How to quantify brittle failure?

[Chen1]

Hello

I modelling the uniaxial compression test for 6 rock samples. the cross-sectional area of the modeled rock samples are 1x1 inch2, 2x2 inch2, 3x3 inch2, 4x4 inch2 , 5x5 inch2 and 6x6 inch2, while the height of all samples is the same which is 1.0 inch.

The material behavior used in the model is concrete damage plasticity, and the model is run in one static general step.

I need to find from the computer modeling what is the critical cross-section where the failure of the rock sample is brittle or violent.

the question now is:

How can i use abaqus to quantify the brittle failure? How can i determine whether the failure is violent or not from the model?


I am thinking of comparing the energy released with material toughness. Griffith mentioned that if the energy released exceeds the surface energy the material will fail violently.

Is there any ideas?
Your idea is highly appreciated.
Thanks

 

[Chen1]

Hello

Can anyone share his idea with me about how to quantify the brittle failure? I used concrete damage plasticity approach and the step is static general.

Thanks

 

[Chen1]

Hello anybody can share his experience with me?

How can i determine from Abaqus that the failure is brittle or not?

 

[amorrison]

Maybe a new post (do not copy this thread) to forum274 will get you some help.

 

[mrgoldthorpe]

Chen,

The computer modelling cannot tell you that. You need tests on samples of your material.

 

[Chen1]

Thanks Guys for your answer,

Let me tell a little bit about my project.

In underground mine, there are pillars similar to the columns in our super structures, these pillars support the load from all overlying rock strata. Due to high stress, these pillars fail violently in a brittle mannar, and rock particles ejected from them and could kill miners. I need to know when these pillars fail violently.

i decided to make a uniaxial test on a rock sample, and apply load high enough to make them fail. However my problem is how to quantify the this brittle (violent failure) from the model.

You know many people depend on the PEEQ to know whether the material yielded or not, however it does not tell me whether the failure is brittle or not?

I used a concreted damage approach for the rock, and the model is similar to the regular uni-axial compression test.


So again i need a parameter like (PEEQ), so it can tell me that the failure is brittle or violent.

Your idea is highly appreciated

Thanks