How to model a crack line inside the geometry?

[eispiata]

Hi everybody,

I would like to model a crack line inside a model. As you can see on the schematic picture that i attached i would like to model a crack inside the geometry and not lying on a boundary.

The file that i attached accounts for a crack lying along one of the principal stress axis. In know that in this case you can simplify the model into its quarter part. But this is just a first step. Cause i would like to be able to model a crack which is not necessarily lying along one of the principal stress axis. And due to the loss of symmetry in the latter case, i need to model a crack line inside the geometry.

My question is not about how to simplify the geometry of the model but it is about how to sketch and model a crack line in side the geometry.

I already know that you can use Special-> Crack -> Seam in the Interaction Module. I read up on that in the Abaqus documentation and it says that the crack line has to lie on a boundary of the model and cannot be located inside.

Do you have any idea of how to use the Interaction module to be able to model a crack line inside the geometry?

Thanks,

Malik

 

[xerf]

You can decompose the domain in 2 parts (made of the same material), with the crack and the ligaments laying along the "interface" (i.e. propagation path) between the 2 parts.

You have to partition the common boundary between the parts at the crack-tips. (Such that you will have a point at the crack-tips.) Then you create TIE constraints for the ligaments and prescribe at least normal contact for the crack edges, if the loading is compressive in order to preclude edge overlapping. If the load is only tensile, such that the crack remains open at all times, then you do not need to assign contact interaction for the crack edges.

As far as I remember, if you use "seam" you cannot preclude the overlapping of the crack edges if the crack closes.

 

[eispiata]

Hi,

Actually i don't need to make the crack grow but i just need to open it or close it since it will be subjected to either tensile or compressive loading.

Thus, all the nodes laying along the "interface" should be completely tied as if the material was continuous except those which are laid within the crack line.

In the case of a crack not laying along one of the principal stress axis, the propagation path won't match the crack line.

I definitely want to avoid any propagation of the crack cause i just want to get the strain and stress field in the vicinity of the crack, whatever the crack orientation is.

In you reply you are talking about ligaments, but i don't know exactly why do you mean by ligaments. Are you talking about an Abaqus feature called "ligament" which helps you tie the nodes at the interface or about the real physical ligaments. It might be a silly question but i wan to make sure that i understood well.

Thanks,

Malik

 

[xerf]

The "ligament" is the (not yet fractured) path in front of the crack tip.

You can use the approach I mentioned with or without propagation.