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

You are not logged in.

#1 2020-11-19 17:10:25

jesuamador
Member
Registered: 2020-10-29
Posts: 9

Differences in convergence with different meshes

Hello to all,

I have a contact model of solids in which one of them has a plastic behavior. Everything has worked well with a mesh (little_mesh).
However, to check the model I decided to change the geometry making it more realistic and changing the mesh but now the calculation does not converge.
I have tried changing several parameters and increasing the number of time steps but it still doesn't work
The mistake I get is:
Failure in the integration of the constitutive law

I don't know if anyone is able to achieve convergence or if it is a mistake in the model.

Anyway, thank you very much
Jesus

*Attached the .comm file


Attachments:
plastic_contact.comm, Size: 4.75 KiB, Downloads: 15

Offline

#2 2020-11-19 17:18:50

jesuamador
Member
Registered: 2020-10-29
Posts: 9

Re: Differences in convergence with different meshes

And here are the two meshes


Attachments:
2meshes.rar, Size: 225.06 KiB, Downloads: 17

Offline

#3 2020-11-21 14:31:52

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 255

Re: Differences in convergence with different meshes

Hello,
there are a few strange things in your model. I changed the followings:
- blocked only the Y direction as the model is axis-symmetric
- added weak springs as the upper part Y direction is not constrained (the stiffness may be too high)
- added a penalty coef. (you should adjust it) for the contact
- changed the contact zone for a larger one
- activated the automatic time-stepping
- changed the ALGO_RESO_CONT to POINT_FIXE in order to facilitate the convergence.

Attached the comm file.

Konyaro

Last edited by konyaro (2020-11-21 14:43:22)


Attachments:
CurrentCase.comm, Size: 5.35 KiB, Downloads: 12

失敗は成功のもと (L'échec est la base de la réussite)

Offline

#4 2020-11-23 11:20:08

jesuamador
Member
Registered: 2020-10-29
Posts: 9

Re: Differences in convergence with different meshes

First of all, thank you very much for the answer and for the work you do to help those of us who are starting to work with Code-Aster.

I tested your code and modified a small detail of the names you gave to the mesh because it gave me an error. I have tried the code with the small geometry and it has worked perfectly. I have tried with different values of stiffness of the springs and the result varies little.
The problem is that when applying exactly the same to the big geometry (I only change the value of the force to -0.1768388) the algorithm does not converge and gives me the integration error of the constitutive law.

I attach the mesh that works and the one that does not and the .comm file with the small modification in case someone is able to see the convergence problem in the big mesh.


Attachments:
CurrentCase.rar, Size: 607.51 KiB, Downloads: 9

Offline

#5 2020-11-23 20:40:28

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 255

Re: Differences in convergence with different meshes

There is something wrong with your mesh. There seems to be a common node to both parts. Can you share the file you used to mesh ?


失敗は成功のもと (L'échec est la base de la réussite)

Offline

#6 2020-11-25 09:16:53

jesuamador
Member
Registered: 2020-10-29
Posts: 9

Re: Differences in convergence with different meshes

I don't know what do you mean with the file I use to mesh.
I´m creating the mesh with Salome-meca mesher. I use Netgen2D-1D with a local size of 0.0003 in contbasel and 0.0005 in contballl and a maximum size of element of 0.2

I use this because I don't know much about this mesher's options. I would like to do an structured mesh with quad elements
whose size increases from the contact corner to the opposite side but i don't know how to do it.

I attach the .brep files of the geometry


Attachments:
Geometry.rar, Size: 1.92 KiB, Downloads: 4

Offline

#7 Yesterday 09:56:51

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 255

Re: Differences in convergence with different meshes

Hello,

Jesus wrote:

I don't know what do you mean with the file I use to mesh.

I meant a python dump file of your study (file --> Dump study).

I corrected the mesh and changed the contact parameters in order to make the simulation converge pretty fast.

Concerning the mesh parameters, you should rather ask your question on the Salome forum.

Attached mesh and comm file.

Konyaro


Attachments:
ball_ok.zip, Size: 49.4 KiB, Downloads: 6

失敗は成功のもと (L'échec est la base de la réussite)

Offline