Minutes of the IR Engineering and Physics Meeting of 15 Mar 96 Quad Testing John Seeman, Hobey DeStaebler, Stan Ecklund, and Zach Wolf made measurements on three quads: a P.M. quad from SSC, a Collin's quad at SSRL, and a HER arc quad with a solenoid wound in front of it. SSC P.M. Quad Measurements: This appeared to be an 8-block quad, with outer support rings and adjustment screws. Hall probe measurements were made along a line parallel to the magnet axis, outside the magnet roughly where the HEB would be if this were Q2. Qualitatively, the By-field dropped off beyond the end of the magnet, with no hump at the end, as predicted by Dave Humphries' 3-D finite-element model. This could be because the magnet was an 8 segment Halbach quad, instead of 16- or 24-segments as modeled for Q2. Just outside the support ring, the By-field was 1000 G. Also, there was a small sinusoidal variation in the curve which correlated with the axial segmentation of the magnet. This could be explained by the small out-of-plane error in magnetization vector orientation Spear Collin's Quad: This quad has a 10 in. bore, with no back legs, but steel plates bolted to each side. It had air gaps inside the bolt-on plate in the horizontal plane. Measuring the By-field with a Hall probe, the field drops off quickly with increased radius. As the probe is moved axially, off the end of the magnet, the field drops smoothly, with no apparent ill effects from the coil turn-arounds. This suggests that the 3-D stray fields in the HEB from the coil turn-arounds may not be an issue. Quad with Solenoid: The quad used was the reference quad for the HER quad refurbishment. It already had a swept-coil harmonic measurement system in-place, and a solenoid was wound and mounted off one face of the magnet. A flux plate with a bore which matched the hyperbolic quad bore, was added for some of the tests. The goal of the tests was to find the correlation between solenoid field strength and octupole harmonic, to try to establish new (and possibly looser) requirements for allowable solenoid leakage field. As expected, as the solenoid strength was increased, the octupole harmonic of the quad increased. No other harmonic seemed to be affected. For this configuration, a 100 G solenoid field produced a change of 4 x 10^-4 in the octupole harmonic. With the mirror plate added, the change in octupole was almost identical. This surprising result could be due to the shape of the mirror plate bore. It is being changed to a round bore, and will be magnetically joined to the core of the magnet around its periphery. Both should render the mirror plate more effective. P.M. Quad Update Dave Humphries presented an updated HEB shielding arrangement, which borrowed from lessons learned from previous iterations. The P.M. quad is shielded by a 1 mm thick cylinder, plus two, 2 mm thick annular rings at the end, one on the face of the quad, and one spaced 15 mm off the quad. As anticipated, the cylinder sees up to a 24 kG field, and the inner annulus a 25 kG maximum field, both varying with the variation of magnetization vector in the P.M. blocks. However, the outer annulus sees a 23 kG max field, which is almost purely a quad field. This shield design produces a 3% reduction in the overall quad field, which is much better than with the earlier thick annulus. However, it still shields the HEB. Dave also looked at a design with a 3 mm thick end annulus, a 2 mm thick HEB shield, and a 1 mm thick septum shield (no cylindrical shield around Q2). This shields the HEB, but produces a 0.6% asymmetry in the main quad field at 4 cm radius. This should be remedied by the addition of a cylindrical shield. Another alternative to reduce the stray field in the HEB, is to add current filaments above and below the septum, to buck the stray By-field. These could be set to tune out the stray field during magnetic. They also provide some safety factor to account for unknowns which may crop up during fabrication. There was some concern about the room this composite shielding design will take. With space being very tight for the P.M. Q2 design, there may not be enough room to add 1.5 cm in front of Q2, plus increase its strength by 3% to offset the loss due to the annular shielding. Furthermore, additional shielding in front of the magnet is needed to protect the thin HEB shielding from saturating due to the stray solenoid field. Dave is working on two fronts: first to model the entire shielding package magnetically, and second, to lay out a realistic mechanical package for this, to show how it could fit in the space available. Iron Q2 Update Based on the solenoid test results presented by John, Fran Younger thought that, to help buck the solenoid field effect on the iron Q2 design, there should be room in the coil pockets to add octupole bucking coils, wound around each pole tip. Fran is developing a 3-D model of the 3/8/96 Q2 design, and will include this in the model. The coils for this design have been spec'd, with total current per coil of 2163 Amps, or 31,230 A/in^2 current density. Total power per coil is 53,320 W, which is cooled with 8 parallel water circuits per coil with a total flow of 10.131 gpm. Water velocity is 16 ft/sec. These minutes, and agenda for future meetings, are available on the Web at: http://www.slac.stanford.edu/accel/pepii/near-ir/home.html