open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello
I have a model that runs fine, with one reach and a few cross-sections as culvert (cf previous post).
When I add one storage area and a spill link to connect to the channel, the run fails quite rapidly with the error message at the bottom of this post. Even if the spill crest level is very high and no flow happens in the spill link, the run fails.
I am struggling to interpret the error message and make sense of this behaviour. Can someone help ?

Thanks
Julien


Erreur n0 104
Arbre d'appel :
=========================================================================
Erreur de calcul.
Erreur sur la largeur : ALARG = 0.000000000 au noeud : 159
pour les indices de planimetrage : 249 et 250

After some more investigation and running the Mascaret engine in debug mode, I have found that the error message is modified when going through the module M_TRAITER_ERREUR_I.
The original error message is as follow:
'Vous depassez le pas de planimetrage maximum : Le code diverge - verifiez le nombre de Courant'.
Hence now the situation seems clearer, the simulation is not stable.

There was some small unstabilities before I introduced the storage area and spill link, but they dampened. Somehow these instabilities now diverge when I add these two elements.

Julien

Hello Julien,

ok. The consideration of storage areas with the super-critical kernel may be unstable.

The next version of the software will only consider storage areas with the subcritical kernel with a new Algorithm from Artelia.

We are working to consider real 2D floodplains with the supercritical kernel.

Regards

Hi Fabrice

Are you able to explain why the engine is getting unstable in this case ?

This is quite a blow as I am using the storage area and the spill link to couple Mascaret with a 2D model using the API. Then in turn the 2D model is coupled with a damage calculation tool. The transcritical engine is one of the main reason why we use Mascaret.

Will the 2D floodplains in Mascaret (3.4?) be integrated with the API ?

Also Nicole had advised me to run the Subcritical engine with the additional keyword 'ATTENUATION DE LA CONVECTION' activated (as there are many transitions to supercritical flow in the river considered) to be able to retain the storage areas. But that didn't work.
Is there anything else you can suggest ?

Thanks in advance
Julien

Hello,

do you use Mascaret v7.2 (provided with FM v3.3). There is a new and more efficient algorithm for the storage areas and the 'reduced momentum equation' can be activated with the GUI?

2D floodplains coupled with the transcritical engine is on the 'to do list'...

Perhaps you can send us your case in order to be "able to explain why the engine is getting unstable in this case" ?

fzao

Hi Fabrice

I use Fudaa-Mascaret 3.2 at the moment (engine v 7.1.7).

What I meant by "Are you able to explain why the engine is getting unstable in this case ?" is the case of the storage areas with the transcritical engine (in general).
Or alternatively why in the latest version the storage areas are inactive with the transcritical engine ?

Thanks
Julien

Bonjour Julien,

Les casiers sont couplés avec cette version de MASCARET (noyau transcritique). Néanmoins, le résultat peut être instable si l'apport de débit lié au casier est trop important. Il faut vérifier la liaison et diminuer le pas de temps et éventuellement passer en implicite.

Si tu veux, envoie nous le cas qu'on vérifie.

Bonne journée

Nicole

Fabrice, Nicole

Here are the *.masc file (remove the pdf extension) and also the text input files zipped.




and



Thanks in advance for looking at this case.

In response to Nicole's last post, it is not an issue with the calculation of the flow over the weir/spill link, as even if I set the weir elevation to a high value that will stay higher than the water level, the run crashes at the very beginning of the simulation.
You can see that this simulation runs properly when the storage areas are deactivated.

Julien