Hello, I am modeling a metal plate under fatigue tension load (direct cyclic approach in low fatigue analysis on Abaqus) even considering XFEM elements.
I have some convergence problems and some questions.
Is there someone who has dealt with this that can help me?
Hello, Danilo I am working on Prediction of crack growth from bolt hole in plate from last 6 months. I am in the final phase of the project. You can apply fatigue loading with XFEM efficiently in ABAQUS Command only. Try to go through ABAQUS Documentation.
Dear @Prajwal S H and @Mohammad Malekan,
Many thanks for your help.
I am trying to apply both XFEM and fatigue analysis, but it is not possible to assign an Hysteretic Energy behaviour (needed for the fatigue analysis) to the material and a MAXPS damge initiation criterion (which is necessary for the XFEM evolution). So I assigned the hysteretic behaviour only to the surface by the fracture criterion, type=fatigue. Is it correct? In this way the material degradation is not due to the number of cycles but to the energy evolution given in MAXPS criterion, but the surface crack evolution will follow the Paris Low or rather the fatigue behaviour.
Please, find in the follow the script related to the material/surface implementation.
@Daniol, Yes your steps are right, but in Fracture Criteria you need to specify
C1,C2,C3,C4,G threshold /G IC ,
This value is very important for the crack to initiate or grow.
Hello, Mohammad Malekan, its a simple Plate with hole problem with crack.
Dear @Prajwal, Thanks a lot. I set c1 and c2 equal zero because I wanted to calibrate them. Moreover, I saw an example in which G threshold /GC and G p /GC was set as zero. In that case, I guess defauls values were assigned by Abaqus. Is it right? Now I am trying with G threshold /GC = 0.7 and Gp /GC= 0.9. I will see if the analysis converges. Do you think these values can be considered as not so bad?
Do you have suggestion for these values (I know that I have to set them very carefully, but at this step I want to set the model and then calibrate the parameters).
Yeah, if u leave blank the software will take its default value. Depending on the material C3,C4,GTH/GIC etc changes.
Dear @Mohammad, I defined the fracture criterion as follows:
*Surface behavior
*fracture criterion, type=fatigue, mixed mode behavior=power
0.5,-0.1,5e-7,1.15,0.1,0.9,42200,42200
42200,1,1,1
Sorry, but I do not understand. I defined a material damage initiation/evolution according to traction separation laws (as it seems to be correct for XFEM). I defined the pre-crack surface (interaction, special, crack, ...) and then a fracture criterion according to Paris Law (the one I posted). Do I need to add another crack initiation/evolution?
Dear Mohammad,
in your opinion, can be correct to implement the fatigue behaviour only in the interaction response between the crack surface and the specimen?
As a matter of fact, I defined the damage for the material differenlty (using MAXPS for the initiation and an Energy criterion for the evolution). So the Paris Law fatigue is present in the model inly in the crack surface response.
Another question, do you think that using XFEM without definition of damage according to MAXPS can be considered as correct, or it is striclty needed to use this specific damage behaviour?
Many thanks in advance.
Dear Mohammad,
I meant that the parameters which govern the fatigue response (Paris Law: c1, c2, ...) have been inserted in the model only for the implementation of the surface behaviour (fracture criterion, ...). As a matter of fact, I assigned to the material a completely different damage response, which is not due to the cyclic fatigue (init: MAXPS and evol: Energy criterion). I do not know why Abaqus does not allow to apply both MAXPS (as you say strictly needed) and HYSTERESIS ENERGY (which, I guess, has to follow Paris Law according to fatigue behaviour) criteria for damage.
Dear Mohammad, Abaqus does not allow to use both maxps and hysteresis energy, but I guess that it is enough to assign fatigue criterion to the XFEM elements.
Do you have any suggestion for the evaluation of c1 to c4 parameters?
I'm doing fatigue analysis using ABAQUS XFEM the only informations i have is the Young Modulus (E) , the poisson ration (v) and also stress intensity factor K1c. With this informations can i model fatigue analysis using ABAQUS XFEM? Or do i need more informations in order to simulate fatigue analysis? If i need more informations from where i can obtain those information (my specimen structural steel E335 with two holes small and big)
Hello,
what you need most is the Paris law parameters (often referred as c and m which are implemented in Abaqus by c3 and c4). You can find an issue I posted on Researchgate more related to this question.
Please, see:
https://www.researchgate.net/post/How_to_find_Abaqus_parameters_for_Paris_Law_crack_initiation_and_growth?view=5a537668615e27d68b2a79a6
- Badges
- Science topic