User login


You are here

Crack Propagation using XFEM ( What happens to dof's corresponding to enriched nodes)

I am working on crack propagation . I am trying to figure what factors should be taken into account when the crack is being propagated using XFEM.

I am especially interested to know what happens to the additional dof's corresponding to enriched nodes. Once the crack is propagated and crack tip is at new location , we add new dof's corresponding to enrichment functions , but what happens to the information stored by the dof's of previous enriched nodes,do we forget them altotgether , or do we map it to the new enriched nodes?


N. Sukumar's picture

You have not indicated if quasi-static or dynamic fracture is the crack
propagation problem of interest. I will assume it is the former; I
am not sure if there is a consensus in terms of best enrichment in time
in regard to dynamic fracture. For quasi-static fracture, a static
problem is solved at each step, so once you advance the crack and know
the new location of the crack surface and front, it becomes a new
static problem. So, the enriched dofs are determined afresh at each
step, and therefore, the dofs  of the enriched nodes from the
previous step do not enter the solution process.

phunguyen's picture

What Sukumar says is exactly correct for linear problems. I think he  assumed that you had been

solving linear elastic fracture mechanics problems. However, if nonlinear fracture problems are solved, then, as in standard nonlinear FEM, you must store the unknowns- enriched dofs included- value at all step since the new solution depends on the previous one.

 Hope that it was oh help.


Subscribe to Comments for "Crack Propagation using XFEM ( What happens to dof's corresponding to enriched nodes)"

Recent comments

More comments


Subscribe to Syndicate