Welcome, Guest
Username: Password: Remember me
Forgot your password? Forgot your username? Create an account
Forum
System Modules
TELEMAC-2D
Probably an error in setting the limits
Flat Threaded
  • Page:
  • 1
  • 2

TOPIC: Probably an error in setting the limits

Probably an error in setting the limits 5 months 3 weeks ago #44844

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
Dear all,

I was given a model ;)
I'm running it without any modifications. B)

I've regenerated the boundary conditions file (*.cli):
- The boundary conditions are not in the same order as the original file (but the correct nodes are taken into account)
- I've adapted the CAS file, the indexing of the rating curves (level/discharge law) ...
The model starts up and stops fairly quickly because some boundaries that are wet at the start of the model become dry.

It's in a river. how it's possible ?
Why is it that in the original model everything is fine, but regenerating the limits causes problems?

I double-checked the flow rates, dimensions and pressure curves several times...
The taring curves are managed in a Fortran. I updated the limits.


Original limit.clm :
2 2 2 0.0 0.0 0.0 0.0 2 0.0 0.0 0.0 759 26429
2 2 2 0.0 0.0 0.0 0.0 2 0.0 0.0 0.0 13196 26430
2 2 2 0.0 0.0 0.0 0.0 2 0.0 0.0 0.0 760 26431

mylimit.cli:
2 2 2  0.000 0.000 0.000 0.000  2  0.000 0.000 0.000       48036       26429   # 
2 2 2  0.000 0.000 0.000 0.000  2  0.000 0.000 0.000       46613       26430   # 
2 2 2  0.000 0.000 0.000 0.000  2  0.000 0.000 0.000       47874       26431   #
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 3 weeks ago #44847

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
I attached the log file (one one them ...).

File Attachment:

File Name: PE00009-00001.log
File Size: 47 KB


I have the famous following error message :
DEBIMP : PROBLEME SUR LA FRONTIERE     20
          DONNER UN PROFIL DE VITESSES       
          DANS LE :                          
          FICHIER DES CONDITIONS AUX LIMITES 
          OU VERIFIER LES HAUTEURS D'EAU.   
          AUTRE CAUSE POSSIBLE :             
          ENTREE TORRENTIELLE A HAUTEUR LIBRE
          METTRE UNE HAUTEUR NON NULLE       
          DANS LES CONDITIONS INITIALES      
          OU IMPOSER LA HAUTEUR D'EAU EN ENTREE.



 PLANTE : ARRET DU PROGRAMME APRES ERREUR
 RETURNING EXIT CODE:            2

It's not because I forgot to put water on my limits. It's because the initial xater leel varies after one time step, and my limit become dry ... But I don't know why.
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 3 weeks ago #44851

  • c.coulet
  • c.coulet's Avatar
  • OFFLINE
  • Moderator
  • Posts: 3722
  • Thank you received: 1031
If the water level varies at the first time step, this mean the local boundary condition is not consistant with the initial condition...
As this model sounds complex with many files and also user fortran adaptation, there is probably something which remains wrong in the changes you made...
Christophe
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44900

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
Hello,

I've taken the time to reply in order to check various points about the model.

I've started from the beginning :
  • Reference: a model that runs normally.
  • I took the mesh from the reference model, and regenerated the *.cli boundary conditions file. Note: the order of the nodes differs between the reference model file and the *.cli file I created. And therefore the order of the boundary conditions.
  • The files containing the liquid limits, the level/discharge laws and the fortran linking the upstream/downstream limits of structures (managed by the level/discharge laws and the fortran) are updated.
  • There must be an error somewhere, because the model is not working normally. But I'm sure I've got the right upstream/downstream limit pairs. After that, have I managed to transcribe them into Fortran? I'd say yes. But the results leave room for doubt.

I still get the error message about the torretiel limit or no water ...

I'm going to try again to generate an initialisation file but I'm going round in circles. I'm sure I'm getting the same error again and again somewhere without identifying it.

