Controlling crack propagation with the XFEM approach

[Lodham]

Hi there,

I am trying to model a quasistatic crack propagation in a solid subjected to bending.
I have inserted a precrack and defined a MaxPs traction separation law for the cracked domain.
I don't understand how Abaqus deals with the crack propagation !

At the beginning, the stress increases at the crack tip and when it reaches the critical value, it propagates and keeps propagating up to the end.
But I don't see any stress concentration at the tip of the crack as it propagates, as if no mechanical computation was made. It's really strange!?
At each new increase of the crack, I would like to be sure that a mechanical computation is made, and I would like to see the stress concentration at the crack tip!
Is there a way to compute SIF or the Strain energy release rate and to compute the crack propagation direction based on these parameters at each increment?

Thnaks
Pierre

 

[Lodham]

Anyone using XFEM crack propagation in Abaqus here ?

 

[Dnov]

Hello Lodham,

I'm not overly familiar with fracture mechanics, let alone how Abaqus handles the calculations, but I have worked with XFEM a bit and have come up with a small tutorial on how to use this method.

This document should hopefully help you with obtaining the contour integral results. The crack propagation direction is given when requesting a certain contour output, but I forget which (SIF, Ct-integral, J-integral, or T-Stress)

Hope this helps,

Dnov

 

[DanStro]

I don't know if will answer your specific question but here is a site with a lot info on using Abaqus with XFEM:

http://sites.google.com/site/matthewjpais/abaqus

HTH,
Dan

Han primo incensus

 

[Lodham]

Thank you guys, this is going to be very useful.
I have also seen that there is no enrichment with singular fields at the crack tips in Abaqus.
This means that tha crack has to cross completely an element at each increment.

I will work on your documents and come back to you if I have additional questions.
Thanks again !!
Pierre

 

[mecheng1234]

hey just reviewed this thread and had a quick question.

i am using XFEM in a 2D analysis and would like to get the SIFs however, the history outputs are not available for crack growth in XFEM.

how do i do this? for some reason ABAQUS doesnt allow me to uncheck crack growth when i choose method as XFEM.

Any advice?

 

[mecheng1234]

any feedback?