Assessment of the Size Effect in Fatigue Analysis of Butt Welds and Cruciform Joints

Reduced fatigue capacity of welded structures for larger thicknesses was introduced in design standards approximately 30 years ago. A significant amount of research on this topic was performed during the following years. In general the presence of a size effect was agreed upon. The size effect is considered to be dependent on the plate thickness at the considered hot spot in addition to size of attachment plate and type of dynamic loading. Only simplified recommendations on the size effect are included in most fatigue design standards. One reason for this is normal scatter in fatigue test data and also somewhat different recommendations based on these data in the literature. This has made it difficult to arrive at full agreement on recommended fatigue analysis procedures. In this paper a review of literature and design standards are presented together with a calibration of analysis method with fatigue test data. The effect of different parameters contributing to the size effect is illustrated. An attempt has been made to use the calibrated analysis model to also quantify the size effect based on crack growth analyses. Finally some recommendations on size effect to be used in fatigue design standards are presented.