| Summary: | Every one to two years, an operating tokamak fusion reactor requires maintenance and replacement of the internal components due to neutron damage. Previous solutions involve sectioning, assembling, and disassembling these components using the ports in between the toroidal field coils. To improve reactor reliability and simplify access, it is highly desirable that the toroidal field coils are demountable; this would also reduce the reactor downtime, when based on the tokamak and related magnetic fusion concepts. Compared to low temperature superconductors, high temperature superconductors have large cryogenic stability margins. In addition, the high thermal stability and improved passive quench protection ability of the non-insulated coil are advantageous, as is the low voltage operation, eliminating the need for high voltage electrical insulation as required in insulated cable coils. By combining the use of HTS in a non-insulated coil, a highly adaptable geometry presents itself for the inclusion of hundreds of demountable joints between coil turns. The joints have stringent requirements; they must be low resistance to minimize power dissipation, and to reduce the non-insulated coil radial current to < 0.5% of the operating current in the constrained geometry, thus ensuring that when accounting for joint variation from coil to coil, the toroidal field ripple limit of 0.5% isn’t exceeded. While pure indium compression joints have demonstrated the low resistances required (∼ nΩ), they are not easily demountable in this environment, and deformation of the joint region is possible. Pb37Sn63 with a melting temperature of 183◦C is used to solder the HTS tapes into the base metal plates of the coil, so a low temperature solder is proposed in order to maintain the integrity of the principal tape matrix in the coil while allowing for the benefits of a solder-based joint for obtaining low resistance. This thesis addresses whether the electrical resistance requirements of these joints can be met, and the associated challenges.
A novel vacuum pressure impregnation method was developed to couple superconducting tape stacks using three low temperature solder candidates: In52Sn48 (MP = 118◦C), In100 (MP = 156.6◦C), Ga100 (MP = 29.8◦C). To predict joint resistances, a finite element model is built and validated by experimental ideal joints; a 10 tape ideal joint has a 2% discrepancy, and a 40 tape ideal joint, a 32 % discrepancy. The VPI solder joints were experimentally tested at 77K; a distributed voltage tap system was used to infer the effective resistances to exit a superconducting stack, cross the solder layer, and enter the second superconducting stack. The normalized total joint experimental resistivities (0.54 – 0.68 µΩ · cm2 ) show good agreement with the model. Nonlinearity in joint I-V traces is modelled and explained to be a result of preferential HTS tape filling; the low current slope indicates the geometric resistance of interest in this thesis. The solder joint layer is then analyzed microstructurally; as a demountable joint is heat cycled, there is continual intermetallic growth between the liquid solder and the solid substrate. For liquid In52Sn48 solder and copper, this growth is quantified across three parameters of time, temperature, and solder joint thickness. This provides a critical joint lifetime, and an appropriate starting joint thickness of ∼ 100 µm. Thermal cycling of soldered superconducting tape stacks is then performed, simulating the heat applied to the tapes in the joint vicinity during demounting and mounting; this results in diffusion of the external tape copper layer into the bulk solder and oxygen-out diffusion from the YBCO layer. Through investigation of Ic, n, and R at 170◦C (operating temperature for In100), it is found that nickel electroplated tapes have higher levels of degradation than unplated tapes; in the latter, cases of simultaneously little to no Ic degradation and strong n degradation are observed, indicating decoupling of the two parameters. Finally, using the finite element model and the experimental results, joint resistances for a possible array of ARC fusion reactor scenarios are predicted; for operating conditions of B = 2 T, T = 10 K, the joint resistivities are 40 nΩ · cm2 for In100 and 124 nΩ · cm2 for In52Sn48. Using a realistic turn to turn joint area of a 100 cm by 2 cm, the electrical joint requirements of low power dissipation and a 0.5% radial current limit can be satisfied in the constrained geometry with low stresses (<150 MPa) in the joint regions. This thesis validates the use of a vacuum pressure impregnation process to couple superconducting stacks with low temperature solders, showing the viability of achieving joint resistances that are required at the operating conditions.
|
|---|