open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,
I have some problem with a model I set up. The domain consists of a river debouching in a big reservoir. The river is flanked by tidal flats.

The tidal flats are 5 m above mean sea level and the water surface in the reservoir is at mean sea level. The river discharge is 1000 m^3/s. All the other 3 open boundaries have a fixed water surface (= 0 m ) condition. I will change the tidal flats elevation in different simulations (so I clearly do not want to get rid of those elements and use wall boundary conditions). In this particular simulation, the tidal flats are not flooded, since the backwater effect raises the water level in the river only of few tens of centimeters. I followed the suggested options at pag 85 of the manual for tidal flats treatment, and I copy them here for your convenience:

VELOCITY PROFILES = 4;0
TURBULENCE MODEL = 1 VELOCITY DIFFUSIVITY = 2.
TIDAL FLATS : YES
OPTION FOR THE TREATMENT OF TIDAL FLATS : 1
TREATMENT OF NEGATIVE DEPTHS : 2
FREE SURFACE GRADIENT COMPATIBILITY : 0.9
H CLIPPING : NO
TYPE OF ADVECTION : 1;5
SUPG OPTION : 0;0
TREATMENT OF THE LINEAR SYSTEM:2
SOLVER:2 PRECONDITIONING:2
SOLVER ACCURACY = 1.E-5
CONTINUITY CORRECTION : YES

and i added to these the option "MASS-LUMPING ON H = 1" that is mandatory since "TREATMENT OF NEGATIVE DEPTHS = 2" (otherwise it stops with an error). The problem is that in this case elements that are at +5 above mean sea level are flooded when they should not. I noticed that everything works perfectly when using only this "personalized" set of command:

TIDAL FLATS = YES
OPTION FOR THE TREATMENT OF TIDAL FLATS : 1 / DEFAULT
TREATMENT OF NEGATIVE DEPTHS : 2
TYPE OF ADVECTION : 1;5
SUPG OPTION : 0;0
CONTINUITY CORRECTION : YES
MASS-LUMPING ON H = 1

but when I add TREATMENT OF THE LINEAR SYSTEM:2 i have the problem again. This in the case of a regular triangular mesh (mesh obtained from a structured mesh dividing each rectangular elements in 2 triangles). When I use a totally irregular triangular mesh, also my "personalized" set of commands fails and tidal flats are flooded. I
attach a snapshot of the velocity field in each case, a figure with the meshes and the files that i used to run the simulations. Any suggestion?

thanks
Alberto

Hello,

Yes, this is a well known problem if you have a fast transition from wet to dry, especially if you have flat dry beds. The point is that if a somewhat spurious flux brings a millimeter of water on a normally dry zone, it should then go back with the effect of the slope. If your point is on a crest water can flow down the other way, if it is on a flat bed the water will just stay. So dry slopes should have a sufficient number of dry elements to avoid such a situation. I hope to have a better solution in version 6.2, if I am given time...

With best regards,

Jean-Michel Hervouet

Hi,

To reduce the risk of occurrence of such problem, here some advices:
1- as said by Jean Michel, try to adapt your mesh in order to obtain better description of tidal flat areas. Give particular attention to steep gradient areas which have key effects on flooding and drying behavior.
2- If you prefer to not readapt the mesh, a smaller time step could give you less preblems
3- Depending on the nature of your application, you can adapt the value of the velocity diffusivity. A high value dissipates large recirculations and tends to diffuse the velocity field which could probably increase the effect of flooding of dry elements in the vicinity of the shore line. A smaller value of diffusivity could be more adapted for your model.

Kind regards.

Riadh.