open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

I am working in a model that blows up after little more than one month. It does not crash but its results become useless after that time.

In these figures is the model before and after its bed evolution -and all sediment results- blow up.





The warning DEPOSITION MORE THAN ONE LAYER THICKNESS appears shortly before the collapse of the model. Here is an image of what happens at two nodes when the warning pops up (21221 at the shore, and 18485 at the open sea).



The question is can this issue be solved without decreasing the timestep? Despite this is what is recommended in the manual I cannot afford to do so, this model is already too CPU demanding.

Thanks in advance for your help, regards, Sebastian.

Hello,
I took a fast look to your files. Firstly, your steering file for Sisyphe looks a bit messy: you are activating all the sedimentary processes in your file! Please comment the ones that you are not using before proceed (it's not clear for me what do you want to solve: mixed, suspended, cohesive/non-cohesive, bedload?)
Secondly, for morphological factor is probably too high for the problem you want to solve.
Please clean up things, try a smaller MF and let me know.

Best wishes,

Pablo

Hi Sebastian,
As your blow ups are happening at the corners of the open boundaries, I suspect it has to do with the boundary conditions. Is the boundary open at the mesh points with shallow water depth?
Other issues:
You have turned of equilibrium concentrations. How are the concentrations at the boundary set?
Why have you turned of the tidal flats in the sediment model (but not in the flow model)? This might lead to very large transport rates if the water depth is small and the velocities are not.

Hello mafknaapen,

The open boundary is indeed close to nodes of high elevation, and several of such nodes are near that eastern corner where the model first blows up. Do you think setting the open boundary farther from the shore could be of help?





At the open boundary H, VX and VY are imposed. Regarding sediments, no boundary concentrations are imposed, thus all sediments on the domain come from what is available at the bed and the initial suspended concentrations.

So far I have tried turning off tidal flats and restricting depths to positive values. However I had not noticed settings where different in both steering files. Thanks for that remark, I will also try fixing this.

Thank you so much for your help,

Sebastian

Hi Sebastian,

yes, do try to avoid wetting and drying on the open boundaries. As you have tidal flats in your model (wetting and drying), don't turn it off for sediments. This does create problems, and might create sediments out of nothing. Turning it on will only marginally slow down the model.

Thanks mafknaapen,

The model seems to run well now, I just excluded beach points from the open boundary and it did the job.