No abstract

No abstract

A 100 kG superconducting magnet with field uniformity better than 1 part in 104 over a 1 cm diam sphere has been constructed from composite Nb3Sn superconducting tape. Data showing the performance of the coil are presented. The field profiles depend both on the magnitude of the field and on the field history; field uniformity is considerably reduced at low fields after an excursion to higher fields, because of trapped flux. Neither the field nor the field profile at a given current agrees with the values calculated for a copper coil of identical geometry. The discrepancy results from the anisotropic diamagnetic properties of the superconducting tape winding, and is semiquantitatively understood from an analysis based on the critical-state model. The construction of the coil is described, particularly the electrical connections between modules and the methods used to obtain field uniformity.

This paper describes some of the experimental methods we have used in building high-field superconducting coils. Field-level measurements have been made using Hall probes, magnetoresistance probes, and search coils. Copper magnetoresistance probes are inexpensive, simple, and reproducible enough for routine testing; such probes can be built into a coil. Reproducibility is better than the normal limits of panel meters or X-Y recorders. The search coil and fluxmeter (or ballistic galvanometer) is used as a primary standard for calibrating other probes. Absolute accuracies of about ½% are attainable by calibration against an NMR probe. We have determined field uniformity by direct measurement of the field gradient. A small coil is oscillated ±1 mm at 10 Hz in the magnetic field. The ac voltage from the oscillating coil, which is directly proportional to the field gradient, is amplified and detected with a lock-in amplifier. Numerical integration of the gradients gives high-resolution field plots. This straightforward apparatus can detect a field gradient of 1 G/cm in a 100 kG field. Our coils are constructed in modules, with each module consisting of a pair of disks or pancakes wound from superconducting tape. It is obviously useful to detect which module in the coil is limiting. For this purpose we use a set of shunt resistors across the modules, and record the voltage across the shunts on an oscillograph as the transition from superconducting to normal state occurs. In this way, one can deduce where the normal region originates in the coil, and can tell something about the way in which it propagates.

No abstract