Your sure there is not an exponential solution to base e?

Is there an analytical way to obtain the function of thermal balancing?
Or other Question: which material property can you change to suppress geometrical nonlinearities as best as possible?

I have doubt that in your model all prestessed regions are able to deform in a realistic way of.
(Compare with theorem of virtual work)

One example: you have a M12 bolt 8.8.
Then the stress cross-section is 84,3 mm^2 (acc. ISO898-1)
In EC3 you will find a pretension force of F p,C = 47200 N.
The normal stress after prestressing is 560 MPa in the stress cross-section, now.

The yield point of the bolt is 640 MPa.

As long as your connected plates are deformationless in the direction of the thickness you will only reach a additional external force of 7000 N till reaching the yield point (in contrary to 47200 N prestress force)...

3D Solid model?

A well-known rail vehicle manufacturer would model it like this:
(without waranty that your costumer wants it like this)
The proof concept for those is the nominal stress concept.

The first step is to derive the mid surface, that you obtain a clean shell model.
Secoundly meshing all as a compound to prevent that you need operator  "liasson-maille" etc..
Mesh zise sould be approx. 20 mm. Also try to suppress all driangle elements.

Use all symmetry conditon.. etc.
Then run time should only a couple of secounds.

I hope it helps you or other "code_aster" enthustiastics

Hi,
here some hints from me:

You use parallelisation method openMP not mpi.
(mpi requires, additional to openMP, domain decomposition what you can do with scotch for instance. You can google after scotch and domain decomposition. Then it becomes more clear…)

My experience with a small model and many time steps (Dyn_non_line) was:
-    starting only one thread
-    switch out hyper threading in the Computer BIOS
-    seek for a computer with the highest tact frequency

METIS produces a shared-memory-object and
SCOTCH produces in my opinion both or at least also one distributed-memory-object ...

here is the .comm file:

Hi Astercode,
again, both numerical calculations are wrong (also Nastran). Change the boundary conditions in both systems that you get 120 MPa in both numerical calculations Aster and Nastran.  This is very important. Your boundary condition produces a singularity. The correct result should be infinitely in Aster and Nastran !!! ...
Kind regards and good luck Volker

Hi astercode,
have you already compared your numerical solutions with the analytical solution?

A numerical solution is a good start point to investigate commercial and open source software.

Kind regards and good luck Volker

I was the next one with this problem.
On my machine helps this, and I don't know why:

LD_PRELOAD='/usr/$LIB/libstdc++.so.6' ./salome
(see thread code-aster.org/forum2/viewtopic.php?id=24078)

Hi All,
here a sketch with my interpretation of the local stresses according the FKM Guideline:

How can we calculate the stress gradient with Code_Aster??

Kind regards Volker
PS

I total agree with you ,
but I got also the same advice from a commercial support-team to introduce skin elements ;-)