Dear All,
I am currently working on a VUMAT implementation for the Johnson-Cook material model. When I run simulations using the Johnson-Cook model without incorporating the damage model, the numerical results align well with the experimental data. The code was working perfectly, and the material behavior is as expected.
However, once I incorporate the damage model into the code, I observe an unusual and unexpected material behavior—material softening. This softening does not align with the Johnson-Cook model behavior, even when using the same parameters. The softening observed is excessive and inconsistent with the behavior predicted by the Johnson-Cook model.
Could someone kindly explain why this might be happening? Are there specific issues related to the integration of the damage model into the VUMAT code that might be causing this discrepancy? I would greatly appreciate any insights or suggestions for addressing this issue.
Thank you all in advance for your help.
Hi Fasil,
I recently released a Johnson–Cook VUMAT that includes damage initiation and evolution, element deletion, and Taylor–Quinney heat conversion. It is fully documented, and it might be useful for your case if you are trying to pinpoint discrepancies in your current implementation. The VUMAT was validated against Abaqus’s built-in JC damage model, and I used it in a machining simulation where it produced good agreement with experimental data.
GitHub repo: https://github.com/mauroarcidiacono/Abaqus-VUMAT-Johnson-Cook
Hope it helps!
Hi Mauro,
Thank you for sharing the VUMAT and GitHub link. This looks very relevant to the issues I’m facing with damage integration and excessive softening. I’ll go through your implementation and documentation, especially the damage evolution and deletion logic, to compare with my approach.
Appreciate your support. This should be very helpful.
Thank you
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators