open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear Community,

I am in trouble with an overflow that disappear.
The discharge is estimated at 200 l/s (0.4 m/s). Mesh size is around 2.4m and roughness at 40 (Strickler).

See picture attached for an overview of my problematic.

I tried to reduce the time step (from 0.1s to 0.01s) with no improvement.

I have searched explanation with no success so far.

would you have an explanation / solution ?

Kind Regards
Maxime F.

Hi
It's not easy to understand your problem with few informations...
Time step is not the only parameter of telemac.
It could be great to have all the file of your simulation to try to understand where the problem comes from...
But at a first basic level, you should probably check all the parameters dealing with tidal flats and associated schemes...

Regards

Thanks for your reactivity !

I add the cas file herein.
I used the solver method 1 (conjugate gradient method) with wave equation.
Types of advection are method of characteristics (1) for H and PSI distributive scheme, mass conservative (5) for PSI.
I set type SUPG option at 2.

I hope those informations would help you to understand.
Best regards

Hi
No files
but takes time to look into the manual and also into some other posts to see all the possibilities you have with the accurate schemes for tidal flats (14 or 15)
There is also dedicated keywords for water depth thresholds ....

Regards

I worked on option 13 or 14 ("CAS-14" file attached). The solver do not converge. I tried different options (time step, solver accuracy, option for the diffusion of velocity...) but I didn't get improvement.

Are there recommendations somewhere for numerical parameters with "edge by edge implementation of scheme" type of advection ?

ps : The first cas file is the one I forgot to attached.

Best regards

Hello

There are still no files.
Your explanations are not very clear. I'm wondering if it is simply a post-processing problem. With the some choices of colours, you can have areas with no color at all (white).

with my kind regards

Riadh

Sorry, here are cas fille.





About post processing, I set color range to only get colored areas for positive depths.
I had a screen capture setting yellow for negative depths.



Anyway we should be able to see velocity vector, isn't it ?

As I said, I tried type of advection 14 with different parameters. The solver didn't converge in 150 iterations (accuracy required at 1E-4). I didn't manage to get usable results.

Best regards

Hello,

What is the maximum Courant number of you model? Have you tried to set the option of VARIABLE TIMESTEP?

Regards,
Costas

I don't know my maximum Courant number. I guess it is much less than 1 as minimum mesh size is around 1.5m and time step is setted at 0.1s.

Using time step variable doesn't seems to solve my problem.
I attached sreenshot of solver information. It looks like there is no discharge in the model...

Best regards

Hello Maxime

If you look at the listing, it seems that you have lots of trouble with your culverts. That's why the solver does not converge and the changes in the advection schemes didnt fix anything.

First of all, remove all the culverts from the model and see if it works well. Then add the culverts one by one and take care about the parameters you are introducing in the culverts files.

I hope that this helps

With my best regards

Riadh