Hi,

please check the attached file for some description about XFEM.

Hi Phu,

As far as I know:

Yes you can do 3D crack growth. We use it through LEFM by treating the cracks as stationary ones and then with an algorithm we propagate the cracks automatically. In this way we observe no divergence issues. Otherwise you might face a lot of convergence problems by using XFEM directly on propagating cracks. I guess the default implemented way for propagating cracks is through cohesive relation. The bulk material can only be elastic.

See you,

Tuncay

Hi,

please check the attached file for some description about XFEM.

Thanks a lot Tuncay and Vahid.

Bests,

Phu

Highly recommend you to use embedded strong discontinuity approach! It is more efficient

Yours sincerely,

DO XUAN NAM

Hi Phu,

Blow are what I know for the XFEM in abaqus.

(1) Yes. It can do 3D crack growth.

(2) Yes. It can do both XEFM and cohesive crack. However, there seems to have convergence issue when I try to have a crack defined by XFEM in bulk material changed to cohesive failure using cohesive surface method in abaqus.

(3) It can deal with material nonlinear as the data input is quite similar to cohesive method, including cracking initiation stress/strain and crack propagation fracture energy.

(4) It seems that currently XFEM can not be used in abaqus Explicit.

(5)Normally PHILSM (Signed distance function to describe the crack surface) and PSILSM (Signed distance function to describe the initial crack front)can be output.

The XFEM in abaqus is quiet like VCCT method, it can only deal with single crack , but having issues with the interactions with many cracks.

I've also found that the convergence is related to the input material properties. Currently it is only valid for first order solid continuum elements.

Regards,

Qingqing

I would like to mention that despite limitations as described above, ABAQUS-XFEM has been used to model the crack propagation and the manuscript is attached for reference.