How ABAQUS quantifies a violent failure or rock 1

[Chen1]

Hello everyone:

Let me first tell you what is a violent failure of a coal (or rock) specimen under compression loading in the laboratory. It is a brittle failure where sudden loss of strength occurs, with very high noise (like bump) moreover, coal particles ejected very fast from the edges and the corners of the tested specimen. this is what i see in the lab.

This violent failure happens at specific conditions (shape and constraints)

My question now is: how can i simulate a violent failure with Abaqus. in other words, when to say using Abaqus that this failure is violent or not?

Your idea or advice is highly appreciated.

Chen

 

[Chen1]

Hello

Is there any idea about how to evaluate a violent failure using Abaqus? what kind of output requests would help to say that the failure is violent or not in Abaqus? I can not tell from stresses or strains.

Thanks
Chen

 

[rogersma]

Hi Chen1,

I would say that if you do not see a sudden change in stress state, then there isn't really much to diagnose here and it seems that your simulation isn't reaching a failure point. In general, a dynamic failure ("violent failure") is characterized by a sudden change in stress state and/or material properties. If you can't see this in your stress/strain records (which you should be able to in the case of dynamic failure) then something must not be working in your simulation. I'm not familiar with any of the built-in dynamic damage subroutines in Abaqus, but I found this tutorial/overview on imechanica:
[URL unfurl="true"]http://imechanica.org/files/l9-damage-failure.pdf[/url]

I personally use a proprietary damage mechanics VUMAT written by a colleague which I am unable to share, but there seems to be many others on the internet if you search. Additionally if you have easily quantifiable failure criteria and a damage evolution law, you may be able to write a Fortran subroutine (UMAT/VUMAT depending) to use with Abaqus.

Best of luck!

-rogersma