Discussion on crab cavity and dry runs for LUMI06, 12.10.2006

Participating: Rama Calaga, Ulrich Dorda, Rogelio Tomas, Joachim Tuckmantel, Frank Zimmermann

Rogelio computed the luminosity scope of the crab cavity option, comparing the luminosity gain vs beta* for nominal crossing, nominal crossing w/o crossing angle, and the crab cavity option with 8 mrad crossing angle. Frank suggested to rather plot the luminosity gain vs. quadrupole aperture, as with separated channels for 8 mrad angle smaller beta* values can be reached.

Rogelio presented his draft talk, starting by listing the advantages (restore luminosity, no LRBB, smaller quads or smaller beta*) and disadvantages (space, noise) of crab crossing. He showed a beam survey across the IR, with 10 sigma beam envelopes. The crab voltage required is 108 MV at 400 MHz for the present scheme, which may require 25 m per beam and side, assuming 4-5 MV/m rf gradient. Rogelio conservatively considered NbTi technology for the low-beta quads and bending magnets, with consistent parameters. He next showed the variation of the tunes with momentum deviation. A large third order chromaticity is visible. Frank suggested that Oliver and Stephane may have specified a tolerance for this term. The dynamic aperture is low (worst seed), which is a common feature of all upgrade optics. The field errors need to get a factor 10 lower than in the present separation dipoles in order for the dynamic aperture to become acceptable. A factor 2 reduction on crab voltage might be possible if flat beams and a doublet are used, via an increase of the beta functions at the IP and at the crab cavity for the plane of crossing.

Rama next discussed the various options of crab cavities (local vs. global), including orbit oscillations and tune shifts. He addressed the luminosity loss due to rf curvature and due to various noise sources. A local chromaticity correction might be facilitated by the separate beam channels. The formula for the amplitude jitter seemed to contain an error (possibly wrong in Frank’s Arcidosso paper?). He next addressed the issue of space, and mode separation by squashing the rf cavities.

Joachim reviewed basic principles and technological issues of crab cavities. Points of concern are the jitter stability required, which he deemed to be a factor 10 beyond the state of the measurement art estimated at 20 ps resolution limit (perhaps Fritz as meanwhile found a solution based on notch filters for the carrier), the bulkiness of the rf system (powering several cavities from the same klystron), the need of noisy klystrons, reliability, and beam loading of the main rf leading to 20 ps shift in the bunch position after the dump gap (perhaps less of an issue if collisions occur only in IP1 and IP5).

Ulrich discussed BBTRACK simulations for the BBLR. Frank suggested that the talk should illustrate how much we can gain with a dc wire as compared with a pulsed wire, mention the information from Steve Hays’, and describe the basic motivation at the outset of the presentation. Rogelio recommended simulating the various upgrade scenarios, Frank simulating PACMAN and nominal bunches for an intermediate wire excitation. For RHIC the cases with chromaticity of 2 and no chromaticity could be shown.