To: Distribution 21 Aug 95 From: Martin Nordby Subject: Minutes of Near IR Engineering Meeting of 18 Aug 95 Q1 Cross-Section Layout Joe Steiber and Martin Nordby presented a layout of all cables and services passing around Q1. This included one of the configurations for SVT cable routing shown at last week's meeting. The SVT cables include real cable dimensions, with 54 ribbon cables coming out each end. Each contains 17 (signal) or 10 (power) twisted pairs per ribbon, and two of the ribbon cables are reserved for the 6 radiation monitors and position monitors. There was a general discussion on the need for shielding the SVT signal cables. Roy Kerth has been working on simulating the cable behavior, but it is not clear yet whether shielding is needed or not. The machine utilities do not include any Q1 services, since they will enter Q1 at the out-board end of the Support Tube. The layout included rough sizes for BPM and pump cables, which need further refinement. Also, the layout double- counted the purge lines, but left out any provision for an alignment system. Removing the extra purge lines will free up some space for alignment tubes. The general consensus on the layout was that space was too tightly packed, considering the overall design is still floating somewhat. Nordby, Steiber and Goozen will open up the Support Tube diameter enough to allow a higher stacking of ribbon cables. This will reduce the azimuthal space by 17%, and give a bigger margin of safety. Q1 Magnet Design M. Nordby presented the most likely material choices for Sm2Co17 and NdFeB permanent magnets options. Both remnant field and intrinsic coercivity are critical to ensure that the quad can tolerate the 15 kG combination of self-, dipole-, and radial solenoid fields. The intrinsic coercivity, especially, varies with manufacturer, but the catalog values quoted for Shin-Etsu R26HS of Hci = 18 kOe, and EEC 2:17-27 of Hci = 25 kOe, look like they should work. The next step is to buy and test these materials. However, this will not be done before BaBar needs the Support Tube diameter set. Thus, we will produce a few realistic options for the Q1 design which include capability of shimming the Q1 quad blocks. This will likely take up more radial space than we have available, but will give everyone a better feel for the price (in radial space) we wil need to pay for the security of being able to shim/adjust/repair the quad harmonics after exposure to the real solenoid field. Support Tube Analysis J. Hodgson reported on the status of the carbon composite tube analysis. His working design was one proposed by Advanced Composites of San Diego, which was a single, 0.060" thick wall of filament-wound carbon, with lay angles of ±10¡ and 90¡. Under gravity loading, this design sags 2.3 mm, with a max stress of 7700 psi in the carbon fiber. The wall thickness and carbon orientations are driven by both the deflection requirement of 5 mm max deflection during an EQ, and the buckling criteria. There was some discussion about the consequences of letting the Tube hit (or "touch") the inner, thin, non-structural wall of the D.C. With no Drift Chamber people in attendance, we shelved the discussion. The buckling analysis centered on three types of buckling. First, the "n=2," or "ovalizing" buckling of the tube is something predicted by Timoshenko to be the most-likely buckling scenario. However, for our length tube (60" for the composite section), it is not clear from either Timoshenko, or Hodgson's ANSYS analysis, that this is (or is not) a real buckling mode. Standard local buckling is predicted to occur next. Hodgson's analysis, and Baker's empirical curves correlating theory with testing, suggest we have a safety factor of about 2 (during an EQ), which should be sufficient (per Baker). No one, including Hodgson, felt thoroughly at easy with "only" a factor of 2 safety on something like buckling. Shear buckling affects only sandwich-type construction. For the CDR design of a 0.5" thick sandwich of Rohacell foam, this appears not to be a problem. Hodgson will try to resolve the thin-wall buckling behavior of the tube. This will determine whether we can use a single-wall tube, or need to stay with the sandwich construction. Q1 Block Tilting M. Zoloterov presented results of theoretical work on tilting the segments of the Q1 magnet. This would reduce the effect of the radial solenoid field on the block material. He found that the quad field quality is apparently not affected by tilting the blocks, such that a segment becomes conical. The field strength is decreased by [1+cos(theta)]/2, and an axial field is introduced. Theoretically, these can be compensated-for. It was generally agreed that this would be a last-ditch choice, since tilting the blocks introduces many complications to the mechanical and magnetic design of Q1. Action Items for Next Week Refine layout of cables and services past Q1 (Steiber, Goozen). Work up rough designs for Q1 with shimming/adjusting capability (Nordby, Mattison, Sullivan, Ecklund). "Finalize" Support Tube O.D. at Near IR/BaBar meeting next Wednesday.