Effect of welding residual stresses on fatigue crack growth thresholds

Abstract Fatigue crack growth thresholds ΔKth were determined for friction stir welded butt joints made from aluminium alloys AA2024 and AA6013. Plotting the thresholds as a function of load ratio R showed distinctly higher amounts for welded joints as compared to those for parent material at small load ratios, but differences became smaller with increasing load ratio, until thresholds became finally identical for the highest R values. Applying Doker’s concept of two controlling parameters, namely ΔK and Kmax [1] , and plotting ΔKth versus Kmax, however, revealed that the effective threshold ΔKth,eff determined at very high R ratios was nearly independent on the alloy and, simultaneously, was identical for parent material and respective welded joints. Thus, differences in threshold behaviour were only caused by the second threshold Kmax,th, which was significantly higher for welded joints as compared to parent material. Differences in Kmax,th coincided with compressive residual stresses determined by cut-compliance measurements in terms of stress intensity factors Krs acting at the crack tip. Based on the analytical approach described by Doker [1] , only one characteristic Krs value was needed to calculate all thresholds of welded joints for 0 ⩽ R ⩽ 1 provided a base material master curve is available.