Thermal Runaway of Conduction-Cooled Monofilament and Multifilament Coated Conductors

Abstract

We experimentally studied the thermal runaway initiating at a low critical current (Ic) part. This low Ic part is determined by the combination of two reasons in a real coil: (a) the unavoidable defects caused by the manufacturing process, which reduce local critical currents (and might not be uniform across the width of a coated conductor) and (b) the magnetic field distribution along the coated conductor. To simulate the thermal runaway using a short monofilament/multifilament REBa2Cu3Oy (RE-123) coated conductor, we artificially created a local defect (low Ic part) in a short sample by pressing using a drill bit (creating a defect close to one edge of a coated conductor) or bending (creating a uniform defect across the width of a coated conductor). The sample of the coated conductor was conduction-cooled to 30 K, and a magnetic field was applied (µ0H up to 2 T) perpendicular to the wide face of the conductor to control its critical current. Transverse voltages in a multifilament coated conductor were measured to obtain the transverse currents among the filaments through the copper layer. Thermal runaway currents (operating currents above which thermal runaway initiates) of the monofilament sample and those of the multifilament sample with additional Joule loss due to the transverse currents were determined and compared to study the effect of the transverse currents on the initiation of thermal runaway in the multifilament coated conductor. Experiments on the protection against thermal runaway were conducted. When a normal voltage (over a preset threshold) was detected, the supplied current would be decreased exponentially. The thresholds for protecting monofilament and multifilament coated conductors from degradation after thermal runaway were compared.

Analysis of critical current fluctuations as a means of checking the quality of high-temperature superconductor tape slitting

“…Such an approach disregards possible lateral distribution of the defects causing local I c reductions that has been revealed in magnetic mapping, optical scanning and Raman spectroscopy [9][10][11][12][13]. It is also clear that defects in different lateral positions could have different impacts on the Tapestar results and on the tape performance [14][15][16]. Nevertheless, in the experiments observing bubble creation in a nitrogen bath, it was found that the overheated zones appear in strips crossing the full width of the tape, probably because of current congestion around defects [17,18].…”