Since tokamak fusion experimental reactors try to sustain 100-200 second D-T burning, helium exhaust from plasma core is one of the major issues of the tokamak reactors. There have been proposed two major concepts about helium exhaust from plasma. They are concepts of magnetic divertor and pumped limiter. In both cases, a very large pumping system is required for the helium and fuel pumping. The cryosorption pump with 4.2K panels is considered to be a prime candidate for these applications. When molecular sieves are used as the adsorbent, it has been recognized that the pump may not be able to accommodate helium-hydrogen isotopes mixtures, because condensed deuterium and tritium will block the adsorbent surface and prevent helium pumping. This means that the cryosorption panels (4.2K) for helium will be surrounded by two chevrons, one at 77K and the other at 4.2K. Recently, cryopumping has intensively studied in the TSTA project in the United States. It has been shown that cryosorption by charcoal and cryotrapping by argon condensed layers will appear to work successfully with tokamak reactors. In the cryopumps, it is necessary to recover the pumped gases as quickly as possible in order to achieve low inventories of tritium. The regencration cycle will be determined by considering several items which include tritium inventory and safety problems.
Frictional sliding occurs on both a microscopic and a macroscopic scale. Sliding on a microscopic scale appears as discrete events called microslips. Microslips are inherent in all sliding events and are quite different from macroscopic instabilities such as stick-slips. It is thought that the “training effect” observed in quench current data from a superconducting braid may be caused by microslips. The mechanism of sliding motion and its effects at 4.2K were studied in detail for a number of metal/insulator pairs that model superconducting magnet windings; the results impact the performance of superconducting magnets. Organic surface coating materials are generally effective in eliminating macroscopic instabilities. Instrumentation used in these experiments includes a high-resolution extensometer and an acoustic emission sensor, both with sensitivities capable of detecting microslips (-1μm).
Energy released following cracks and bond failures were measured for an EPON epoxy near 4.2K. Crack events were monitored with an acoustic emission sensor; the energy released by each crack or bond failure was calculated from the temperature rise measured with thermocouples. Cracking was observed to be load dependent; this may account in part for the training phenomenon observed in bringing epoxy-impregnated superconducting magnets to full design field.
Lorentz force induced strand motion in a superconducting cable (or bundle conductor) was investigated by a combined technique of voltage and acoustic emission measurement. Both types of motion, reversible and irreversible, were observed. Dissipation energy for one slip event was less than a few mJ.
A superconducting bobbinless solenoid has been constructed and successfully excited without training up to the critical current of the conductor. The central field is 5.86T at the maximum overall coil current density of 300A/mm2. The inner diameter of the windings is 19.0cm, the outer diameter is 23.4cm, and the length is 26.7cm. The conductor is a NbTi/Cu keystoned compacted strands cable whose critical current is 4.6kA at 6.3T and 4.2K. The magnet consists of 2 layers with 345 turns of edgewise windings, a GFRP outercylinder and two sheets of cooling channel between them. It has no metallic part except the superconducting cable. Therefore, it is suitable for pulsed operations. Each component has been bonded with epoxy resin and the magnet has been shaped in one body with hydraulic presses under the radial pressure of 14.7MPa and the axial pressure of 29.4MPa. The pressures were determined so that the stresses in the windings become almost the same in the order of magnetudes and directions as thosewhich will be produced by the Lorentz force. The local wire movement considered as one of the causes of training has been reduced in this magnet. During e ci ation, however, the magnet deforms elastically. These strains of the windings in the axial and circu f rential directions were measured by strain gauges at the center turn of the magnet. At the same time the voltage oscillations considered to synchronize with mechanical vibrations of the magnet were measured by voltage taps. These voltage taps having intervals of 2, 5, 10.5cm were attached to the cable at various positions in the windings. From these experiments, it was found that the change of the axial strain was not uniform but stepwise for the change of the magnet operating current. In addition, the mechanical vibrations were induced by the sudden deformation of the windings. The thermal stabilities of the magnet were investigated by the heating experiments. A heater was placed on the inner surface of the bobbinless solenoid. The energy dissipated at the heater to quench the magnet and the propagation velocities of normal zone were measured. The static stresses and deformations in the magnet under the pressurized curing, cooling down and excitation were analyzed by a finite element method, using the program ISAS II which is the Japanese version of NASTRAN. The principal vibrations of the magnet were also analyzed, and the deformation modes were studied. From these data of stresses, strains and vibrations, the magnitudes of mechanical disturbances were estimated. On the other hand, the minimum quench energy of the magnet was estimated by the computer simulations of heating experiments. As a result, the following things become clear. Even in the case of this magnet, the mechanical disturbances have not been completely removed. These disturbances are induced by the irregular contractions of the windings in the axial direction, and they can trigger the quench, if the cooling condition of the magnet is getting worse.