Need help: Failed Step 2 - forging

General Questions
1

Hi :slight_smile:
I, once more, need your help. My step 2 “Auszug” fails. I tried it with another solver PaStix and I still does not work. Can you take a look at my results/monitor?

2

Can you share your files (possibly the .pmx file) ?

3

Yes, sure. It took a bit to load. I tried it with 2 different constructions. Here you can look at both. I think “Versuch 2” failed because of the long cone. PrePoMax - Google Drive
But why didn´t step 2 try to start?

4

You shouldn’t mix static and dynamic explicit steps in one analysis. It’s unlikely to ever work and impossible to do in Abaqus.

But apart from that, the press is just a single C3D8 element. Even if it’s rigid, you should use more elements for contact.

Also, can you briefly describe how this process works in real life ? Because it’s not clear for me from your setup. The press moves up, pushes the “Zahnplate” which would move away from “Prallplate” but is partially fixed (isn’t the model a bit overconstrained by the way ?). So is “Zahnplate” supposed to bend at the bottom and somehow push on “Prallplate” ? Then “Zahnplate” moves up in the second step.

5

Also, can you briefly describe how this process works in real life ?

The model represents a two-step forming and testing process of a mechanical interlock joint:

  • Step 1 – Forming (Joining): A rigid punch (representing the press) drives the pin or “tooth” of the upper plate (“Zahnplatte”) into the lower plate (“Prallplatte”). The goal is to plastically deform the pin so that it creates a form-fit connection within the receiving hole of the lower plate.
  • Step 2 – Pull-out (Testing): The deformed pin is then pulled back out of the lower plate to determine the required pull-out force. This step simulates the mechanical strength of the formed joint.
    I will try to handle both steps within static analysis only. Do you think it will work out?

But apart from that, the press is just a single C3D8 element. Even if it’s rigid, you should use more elements for contact.

What should I do here? Should I use different meshing parameters?

isn’t the model a bit overconstrained by the way ?

Thank you for bringing this up. The model uses a quarter symmetry approach, and the boundary conditions are based on real-world experiments with some simplifications to reduce computation time. These include:

  • Symmetry constraints applied in the x- and y-directions.

  • The press tool is constrained in all directions and rotations except for z-translation.

  • The “Zahnplatte” (toothed plate) is fixed in x and z directions to replicate the clamping effect; a previously included rigid clamping body was removed to simplify contact definitions.

  • The “Prallplatte” is:
    Fully constrained at its outer rim (x, y, z) to represent bolted conditions.
    Fixed in z-direction at its bottom surface.

  • In the pull-out step, the tooth is pulled in z-direction, while the “Prallplatte” remains fixed.

These constraints are designed to reflect experimental conditions where the base plate is firmly held in position by clamps, and the upper element is actively deformed or pulled. Do you think I should delete some constrains?

Thanks again for the helpful insights – I really appreciate your time!

6

Thanks for the explanation, now it’s much clearer.

You may have some convergence issues but it’s still more robust than CalculiX’s dynamics, especially explicit.

Yes, change the meshing parameters for that part.

At least Z symmetry overlaps with Prallplate’s clamp. Also, Z symmetry is not applied to the entire Zahnplate’s face but I assume it’s intentional (like the lack of U3=0 constrain on its upper part). But it seems that you know what you are doing with those BCs so I assume they do represent the real-life conditions.

7

Okay, thank you. I will try step 2 as static and update you.