Hello Everyone, I’m currently working on a concrete model using the Concrete Damaged Plasticity (CDP) model in Abaqus. However, the peak load capacity I’m obtaining is around 300 kN, whereas the experimental results indicate a peak of approximately 500 kN.
I’ve already conducted a sensitivity analysis involving boundary conditions, kinematic coupling constraints, CDP flow parameters, mesh size, etc., but none of these adjustments have resolved the discrepancy.
Could anyone share a concrete UMAT subroutine or point me to any resources that could help me develop a UMAT related to concrete behavior or the CDP model?
Thank you in advance!
https://github.com/marioruiarruda/Mazars_UMAT?utm_source=chatgpt.com
https://github.com/seungwookseok/ABAQUS-version-CDPM2?utm_source=chatgpt.com
https://data.mendeley.com/datasets/nph64c5pvy/1?utm_source=chatgpt.com
https://github.com/jpsferreira/UMAT-ABAQUS?utm_source=chatgpt.com
https://caeassistant.com/blog/start-umat-abaqus/?utm_source=chatgpt.com
https://caeassistant.com/blog/abaqus-fortran-subroutines/?utm_source=chatgpt.com
I hope these references are helpful to you.
In my work on LPBF steels, implementing a custom UMAT helped reconcile simulated versus measured stress–strain curves. I found that:
- Accurate Tangent Modulus: Deriving the consistent tangent by hand (instead of finite differences) cut my iteration count in half.
- Parameter Calibration: Fitting damage evolution to direct tension tests (unlike bending tests) provided a more reliable softening slope.
- Code Verification: I wrote unit tests in Fortran to verify the stress update in isolation before embedding it in Abaqus, saving countless Abaqus runs.
For your CDP model, consider starting with the Mazars UMAT as a template, then progressively replace damage laws with those in the official Abaqus manual to match your experimental curves
Ali Behbahani thanks for your response. It's really helpful, and I appreciate it. I will check this out. Thank you so much for your help.
You can use USFLD to integrate damage , it is easier then UMAT . Please expire that option. If require any person assistance regarding USFLD you contact me [email protected]
Means I have done for CDP but what understand is that because on some potential function of stress or strain it might calculate the damage , so to modale the damage based that you just need update you variable depending on the potential function and visualize as output , I have implemented USFLD to model mohar column softening of soil
https://www.linkedin.com/posts/saurabh-rai-80_abaqus-subroutinedevelopment-usermaterialmodel-activity-7277567502044278784-fu0z?utm_source=share&utm_medium=member_android&rcm=ACoAAA3IM9UBZGgp31Bq7i8YGiHUDkI9hwlN1gE
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology