Reproducing stress drop (brittle failure) in 3DEC UCS tests

Hello all, new user here. I would like to kindly ask for advice on how to reproduce the realistic stress drop from brittle failure in UCS tests. I am simulating a UCS test with Mohr-Coulomb constitutive model and I am using this simulation to calibrate my rock strength for the next step of large block scale simulations incorporating joint sets and DFNs.

My input M-C parameters are the actual values obtained from lab triaxial tests, and in terms of the generated UCS, the strength in the simulation matches quite well with the average lab UCS values. However, rather than showing the realistic sudden stress drop, the results show a plateauing residual strength. The code for the UCS simulation refers to the Voronoi UCS example but without the Voronoi generation due to the simulation goal mentioned above.

I have tried adding the flag brittle true but it does not have much effect.
Thank you in advance for your kind help.


UCS sample generation.dat (456 Bytes)
UCS test-without flag brittle.dat (1.3 KB)
UCS test-flag brittle.dat (1.3 KB)

If you are using the Mohr-coulomb model, then the default residual cohesion is 0, and the default residual friction is equal to the peak friction. So you see some stress drop, because the cohesion decreases, but there is still some residual strength because of the relatively high friction (maybe). Trye reducing the residual friction property and see what happens.

Dear Jim,

Thank you for your reply and I appreciate your input.
So just to confirm, this means I would refer to the guidelines for strain-softening Mohr-coulomb model?

Thank you.

Oh - I guess so. I though there was a residual cohesion and residual friction for Mohr-Coulomb model, but I was wrong. So probably the friction AND the cohesion stay at the peak values when failure occurs. So you need strain softening if you want them to reduce.

Noted, I will give it a try.
Thank you!