Interpreting fatigue test results using a probabilistic fracture approach

The foundation for the structural airworthiness of many military aircraft is an airframe fatigue test conducted using representative loading. Traditionally, the results from the fatigue test are scaled by empirical scatter factors to derive a safe economic life of the airframe aligned to an acceptable probability of structural failure. This paper illustrates a relatively novel approach to the estimation of the life of the fleet, utilising a probabilistic approach, and considering the wing attachment bulkheads of a specific fighter aircraft. Probabilistic fracture mechanics is used as the basis of the analysis and where possible, all relevant data, including crack growth and its variability, have been collected directly from the aircraft's fatigue test supplemented by laboratory coupon data. The example uses available initial crack size data and associated crack growth rates to estimate the stress levels at critical locations. The benefit and accuracy of the fleet aircraft fatigue monitoring system is also considered. The analysis attempts to include and quantify the effect of all parameters that significantly influence the growth of cracks in metallic airframes, and illustrates the derivation of a probability of failure versus flight hours. The paper also identifies deficiencies in the data required to fully populate the model, and describes current test and teardown activities which are intended to provide information about the defect types which might affect failure of fleet aircraft.