To: Distribution 1 Sep 95 From: Martin Nordby Subject: Minutes of Near IR Engineering Meeting of 1 Sep 95 Q1 Cross-Section Layout Fred Goozen showed the latest incarnation of the Q1 cross-section, with cables and utilities. He modified this to better distribute the SVT cables around the periphery, by moving one stack of signal- and one of power-cables. This layout still maintains as much radial space as possible, with none "squandered" on poor packing. The extra radial space gained from the 1 cm increase in Support Tube radius will only be used as necessary. The next step will be to lay out the major cables and utilities in 3-D to look at routing around the B1/Q1 gap and look for gross interferences that are not apparent in the cross-section. Joe Stieber will start to investigate this. Also, with information from the 3-D CAD model, the SVT group will build a geometrically-correct prototype to mock-up the routing in real life. Access Holes in the Support Tube With cable routing settling down somewhat, the next issue regarding the Support Tube is the need for access holes. At first guess, Roy Kerth did not think that any/many would be needed in the stainless steel part of the tube, and possibly a few could be used in the carbon-composite tube. For the machine, the opposite is probably the case. Action on two fronts was agreed on: first, Roy would talk with the other SVT collaborators about the possible use of holes, and locations/sizes. Second, Martin Nordby will try a first-cut at an assembly sequence for the Support Tube. This should point to hole requirements from the machine side. Alternatives for Q1 Mike Sullivan presented results to-date of his investigation into removing parts of the backward Q1 to avoid the possible de-magnetization problem of the radial solenoid field. He looked at removing 20cm, 30 cm, and the entire out- board end (40 cm). To compensate, he increased the strength of the in-board part of Q1 by reducing the inner radius of the quad blocks. His goal was to produce a magnet which gave the same beta star as the original design. As more material was removed, the beam separation at the Q4 and Q5 septum magnets decreased, the synchrotron radiation produced increased, and the inner radius of Q1 got smaller. The numbers are: Baseline -20 cm -30 cm -40 cm Chamber inner rad. 60 mm 53 48 46 Separation at Q2 90.4 mm 90.4 90.5 89.2 Separation at Q4 143 mm 141.1 140.1 135.1 Separation at Q5 266 mm 259.4 255.6 242.5 Synchrotron rad. 3209 W 3378 3480 3086 The inner radii of the beampipe is approaching the 15 sigma + 2 mm beam stay-clear, and it could be very difficult to produce a magnet with good multi- pole quality at this large a fraction of its bore radius. Furthermore, the beam separations at Q4 and Q5 are already making these magnets difficult to design. So reducing the separation is not an option. Finally, the upstream LEB SR fan may end up hitting the smaller diameter beampipe. This would mean a mask may be needed. In summary, Mike said that the optics looked like they may be do-able, but the three problems this alternative design produces are all potentially show- stoppers. Two other options were brought up as alternatives: Tilt BaBar's solenoid--it is not clear that this would help, especially since it would hurt the forward Q1. Extend B1 to the front face of Q1, and only remove the dipole ring from the 20 cm of space in Q1. This may give the desired optics effect, but would put a strong kink in the HEB which would produce SR with an unkown destination. These options will be held in abayance, since any future work would require a big time investment. Cornell Visit Martin Nordby summarized the observations from his and Stan Ecklund's trip to Cornell, and discussions with CESR people regarding their permanent magnet quads. To summarize: Operational Experience REC (Q0) Quad -REC Quad (Q0) is mounted to end face of iron flux return -Q0 positioning is repeatable to 0.010" after removal/re-installation -They still need to re-establish a gold orbit after work in I.R., despite relatively good repeatability -Q0 rotates as part of solenoid compensation scheme, so it is held in the middle by a rotatable collar, and its vacuum chamber is held independently Old Q0 Chamber -Q0 chamber has a built-in, stepped mask -Chamber was not aligned closely with respect to Q0, and also tended to distort from thermal loads -Chamber distortion pushed Q0 by 10 microns -Effect on beam orbit seen by downstream BPM's -Axes of Q0 quad field only known to +/- 0.1 degrees -10 micro-radian change in Q0 rotation affected I.P. luminosity New Q0 Vacuum Chamber -Chamber is copper, with built-in stepped masking, and a move-able tantalum radiation shield to protect SVD during injection. -Chamber has a stainless steel jacket, and is cooled with PF200 coolant -TSP pump outboard of Q0 chamber -Their tests showed no dust production from TSP - Nervous about using NEG's, based on ESRF experience -Cold cathode vacuum guage mounted off chamber at TSP -RGA head at 6 m from IP, not remotely read-able (yet) New Be Vertex Chamber -Chamber has concentric Be tubes, with water flowing between them -Entire chamber built by Brush-Wellman (Electrofusion) -Production process: Braze Be tubes to st steel end collars Braze tubes together Gold plate at Cornell (glow-discharge sputter in Ar atmosphere) Blank-off vacuum surfaces and immerse in low-viscosity epoxy (B-W proprietary epoxy) to coat. Coat with two layers, inside and out REC Quad (Q0) Refurbishing -Multi-pole content = 5 e-4 at 40 mm radius -Only outboard end of Q0 was temp-stabilized, since this is where beta function is highest -Quads were lengthened 20 cm by using segments of unused REC Quads from second I.R. -Quads are 16-segment Halbach configuration quads -Each segment measured in a vertical rotating-coil set-up Precision of vertical coil: 2 e-4 Bucking coil reduced quad-component by 30-fold Measurement done in a warehouse, with no temperature control Labview used for data acquisition and numerical integration of voltage signal from rotating coil Individual blocks moved radially to remove multipole content of segment and to rotate quad field to remove skew with respect to pins on collar -Segments were paired and re-measured in vertical-coil set-up -Entire magnet assembled and measured in horizontal rotating-coil fixture, with bucking coil. Precision of horiz-coil was 1 e-3 Refurbishing Results: One segment was measured, then re-heated to 120 degC and re- measured (Steve Herb heated all segments to 120 degC when he made quads in 1984). No change in multi-pole content seen down to 2 e-4 precision of apparatus Some segments saw no change in multi-pole content after 6 years of operation in axial solenoid field Segments that were in the 2 kG radial solenoid field saw 1 e-3 changes in multi-pole content