Multi-step analysis of thin shell

General Questions
1

Greetings,
I recently started using PrePoMax to analyze the deformation and torque generated by a thin film.
My goal is to first pre-stress the film and then measure the torque generated during rotation.
The film is modeled using shell elements.

Current workflow:

image
image844×765 80.1 KB

image
image940×456 47.4 KB

image
image940×456 53.2 KB

image
image501×571 39.5 KB

Step 1
The straight film is deformed by moving F_initial to an intermediate arbitrary point before reaching the F_inner position.
(The film twists if F_initial moves directly to F_inner, so I split the process into two steps.)
The surface is fixed to F_initial using a rigid body constraint.

image
image940×456 60.3 KB

Step 2
The deformation is completed by moving F_initial to the F_inner position.
This step creates the initial pre-stress in the film.
This becomes the initial condition of my device.

image
image938×532 90.6 KB

image
image938×506 45.3 KB

Step 3
I would like to rotate the surface around the axis defined by reference_point_1 to measure the torque as a function of rotation angle. However, I could not achieve this using a multi-step analysis. The screenshot above shows a separate analysis to illustrate an example. In this case, the film did not start from the pre-stressed state.

I encountered the following difficulties.

Questions:

  1. The F_initial position is reset when a new step starts.
    I currently need to manually reposition it using displacement/rotation boundary conditions and amplitude definitions. Otherwise, the analysis crashes.
    Is there a way to continue the analysis using the deformed configuration from the previous step without redefining boundary conditions?

  2. I cannot change constraints between steps.
    The surface should detach from F_initial and then be attached as a rigid body to reference_point_1.
    However, I could not find a way to change rigid body constraints between steps.
    Running the steps separately is possible, but the film must remain pre-stressed before the rotation in order to measure the torque correctly.

  3. Additionally, I suspect the film may exhibit bistable buckling behavior depending on the fixed position. Can PrePoMax handle this kind of problem?

Unfortunately, due to company security policies, I cannot upload the PMX file.

PrePoMax version: 2.4.0

Thanks for reading.

2

Check this thread (especially the last point there): Understanding multistep analyses

There’s no such option. Only contact pairs can be deactivated.

Yes, but you need to enable Nlgeom for that and avoid shells + rigid body constraints + Nlgeom because it leads to non-convergence. As a workaround, you can just use solid elements at least for the regions with rigid body constraints, but it seems that you already did that.

Even via a private message ?

3

There’s no such option. Only contact pairs can be deactivated.

Then do you have any suggestions for Step 3? Like exporting the stress state and the deformed mesh and importing them into a new PMX file?

Even via a private message ?

Yes, the company security app on my PC prohibits any kind of file transfer (uploads, cloud storage, messaging, etc.).

I managed to upload images by working around the restriction, but uploading other files is impossible.

Email might be possible, but it would require approval from my manager.

4

So it’s fully elastic deformation ? Perhaps you could use a single step with amplitudes to define non-linear time variation of loads/boundary conditions.

Unfortunately, CalculiX doesn’t have the import functionality known from Abaqus. It only has restart (continuing previous analysis, but it works like multiple steps in separate runs). You could easily export the deformed mesh, but not the material state (you would need to apply the initial stresses/strains manually).

Check this example too: CalculiX-Examples/Streifen at master · calculix/CalculiX-Examples · GitHub

I will reach out to you via pm and we’ll see what we can do.