[Salome Meca Tutorial] Simulating a contact solid-plate

In this tutorial, you will learn:
How to imprint a solid on a plate
How to mesh more finely around a contact area
How to create a contact normal
How to assign a contact between a solid and a plate
How to assign a forced displacement
For more tutorials and other FEA related articles, check out my blog:
feaforall.com

[@gaebu68]

Thanks so much for the video! Did you find a solution for the stresses? Because refining the mesh did not help. I am currently trying to model sandwich panels using sheets and solids with different materials. I use almost the same approach as you do, except I create a compound mesh, because I wasn't able to assign different materials with your method. The displacements always look fine, using SIGM_ELGA instead of ELNO, the stresses in the field around the contact look fine as well, but not at the contact. Would be interesting to hear your take.

[@mohsendaneshmandi9213]

Great tutorial, very useful. Would you please show us how we can simulate collision between thousands or even millions of sphere particles at the size of around 0.8mm made of hard materials and the mechanical surfaces to improve fatigue resistances of these surfaces, considering dynamic parameters of the particles like speed, density; this surface treatment operation is known as shot peening.

[@stefanocappeller2821]

That's valuable, thanks! Could you make the same example pointing on VM and principal stresses?

[@MrGrapefruit77]

Hi Cyprien, many thanks for your videos. This one helped me to define the contact between solid and shell elements. I have one question concerning the crea_champ that defines the normal vector of the contact for the liaison_mail command: I know from other software (e.g. Abaqus) that the direction of this normal field is very important. How do you check this and make sure the normal field direction is correct in Salome/code_aster?

[@JPCCastas]

Great video Cyprien. I was wondering if is possible to do two videos, a Random Vibration study and a Shock Response Function study? Thanks and keep the good work!

[@McHumaty]

This type of contact can be used for any solid-solid contact? For example, the contact of a hook with a eye bolt...

[@Feaforall]

If this is a glued contact without friction, I think yes. There are other types of contacts available for more advanced cases.

[@cwf0073]

Hi Cyprien. Great explanation to help a beginner like me. May i ask what is the difference between LIAISSON_MAIL and DEFI_CONTACT? Thank you.

[@Feaforall]

From what I remember: Liaison_mail is to link the nodes of 2 distinct mesh sets and Defi_contact is for real contact definition using specific contact criteria.

[@cwf0073]

@@Feaforall Hi cyprien, yes you are right. thank you!!!

[@jandykman7851]

Thank you very much.

[@raymiller4453]

Cyprien, What is wrong with this? I follow your operation,and run the satge. Then.... Unexpected error Type: UnicodeDecodeError Value: ('ascii', '\xe4\xba\x8c-09-180326', 0, 1, 'ordinal not in range(128)')

[@ing.nicola1706]

Nice Video ! - How to take into account friction coefficient whit lateral displacement? - If the plate has a thickness, the box, and the plate are penetrated each with other.. is it an error? the is bo gap... Not for video : - I hope in more tutorials about LIAISON ... like to simulate connection with a lot of bolts since creating the mesh of the bolts ...... - How to run multiprocessor-simulations with Aster MPI ....

[@Feaforall]

Thanks for the ideas, Nicola, I am taking note of all of that for the next videos... you will have to use DEFI_CONTACT for a contact with friction and check in the doc which of the contact types support friction

[@fabricekevinefeutang1034]

thank you cyprien for this tutorial.

[@joelHernandez-qz6vh]

hello how can I make a model of a reinforced concrete beam?

[@McHumaty]

Is it possible to do impact analysis in the Aster Study?

[@Feaforall]

Not really, code_aster doesn’t have a nonlinear explicit solver. Mostly static, NL Static and linear dynamic (NL dynamic implicit works I think, but that’s not suitable for impact)

[@McHumaty]

@@Feaforall the interesting thing then would be to implement the algorithm (nonlinear explicit solver)... very hard work, haha.

[@pagna9448]

salut cyprien, merci pour la vidéo, très instructive, peu tu me citer un exemple en mécanique ou en a besoin de ce type de simulation ? par exemple le contact entre deux dents d'engrenages? ou plutôt des cas d'emboutissage de tôle?merci

[@Feaforall]

Ce que j’utilise dans cette vidéo est plus un « glued contact » ou les deux pièces sont fixées l’une par rapport à l’autre. C’est utile dans tout système qui a des solides ainsi que éléments coques! Ils y a aussi d’autres façons de simuler des contacts dans code_aster mais c’est non linéaire et j’en parlerai sûrement dans une prochaine vidéo!

[@pagna9448]

merci

[@thierry003cantal9]

Hi Cyrien, Thank you for this tutorial :) I try to simulate this example in salome_meca 2020 but i have a litle error as : Erreur d'utilisation : ║ ║ La norme du vecteur normal (moyenne des normales des éléments concourants) est presque nulle. ║ ║ Les facettes concourantes au noeud N571 ne définissent pas une normale fiable. ║ ║ Il y a un problème dans la définition des mailles de bord . ║ ║ ║ ║ Suggestion : ║ ║ Pensez à réorienter les mailles de bord avec l'opérateur MODI_MAILLAGE. in your example you don't use MODI_MAILLAGE is it normal ? Thank :) Thierry

[@Feaforall]

Maybe something has changed in salome_meca 2020 and now it needs the MODI_MAILLAGE... I haven’t tried on the newest version

[@braiandesia866]

I had the same problem using 2019.0.3 version. I solved it by applying the command "MODI MAILLAGE" to the plate mesh: Pmesh = MODI_MAILLAGE( MAILLAGE=Pmesh, ORIE_NORM_COQUE=_F( GROUP_MA=('Plate_surface', ) ) ) where "Plate_surface" is the name of the whole plate surface. Then I used this modified mesh in the command "ASSE_MAILLAGE" . I hope you find this useful!

[@joseantonioschiavon1920]

@@braiandesia866 Thanks a lot, Braian. This worked for me!

[@bago3191]

Bonjour Cyprien, merci pour cette vidéo. J'ai suivi le tuto , sauf qu'à la fin j'ai un message d'erreur : ! ! ! ! ! L'angle formé par le vecteur normal courant à une face et le vecteur normal moyen, au noeud N3, est supérieur a 10 degrés et vaut ! ! .18E+03 degrés. ! ! ! ! ! ! This is a warning. If you do not understand the meaning of this ! ! warning, you can obtain unexpected results! ! !------------------------------------------------------------------------------------------------------------------------------------! !------------------------------------------------------------------------------------------------! ! ! ! ! ! Erreur d'utilisation : ! ! La norme du vecteur normal (moyenne des normales des éléments concourants) est presque nulle. ! ! Les facettes ! ! concourantes au noeud N33 ne définissent pas une normale fiable. ! ! Il y a un problème dans la définition des mailles de bord ! ! . ! ! ! ! Suggestion : ! ! Pensez à réorienter les mailles de bord avec l'opérateur MODI_MAILLAGE. J'ai tenté d'utiliser l'opérateur MODI_MAILLAGE, mais rien n'y fait... Saurais-tu d'où vient l'erreur et comment la corriger ? Merci beaucoup pour ton aide Bastien