Research report | A Demonstration using Low-kt Fatigue Specimens of a Method for Predicting the Fatigue Behaviour of Corroded Aircraft Components

Abstract

Corrosion is well known to reduce the structural integrity of aluminium alloy aircraft components. In addition, it can cause early fatigue failures in components in which fatigue is not considered to be a life limiting factor. This is because corrosion damage, such as corrosion pits, is up to 100 times the size of the inclusions intrinsic in most aerospace aluminium alloys. The trailing edge flap lug of the F/A-18 Hornet aircraft is an example of an unexpected failure due to corrosion damage. In this report a Monte Carlo model is developed to simulate this phenomenon. This model predicts the fatigue lives of corroded and uncorroded specimens of the aluminium alloy 7010-T7651. It does this using high-quality fatigue crack growth data for this alloy from a previous research project (SICAS) combined with probability density functions for size of the corrosion pits and inclusions in this alloy. The distribution of the predicted fatigue lives is an excellent match for that observed in the SICAS project. The model was then extended to predict the location of fatigue failures. It showed that with good laboratory data the model could very accurately predict the location and life of pittinginduced fatigue failures.

Executive Summary

The unexpected failure due to fatigue of the trailing edge flap (TEF) lugs of a Royal Australian Air Force (RAAF) F/A-18 in 1993 showed that, if neglected, corrosion can severely reduce aircraft structural integrity. It has become apparent, since at least the 1980s, that corrosion is a major cost in maintaining fleets of aircraft. The Defence Science and Technology Organisation (DSTO) has accordingly conducted a great deal of research into this issue. This research has principally concentrated on how corrosion reduces the fatigue endurance of aircraft components.

 The F/A-18 TEF lug failure showed that corrosion can create new modes of structural failure. Specifically, the lug that failed had been designed to have an effectively infinite life. There was therefore no expectation that it was a fatigue critical part. Despite this the presence of corrosion pits in the lug made it fatigue critical and its failure due to fatigue was unexpected and nearly catastrophic.

 The research described in this report is intended to address this issue. The approach taken was to develop a Monte Carlo model of the fatigue life of corroded specimens of aluminium alloy 7010-T7651. The model simulates both the alloy’s metallurgical inclusions and corrosion pits using extreme value statistical distributions. The inclusions are spread randomly across the specimen’s surface while the corrosion pits are contained in corrosion strikes of set size and location. The model was then used to predict the fatigue life distribution of corroded and uncorroded 7010-T7651. 

The model’s fatigue life predictions were found to be very accurate when compared to the results of an earlier research program conducted at DSTO. The model’s prediction of failure locations were compared to experimental results from a small trial conducted as part of the current work and were again found to be accurate. 

It is concluded that the model developed here should be expanded to deal with real aircraft components and more complex corrosion conditions. It should also be combined with models for corrosion nucleation and growth to create an end-to-end corrosion prediction model for the RAAF’s fleet of aircraft.