This paper describes an experimental study on fatigue strength of welded tubular T joints, performed in small and large-scale models fabricated from mild steel tubes. Fatigue tests were conducted under three kinds of loading patterns: constant amplitude loading, block programmed loading and random loading. In block programmed loading tests, the period of load sequence was varied, while the load spectrum was the same. Thus, the effects of the difference in load sequences upon fatigue lives were examined independently from the effect of load spectrum. Crack initiation life was hardly affected by the load sequence patterns. However, the crack penetration life or failure life were largely affected by types and periods of load sequences. The limit of validity of the Miner’s rule are discussed together with possible interpretations on the effects of the difference in load sequence patterns.

This content is only available via PDF.
You do not currently have access to this content.