Fatigue Strength Reduction Factors for Socket Welds as a Function of Leg Length

Socket welds in nuclear power plant piping systems have had a history of failure due to high cycle fatigue. Cyclic bending loads, typically from vibration, have caused cracking either at the weld root or at the toes. A testing program was conducted to measure the effectiveness of proposed changes in socket weld geometry and fabrication [3]. One of the significant results of that testing program was that if the socket weld profile was modified such that the weld leg length along the pipe side is at least double the Code minimum requirement the fatigue life of the weld becomes approximately as good as that of a butt weld. The test results demonstrated that the fatigue strength reduction factor (FSRF) is about 37% smaller for socket welds with the longer weld leg, than for a standard size weld. Many plants have been replacing or building up their socket welds in locations susceptible to vibration to the 2×1 geometry to take advantage of the improved fatigue strength. However, there have been situations where welds were built up from the Code minimum, but inspections determined that the leg length was somewhat less than twice the minimum required and the weld was not accepted as meeting the 2×1 geometry. It is not known how the fatigue strength reduction factor would vary as a function of leg length for intermediate sizes. It is also of interest to determine the effect of increasing the weld leg length on the fitting side. For example, is it possible that increased fatigue resistance can be provided by adding additional leg length on the fitting side instead of the pipe side, or can a greater benefit be achieved if they were provided on both sides?