For many years the fatigue behavior of welds and weldments has been studied in the laboratory in terms of the geometry of the members, the stresses to which the members have been subjected and the materials of which they are fabricated. Such information has served as a basis for the fatigue design of numerous welded structures. However, because of the simplicity of the laboratory tests when compared with real service conditions and many simplifying design assumptions, the resulting design provisions have been very general and often overly conservative.
The recent developments in structural safety and reliability have provided bases upon which random loadings and reliability criteria can be used to develop simple but more realistic fatigue design criteria. To provide an indication of the applicability of some of these new hypotheses, laboratory tests conducted on butt-welded joints in structural steel and tested under a tension-biased random loading have been compared with a proposed random loading fatigue design criteria. Although of limited scope, the excellent correlation of the data indicates that more realistic design procedures for random loadings now are possible and can be used readily for the fatigue design of welded structures.
Although fatigue has been studied for nearly 150 years, research on the fatigue behavior of weldments and welded structures is much more recent in the first fatigue studies of weldments were introduced only about 50 years ago. Since then numerous laboratory fatigue studies of welds and weldments have been conducted, the results of which have been used in the development of fatigue design requirements.
The laboratory fatigue data is only one of the factors that must be considered in developing design specifications and in providing structures with adequate fatigue resistance. The expected quality of the fabrication of the structure should be included, the loading history for the total life of the structure predictable, a fatigue factor of safety or level of reliability selected, and the uncertainty produced by the many variables that affect the fatigue behavior taken into account. Unfortunately, the effects of many of the fatigue factors have not been sufficiently well defined to provide accurate quantitative evaluations; in many instances only qualitative indications can be introduced at this time. Nevertheless, theoretical developments now have been introduced that make possible in design, the inclusion of estimates of the effects of many of the fatigue factors.
This paper provides a summary of some of the theoretical developments introduced into fatigue analyses in recent years and a simple fatigue design criteria based on these theoretical relationships. The resulting design criteria incorporate the results of the laboratory tests and approximations of the factors for which there is limited information. As additional data are developed, the assumptions that have been made can readily be adjusted.