Simscale Results different from Code Aster Result


#1

SO I have a dental project where I have to extract stress at gums (gingiva) because of the movement of tooth which is caused by aligner. I have to do Static Analysis of this arrangement.

So I made a simple geometry where Cuboid represents teeth and the plate represents the Aligner. In the geometry itself i penetrated the plate geometry into the cuboid so that after giving contact it would push back the cuboid and I would get the stress at the base part as shown in the picture. I ran this model in Calculix and I got the results which is shown in the same picture. I tried this on Simscale and I gave penalization contact between those faces and kept a high pressure overclosure constant and it gave some result -https://www.simscale.com/projects/swa/column-100-10-0-1/. I tried to do this in code Aster but its not giving me any result. I gave a Continuous Contact with penalization but no legends of stress. Link to the input files-https://drive.google.com/file/d/1Df6lWIpKoI-MWY2jqtN6SLqjO83bancq/view?usp=sharing
@rszoeke Please help me out of this
@jousefm


#2

Hi @swa,
your comm file uses a static linear analysis without any contact definition - neither bonded nor physical. You define a physical contact (DEFI_CONTACT) but it is never actually referenced in the solution calculation (MECA_STATIQUE).

Why do you even try to model this locally with Code-Aster, your case seems to work fine with simscale?

Actually, the input file which is used for the analysis is shown in the solver log - you can copy all settings from there if you need to.

Best,
Richard


#3

To answer "Why I am doing such an annoying thing to do the same thing in Code Aster when its already done in Simscale.
So our Instituion is planning to buy license of Simscale. They want us to know the basis of FEA software so that we wont be some dummies clicking buttons.

Now, I corrected my mornoc mistake and I followed the simscale solver log so that I know what was actually going on. I found out about auto time stepping and how to extract results using
(CALC_CHAMP). This are the following questions I have.
1)In the Defi_Con line of my simscale project-https://www.simscale.com/projects/swa/column-100-10-0-1/
What does this line represent
SANS_GROUP_NO=(‘master_0’,‘up_cc_2’,‘up_cc_0’,‘up_cc_1’),
2) What does this variable represent in ECHEC
SUBD_NIVEAU
PCENT_ITER_PLUS
SUBD_PAS
ACTION
3) SO I tried to extract Deformation as NON-LINEAR STRAIN in COMP_MED in code Aster but it gave me error besides following exact way the log did it. Here is the link to the .comm file- https://drive.google.com/file/d/1nBweCOwE8gYv7yIqpkYd5yINobPCTkbl/view?usp=sharing


#4

Hi @swa,
ok I see. I’ll try to answer your questions one by one:

  1. SANS_GROUP_NO is used in DEFI_CONTACT to explicitly remove certain nodes from the contact definition. We do this automatically to prevent over-constraining the system if the same node is used twice in a contact and a contact constraint as slave node or if there are shared nodes between master and slave surface (which is not allowed).

  2. You can read about these in the documentation of DEFI_LIST_INST, they basically control the automatic time step subdivision.

  3. Seems like a typo, instead of

  • NOM_CMP=(‘EPXX’, ‘EPYY’, ‘EPZZ’, ‘EPXY’, ‘EPXZ’, ‘EPYZ’)
    you wrote
  • NOM_CMP=(‘EXX’, ‘EYY’, ‘EZZ’, ‘EXY’, ‘EXZ’, ‘EYZ’)
    You can also find out here about the meaning and nomenclature of the stresses and strains etc. in Code_Aster.

Best,
Richard