Bellows-Welding Study

摘要

Standing lip welds in circular discs of 0.005 in. thick Type 304L stainless steel were gas tungsten arc welded to investigate low fatigue life in welded bellows. Eleven variables, such as arc gap, alignment of the tungsten with the weld centerline, mismatch, etc., which are encountered in production welding, were simulated to study their effects on weld penetration. Weld current and arc voltage were monitored to determine if they were sensitive to variables that resulted in unacceptable weld penetration. The study was performed to minimize random variables. Precision chills were used to obtain uniform weld chill. A jewelers lathe was used to provide weld travel. A live center in the tail stock of the lathe was used to apply chill pressure. A microscope was used to align the tungsten electrode with the joint and to observe the weld. A welding power source, with precise control of welding current was used. Discs were deburred. Eight cross sections were examined to measure weld penetration. Weld penetration values varied considerably within individual welds. Minimum values were 25 to 50% lower than maximum values. Also photographs of weld cross sections from fatigue tested bellows were examined to measure weld penetration. The range of penetration values from these bellows was the same as found within individual welds. Variable weld penetration was attributed to variable weld chill and arc wander. An attempt was made to study the effects of weld chill, and a trend was observed. No attempt was made to study arc wander, but it appeared to be less in the last series with burr-free discs. Weld current arc voltage could not be correlated with production variables or variable penetration. (ERA citation 08:007922)