(larger image)The NLC beam is extremely intense, ~12MW average power (750 GeV beam) in a spot 10mm x 80mm. An errant beam would instantly destroy the surface of a conventional collimator. Some far out possibilities are being considered, such as collimation by liquid metal jets, which, in essence, would create a continuously renewed collimation surface. The appeal of this renewable collimator would be the economy of relaxing tolerances on the beam location, but the technical challenges are formidible. A level lower in technical difficulty is the consumable collimator. This would be a solid surface, which, if damaged by an errant beam, could be indexed to present a fresh surface.
CONSUMABLE COLLIMATION
Two approaches are being considered. One is a rotary "jaw" which can be rotated to present a fresh collimation surface. We assume that ~ 1mm of offset will yield a sufficiently clean surface. Thus, a 0.32m diameter jaw could absorb about 1000 hits before requiring replacement.
The higher risk concept is a thin tape passing over a stationary jaw. If hit, the tape could be advanced a few mm. Fresh tape would be supplied from a feed reel and damaged tape would be accumulated on a take-up reel. This concept's greater storage density of collimation material offers the possibility of a longer interval between replacements.
Preliminary Design Considerations
Concepts and Specs - first, we looked at a thin Ti collimation surface (per ZDR). This material was presentated to B-D and Accelerator Physicists on 7/27/99. The concepts are illustrated and some issues are identified.
Heat deposited by image current in collimation surface is estimated at .42W.
Dissipation of heat from the collimation surface is readily accomplished by radiation for tape concept and rotor concept. Heat due to continuously absorbed beam halo for the current design is about .23W, which is added to the image current heat.
Current effort is concentrating on bi-metallic collimation rotors, which use ~.25 X0 of copper for beam interception in conjunction with a "ramp" of low-Z material to provide a smooth path for the image current. Aluminum and Copper plated on Beryllium are being considered for the ramp. We're looking into fabrication details.
Rotary Collimator Model
Here's a conceptual solid model of the rotary device showing vacuum chamber and internal components. To center the aperture on the beam, the chamber is continuously aligned to the beam via BPMs and magnet movers. The width of the aperture is controlled by the internal mechanism.
This is the internal mechanism, including two rotors and a central datum structure containing the flared beam pipe. In addition to rotation, the lower rotor has one degree of freedom (vertical) determined by a flexure-hinged parallelogram support. The upper rotor has an additional degree of freedom (horizontal) provided by a flexure in its vertical support. The rotors are preloaded toward each other by springs (not shown). Separation of the rotors (i.e. the collimation aperture) is controlled by two roller-stops mounted on the datum structure.
This close-up shows the stops for positioning the rotors. The stops are supported on flexure-hinged parallelograms. The distance between them is controlled by a connecting shaft, which has opposing left and right hand threads. The pair of stops can float horizontally to mate with the lower rotor. The upper rotor can float horizontally to mate with the stops, thus aligning it horizontally with the lower rotor. To widen the aperture, the screw is turned so as to bring the stops closer together, thereby forcing the rotors apart.
Bearings for UHV Service
Traditionally, SLAC collimators have been designed with guides and motion-generating mechanisms located outside the vacuum envelope, with motion transmitted through chamber walls via bellows. Due to the extremely tight tolerances, this isn't practical for the NLC collimation system. Recognizing that our environment is similar to the environment in which space vehicles must survive, we've looked to NASA for information. Given our very low speeds (essentially static) and low number of revolutions (~1 for the rotor support bearings), the problem boils down to minimizing bearing contact stress and preventing metal-metal contact.
Based on space vehicle expreience, bearings with ceramic balls or standard bearings lubricated with solid lubricants seem to be worthy of consideration. Solid lubricants should be thin film to minimize the risk of loose lubricant jamming the bearing. The two most promising approaches are standard precision ball bearings with ion-sputtered MoS2 lubrication or fitted with Si3N4 balls. A brief summary of our literature survey is attached.
Modified 08/14/00
