Fracture mechanics versus Damage mechanics

[ChrisHarvey]

Hello,

I was wondering if someone could clarify the major differences and limitations of Damage Mechanics and Fracture Mechanics.

I'm more familiar with the fracture mechanics approach, but have never done any damage mechanics before. I'm just trying to sort out in my head how fracture mechanics is situated relative to damage mechanics.

I appreciate the fundamental differences such as:
- damge mech = does not treat individual cracks but instead uses constitutive laws to give defect density variables.
- fracture mech = considers individual damage modes.

However, having compared quite a few sources on the subjects, I'm left quite confused. One says the damage mechanics does not consider individual cracks, but another says it is possible to track a crack's propagation using this method. One source describes the fracture mechanics approach as a boundary value problem limited to simple cases (and thus making it more generic). Surely both approaches depend on the boundary conditions of the problem!? Also, fracture mechanics is in one case described as being a more global approach. How can this be true when it targets individual defects?

I suspect there's some significant bias depending on author's own preferences.

Any guidance will be gratefully received,
Bet regards,
Chris

 

[PerennialII]

Yeah, would say the longer the discussion about the topic the harder the differences are to identify. Actually would say (personally ) that the differences have little to do with respect to what is being modeled, since both do apply quite nicely to similar 'subsets' of problems (be that for example cracks in fracture mechanics), damage mechanics overall having a broader scope & range than fracture mechanics.

The 'true' difference would say is in what is being "done" to control the event (be that then for example crack initiation and propagation). Fracture mechanics utilizes balance laws (J-integral for example) or state variable derived criteria (SIF for example might be categorized this way, or the whole "stress - based" approach to fracture) to come up with means to control the problem of fracture, and it does this without modifying (introducing additional internal variables) the underlying constitutive model(s) but rather works "on top" of them (in general using thermomechanics). Damage mechanics, however, introduces a new state variable "damage" (be it scalar, vector, matrix, tensor then) to the constitutive model along with it's evolution equation (and failure & initiation criteria etc.) and coupling to other state variables to do its thing.

Some models which persist within fracture mechanics have more to do with damage mechanics (or borrow "elements" from it) and in many cases both interact plenty (and both are needed for solution of certain problem, take ductile crack propagation for example when done the "proper" way ). An interesting animal no matter, although don't know how many years have spent solving related numerical convergence problems .

 

[ChrisHarvey]

Thanks PerennialII... just the kind of overview I was after.