The effects of potting on training and quench propagation in a large stored energy superconducting dipole coil

Abstract

A superconducting racetrack dipole coil was constructed to compare directly training and quench behavior in potted and non-potted coils. The stored energy of this coil was 175 KJoules at the conductor's short sample limit of 238 Amp with a peak field in the coil of 7.6 Tesla. The outward magnetic forces were restrained by rows of tie rods between side plates. Comparisons of training behavior were made for both steel and aluminum tie rods. Helium flow was provided by channels in the fiberglass cable tape allowing one-quarter of the conductor surface direct access to the helium supply. After training the coil to 90% of short sample limit, the tie rods were relaxed and the entire coil was vacuum impregnated with a standard clear magnet epoxy. After potting, the previous tie rod preloads were re-established. This resulted in a shallower training curve, and required retraining after thermal cycling. The unpotted coil showed no evidence of internal quench propagation below 80% of short sample, whereas the potted coil exhibited good quench propagation and energy dissipation at all currents, simplifying protection strategies.