To: Distribution 29 August 97

From: Martin Nordby

Subject: IR Engineering and Physics Meeting Minutes: 15 August 97

Mike Sullivan presented a methodology for determining polarity of the six permanent magnets in the I.R. Since these are, indeed, permanent, their polarities depend on the physical construction of the magnet, and can not be changed after the magnet is built/installed.

(All coordinates, below, are assumed to be with respect to the IR Coordinate System, unless specifically stated otherwise.)

For B1 and Q1, looking from the I.P. out in the +Z direction: B1 force is to the left (+X), and the B-field lines are up. For Q1, the quad north poles are in quadrants 1 and 3, and the B-field is up. This moves the magnetic center of Q1 to the left ( +X).

Looking from the I.P. in the –Z direction: B1 force is to the left (-X), and the B-field is down. For Q1, the quad north poles are in quadrants 2 and 4, and the B-field is down. This moves the magnetic center to the left (-X).

For the SK1 in the forward direction (+Z), (a.k.a.: SK1R), skew rotation is positive (+X towards +Y), and when viewed from the I.P., the normal quad has north poles in quadrants 2 and 4.

For the SK1 in the backward direction (-Z), (a.k.a.: SK1L), skew rotation is negative (+Y towards +X), and, when viewed from the I.P., the normal quad has north poles in quadrants1 and 3.

This means that, for SK1, the rotating mechanism for both magnets can be identical, but the ring of magnetic material must be built with reverse polarity.

Martin Nordby reported on the Support Tube design and analysis. There will be a design review for the S.T. coming up, and this provided an opportunity to bring up any issues before the review. The final design for the Central Tube is a single wall carbon-fiber composite tube, made from filament wound carbon. It is 0.08" thick, and 0.8% radiation length. Analysis of the CFC Central Tube shows that deflection under gravity loading is 2 mm, and it has a safety factor against buckling of 2.2, with respect to empirical test data for similar tubes. The Central Tube has filament windings oriented [+10,-10, 90, -10, +10].

Analysis assumed a maximum earthquake loading of 1.8 g, which occurs in the vertical direction. This originates from a SLAC peak ground acceleration of 0.6 g, a vertical scaling factor of 2/3, and a magnification factor of 2, given the natural frequency of the Support Tube of around 10 Hz. Thus, 1.8 g = (1g gravity) + (0.6 g PGA)(2/3)(MF = 2). Given this worst-case loading, the clearance needed between the S.T. and the Drift Chamber is 6.6 mm. This includes needed EQ motion for the S.T., Raft, and Drift Chamber/DIRC with respect to the detector steel. Total available space is 7 mm.

The Support Tube is being held at each end with Movers which allow remote vertical positioning of +/- 3 mm. This is done with an L-shaped bracket which pivots, converting an axial motion of the actuator into vertical motion of the S.T.

Jim Krebs reported on progress in finalizing the extraction mechanism for the back end Q2 Shielding Plug. The original design did not account for the magnetic forces acting on the Plug. These push the remove-able horseshoe down on the in-board end, and up on the outboard end. Also, the Plug was not restrained vertically. The extraction system was modified to solve these hold-down problems, by making the rails more like a T-slot, preventing uplift. The rail mounts to the DIRC magnetic shield cylinder were changed to allow for more error in the fabrication of the cylinder.

The horseshoe-shaped Plug now sits on bronze sliders, instead of cam-followers. The sliders are backed up by pre-loaded Bellevillle washers, so the weight is distributed evenly over the 8 sliders on each rail. Each slider is 1" diameter, and will be radiused to prevent it from catching on the seam in the rails at the end of the DIRC S.O.B. The lead screw to drive the Plug is ¾" diameter, with a 0.2" pitch acme thread. The rail opposite the screw has a cam follower to keep the Plug in-line. There was concern that this would tend to allow the Plug to rack, and jam on the rails, so Jim will look at modifying the cam-follower, and moving it so it is in-line with the screw.

To accommodate the T-slot in the rails, they now stick up above the fixed portion of the Plug by 29 mm.

Stan Ecklund looked further at the effect of non-unity permeability of the SmCo material, both parallel and perpendicular to the magnetization direction. For a given mu-perp and mu-parallel, they produce a b3 harmonic. However, the magnitude varies with the ratio of radii of the magnetic material (r2/r1). The effect actually crosses 0 for a given r2/r1. Thus, for the B1 Magnet, this effect varies by slice, with some of the slices having a predicted effect very near 0.

According to Mike Sullivan, the as-built b3 harmonics for B1 slices and Q1A dipole rings tracks very well with the curve predicted by Stan. However, their magnitude is less than predicted. This may be due to the 3-D nature of the real magnetic fields. Thus, when slices are stacked, the measured b3 for the assembly should start to approach the predicted value.

If this is true, it may bear on how we tune the magnet. For Q1A, the quad blocks are used to tune out harmonics generated by both dipole and quad blocks. However, if the b3 from the dipole actually grows when slices are stacked, we can not tune this out, since the blocks will already be potted in epoxy. Two options to correct for this are to either build in an intentional b3 harmonic, or to use the Q1 Harmonic Correction Ring to remove the final as-assembled b3 harmonic. Stan plans to look at this more before we decide on a direction.

These minutes, and agenda for future meetings, are available on the Web at: