Traction Separation Law in Dolfinx

General
1

All,

I am trying to leverage the work from @bleyerj from Cohesive zone modelling of debonding and bulk fracture to approach traction separation laws using dolfinx. I have worked to replicate the original code in the updated framework but am facing numerous challenges in its implementation, primarily in selecting appropriate function spaces and solver settings.

Has anyone done similar work in dolfinx?

2

Hi, sorry I did not have time yet to port this demo to dolfinx. I will try to do it in the near future though…
Can you be more specific on what is your exact issue ?

1 Like
3

@bleyerj , Thanks for the reply. My intent was just to ping the community, happy to provide specific details if you are interested.

Selection of Function Spaces
I have done little work with quadrature elements and spaces other than CG/DG. As per your comment in the above example

The ideal solution would be here to use Quadrature elements on the mesh facets, which are unfortunately not available in FEniCS. We will instead use a Discontinuous Lagrange Trace (also known as HDiv Trace ) space which represents polynomials living on the facet of the meshes and which are discontinuous at the vertices.

The issue hence is deciding the element family for V_f. I see from hdiv-trace-elements that DGT is not supported yet in Dolfinx. I am not familiar with using Quadrature elements in Dolfinx though I am trying to learn. Any advice is appreciated.

Thanks,

-Sven

4

Hi. Is there any update on this topic (CZM in FEniCSx)? I’m also interested. Thank you.

5

A first alternative is to update the code by Jeremy Bleyer using CR (Cruzier-Raviart) element for the jump field, with somenthing like

    "CR", msh.basix_cell(), surface_degree, discontinuous=False
)

I have done somenthing few months ago using this.

We started working on a better version with @bleyerj and @dokken two weeks ago. We should end up with somenthing hopefully for the end of the year. The plan is to use facet meshes and formulate the problem as a convex optimization problem, using Mosek as a solver.

1 Like
6

Excellent, thank you @cmaurini. I will keep exploring this. Eventually I aim at coupling it with some contact mechanics & friction to model a tool / obstacle, in the spirit of a cutting tool.

7

How is it going, do you think you will finish before the end of this year?

I want to define a hertzian-model between two cylinders and I think this would be a great help. Thank you in advance for pushing this project forward!