A B S T R A C T Experimental fatigue data for butt-welded joints in as-welded condition and under constant amplitude tensile loading were analysed using the effective notch stress system and a new master curve analysis that takes the local stress ratio, R local , into account. The local stresses needed for computation of R local are calculated with the notch strain approach in conjunction with the reference radius concept. The main focus was to predict with the derived master curve the fatigue strength of peened butt-welded joints. The lowest surface residual stresses after peening were first estimated based on reported measurements and an analytical lower bound result. The predictions showed quite similar strength dependences and FAT values as reported for high-frequency mechanical impact treated welds for applied stress ratio R = 0.1. The benefits of peening reduce faster for higher strength steels when R increases. When R = 0.5, the FATs are practically the same for all steel grades.Keywords butt joint; effective notch stress; fatigue; improvement; local stress ratio; residual stress.
N O M E N C L A T U R EASW = as-welded condition ENS = effective notch stress approach FAT = IIW fatigue class, fatigue strength corresponding to two million cycles GMAW = gas metal arc welding HAZ = heat affected zone HFMI, HFP = high frequency mechanical impact treatment (generic term) HSS = high strength steel IIW = International Institute of Welding MSSPD = minimization of the sum of squared perpendicular distances from a line NS = nominal stress approach QC = quenched and cold-formable steel SINTAP = structural integrity assessment procedure SWT = Smith-Watson-Topper approach TIG = tungsten inert gas welding UP = ultrasonic peening (device name) C = fatigue capacity E = Young's modulus f u = ultimate strength, nominal f y = yield strength H = cyclic strain hardening coefficient K f = fatigue effective fatigue stress concentration factor between notch stress and nominal stress K HP = improvement factor K m = structural stress concentration factor Correspondence: T. Nykänen.