the boundaries file, the CAS file and the result file of my last test.
The semi-circular southern part of the estate is entirely in the Atlantic Ocean.
I am sorry to not take into account the rule 7 before.

Thank you and Best Regards,
Marleine

Hello,
I am unable to post the result file. Maybe it is due to the size of the file which is about 9 Go.

It seems that the geometry file was not loaded. So I add it here.

Thank you.

It seems that the geometry file was not loaded. So I add it here.

Thank you.

Hi Marleine,

I would recommend specifying the keyword STOP CRITERIA that is commented out in your .cas file. As mentioned in the manual, the default values are 1.E-4 which is 0.1 mm or mm/s (depending on the case), which seems to me a tight tolerance when modelling the ocean.

I would suggest using 1.E-3 (1 mm) or even higher across all values (as per the T2D user manual section 5.1), but anyone with more expertise may suggest otherwise.

Hope this helps,
André Renault

Thank you Renault,
I have taken your suggestion into account. I will come back to share the result.

Best Regards,
Marleine

Hello Marleine,

I have never used this keyword but my first idea is the same as André's.
If you are still not satisfied, I would suggest you to comment this keyword and run a long simulation and decide after when the steady state appears depending on criteria you choose.

Moreover, some extensions are not allowed on this forum (e.g. .exe) and a few are allowed (e.g. .zip, .tar, .txt, .cas, .cli, .slf I think, maybe others, I do not have the whole list in mind).

Hope this helps,

Chi-Tuan

Hello,
Thank you Chi-Tuan and Renault,

The files that I could not load are selafin (.res and .slf).

About the simulation, I tried with two values for STOP CRITERIA.
The first value 1.E-3, does not change my result compare to the previous running : the simulation reaches the end of time loop without the steady state having been declared; however, when i check the graphs of the different outputs, i notice that they become more or less constant after the 18th hour of simulation (much earlier for some outputs) but there is no convergence between inflows and ouflows.
With the second value, 1.E-2, the steady is declared at the 38th minute of simulation, whereas the graphs of outputs show no stabilization.
Currently, I am running a 10 days simulation with 1.E-3 as STOP CRITERIA.

In my case can I consider the steady state reached once the outputs get stabilized despite the non-convergence of ouflow and inflow ?

Thank you and Best regards,
Marleine