Back to B. Parker's Home Page.

Presently under Construction - text/annotation will be added as time permits!

                                                 30 January 1996

                  HERA Betriebsseminar 1996
                        St. Engelmar

              Neue Optiken fuer PETRA und P-Weg

                        presented by
                        Brett Parker
                 DESY - Machine Physics Group
                               

Viewgraphs from B. Parker's St. Engelmar Presentation

• View the title Page from Neue Optiken fuer PETRA und P-Weg as presented by B. Parker on 30 January 1996 at the Yearly HERA Operations Conference held in St. Engelmar.
• Outline of motivations for this study.
  There is, as outlined on the next viewgraph, a problem with satellite bunches due to inadequate longitudinal matching between PETRA and HERA for beam transfers made at high bunch current. Presently the matching into the HERA rf-bucket is facilitated by bunch rotation in PETRA involving a sudden rf-phase jump followed by a delay of 1/4 a synchrotron period. At high beam current the particles in the tails of the longitudinal distribution "rotate" differently than the core particles and the resulting S-shaped sheering of the phase distribution presents an effectively longer distribution for capture in HERA and some small fraction of these particles are captured in nearby rf-buckets.

  A partial remedy was proposed by R. Brinkmann which involves slightly reducing the longitudinal spread via a trick involving a change of the transverse optics. An obvious way to get shorter bunches would be to increase the circumferential rf-voltage; however, Brinkmann noted that a reduction of the average dispersion in the dipoles (i.e. lower momentum compaction factor) would have the same effect. The bunch length could be shortened approximately as the fourth-root of the dispersion reduction factor and even a small change in total bunch length might manifest itself as a significant reduction of the number of particles trapped in the satellite positions. The transfer procedure would then involve making an optics change at the PETRA to energy, 40 GeV, before attempting bunch rotation and subsequent extraction. It is therefore important to examine what range of optics transitions are possible.

  A second motivation for these studies is finding means for understanding and combating transverse emmitance growth and beam loss in PETRA during the ramp.

  Finally it became obvious during the course of the study that there was room for improvement in the PETRA to HERA beam transfer line optics which might be achievable with a different optics solution at the PETRA top end, 40 GeV.

• PETRA / HERA layout schematic.
  Note definition and placement of Long (L, purple), Short (S, brown) and Bypass (B, green) straight sections between eight Arc (A, red) sections. Proton transfer line to HERA is highlighted in blue.
• A few facts and constraints.
• Outline of a simple approach to PETRA optic matching. There are three combinations of arc to short, long and bypass straight sections and therefore three "matchings" which have to be done. Here one assumes mirror symmetric solutions and starts with a FODO-solution in the center of an arc and propagates this into the straights with an overall constraint that the total phase advance is kept constant. Unfortunately one soon runs into trouble with the extraction kicker to septum phase advance as outlined on the next view graph.
• Plot of lattice functions for transition from short straight section to arc to long straight section with positions of extraction kickers (K1, K2) and extraction septum (S) are indicated.
  The extraction kickers just fit in the space created by removing sextapole corrector magnets. With a naive matching procedure, the phase advance difference between them and the extraction septum increases significantly beyond the ideal 90 degree relative phase difference. While it is conceivable to move the kickers to another location with a more favorable phase advance (as indicated on the plot), such a solution, involving a major hardware modification, is unattractive. In addition the phase advances for the closed orbit septum bump dipoles are also spoiled and given the existing dipole strength limits would also require a similar hardware change.
• First step away from FODO solution is use of "A-B" quadrupole circuit asymmetry.
• Outline of systematic fitting procedure.
• Review of present (original design) optics solution.
  Note relatively large, greater than 14m, peak dispersion and 9.5 m average dispersion.
• Matching between PETRA extraction point and front end of proton transport line (P-Weg) to HERA. Note choke point a first vertical dipole bend.
• Improved P-Weg solution .
• Plot of compromise solution .
• Summary of compromise solution.
• Results from machine studies with compromise solution.
• Natural solution with lowest average dispersion.
  Note change of machine tune and need to cross transition.
• First imaginary transition gamma lattice.
• Outline of transition crossing and sample of an imaginary gamma transition lattice.
• Improved imaginary transition gamma lattice.
  Note relatively large dispersion slope at extraction matching point makes matching into P-Weg problematic; however, imaginary transition gamma solution is still useful for generation, via k-value interpolation, solutions with arbitrary transition gamma.
• Solution with smallest beta in arc dipoles.
  Note however to achieve this solution had to cheat on quadrupole strengths for main focusing (QF) and defocusing (QD) quadrupoles in the center of the arc sections. With many magnets in series, achievable quadrupole gradients are limited by voltage drop across the quadrupole strings (i.e. maximum power supply output voltage). The solution shown can only be ramped to about 35 GeV and is thus short of what would be needed for extraction at 40 GeV.
• Hardware solution to QF- QD-string voltage drops could be to split up strings and put on two new power supplies. With additional quadrupole gradient available in the QF and QD circuits an even wider range of solutions (reduced dispersion, imaginary transition gamma and reduced beta) are possible. Technical problem of lack of space for new current bus might be solved by taking over bus from skew-quadrupole circuits as shown.
• Summary of optics solutions and possible PETRA ramp scenarios.
• Wish list of parameters for electron-cooling in PETRA.
• Summary of achievable parameters, problems and associated constraints/conflicts.
• Personal assessment of near-term future.
  Note degree of confidence is expressed here via shading of the print. Those developments deemed most likely are shown in the darkest print.

Last Update: 14 February 1996