Shown above, Fig.1, is the first automated precision magnetic field profiler. Initially the pole tips were blanchard ground to obtain a very parallel gap. Doing so has removed any radial component of the magnetic field, which is known to provide focusing - weak focusing. In the Spring semester of 2002 two students modeled several pole tip types and their focusing effect. We have chosen one of their designs and our machine shop produced two "fully annealled" 1006 soft steel pole tips to our specifications. To quantify the radial component and to verify modeled expectations a micro-stepper motor controlled radial positioner was mounted to the magnet. The stepper moved a Hall-effect probe along a radial path to the center of the pole piece. Since the positioner was completed before the new pole-pieces were ready, we tried it out on the existing "perfectly" parallel gap. It was good to verify that indeed the field was uniform:
Above, Fig.2, is the vertical field profile as a function of radius, Bz(r) (in Tesla), of the parallel pole pieces. Indeed the field was very uniform in the region of the beam, thus the magnetic field provided very little focusing (if any at all).
Perhaps to illustrate the focusing effect better, the following figure exaggerates the magnetic field lines that give rise to a restoring force, tending to push the ions back to the median plane.
The most important feature of the new pole tips is the most imperceptible. From the center of the pole to a 5 inch radius, the height drops 0.020 inches. Installed, this creates a 2% increase in the gap (from center to 5-inch radius), thus creating a decrease in the field of about the same, ~2 %, But the 2-dimensional curl of the B-field requires that decrease be made up in a radial component, thus creating the needed field for a restoring force.
Two students spent the spring semester of 2003 characterizing the new pole pieces. The measurement utilized the same system that measured the flat pole pieces. First, to calibrate the probes position, a small iron "needle" excited by a small copper coil was mounted on the positioner's bracket at a precisely known distance from the pole tips' center. The probe was run in and out with only the needle energized - a fit to the peak determined the radial 8.000 inch position from the pole tips center. Residual magnetization of the 12-inch magnet produced a significant background that needed to be subtracted, this was accomplished by first measuring the residual field of the magnet, then the iron "needle" field and the residual 12-inch field, then subtracting the former from the latter.
With the probe position calibrated, field profiles of the energized 12-inch magnet were taken at three coil current values. 20 Amps, 30 Amps, and 40 Amps were chosen as to compare the profile between to points in the non-saturated regime (i.e. between 20 and 30 Amps) and to compare a high-field non-saturated point with the onset of saturation. It was a concern that the desired field profile would not persist into the saturated regime. As can be seen in the following plot, that was not the case, and the profile remain fairly consistent even into saturation. I refer the interested reader to the work of Rob Friedman and John McClain (listed on the left hand side page) for the details of the analysis.
After testing the with the new pole pieces, the beam intensity was only marginally increased. After great consideration, we felt that there must be an azimuthally varying field destroying the beam. This could come from the magnet's vertical return yokes locking the field lines in on the sides, while allowing field lines to "leak" out the front and back. As a result, during the Spring Semester of '04 and continuing into the summer two students have been working on the design of a 2-dimensional automated field mapper. When complete, this will allow a square area of 16-inches by 16-inches to be mapped out in the median plane. Not only will this system indicate if there is an azimuthally varying field, it will tell us where to correct with shims. This will provide an important step in the improvement of the cyclotron beam.