27 Oct. 95 To: Distribution From: Stan Ecklund Subject: Minutes of Near IR Engineering Meeting of 27 Oct. 95 Permanent Magnet Block Tolerances: Mike Sullivan gave a status report on the tolerance study of B1 and Q1. He is using Gordon Bowden's program MBUILD to calculate harmonic errors with input errors for remnant field magnitude and direction, block rotation, block angular and block radial sizes. He has extended the code to handle a 32 sector magnet slice. He uses another program, MULIPOLES to calculate the harmonics around the beam axis. He checked that the program gives the previous answer for harmonics, which the LATTICE group has been using for Q1. This had what Martin and Dave Humphries say are tight tolerances according to material vendors (0.001" mechanical 0.5% in field strength and 0.5 degree in direction). Mike next looked at the dipole ring of the hybrid Q1. There he finds he can meet harmonics as specified previously for Q1 with relatively loose requirements (0.080" radial 2% strength, and 2 deg direction). Tighter but reasonably achievable tolerances can do 5 times better. It looks as though the dipole ring will not be a problem, even without block sorting or harmonic correction. Next he looked at the B1 dipole, with the goal of having a 1E-4 magnet at the largest beam stay clear. Unfortunately harmonics are an order of magnitude greater than this goal. Uli pointed out that the field uniformity may not be as bad as 1E-4 as harmonics can cancel each other. We need to determine the tolerance from more LATTICE studies; for example, can we relax the sextupole tolerance. Jack Tanabe suggested a correction slice to cancel the integral error. Space for it may be a problem. Q2 Permanent Magnet Design: Dave Humphries presented a preliminary cost estimate for a Sm2Co17 Q2 combined with SK1 skew quad. Previous EAC for conventional Q2 and SK1 totaled 396.7K$, new estimate is 439.5 K$, however, the estimate will be refined with more numbers from vendors coming in. Dave reported that a general cost for SmCo material is $4.25 per cubic centimeter. Tighter than usual tolerances, or special requirements will run the cost higher. The above estimate has 75K$ for magnetic materials. NdFeB may be a better material because of its higher field and less brittle properties, but it is more sensitive to temperature and radiation exposure. For radiation exposure, Hobey suggested calculating the effect of 100 neutrons per microsec coming out of the Q2 septum as a start to get an order of magnitude on the radiation problem. The Q2 could either have a dipole ring or be offset to give the required separation of beams at Q4. A dipole ring seems to be preferred. Mike Sullivan is working on a note to summarize all the properties (offsets, strengths,...) for the near IR magnets. Dave will work on a summary spec sheet for Q2, SK1 giving properties of these magnets. The gradient z-location of Q2 will change slightly with the PM design which will change the strengths of Q1 and Q2 slightly. Luminosity Monitor: Hobey reported on the Fan Stay Clear for the Luminosity monitor. It originally was desired to be about 8 sigma of the beam divergence (at the IP clearance) for the photons coming from the IP. Q5 and Q4 apertures at the outboard ends want to reduce that, but with some effort it looks as if 6.7 sigma can be kept clear. The robustness of the signal will depend on not letting this get too small. The luminosity monitor could be on either side of the IP; the downstream LEB is chosen, assuming we have only one, because of the lower energy for synchrotron radiation. However the vacuum pipe should be built to allow a luminosity monitor in either or both locations. Transparencies are attached in hard copy version.