open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,
I want to simulate overtopping of river bank and generated a mesh with river bed at 0m and the wall 0.2m high as shown in Fig. 1 and 2. The initial water level is 0.15m and the simulation runs without errors. However, the water level seems to pass through the wall rather than overtopping as shown in Fig. 3 and 4. Is it a problem because of the mesh? I have attached cas and slf files for perusal. Thanks!

Ananth.

Hi Ananth,

I am not sure what is happening in your case, but you certainly can improve your mesh to model water overtopping. I suggest the following:

1. In BK, use 4 parallel softlines to fix elements at the edges of the wall (front foot, front crown, back crown, back foot). The element size should be appropriately small to fit between the lines.
2. Use at least 3 points across the crown width and make sure that the middle point has slightly bigger elevation (like a caret ^) to make sure that water flows back or forth.

Best Regards,
Costas

Hi Costas,

I have now tried defining the river wall with 5 softlines, two at the edges and three at the crown. But the results do not improve. Is this because of the steep slope?

If this doesn't work, My next option is to use a weir as the river wall.

Is there another way to simulate overtopping using islands or hardlines?

Can you post the new geometry file to have a look?
Costas

Hi Costas,

I'm attaching a simple case of dike I tested with 5 or more softlines at the structure. Still the results were similar with water dropping through the structure as shown in below screenshots. Please suggest me any improvements. Many thanks!

Ananth.

Also, I only ran a case without any inflow and outflow expecting no change in water levels.

Hi, the problem is clearly with the mesh. At the ends, you have poor definition and 'holes' appear. Also when I meant the crest should be like a caret '^', I meant that the middle nodes should be something like 1% higher than the adjacent ones.
Here is an example I drafted having my morning coffee:



Here are the files used:



In order for it to work, I had to change the interpolation method to linear. Nearest neighbor should also work, but 'inverse distance' doesn't. As a matter of fact, I never use it.
Costas

Hi Costas,

Many thanks for the tip to use linear interpolation, the mesh definitely looks clean with that. However, in my case I was able to find out why there is a drop in water surface at the dike during simulation. Defining initial condition as "Constant depth" forced the water depth at the dike to drop to zero during the simulation, whereas using "Constant elevation" did not cause this issue and by default the water depth at the dike was zero at the beginning of the simulation.

Cheers,
Ananth.