Probabilistic fatigue damage assessment for short crack growth

Fatigue and fracture processes are characterized by randomness due to the uncertainties that are associated with the governing parameters. Typical sources of these uncertainties include variability in material properties, initial flaw sizes and shapes, geometry, loads, and other environmental factors. Probabilistic fatigue crack growth models that are suitable for the assessment of long fatigue cracks have been developed and applied by several researchers. The case for short fatigue crack growth, however, has not received the same level of attention. This paper presents a robust probabilistic model for short fatigue crack growth, which accounts for the effects of crack aspect ratio, stress ratio, as well as crack closure and retardation. Advanced fast probabilistic integration techniques in conjunction with the response surface methodology are used for the probability and sensitivity computations. The results obtained from these methods are compared and successful verified using the direct Monte Carlo simulation scheme, thereby confirming the robustness of the advanced techniques employed. The work reported herein is a valuable contribution of aging structures that are subjected to cumulative damage degradation mechanisms.