Thermal diffusion and quench propagation in YBCO pancake coils wound with ZnO and Mylar insulations

The thermal diffusion properties of several different kinds of YBCO (yttrium barium copper oxide) insulations and the quench properties of pancake coils made using these insulations were studied. Insulations investigated include Nomex, Kapton, and Mylar, as well as insulations based on ZnO, Zn2GeO4, and ZnO–Cu. Nomex, Kapton, and Mylar, chosen for their availability and ease of use, were obtained as thin ribbons, while the ZnO based insulations were chosen for their high thermal conductivity and were applied by a thin film technique. Initially, short stacks of YBCO conductors with interlayer insulation, epoxy, and a central heater strip were made and later measured as regards their thermal conductivity in liquid nitrogen. Subsequently, three different pancake coils were made. The first two were smaller, each using one meter total of YBCO tape present as four turns around a G-10 former. One of these smaller coils used Mylar insulation co-wound with the YBCO tape, the other used YBCO tape onto which ZnO based insulation had been deposited. One larger coil was made which used 12 total meters of ZnO insulated tape and had 45 turns. Temperature gradients were measured and thermal conductivities were estimated from these coils; the results obtained were compared to those for the short stacks. Quench propagation velocity measurements were performed on the coils (77 K, self-field) by applying a DC current and then using a heater pulse to initiate a quench. Radial NZP (normal zone propagation velocity) values (0.02–1 mm s − 1) were two orders of magnitude lower than axial values (~10–20 mm s − 1). Nevertheless, the quenches were generally seen to propagate radially within the coils, in the sense that any given turn in the coil is driven normal by the turn underneath it. This was due to the fact that while the radial NZP is much lower than the NZP along the conductor (~100 ×) the distance by which the normal zone must expand longitudinally is much larger than the distance by which it must expand radially to reach the same point; in our case this ratio is ~ 1600.