open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am in the process of completing three way coupling between Telemac-2D, Tomawac and Sisyphe for a small nearshore domain around an existing harbour. Boundary conditions for hydrodynamics and waves are obtained from global models, and vary in time and space. The coupling between Telemac-2D and Tomawac works fine (I can see that wave induced currents are properly added to wind generated currents). So far so good.

Simulations for three way coupling for my nearshore domain for three different wind storms are ok for two of my three cases. The problem seems to lie with the storm that causes a set-down in the nearshore domain of the lake (water levels drop from +0.6 m to -0.25 m during the storm peak, then goes up to +0.6 m at the end). The error seems to come from Sisyphe (as coupling Telemac-2D and Tomawac for this case worked fine). The hint of the error is given below:

I have checked, and I see that I don't have negative depths on the boundary of my domain. I have even artificially lowered the bathymetry in the domain to make sure the maximum bottom elevation is at -0.5 m, thus ensuring water on all nodes in the domain, even during peak storm. Same error again.

I read on the forum older posts related to negative depth prescribed on boundary. Older posts talk about changes in keywords from an older Sisyphe version, and an improperly defined rigid bed.

I have tidal flats turned on in all models, so I think it can't be the source of the problem. My Sisyphe model has a non-erodable zone only on the offshore boundary nodes. My Sisyphe steering file is attached, as is the boundary conditions file.

Does anyone have a suggestion on how to address this problem?

Thank you,

Pat

Hello,

The subroutine that "speaks" in this case is positive_depths from library BIEF. When called by Sisyphe, the "depth" is the sediment height, that cannot be negative. As the sediment height is (bottom - non erodable bed), I would look how is defined the non erodable bed, there is something strange here, maybe the boundary condition does not respect this principle of sediment height that should be positive (or we have to clip it somewhere).

With best regards,

Jean-Michel Hervouet

I missed the comment
BEDLOAD_SOLVS_FE
So the remark below is not correct apart from the advise to avoid a changable bed combined with prescribed velocities.

Not sure how the boundary condition deals with the freedom to change the bed, but not erode it because the bed is non-erodable.

Hi Pat,

I am not sure about this, without looking at results file. The message suggests that part of the boundary accretes over time, and then reaches zero water depth, whle you are still prescribing a flow velocity through that boundary.

However, you say that that is not the case. Could it be that this happens after the last time the results are stored?

In general, I would try to avoid prescribing velocities on a boundary, while allowing the bed level to change as ou are doing with litbor=4 in the cli-file. If the velocity gradients allow sediment to settle, the bed will accrete and there is noting to stop this to continue.

Hope this helps,

Michiel

Hello JMH and Michiel,

Thank you for your help.

I originally assigned all offshore nodes as non-erodable (i.e., locations where velocity is imposed). I would also like to make the offshore nodes not able to accumulate sediment (in addition to being non-erodable). Do you know how I can do that?

I thought setting LITBOR=5 might do the trick, but no.

Pat