open TELEMAC-MASCARET The mathematically superior suite of solvers

Good morning

I am modelling sedimentation in a reservoir due to suspended transport of cohesive sediments. I noticed that in some consecutive instants of my simulation I have an impressive amount of resuspension (see figure 1 in the attachments, the print screens are taken 2 days apart). First I thought that it was due to the erosion stress, but doing several tests varying this parameter I noticed that I always have this problem (in more or less extend depending on my parameters).

Analysing the results I noticed that I have a huge peak in my discharge inside my reservoir, but this peak is not caused by my inflow boundary condition. In figure 2 in the attachments you can notice that the hydrograph in my boundary condition (red line) is perfectly fine, but in a cross-section 700 m downstream, these huge unphysical discharges are presented (green line). These peaks coincide with the resuspension of my sediments.

I couldn't find any triggers for this peak that looks almost like a dam break. As far as I noticed, there are no water level differences right before the peak triggers, nor a continuous sediment barrier across my river that make the effect of a dam. Also, I noticed that the peak increases a lot in sections downstream, reaching almost 3000 m3/s.

I ran only the hydraulic model and it is working fine, the problem arises when I couple telemac2d with Gaia.

I was wondering if you have any suggestions of why this happens. I attached my steering files.

Thank you for your time.

Kind regards,
Eduardo

Good afternoon

I still have not been able to solve the problem but I have some updates that maybe can help.

I think I have discovered the cause of these oscillations, as it is related to a second problem the model has. As it can be seen in attachment 1, in some loose nodes of my model I have a lot of deposition, with unphysical values.

From a post I read in the forum (www.openmascaret.org/index.php/assistanc...sisyphe?limitstart=0) they say that this unphysical sedimentation is due to an overestimation of velocities when we have an element with wet and dry nodes, and we use the finite element method. This problem per se is not the main limitiation for me, since I am interested in the general sedimentation pattern and this phenomenon is limited to some loose nodes, but I think they do induce the peaks in the hydrograph because of the huge topographical gradient they create between the nodes of the element.

I tried to refine the mesh in the areas of the tidal flats as suggested in the forum to reduced this sedimentation, but it didn't work. I also try changing to final volumes but the increase in computational time make it completely unfeasible for the project.

I tried to reduce the free surface gradient compatibility to zero, and the peaks in the hydrograph reduced a lot (see second attachment) but they have not disappeared completely. I also noticed that when I used the scheme for advection of velocities of 1 my results are better than with 14.

So I think there are two possible ideas to solve the problem: avoid the excessive deposition in the loose nodes, or somehow avoid that the topographical gradients created by these depositions affect the hydrodynamics, but so far I have not been able to implement these ideas in solutions.

I was wondering if there are any suggestions of how to solve this problem. Any ideas will be highly appreciated. I attach my steering files.

Thank you very much for your time!