Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2011-09-26 20:24:58

Registered: 2008-01-19
Posts: 781

Warning: CALCULEL_7

I am encountering a warning:

   ! <A> <CALCULEL_7>                                                                   !
   !  Erreur de maillage :                                                              !
   !    La maille M93754 de type MECA_TETRA10 est trop distordue.                       !
   !    Le jacobien de la transformation gmrique n'a pas le meme signe sur tous les !
   !    points de Gauss.                                                                !
   !  Risques & conseils :                                                              !
   !    Le maillage a-t-il produit par un mailleur ?                                !
   !    La connectivitrespecte-elle bien la convention Aster ?                        !
   ! Ceci est une alarme. Si vous ne comprenez pas le sens de cette                     !
   ! alarme, vous pouvez obtenir des rultats inattendus !                             !

I am using CA_11.0.10 when running a modal analysis.  This occurs on a 3D tetrahedral mesh generated by Salome_6.3.1 (64bit binaries) which was converted to quadratic mesh using the same.  I did ask that the median nodes be aligned with the geometry.  I cannot post the mesh because of the proprietary nature of the study but the *.mess file is attached.  Can somebody shed some light please?  Is this a CodeAster bug as mentioned in the latter of the following topics?


Please note I am trying to understand by Google translating the above topics and therefore may have missed something.  The study using only linear elements is okay, only that it over predicts the modal frequencies.  Thank you.

Regards, JMB

PS: Will attach the *.mess file as soon as the job has finished running...

SalomeMeca 2021
Ubuntu 20.04, 22.04


#2 2011-09-26 20:46:50

Thomas DE SOZA
From: EDF
Registered: 2007-11-23
Posts: 3,066

Re: Warning: CALCULEL_7


This warning is raised because the Jacobian computed at each Gauss Points of one the TET10 element of your mesh changes sign. This usually is not good because it may mean that one element is distorted.
The aforementioned post pointed out one modelization of Code_Aster where this check was not needed since this modelization (a joint) relied on flattened HEXA or QUAD.

Unless you use such a special modelization, you should pay attention to your mesh and check if there is not an element that is distorted. To do that, either display the aspect ratio in Salomé or (better) use the MACR_INFO_MAIL operator in Aster (which calls the external software HOMARD) and have a look at the quality of the elements.

In order to check the element pointed out by the message, you first need to convert your mesh to Aster format (.mail) using right click -> Mesh converter in ASTK and then open the ascii file generated and look for "M93754", you will find the nodes belonging to this element and you can look them in Salomé (node number in Salomé and Aster are the same but not element number).