Does anyone know about the Abaqus XFEM based on Linear Elastic Fracture Mechanics? The question is about how the fissure initiates under this finite element analysis framework. It's not like the XFEM is based on the cohesive behavior, where there is a phase damage evolution after the damage initiation criterion is met and before the element cracks into two parts; when the damage parameter D reaches 1, the crack takes place. But with XFEM LEFM, how does the crack initiate? It is not clearly explained in the documentation.
Zhao
XFEM uses CZM and LEFM based on VCCT for cracks modelling. For both the approaches a stress raiser or concentration is needed to initiate the crack as highlighted by John. In both the cases, crack will initiate when the stress reaches a maximum value at the notch. Modelling procedure, which I searched from XFEM notes is given as under:
Two distinct types of damage modeling within an XFEM framework
•Cohesive damage
•Linear elastic fracture mechanics (LEFM)
a. Cohesive damage
•Based on cohesive damage mechanics
•Uses traction-separation laws for damage initiation and evolution
•Damage properties such as initiation and evolution are specified in the material definition
b. LEFM-based damage
•Based on linear-elastic fracture mechanics
•Uses the virtual crack closure technique (VCCT) for damage evolution
.Specify damage initiation as part of the material definition such as Max principal stress criterion similar to CZM approach.
•Damage evolution and stabilization are specified as part of the interaction property definition.
In my view, this detail will help solving your problem.
The fracture is usually predefined, then grows depending on the damage criteria used, this can be displacement at failure maximum principal stress etc:
Definitions:
Mechanical -> Damage for Traction Separation laws. + Damage Evolution
If using a complicated model and not knowing where the fracture will start you can do an analysis of a stress based model, then apply the fracture to the area you identified as the likely point of failure. Usually this is the point of maximum stress, but using the point for maximum energy release rate is often a more accurate representation.
Hello, dear Iskandar, thank you for your anwser. You talked about the fracture mechanics based on the cohesive segment method. But I want to know more clairly the fracture using the XFEM based on the linear elastic fracture mechanics in Abaqus. By the way, the predifined crack is not necessary, to my knowledge.
Yes Zhao, unlike the cohesive zone model, predefined crack path is not necessary for XFEM. Still for damage initiation a stress/ strain based criterion such as quadratic stress/strain criterion is needed as highlighted by Iskandar.
Hello Himayat. In fact, what I want to know is about the initiation of crack by XFEM based on the linear elastic frature mechanics in Abaqus. Abaqus proposes two approches for XFEM, one is based on the cohesive segment method, another based on the linear elastic fracture mechanics. The first one is easy to understand. But regarding the second, it's not very clear, mainly concerning the crack initiation. Knowing that linear elastic fracture mechanics does not consider the damage or the damage evolution. So I just can't understand at which moment the element cracks into two parts. It's not likely that the crack takes place just when the stress or strain crirerion is met, but I can not say more about it.
Best regards
I have not used the LEFM based XFEM for fracture simulation. However, in my view it should be similar to VCCT approach in Abaqus which is based on LEFM.
Yes, exactly, it is very similar to VCCT approach. But the latter needs un prefiened crack so that the calculation can be performed, because it is for the crack propagation. But for XFEM based on LEFM (VCCT), the crack can initiate it-self. So I can't understand how it initiates? For XFEM based on cohesive segment method, un real crack acrosses un element when the damage parameter D reaches one, not immediately after the criterion is met. However, for XFEM LEFM, I want to know when un element cracks really, so that the VCCT approach can continue to calculate the propagation based on the self-initiated crack. I don't know if you understand my question. Looking forward for your anwser, thank you.
To the best of my knowledge, LEFM always requires a pre-existing flaw. The concepts of LEFM, such as the stress intensity factor and the strain energy release rate are based on describing the stress or energy state in the vicinity of a crack tip. If there is no crack tip these concepts make no sense. Thus LEFM can not predict initiation from 0 crack length. However you could assume the presence of a very small flaw and assume that a crack will initiate under the same conditions that would cause that flaw to propagate.
Zhao
XFEM uses CZM and LEFM based on VCCT for cracks modelling. For both the approaches a stress raiser or concentration is needed to initiate the crack as highlighted by John. In both the cases, crack will initiate when the stress reaches a maximum value at the notch. Modelling procedure, which I searched from XFEM notes is given as under:
Two distinct types of damage modeling within an XFEM framework
•Cohesive damage
•Linear elastic fracture mechanics (LEFM)
a. Cohesive damage
•Based on cohesive damage mechanics
•Uses traction-separation laws for damage initiation and evolution
•Damage properties such as initiation and evolution are specified in the material definition
b. LEFM-based damage
•Based on linear-elastic fracture mechanics
•Uses the virtual crack closure technique (VCCT) for damage evolution
.Specify damage initiation as part of the material definition such as Max principal stress criterion similar to CZM approach.
•Damage evolution and stabilization are specified as part of the interaction property definition.
In my view, this detail will help solving your problem.
Hello, i am having also some problems with cohesive segments and lefm approaches. Because when i try with both of them, LEFM results are always much bigger than cohesive segments approach. I am questioning if i am able to apply the things correctly. For cohesive segments i am selecting Maxps for initiation and energy modes for evolution in material part. But for LEFM, i use also Maxps for initiation in the material module, but for evolution i defined it in interaction module (fracture) and asssigned it to crack. Can anyone say if it is a correct way for both approaches?
Hi
from a pp by zhen-zhong Du:
Specify damage initiation in the material property definition and VCCT becomes active when damage criteria are met.
from documentation:
If there is no pre-existing crack for a given enriched region, the XFEM-based LEFM approach is not activated until a crack nucleates. The crack nucleation is governed by one of the six built-in stress- or strain-based crack initiation criteria or a user-defined crack initiation criterion.
Hi,
Is there any conclusion? What does the damage mean in XFEM-based LEFM? From my understanding, there are only two status, cracked or uncracked, STATUSXFEM should be either 0 or 1. I don't know what a decimal (STATUSXFEM=0.4) means near the crack tip.
Chen Yaxiong Typically damage variables do not jump from 0 to 1 suddenly. XFEM damage variable (also phase field damage variable) is typically a defined as a function of material attributes. For example, the damage variable can be defined as a function of elastic modulus. When damage initiates in an element the modulus of that element is set to zero. And until damage occurs the damage variable evolves following a function, which say defines increasing damage with decreasing elastic modulus. Hope this clears things a little bit.
Tousif Ahmed If it is cohesive-based, damage can be defined by the post peak stress status, but how it is defined in LFEM-based XFEM, a ratio of energy release rate to critical value?
Chen Yaxiong After damage is initiated (say, after the post peak stress ), one has to define the damage evolution. It can be total displaencemt after damage initiation or a critical energy value. You can find details in here: http://130.149.89.49:2080/v2016/books/key/default.htm?startat=ch04abk03.html#usb-kws-mdamageevolution
Does this answer your question?
- Badges
- Science topic