Minutes of the IR Engineering and Physics Meeting of 1 Mar 96 Envelope Dimensions of Machine Elements Scott DeBarger showed updated envelopes for machine components inside BaBar. The new Support Tube envelope is 46 cm diameter, which includes the 45 cm diameter tube, plus 5 mm radial space for deflection and movement range. The S.T. has also been lengthened on the Forward end to 2335 mm from the I.P. (from 2100 mm). On the Forward end, the envelope for the tungsten shielding interferes with that of the new Drift Chamber forward endplate. The physical interference is mild, but the shielding (and the Cantilevered Raft to which the shielding is attached) will block access to wires at small radius for diagnosing or removal.. This needs to be addressed. Also, with the current design of the Cantilever Raft, the SVT cables will be routed out the hollow arms of the cantilever. However, since we may need to fall back to a remove-able support design, if the Cantilever Raft allows unacceptable flux leakage, there must be space available to route SVT and machine cables out the face of the barrel IFR. This will be included in the space requirements, per Harvey Lynch. Radiation Monitoring David Kirkby presented current thinking regarding the radiation monitors in-board of B1, near the SVT cone inner radius. This was the first serious look at these monitors, what they would be used for, and what was required of them. The primary requirements for this system are: fast (100 micro-second) detection of large (>1 rad) doses; providing a trigger for the beam- abort system; interlocks to high-voltage systems (SVT and DC); and possibly triggering limited-rate injection. These all have the goal of protecting the SVT from radiation damage Possible secondary requirements are geared more towards machine diagnostics: monitoring slowly varying (1 Hz) radiation dose rates; monitoring total dose. The monitors must occupy a very small volume near the IP, with little room for cables, and expected cable run length from 5 m, minimum (if pre-amps are mounted just outside the detector) to 35 m, max (if pre- amps and other electronics are in SVT racks outside shielding wall). The front-end electronics must be compatible with both PEP-II and BaBar control systems, and the beam abort system. The signal could be synchronized with the RF clock for turn-by-turn or bunch train monitoring. The radiation environment near the I.P. is due largely to lost particles and synchrotron radiation. During injection, the backgrounds at the I.P. were assumed to increase 5X for the HEB, and 11X for the LEB. However, because injection will be relatively infrequent, and its duration short (6 minutes to fill and 3 min/hour top-off), the total contribution to radiation dose is small. Lost-particle simulations of radiation dose in the SVT show that for layer 1, the average dose is 19 krads/yr, and the max dose (in the horizontal plane) is 48 krads/yr). At the hybrid, the average dose is 16 krads/yr, and the max dose is 44 krads/yr. At the locations of the radiation monitors, the max dose (on the horizontal plane) is 2-3 times higher than the peak dose in the SVT layer 1 (100-200 krads/yr), so these monitors should provide a meaningful measurement for silicon dose. There was considerable discussion regarding the needed response time for these monitors, and how they would protect the detector from machine accidents. First, it was generally accepted that nothing could be done to protect the detector from a kicker failure, since its response time is faster than 10 micro-seconds. In the 10-100 micro-seconds range, there was much discussion about what beam-failure scenarios could occur, and how they are being prepared for. The response time of the beam abort system is around 20 micro-seconds (worst-case) according to Alan Fisher, which will be fast enough to abort a beam due to a klystron failure (100 micro-second time constant) or a magnet failure (100 msec). However, the abort system is triggered by radiation monitors in the rings (far from the I.P.), and the threshold for these may be higher than needed for detector protection. Furthermore, even if the threshold were acceptable, could radiation in the detector climb fast enough while the beam was becoming unstable and being aborted. If these are, indeed possible problems, than a fast (10 micro-second) integration and response time for these radiation monitors may be needed to ensure that the beam is aborted quickly enough to protect the detector. For now, the integration time for these monitors is assumed to be 100 micro-seconds, which may be sufficient to protect the SVT. If a faster response time is needed, the pre-amps would most likely have to be moved to within 5 m of the monitors, but no other hardware changes would be needed. The damaging dose (abort threshold) is defined as >5 rads at >0.1 rad/sec. The dose would be recorded in two ways. First, an interval measurement would measure the time interval during which a 1 rad dose is absorbed (100 micro-seconds to 2 seconds). Second, a dose measurement would check the dose absorbed over a 1 second interval (1 mrad to 2 rad). These would both be used to monitor total dose and dose rate. Q2 Chamber Thermal Analysis Lou Bertolini showed results of thermal analysis of a new LEB chamber cross-section for Q2. This includes an aluminum outer yoke and inner ring around the permanent magnet, and new (lower) power inputs from synchrotron radiation fans, to correct an error in previous models. The epoxy between the P.M. blocks and aluminum outer yoke is modelled, with varying thermal conductivities. This variable affects the azimuthal temperature variation in the P.M. material. The water temperature is assumed to be 25 degC, and the outer yoke sees natural convection to air at 20 degC. The minimum gap between the chamber and magnet is 2 mm. For the higher-conductivity epoxy, the azimuthal temperature variation is 0.5-4 degC, with the average temperature rise of the magnet being 1-3 degC above the water temperature. These results are considerably better than the previous model, showing that the increased conduction paths in the chamber and magnet collars is helping considerably. The next step is to further investigate real epoxies, and refine the model to include the actual geometry of the outer yoke, especially including the cut-away for the HEB beampipe. Martin Nordby will track down information on the B1 Chamber analysis, which looked at effects of the epoxy joints. Lou will also work to optimize the thicknesses of the chamber wall and the inner collar of the magnet. These minutes, and agenda for future meetings, are available on the Web at: http://www.slac.stanford.edu/accel/pepii/near-ir/home.html