Dear connections,
I encountered a problem in that the results of the stress intensity factor SIF in the two methods did not match. And I think the fault lies in abaqus (EFM intégral J).But I do not know where exactly!
So really I need your help if it is possible.
You find in attachment file which contains more information.
https://drive.google.com/file/d/112TTvbYqiU5KBAz5J0Sh9hRGSlAYa2wh/view?usp=sharing
Please see our article:
Stress Intensity Factors Evaluation for Rolling Contact Fatigue Cracks in Rails
It is impossible to give any answer without knowing details of crack model (cae or inp file). The only thing that can be noted is that the critical stress intensity factors for real materials are much more lower than result obtained with abaqus: https://en.wikipedia.org/wiki/Fracture_toughness
Check J - Integral again.
f(lamda) = 1.105 to 1.2 for double edge crack , so that, K1 ranged from 55.25 to 60.
Hi Bounedjar Adel
I had a look at the attachment and noticed it might be in French. But at least I think I understand what you are working on.
I think you have used a quarter model. Since it is in symmetry, did you half the stress or force applied on the boundary condition to the left? If not, you will will need to double the stress you apply in the analytical solution.
Next is on dimensional consistency. Not sure what parameters you used in the geometry, but Abaqus does not specify dimensions, and you will need to be careful when inputting them. Example, if you used mm in dimension-ing the model, stress in Nmm^-2 will be in MPa instead of Pa and so on.
Try doing a quick check, might be a simple fix. Hope it helps.
Regards,
Wee
The example of quarter model of the double-edged notch specimen is available in Abaqus Benchmarks Guide sect.1.16.1 Contour integral evaluation: two-dimensional case: http://wufengyun.com/v6.14/books/bmk/default.htm
The *.inp or *.py files presented there can be used as templates. The convergence to theoretical result is very good.
The error might be because of the following reason:
1. If you are using quarter model then you need to divide the forces by a factor of 4.
2. Dimensional inconsistency. Actually Abaqus is dimension-less so you need to be careful while inserting the properties. It's always better to work with SI units.
3. And lastly it goes without saying that if you are using coarse mesh you will have to reduce the element size to reduce error.
Hope this will help you.
Regards
Ateeb
Please see this paper
Article Stress Intensity Factors Evaluation for Rolling Contact Fati...
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress