Circular fixed membrane with initial stress

  • H. Ebb
  • Topic Author
  • Offline
  • Fresh Boarder
  • Fresh Boarder
More
3 years 1 month ago #9196 by H. Ebb
Hello,

I'm trying to simulate the bending of a circular fixed membrane. The membrane is prestressed with ca. 50MPa. I've found out how to apply the prestress in Code-Aster by first creating a field (CREA_CHAMP) with the given prestress in X- and Y-direction, then applying uit with ETAT_INIT in STAT_NON_LINE and then using it in a second STAT_NON_LINE with ETAT_INIT and EVOL_NOLI. A pressure of 1Pa will be applied on the lower surface of the membrane, which is called 'Pressure' in the *.comm-file and mesh-file. The simulation runs and it shows me a result but in my opinion the result isn't right due to the fact that I have an analytical formula to calculate the bending of a circular fixed prestressed membrane. The result of the analytical calculation is about 3 nm and the result of the simulation is about 430 nm (with the finest mesh I could use without getting a JEVEUX-error for more RAM), which differs with a factor over 129. In Paravis the stress at the beginning can be illustrated with 50MPa.
The mesh is created with GMSH. I've tried to refine the mesh and with finer mesh, the bending increases and there is no point of convergence.


I'm thankful for any help,

Hanna

Please Log in or Create an account to join the conversation.

More
3 years 1 month ago #9197 by Joël Cugnoni
Replied by Joël Cugnoni on topic Re: Circular fixed membrane with initial stress
Hello

Maybe you are not considering the geometrical non linearities in your analysis.
You should use a large displacement rotation formulation in Stat non line. Use the keyword COMP_INCR or COMP_ELAS with DEFORMATION set to GROT_GDEP or SIMO_MIEHE or in the worst case PETIT_REAC

What it does is to enable geometrically non linear effect by using an appropriate geometry updating and strain tensor calculation (Green Lagrange and Piola-Kirchoff for GROTGDEP for example).

You can also experiment with follower or non follower forces types with the load option SUIVEUSE

That would be my first guess for a membrane problem.

Cheers

Joël Cugnoni - a.k.a admin
www.caelinux.com
The following user(s) said Thank You: H. Ebb

Please Log in or Create an account to join the conversation.

Moderators: catux
Time to create page: 0.128 seconds
Powered by Kunena Forum