Crack growth retardation under aircraft spectrum loads

Abstract In the past, the prediction of the fatigue life of a part was often compared with Miner's rule (Σn/N = 1) wherein damage was computed by a linear accumulation of growth under individual loads. The difference between the predicted and observed damage for variable amplitude tests showed up as either retarded or accelerated fatigue life depending on test conditions. In this program, fatigue crack growth tests were performed with the compact tension specimen under aircraft spectrum loading to develop techniques for predicting fatigue life. Two models were investigated for their application to retardation behavior—linear cumulative growth and Wheeler. Constant amplitude crack growth rate data represented by the Forman equation was used as source data for the computer models. These models appear suitable for comparison with actual test results for diffusion-bonded Ti-6A1-4V and HP-9Ni-4Co steel at 0.3 and 0.2 carbon contents for the aircraft spectrums investigated. The 2219-T851 aluminum alloy, on the other hand, exhibited accelelerated behavior in relation to the linear cumulative growth scheme. An explanation is proposed by relating recent crack closure and spike overload studies in aluminum with the format of the applied spectrum. Also, the Wheeler model was found dependent on material and spectrum variables, and the Forman equation did not accurately represent crack growth rate below a rates of 10−5 in/cycle.