Talk:Contrib:Claws/Code Aster/10 x cases/liaison mail

From CAELinuxWiki
Jump to: navigation, search

Hoi Claus

Very nice! This coupling with liaison_mail is a item that has been interested me for long but never had time to study it. This makes it all much easier.

One question about your calculation: how do you prevent the rotation of the component with the pressure applied to it? Maybe I just didnot read carefully enough and I overlooked the obvious.

Thank you once again.

kind regards - kees


Thanks for the input.

Bear in mind that the page is not done yet (feb 10) so it will change a bit when I pull myself together and get it done.

About your question: Damn good question! At first I thought it was because I used LIAISON_UNIF; this makes the 'press' face not rotate and deform, like it was welded to the source of the pressure. But! and this is a big but, if I remove LIAISON_UNIF and and apply e.g. PRESS_REP instead of FORCE_FACE (-Fz), the pressure face starts to bend - and that just doesn't make sense. I get some errors on the calculation that might be related. As far as I can tell the two meshes are too different and the 'calculation might produce incorrect results' - I however, haven't had time to refine mesh and play around with it recently. Do you have a suggestion?

I will investigate more when I get the time - right now Im playing with beams and shells for some semi-actual work :)

  !  l'angle forme par le vecteur normal courant a 1 face et le vecteur normal moyenne, au noeud  N6559 , est superieur a 10 degres et !
  ! vaut  16.  degres.
  !  PROJ_CHAMP (ou LIAISON_MAIL) :                                               !
  !  Nombre de noeuds projet�s sur des mailles un peu distantes : 22.             !
  !  (la distance � la maille est sup�rieure � 1/10i�me du diam�tre de la maille) !
  !                                                                               !
  !  Le noeud NS22841 est projet� le plus loin � la distance 0.154085             !


Two images: with and without LIAISON_UNIF - as you can see, without unif the hinge applies torsion to the pin that it sits on, and it shouldn't do that. Something is amiss.

Claws assm without UNIF.png

without_UNIF.png

Claws assm with UNIF.png

with_UNIF.png


Hoi Claus Alright, I guess I understand what happens when you use LIAISON_UNIF, with restricted directions DX, DY and DZ. This effectively prevents the pressure part from rotating, since eg all DX values of the part are the same.

Mot-clé LIAISON_UNIF 4.20.1 But Mot-clé facteur permettant d'imposer une même valeur (inconnue) à des degrés de liberté d'un ensemble de nœuds. Ces nœuds sont définis par les groupes de mailles, les mailles, les groupes de nœuds ou la liste de nœuds auxquels ils appartiennent.

What I indeed donot understand what happens if you omit this key word. I guess the pressure part should rotate around the axis. Right? I am trying to run an example at the moments. Not running yet. To be continued. - kees


Hello Claws,

The hdf file which is supposed to be on the bottom, is not there. Would it be possible to at that file?

regards


I've attached the whole case to the wiki now, at the bottom, it is still a work in progress though.

/C