Extended Flattop Operation of Bevatron Results from Studies of Generator Dovetail Failure

In mid-December 1966, severe fracturing of the Bevatron generator pole dovetails was discovered before the failure became catastropnic. New poles had to be procured and installed on both generators. This refurbishing consumed six months during which the Bevatron was inoperable--raising doubts in the scientists minds whether the failure could indeed be termed noncatastrophic. The failure was eventually attributed to a fatigue condition arising from increased alternating stress loads associated with generator speed changes during flattop. When the generators were returned to service a restriction was placed on the length of flattop. Cracks again began to be seen on the outermost laminations after several months. Many means were investigated to make corrections by mechanical changes, but all such schemes were repugnant because of large costs and extended shutdown for installation. In parallel with the mechanical studies, electrical engineers were investigating different electrical approaches - schemes that would achieve flattopping but minimize generator speed change. In June 1968 a test installation was made of the most promising method. The restricted flattop then being allowed was achieved with less generator speed change than occurs during normal full-energy pulsing. Flattop times have gradually increased from 0.6 to 2 s, while generator speed change has still been held to no more than normal pulsing. No propagation of existing cracks or initiation of new cracks has been observed, but careful surveillance continues.