Application of a strip-yield model to predict crack growth under variable-amplitude and spectrum loading – Part 2: Middle-crack-tension specimens

In previous work, fatigue-crack-growth tests were conducted on middle-crack tension, M(T), and compact, C(T), specimens made from the same D16Cz (clad) aluminum alloy sheet. These tests were conducted over a wide range of stress ratios (R = Pmin/Pmax = 0.5 to 0.75) to generate crack-growth-rate data from threshold to near fracture. These tests were used to generate the effective stress-intensity factor range (DKeff) against rate curve using a crack-closure model. The analyses collapsed the rate data from both specimen types into a fairly narrow band over many orders of magnitude in rates using proper constraint factors. Constraint factors were established from single-spike overload and the constant-amplitude tests. Herein, the life-prediction code, FASTRAN, which is based on the strip-yield model concept, was used to calculate the crack-length-against-cycles under constant-amplitude (CA) loading and the single-spike overload (OL) tests; and to predict crack growth under variable-amplitude (VA) loading and simulated aircraft loading spectrum tests on the M(T) specimens. The calculated crack-growth lives under CA and an OL tests were generally within ±20% of the test results, but slower crack growth under the double-shear fatigue mode, rather than single shear, may be the reason for some of the larger differences. The predicted crack-growth lives for the VA and Mini-Falstaff spectrum tests were also short by 25–45%. A modified model with some assumed notch constraint effects matched the spectrum tests quite well. Issues on the crack-starter-notch effects under spectrum loading are discussed, and recommendations are suggested on avoiding these notch effects. 2011 Elsevier Ltd. All rights reserved.