Evaluation of spectrum fatigue crack growth using variable amplitude data

The ability to understand and predict fatigue life remains a key technical factor in maintaining aircraft fleets which are required to safely operate up to their design lives, and sometimes beyond. In particular military aircraft are often required to achieve long lives under demanding operational conditions. The load spectra applied to these aircraft are complex and highly variable, and experience has shown that traditional fatigue prediction tools do not always perform well in calculating the lives of modern, highly optimised airframes. This paper summarises a recent semi-empirical model that appears to be capable of producing more accurate fatigue life predictions using flight load spectra based on realistic in-service usage. The model relies on experimentally derived variable amplitude crack growth data. The model itself is founded on the observation that, as a first approximation, for lead cracks in many materials the log of the crack depth is proportional to the number of cycles or spectrum loading blocks. This relationship can also be used to define the effective crack-like size of the initiating defect or discontinuity. The implications of this exponential growth, including the crack depth dependency of the crack growth rate, are discussed leading to a reconciliation of a Paris-type growth model with that of the earlier Frost and Dugdale stress-cubed growth relationship.