MAT_072R3 (KCC) for UHPC: Single element softening matches, but full Direct Tension (DT) specimen shows sudden drop

[Jubin Adhikari]

Hello everyone,
I am trying to simulate the uniaxial tensile behavior of an Ultra-High-Performance Concrete (UHPC) with an unconfined compressive strength of 175 MPa using MAT_072R3 (MAT_CONCRETE_DAMAGE_REL3) in LS-DYNA. I am referencing the methodology from a recent paper on wind debris impact on UHPC panels
, but I am running into a severe damage localization issue when scaling up to a full Direct Tension (DT) specimen.

My Setup & Single Element Calibration:

I iteratively calibrated the b1 value to match the compression stress strain result and also b2 value to match the tensile stress strain result. The single 5 mm unit cell behaves well in uniaxial compression and tension tension.
The Problem: When I apply these exact same, calibrated parameters to my full multi-element Direct Tension (DT) specimen, the global response completely fails to match the experimental results. Specifically, when I extract the strain over the experimental gauge length, the initial slope rises too sharply, and immediately after the peak stress is reached, the global stress-strain curve exhibits a sudden, near-vertical drop with virtually zero softening.

Troubleshooting I have already done:

Element Erosion: I realized the elements in the crack band were hitting the principal strain erosion limit (0.015) and deleting instantly. I have since increased/disabled the erosion criterion, but the vertical drop in the gauge-length stress-strain curve persists.

In my DT specimen, I computed the strain from the coordinates of two points within the gauge length and the stress by dividing the reaction force by the gauge section area. Is this the correct approach? If there are any errors or shortcomings in my methodology, I would greatly appreciate your feedback.  I have attached .k file for your reference. 

Thank you! 

[Christopher Chung]

Hi Jubin,

A quick check: With calibration, are you using nodal forces for the single element and are you consistently doing that with your full scale component test too? I have been bitten so many times by overlooking engineering stress vs true stress (It might just be me living in a plastic world).

- CJ

--
You received this message because you are subscribed to the Google Groups "LS-DYNA2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ls-dyna2+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ls-dyna2/f2884710-46e1-4c6a-8379-cd0cb52d1f2cn%40googlegroups.com.

[Christopher Chung]

I just read the last part of your email! Are you dividing by the undeformed CSA or the real-time (current config) CSA to compute for stress?