Abstract
1-Introduction
2-Specimens
3-Experimental equipment and loading
4-Theory and methods
5-Results and discussion
6-Conclusions
Acknowledgements
References
Abstract
An experimental mean curve and a design fatigue curve corresponding to 95% survival probability were derived from realistic fatigue experiments on a non-welded water pressurized piping component with primarily focus on high cycle fatigue. The components were subjected to a synthetic variable amplitude bending deformation. Comparison with the results obtained for a similar piping component with a circumferential butt weld allowed the determination of an experimental fatigue strength reduction factor. Comparison with the fatigue procedure and design curve in ASME BPVC Section III allowed to quantify its conservatism with regards to accounting for the presence of a weldment and more generally transferability.
Introduction
Weldments are considered critical for the fatigue strength of structures or components. Fatigue cracks do namely tend to occur in the vicinity of welding joints rather than in the smooth base material. Welds represent indeed a local structural discontinuity or stress concentration which results in a general fatigue strength reduction, i.e. the load level inducing a given fatigue life will typically be lower for the component or structure including a welding joint. The stress concentration introduced by the weld can be related to geometrical notches in for instance weld toes or different local material properties resulting from the welding process. In design, the fatigue strength reduction factor (FSRF), here denoted Kf, quantifies this detrimental effect of a stress concentration. This quantity is also designated as the fatigue notch factor or fatigue effective stress concentration factor. In the case of a welding joint, it is defined for a given number of cycles as the ratio of the fatigue strengths of the smooth or plain component and the welded component. It is often approximated conservatively by the stress concentration or notch factor denoted Kt. Different formula relating both factors have been proposed in the literature [1], but a classic approach introduces the notch sensitivity factor [2], q, defined in Eq. (1)