GMSH and contact / Excavation

Padr

Hello,

I am trying to simulate a 3D excavation of a tunnel. I have created a mesh using GMSH where excavation will be performed by a sequential removal of elements (deactivation at each step from the corresponding model).

In the same time I am trying to add a support which will be actually an activation of another volume element (ring) with different properties at each excavation step.

The problem comes for the adjacent volumes-rings of the support. Even though they are defined completely elastic, I see an accumulation of stresses on their interface, seeming like the act independently.

Since both elastic material, aren't they supposed to be behaving as one? Are there any suggestions?

Thanks!

jeanpierreaubry

Padr wrote:
Hello,

Are there any suggestions?

yes, one!

supplying the files of the problem!!

Padr

U will find attached the case comm

Thank you!

Attached file:
CurrentCase.comm, 24344kb

jeanpierreaubry

i get this message

La bibliothèque MFRONT n'a pas pu être chargée.

some files are missing how to load them

Padr

You will find attached a new comm file. Just changed the mfront file with a law from inside Code Aster.

It converges fine, however I still see the same issue.

Attached file:
exc.comm, 24079kb

jeanpierreaubry

cannot see any operator printing a result in the command file

which results are you talking about?

Padr

I just added a IMPR_RESU to get some stresses.

I attach the picture. U will see an accumulation on the interface between two adjacent rings.

I have tried with finer meshes, still the same issue.

My question is, is there any need of defining a contact? For adjacent volumes that are both elastic.

Thanks for your responses.

Attached file:
stresssiyy.png, 111110kb

jeanpierreaubry

i must say that i do not understand what you are trying to do

i cannot see any loading in your command file

and tell me where are the groups i should look for to see the interface and stress you are talking about

and i do not understand the question about CONTACT

Padr

Hello,

I really believe that understanding the whole model/mesh is relatively dificult.

The main issue is that in every step I am activating new elements (volumes), that are having an elastic behaviour.

The mesh is fully conformal yet I get to see stress accumulation on their interface. You will find attached a picture of the magnitude of stresses.

Since I do not define any contact between them, aren't they supposed to behave as one?

Attached file:
rings.png, 1111563kb

jeanpierreaubry

i start understanding your problem a bit better

which result is shown on this pic ResR4 or another

which "rings" are shown

i am running it by myself, on the linear mesh and i get similar results

did you look at the displacement and what do you think about it

CONTACT is for 1 moving body coming in contact with another one at some step in the calculation

here we have common node so it should not be used

attached is a view of this in Gmsh

_F(RESULTAT=resR4,NOM_CHAM='SIGM_ELNO',NOM_CMP='SIYY',NOM_CHAM_MED='r4Rc',GROUP_MA=('R1c','R2c','R3c','R5c',),),

i wonder whether the trouble does not come from the cascade of STAT_NON_LINE which adds results in a weird manner

Attached file:
r4Rc.png, 28685kb

Padr

Thank you Jean Pierre for replying.

Yes, what you attached is exactly what I am facing.

Seems relatively strange that is not a contact issue obviously, neither a mesh one (in my view at least).

jeanpierreaubry

Padr wrote:

Seems relatively strange that is not a contact issue obviously, neither a mesh one (in my view at least).

yes i agree

strange indeed