Experimental Plan for the Development of Equivalent Crack Size Distributions and a Monte Carlo Model of Fatigue in Low and High-kt Specimens of Corroded AA7050-T7451

ABSTRACT

Fatigue life data from low and high-kt specimen are difficult to compare due to the smaller sampling volume of high-kt specimens. As a result, equivalent crack size (ECS) distributions developed using low-kt specimens cannot be used in high-kt conditions such as near fastener holes. This experimental plan addresses this issue for corroded AA7050-T7451. Fatigue crack growth (FCG) and fatigue life tests will be conducted on low and high-kt specimens which have been designed to minimise the differences between the specimens other than the different stress field. FCG data and ECS distributions will be developed for both specimens types. A Monte Carlo model will then be used to test if the low-kt ECS distribution can accurately predict the fatigue lives of the high-kt specimens. If successful, this will allow low-kt ECS distributions, which are easier to derive, to be used under high-kt conditions.

EXECUTIVE SUMMARY

Corrosion of the fastener holes in modern aircraft, such as the fleet of the Australian Defence Force is a widespread and persistent problem. It is desirable to be able to accurately predict how this corrosion affects the fatigue performance of aircraft. However, it is difficult to compare fatigue life data from low and high-kt specimens due to the greatly reduced sampling volume of high-kt fatigue specimens. This means that equivalent crack size (ECS) distributions for corroded material developed using low-kt specimens are typically of limited use in high-kt conditions such as near a fastener hole in an airframe. This is problematic as ECS distributions are easier to develop under low-kt test conditions.

The experimental plan addresses this issue. Fatigue crack growth and fatigue life tests will be conducted on low and high-kt specimens of corroded AA7050-T7451. These specimens have been designed to minimise the differences between the specimens other than the difference in stress field. Fatigue crack growth data and equivalent crack size distributions will be developed from both the low and high-kt specimens. A Monte Carlo simulation will then be used to see if the ECS distribution from the low-kt specimens can accurately predict the fatigue lives of the high-kt specimens. If this approach is successful, it will be possible to use low-kt equivalent crack size distributions, which are easier to derive, to predict the fatigue behaviour under high-kt conditions.