Straight Merger Updated Andrew Hutton Straight Merger with

“Straight” Merger Updated Andrew Hutton

Straight Merger with No Bending of Low Energy Beam Recirc ulated Beam Injector Cryomodule Septum Magnet RF Separator with Superposed Magnetic Field ERL Cryomodule

Concept (1) • Use an RF Separator to separate the injected beam from the recirculating beam • Immerse the RF Separator in a DC magnetic field • Arrange timing and relative amplitudes so that the injected beam is not deflected • Bunches are at maximum of RF deflection – bunch center has zero slope • This means that the kick seen by the recirculated beam is doubled • Needs to be sufficient to provide adequate separation at the septum

Waveforms 2, 5 RF 2, 0 DC RF+DC 1, 5 1, 0 0, 5 0, 0 0 100 200 300 400 500 600 700 800 -0, 5 -1, 0 Injected Bunch -1, 5 Recirculated Bunch 900 1000 1100

Concept (2) • Use an RF Separator to separate the injected beam from the recirculating beam • Immerse the RF Separator in a DC magnetic field • Arrange timing and relative amplitudes so that the injected beam is not deflected • This means that the kick seen by the recirculated beam is doubled • Bunches are at maximum of RF deflection – bunch center has zero slope • Needs to be sufficient to provide adequate separation at the septum • The injector beam and the recirculating beam are further separated by a septum magnet • Injected beam sees zero field • Recirculating field sees deflecting field to provide adequate separation at the last injector cryomodule

Separator Frequency • The separator should be designed for the CCR parameters • Assume that recirculated beam is timed to arrive midway between injected bunches • Initially, the beam frequency in the injector will be 476/11 = 44. 3 MHz • Later, the frequency will be 976/11 = 88. 6 MHz • A single RF Separator will not work for both frequencies • Injected bunch separation at 44. 3 MHz is 22. 6 ns, the recirculated bunches are interspersed 11. 3 ns from injected bunches • Phase of RF must change sign every 11. 3 ns • Injected bunch separation at 88. 6 MHz is 11. 3 ns • Phase of RF must be the same every 11. 3 ns • Fundamentally incompatible requirements!

CEBAF RF Separator • CEBAF 500 MHz separator cavity • Provides a kick of >15 mrad for a Me. V beam using 3. 5 k. W RF power • The magnetic field adds an equal kick 55 • Field ~100 G • Total kick for 55 Me. V beam is >30 mrad • Assume a drift length of 2 meter • Separation at the septum is >60 mm • Aperture of CEBAF RF separator is 14. 3 mm • Injector beam is large, but not deflected • Recirculating beam is deflected ~18 mm within separator • Needs more aperture at downstream end for recirculated beam

Separator Frequency Options • With the ERL RF System at 476. 3 MHz and the harmonic kicker injecting/extracting every 11 th bunch, the injector rep rate is 43. 3 MHz • Possible frequencies for the separator frequency are odd multiples of the bunch frequency • 43. 3 x n MHz where n = 1, 3, 5, 7, 9 or 11 • 43. 3 MHz, 129. 9 MHz, 216. 5 MHz, 303. 1 MHz, 389. 7 MHz or 476. 3 MHz • With the ERL RF System at 952. 6 MHz and the harmonic kicker injecting/extracting every 11 th bunch, the injector rep rate is 86. 6 MHz • Possible frequencies for the separator frequency are 86. 6 x n MHz where n = 1, 3, 5, 7, 9 or 11 • 86. 6 MHz, 259. 8 MHz, 433 MHz, 606. 2 MHz, 779. 4 MHz or 952. 6 MHz • Separator design can be scaled from the CEBAF 499 MHz kicker • Should be low enough frequency to provide sufficient aperture • Should be high enough frequency so it is not too large • Can scale number of 2 -rod half-cells to get appropriate length (~60 cm)

RF Separator Options • Parameters scaled from the CEBAF RF Separator

Curvature of Injected Bunch (Banana effect) • RF deflection of bunch center ≈170 mrad • Consider the 2 half-cell RF separator at 216. 5 MHz • Wavelength = 138. 6 cm • Bunch length = ± 1 cm • RF deflection at bunch ends = 170 cos (2�� /138. 6) mrad = 170 x 0. 99897262521 = 169. 825 mrad • Maximum deviation at bunch ends = 170 x. 001 mrad = 175 µrad • A factor twenty less than at 952. 6 MHz • Banana effect is proportional to the square of the frequency • This could be further reduced by a factor two: • Adjusting the dipole field to be centered on the “banana”

Separator Layout • Separators are made up of quarter-wave cells • We need two of these cells • There are two possible orientations • The orientation on the left is better for this application • But, harder to arrange cooling channels

Optimization of the RF Separator • A four–rod SRF cavity was developed for CERN by Ben Hallas as a Ph. D thesis at Lancaster University in England • It was built by Niowave and will be tested in the SPS • A lot of effort was spent to optimize the rod shape • The best structure had kidney shape rods, which increased the aperture • A similar optimization could be done for the RF separator

Conclusion • Adopting a lower frequency has several advantages • Larger aperture • Less RF curvature • Power density on the rods is reduced (more surface area) • Cavity is larger and heavier • But the size is not unreasonable • Initial cavity installed to work with the 476. 3 MHz RF System will need to be replaced when the RF is upgraded to 952. 6 MHz • Initial RF Separator operates at 216. 5 MHz • Upgrade Separator operates at 259. 8 MHz • Opportunities exist for optimization of the shape