Report, Test of HIFROST Longitudinal Holding Coil (7/19/6)

Transcription

1 Report, Test of HIFROST Longitudinal Holding Coil (7/19/6) D. Crabb 1, D.G. Haase, B. Jelinek, P. Kingsberry 2, S. Kucuker 1, R. Timberlake and C. Westerfeldt Testing has been successfully completed at TUNL on the longitudinal (solenoidal) superconducting polarization holding coil for the HIGS FROzen- Spin Target ( HIFROST ). This coil was obtained along with the HIFROST cryostat that was originally designed and built by Tapio Niniikosky at CERN and is currently being rebuilt at the University of Virginia under the direction of Don Crabb. Hall-probe measurements in liquid helium indicate a central field of.65 tesla at 2 amperes, with the Hall-probe rotated 9 degrees 3 times to check that the measurements were not dependent on probe directionality. The maximum variation of this field along the axis of the 1-cm target region, measured in 1-cm increments, is 2.4% from the average value. A jig to suspend the coil in liquid helium for the tests was designed and built at TUNL. It consisted of two circular stainless steel plates connected by 3 support tubes. The coil shell was mounted on the smaller, lower plate, while the upper plate was clamped as a lid to the top flange of a 42-inch-high Cryofab helium Dewar. Holes were drilled in the lower plate to allow liquid helium to fill the coil shell, and in the supporting rods to eliminate Taconic oscillations. The coil leads were screwed to 2 brass rods that were held by a swinging Teflon bushing anchored on one of the coil mounting screws. The rods extended through another Teflon bushing clamped to a T-connector on the lid. The inside of the hole in the lid was insulated with Kapton tape. 2 heavy copper clamps were fixed to the outer ends of the connecting rods. 12-AWG solid copper wire was soldered to the clamps for connection to a 4-V, 5-A Hewlitt Packard 6269B power supply configured for current control. A 36-inch-active-length helium level sensor ran through a Quick-Connect fitting welded to the top of a hole drilled in the upper plate, extending down outside the coil. The level read-out unit was calibrated for the full length of the sensor and zeroed with a 41-ohm load resistor to simulate the length of the probe outside of liquid helium (non-superconducting). 1 University of Virginia, Charlottesville, VA. 2 Massachusetts Institute of Technology, Cambridge, MA and TUNL 1

2 A guide tube for the helium transfer line was welded through another hole drilled in the periphery of the lid, with another Quick-Connect on top to hold the transfer line. Finally, another Quick-Connect butt-welded over a hole drilled in the center of the lid allowed positioning of a longitudinal Hall probe along the axis of the coil via a flaired tube extending to within 1 inch of the lid and welded through the lower plate. The tube extended to within 2.3 cm of the upper edge of the coil. Before testing, the helium transfer line and test Dewar isolation chambers were pumped out overnight with an Alcatel ASM leak-check pump. A nitrogen relief valve and vent line were attached to the two vents on the Dewar nitrogen jacket, which was then filled with liquid nitrogen overnight. To begin testing, helium gas was blown through the Dewar. The distance from a point roughly at the longitudinal midpoint of the coil to the top of the Hall-probe Quick-Connect was measured, and the Hall probe marked at that distance from its lower end for reference. The resistance of the coil circuit was measured at 1438 ohms. After installing and clamping the test rig, the 1-liter helium-supply Dewar was pressurized with a line from a helium gas tank. The test Dewar was filled with helium, and the level monitored throughout to maintain a minimum sensor level of 2, which kept the surface of the helium several inches above the coil connections. The coil-circuit resistance was monitored as the resistance went to zero in superconducting mode. The Hall probe was initially positioned with the active end roughly at the longitudinal midpoint of the coil (at the point referred to above). The current was ramped up incrementally from zero, and the field strength recorded. A quench occurred at 14 amperes. After disassembly, we discovered one of the coil leads had broken at its connection point on the connecting rod. These leads are not only fragile (1-micron niobium-titanium?), but had apparently been previously soldered, damaging the core of the wire. But sufficient extra wire had been wrapped on the coil to allow us to re-wrap the broken end of the coil wire on the small brass screw on the connecting rod. Indeed, sufficient extra wire still exists that similar breakages in the future should not affect the performance of the coil. It should be noted that Kapton tape was used to support the coil leads on the mounting shell, but this become detached in the helium, unlike the Kapton in the T-connector isolating the connecting rods, which held up well. The current was then successfully ramped up incrementally to the design current of 2 amperes, which was expected to produce a field of.6 T. The actual central field strength was.646 T., which surpassed our expectations 2

3 and the requirements of our experimental program. The values obtained are plotted in the first figure below, which clearly demonstrates the linearity of the current-field relationship. Field v (increasing) Current I(A) Figure 1: The current was then ramped down via identical current values. This roughly reproduced the original field strengths, thus ruling out any hysteresis effects. The field values with decreasing current are plotted next (Fig. 2). Since the linearity of the current-field relationship was established, the current was held to 1 amperes for the field mapping. We began at the bottom (the downstream end) of the interior of the coil shell, or.8 cm beyond the bottom edge of the coil. Readings were taken every centimeter up to 23 cm from the bottom, well above the upper edge of the coil (and indeed, into the lower edge of the Hall-probe guide tube, which extended to within 21.1 cm of the bottom). The values obtained are shown below. On the plot, the coil extends from x=.8 cm to x=18.7 cm. If, as expected, the 1-centimeter-long target we plan to use is positioned judiciously near the midpoint of the coil (at x=9.75 cm), the maximum variation in the field homogeneity over its length along the axis would be 2.4% from average. This is more than adequate for our 3

4 Field v (decreasing) Current I(A) Figure 2: frozen-spin target program. The probe was then returned roughly to the midpoint and rotated 9 degrees 3 times, and showed no variation in the field strength due to the directionality of the probe. The possibility of constructing saddle coils on the wall of a duplicate refrigerator target chamber, to provide a holding field for transverse polarization, is now being explored at TUNL. It should be noted that the test system described here has been engineered for dual use, in anticipation of the different geometry of our transverse-field Hall probe. 4

5 Field v Position Distance (cm) from downstream end, target chamber Figure 3: The coil extended from x=.8 cm to x=18.7 cm. 5