Full text
| |
| Author(s): |
Diego Felipe Sarzosa Burgos
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Escola Politécnica (EP/BC) |
| Defense date: | 2012-04-18 |
| Examining board members: |
Claudio Ruggieri;
Theodoro Antoun Netto;
Carlos Eduardo Chaves;
Gustavo Henrique Bolognesi Donato;
Leonardo Barbosa Godefroid
|
| Advisor: | Claudio Ruggieri |
| Abstract | |
This thesis investigates the effect of crack closure on fatigue life predictions in homogeneous material and welded joints. It presents a methodology for numerical simulation of the phenomenon of plasticity-induced crack closure by using the finite element method. Experimental tests of fatigue crack propagation on ASTM A516 Gr 70 steel at room temperature were performed using typical specimens of fracture mechanics in two conditions: welded and homogeneous; these results were used to validate the numerical methodology for calculating the crack opening load. Comparisons between experimental propagation life and estimations obtained by numerical integration of propagation law, including the effect of crack closure, have been made using various criteria to calculate the opening load .The numerical analyses allow to predict the opening load by analyzing the evolution of stresses, deformations and displacements fields in the region near the crack which is growing and is influenced by the contact of the crack flanks. Lower levels of protection due to crack closure were obtained for the specimen with mechanical heterogeneity when compared to the homogeneous specimen in the numerical analyses. During the test of fatigue crack propagation the homogeneous specimen had smaller crack growth rates when compared to the welded specimen. Fatigue life predictions without including the effect of crack closure result in conservative estimations around 20% for all studied cases. Three-dimensional analysis of the phenomenon of crack closure shows that the fatigue crack growth in stage II appears to be controlled by the delay that occurs on the specimen surface. Including the effect of crack closure in life predictions, the estimates of fatigue life were quite close to those obtained experimentally, with conservative differences between 1% and 15% depending on the method of opening load calculation. The results show that the fatigue life predictions can be made more accurate and yet remain conservative when the effect of crack closure is included on the methodology for calculating the remaining life of structural components. (AU) | |