Minutes of the IR Engineering and Physics Meeting of 8 Mar 96 Frangible Link Design Donn McMahon presented the latest design for the vacuum Frangible Link. The "Frangible Link" is the break-able link in the vacuum system located just outboard of Q5 in both beamlines. During an earthquake, as the Q2/4/5 Raft pivots with the detector, the Frangible link either distorts, or, eventually, breaks. The Link uses a standard welded bellows, with an inner RF shield similar to the HER straight section bellows design. It is designed to take 1" travel during bake-out of the neighboring chambers, and +/- 8 in. axial travel during an earthquake. Transversely, the Link can take +/- 4 in. of motion. The bellows was designed to hold vacuum under all of these conditions, but this is not needed. Donn will investigate whether this simplifies the design. Since the RF shield is based on the straight section design, it can only tolerate 0.06 - 0.1 in. of transverse motion before the RF fingers are damaged. This offset is very marginal for the IR, where expected motions during alignment, or even a small earthquake, could be up to 0.25 in. The Frangible Link is 24 in. long, and uses 16.5 in. Conflat flanges. It fits between the end of the Q5 chamber and the collimator on the HEB, and inboard of BV1 on the LEB. This may require moving some correctors and a pump on the LEB. Lou Bertolini and Donn are looking at that. The Link is being designed to take 2 W/linear cm of power, with water cooling at the ends. The next step in the design is: finish thermal analysis of RF screen; finish buckling analysis, so screen thickness and slot length can be set; look at modifying the RF finger design so it can tolerate at least 0.25 in. of offset without damage to the fingers. Martin Nordby will produce a design specification for the Frangible Link. Iron Q2 Design Update Fran Younger re-sized the iron Q2 design, using beam stay-clears from MIke Sullivan and Martin Nordby for the HEB, LEB, and Luminosity Monitor trajectories with the iron Q2 (dubbed the 3/8/96 design). The bore of the magnet was increase to 47.9 mm radius to clear the actual LEB BSC's. This required a 28% increase in Amp-turns for the coil. The new magnet also has smaller coil pockets, and as small a cross-section as practical.. It is 28 cm high by 36 cm wide, with 14 cm from the bore centerline to the LEB side of the iron. The slightly wider septum allowed an increase in coil thickness to 18 mm, so the current density is similar to the previous design. The coils are not symmetric, since the LEB BSC is not symmetric. This produces higher n = 6, 10, and 14 poles, but they should be reduce-able with optimization. Fran also ran a first pass of a 3-D model of the old (2/29/96) design. With no HEB shielding or mirror plates, the field at the HEB centerline was 38 G. This was high, and clearly warranted addition of a shield. The shields for Q4 and Q5 on the LER are 3 mm, but there was probably only room for 1-2 mm of shielding here. Fran will set up a new 3-D model with the new (3/8/96) design, since it is much closer to the "final" geometry. Q1 Trim Coils Update Stan Ecklund reported on the latest concept for the Q1 Trim Coils. Previous designs with uniform coil current produced quad fields with high harmonics, so Stan looked at better approximating a cos(2*theta) current distribution by spreading the conductors around 360 degrees, and stepping the current down from 245 A, down to 49 A, with 2 conductors at each current, with 8 conductors per coil. This produced a peak field error of 0.03 G, at the maximum radius of the LEB BSC (at 2.1 m). Coil current density is 1790 A/cm^2, in the high-current conductors, with total power dissipated in all conductors of 5213 W (per coil). With a 0.32 gpm flow in each of 2 circuits per coil, total temperature rise in the water is 7.4 degC. Since the coils use all the azimuthal space inside the inner radius of the P.M. Q1, there is no longer any room for a temperature-stabilizing water circuit. Thus, the 7.4 degC increase could be a problem. Also, with four current settings per coil, the power jumpers at the end get complicated, and could take up considerable space. If the turn-around remains in the radial space of the coil (as opposed to dog-earing) they may contribute significantly to the harmonics of the coil. However, there may not be room for any other turn- around method. These issues need further investigation. Nonetheless, this design shows that there is a solution for this coil design which uses discrete-current conductors. These minutes, and agenda for future meetings, are available on the Web at: http://www.slac.stanford.edu/accel/pepii/near-ir/home.html