open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello to all,

I am not yet familiar with details of Telemac and now I have a problem with launching a simulation. The simulation starts good, but get stopped after a short while. There appears the following error:

a la ligne 12 le mot cle suivant: velocity profile est inconnu...
arret de damocles
erreur dans le fichier des parameters
damocle:trying another language

I don't know if that's the problem.
The simulation stopps not at that point. It continues:

No compilation/linking/file errors detected
No execution errors detected
Returning exit status 0

then the system launches Perl and it stopps.

I think Perl is correctly installed, because other calculations are running perfectly. What could be the problem?
I hope anybody can help me.

Thanks in advance.

Hello,

It is just that this key-word is:

VELOCITY PROFILES : 1;1;....

with a S at the end, there is actually one option per liquid boundary, you must give an integer for all these liquid boundaries, separated by a semicolon. The numbering of liquid boundaries is the order of appearance in the boundary conditions file, more precisely liquid boundaries are numbered starting from the most south-west point in the mesh, and then going along the contour in a counterclock-wise way.

With best regards,

Jean-Michel Hervouet

Hi,

I defined for 2 boundary conditions 2 velocity profiles. For the inflow with prescribed flowrate a "normal velocity given in ubor in the colim file" and for the outflow a prescribed elevation with a constant velocity profile.
So, I think the definition is okay. It's like in the user manual. But the simulation still stops...
What could be the problem?

Thanks in advance.

Hello,

Please tell us what message you get or send your case or at least the parameter file, so that we can have a look at it.

With best regards,

Jean-Michel Hervouet

Enclosed I send you the file.

Thanks

With best regards

Hello,

I tried your case. Telemac-2D tells me that you have an element with negative determinant (i.e. which is not correctly numbered). I had a look and I find that:

element 6650 has points 3570 3586 3584
element 6651 has points 3570 3584 3594

The problem is that segment 3570-3584 is a boundary segment, it cannot belong to 2 elements.

So well, sorry, your mesh is not correct.

With best regards,

Jean-Michel Hervouet

Hello,

thank you for your help. I am trying to figure out, what's the problem.

I created the mesh with Blue Kenue following the indications of the manual. So there must be a mistake in the mesh generation. For creating the Selfin file in Blue Kenue a 2d Line Set has to be created and dragged into the icon of the Selfin file, hasn't it? Does this 2d Line as boundary of the modelled area lie directly on the border nodes of the modelled area or don't the line and the border nodes coincide?
I had a look on the Selfin file in Blue Kenue and the ipobo part of the file appears in the view as a zigzag line and doesn't lie directly on the boundary. Could that be the problem?

With best regards

Hello,

Yes, IPOBO is an array that is 0 for inner points and gives a numbering for others.

Unfortunately our Blue Kenue developer is here in Paris for the Telemac User training day on ... Blue Kenue, so he won't answer on the forum immediately.

With best regards,

Jean-Michel Hervouet

Hi,

thanks for the quick answer.
Does that mean that IPOBO should be 0 or should be numbered for creating a correct mesh?
I send you some screenshots where you can see how the IPOBO of my mesh
look like.

Best regards

Hello,

I think we have to wait an answer by Martin Serrer. In the guide for programming in the Telemac system you find that in Selafin format IPOBO is :

1 record containing table IPOBO (integer array of dimension NPOIN); the value is 0 for an internal point, and gives the numbering of boundary points for the others. This array is never used (its data can be retrieved by another way).

I do not think that the user must fill this array himself. It is just a matter of numbering boundary points in a counter-clock-wise way but this must certainly be automatically done.

With best regards,

Jean-Michel Hervouet