Can I share my model (the reference and the one for which I just used STBTEL to renumber the nodes and therefore redo the limits), so that someone can look at what might be wrong?
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44903

  • c.coulet
  • c.coulet's Avatar
  • OFFLINE
  • Moderator
  • Posts: 3722
  • Thank you received: 1031
Hi
Hard to spend time debugging your model for free!

Firstly a remark:

Why renumbering the nodes and by consequence having to regenerate the boundary condition file?

By the way, there is one point I don't understand in your trial:
You generate a .cli file from the reference geo file. as a first result, the reference .cli file and the new generated .cli file are different. Is it right?
Could we see the differences?

Regards
Christophe
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44906

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
I understand. I aksed ... :)

Why renumbering the nodes ?
Because I have to change something in the mesh. this updating implies to add a couple of boundaries. But it doesn't work.

Therefore, I chose to apply the steps I have to realize for my new mesh, but for the old one, permitting me to check some error ...



To answer you : yes, the reference *.cli and the new one are different :
--> without using STBTEL, if I create a *.cli for the reference modele, by myself, the node ID are not changed, but the order oh the boundary Nodes evolves. (the boundary 1 become the boundary 14 for exemple).
--> Using STBTEL first, the node ID changes ...

Whatever I do, I have the error.


This morning I started a "new" method. I'll see.
I'll let the community know about the approach that helped me. If I can find it! Or if it finds me! :p
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44913

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
About the differences, I attached the "REF" boundary file and my own.

No STBTEL thing. Just new > Boundary Conditions in BlueKenue.
Doing this, limits are mixed up.


Breaking news ... : I'm currently verifying the order of boundaries in my modified project. And the boundaries doesn't follows the numerotation order of the CLI file ! But the order of the reference file !
To check this, I activate boundary by boundary (29 ...), to see which one get water.
Until then, I exported the model nodes to QGis. I matched the nodes with the CLI (attribute join, using the node ID). So I was able to locate the limits. ...


How is this possible? Are the limits renumbered in Fortran? In this case, which subroutine, or whixh line ? I didnt see it.


Thanks a lot


laurie
Attachments:
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44914

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
Ok, I finish.

It appears that the problem is not the renumbering of the boundaries. In fact, it is the consequence of the ‘problem’.


This time I worked on the modified model. The one to which I added a pair of boundary conditions (Himp upstream and Qimp downstream of a structure). These two boundary conditions have been created in my *.b2c and *.cli. But the model ignores them. It only takes into account the limits of the reference model.

In the fortran, I read ‘MASK’ ... I understand that the limits are defined via a MASK. But I have no idea how to visualise these masks (are they shapefile polygons that locate the boundaries?), or how to modify them ...

Without this knowledge, I can't use the model!

Can you try to explain me ? Or help me to identify the section in an user guide I can find my answer ?
In a french version for V7P0, section 4.2.10, there is a ligth paragraf about MASKOB and MASKEL.
I don't understand how to use this tool :(
The administrator has disabled public write access.

Probably an error in setting the limits 5 months 2 weeks ago #44915

  • c.coulet
  • c.coulet's Avatar
  • OFFLINE
  • Moderator
  • Posts: 3722
  • Thank you received: 1031
Don't see any MASK reference used by user programmation in the fortran...
But in the beginning, there is some reference to boundaries (number and pairs)

Regards
Christophe
The administrator has disabled public write access.
The following user(s) said Thank You: Laurie

Probably an error in setting the limits 5 months 2 weeks ago #44916

  • Laurie
  • Laurie's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 97
  • Thank you received: 5
Sorry, I think I need to eat and sleep !

There was a mistake in the used CLI file. 2 boundaries was named but no associated to a Himp (or somethin else).
My mistake.

I have to verify again, one by one, the order of boundaries, cause it changes now ... But I think I can carried out this work.

I don't understand why the limits are not organized in CLI order. But it does not matter. If I can get away with it like this, that's fine with me.

Thanks for your time and advice.
The administrator has disabled public write access.
Flat Threaded
  • Page:
  • 1
  • 2
Moderators: pham
Forum
System Modules
TELEMAC-2D
Probably an error in setting the limits

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.