Brief summary of discussion on crab cavity, 11.08.2006

Participating: Rama Calaga, Rogelio Tomas, Frank Zimmermann

Rama discussed geometrical constraints for a squashed-cell cavity. The optimum squash ratio of 0.75 provides the largest separation in frequency between the crabbing mode and the two next higher mode(s). The squashing implies a slight increase in the larger dimension of the crab cavity, unfortunately. The fundamental mode can only be damped with a coaxial device. The higher modes could in principle be coupled to the adjacent vacuum chamber. This might reduce the shunt impedance of the crabbing mode. Possibly some additional higher-order mode dampers may be required depending on the choice of the outer beam-pipe diameter. The half aperture in the plan of crabbing is 53 cm.

One novel idea by Rama is to couple two cavities tilted by 45 degrees in opposite directions, and to use the vector sum for the crabbing. This reduces the half dimension in the crabbing direction by 10-15%.

A simplified/optimized version may be star shaped cavity with flattened edges.

Frank asked whether we could conceive a self-compensation between the jitter in two coupled cavities, so as to relax the noise tolerance.

A spoke-cavity design is interesting but time consuming with uncertain success. It would be the first ever deflecting spoke cavity. Argonne has some experience with spoke cavities for RIA.

Rogelio pointed out that two crab cavities per beam (e.g., located in the rf section and/or Alice and/or LHCB) would allow fine tuning for the two IPs without need to perfect optics control. However, with a single crab rf section per beam, we could use the LHCB and Alice straight sections for optics tuning and adjustment of the phase advance between IPs. A length of 10 m in the rf section would be required. According to Joachim in the previous meeting only 1 m Is available presently. However, some moderate optics and layout change could be imagined for an upgrade.

Noise data and spectral shapes would be of interest for more precise predictions of the crab cavity noise effect. Low level rf may be optimized for noise suppression.

Rama will contact LBNL folks for information and help. KEKB (Akai) could also be contacted and might provide noise data from the KEKB cavity.

MAD-X or SIXTRACK simulations could study the emittance growth due to the crabbed closed orbit. The crab cavity should be symplectic (energy change together with transverse deflection).

Rogelio will first model the crab cavity with a kick matrix. Later noise could be added (is it possible in MAD-X?). Ideally we would like to have a crab cavity element in MAD-X. Rogelio will contact Alex Koschik.