open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello Laurent,

thanks for this very detailed answer it helped me a lot. Right now I'm facing issues regarding wave breaking, since this seems quite unrealistic to me.
I tried to use a different time step to improve the situation but without any success. Do you think it would help to work on the mesh?

I attach the model in any case you have the time to take a look at it.

Kind regards,
Clemens

Hi Clemens,

First, your mesh as many points with only 2 neighbours. It can be a problem: for instance, your configuration doesn't run on 16 processors on my computer, due to partitioning problem.
A large area of the northern part of your mesh is useless for this case.

Second, you're not considering breaking effects since the keyword "CONSIDERATION OF SOURCE TERMS" is not turned on. Then you'll be able to play with the maximum ratio Hs/D and with the formulation of the depth-induced breaking dissipation.
Maybe you can also decrease your time step (10 seconds should be ok) and I think only 30 minutes of simulation are enough to propagate your stationary conditions over your domain.

Regards,
Laurent

Hi Laurent,

thank you for the answer. I could improve my mesh in this area with a higher detail, but actually this area is not that important for my research. It runs on my computer, but do you think it could cause problems close to the inlet?

Regarding the keyword "CONSIDERATION OF SOURCE TERMS" I totally missed it but its turned on and I will work my way through from here.

Another question regarding the wave direction. I have 36 as NUMBER OF DIRECTIONS and wonder if it is possible to start with e.g. 34 for 340° as θ1 and 2 for 20° for θ2?

Kind regards,
- Clemens

Hello Clemens,

you can't choose the directions that way your circle is divided by ND (number of directions) thetai = (i-1)*360/ND as you can see in the manual p 42 and 43.

regards

T.

Hello,

thanks for the help. I already read through this but still I am not sure how to tell TOMAWAC that my waves propagate in between 330° and 20°, since this would go through 0°.
Thanks for help.

Regards,
Clemens

Hello

Unfortunately you have to discretise directions even if you know that there won't be energy in those. The code has been thought with a uniformely discretized directions and it would be a real big work to change that.

That is about discretisation, but if you want to set direction in your boundary spectrum or initial spectrum you can play with the ANGULAR DISTRIBUTION FUNCTION (3 choices) and the DIRECTIONAL SPREAD (s) to be more or less concentrate around your central direction.

I attach a python script that allow to see all the choices. s=1 for the second function shouldn't be allowed because it is assumed that you have 0 on the boundary.

Hello,

thank you very much. Just another question. How does the boundary conditions interact with initial conditions?
Are the initial conditions set for the whole mesh when the model starts and the boundary conditions just act at the set boundary?

Clemens,

If you read the whole discussion of this thread, the answers to your questions have already been given in the thread #29444 and detailed in the user manual.
Moreover, the terms "initial conditions" and "boundary conditions" are rather explicit to understand what they represent.

Laurent