Triple-spoke compared with Elliptical-cell Cavities

Transcription

1 Triple-spoke compared with Elliptical-cell Cavities Ken Shepard - ANL Physics Division 2th International Workshop on RF Superconductivity Argonne National Laboratory Operated by The University of Chicago for the U.S. Department of Energy Office of Science U.S. Department of Energy

2 RIA Driver: Elliptical Cell or Triple Spoke Option? 345 MHz at 4K Frequency Voltage Cryogenics Microphonics 805 MHz at 2K 2 SRF 2005 July -5, 2005 Ithaca, NY

3 Triple-spoke vs E-cell for the RIA Driver Performance assumption is Epeak = 27.5 MV/m Triple-spoke cavities Elliptical 6-cell Cavities 3

4 Triple-spoke vs E-cell RIA Driver: High-energy Section CAVITY TYPE: E-CELL TRIPLE SPOKE Operating Temperature K Beta Geometric Frequency MHz Active Length cm QRs ohm R/Q ohm Epeak MV/m Bpeak Gauss RF Energy mjoule

5 345 MHz β=0.63 Triple-spoke cold tests Frequency 345 MHz β L(3βλ/2) 82 cm QRS (G) 93 Ω R/Q 549 Ω below for EACC =.0 MV/m RF Energy J BPEAK 90 G EPEAK 2.93 MV/m 5 SRF 2005 July -5, 2005 Ithaca, NY

6 345 MHz β=0.5 Triple-spoke cold tests.e+ At 4.2 K RIA Goal At 2.0 K 0 Watt Load.E+ Q.E+09.E Eacc (MV/m) 6 SRF 2005 July -5, 2005 Ithaca, NY

7 JLAB/MSU (2) β=0.47 Elliptical-cell at 2 K 7

8 Present performance level bottom line Parameter Six-cell Elliptical Three-spoke β = 0.47 β = 0.62 β = 0.50 β = 0.63 Frequency (MHZ) Length (cm) E A (MV/m) E PEAK (MV/m) B PEAK (Gauss) R/Q (Ω) Q at E A 9.50E E E E+08 Voltage (MV) Temperature (K) Heat Load* *Watts per MV - at 4.2 K 8

9 Longitudinal Acceptance: Spoke vs. E-cell Triple-spoke resonators 345 MHz E peak = 27.5 MV/m Baseline Design: 6-cell elliptical cavities 805 MHz E peak =27.5 MV/m 9

10 ANL: Triple-spoke option is favored for RIA The beam dynamics are better Can operate at 4 K The mechanical stability is excellent linac costs will be less than for SNS, probably less than E-cell with re-designed cryostat

11 RIA Driver Partial beamlist: r-process beams Ion (%) Q source Q strip Q strip2 I published (pµa) Energy/A Power (kw) H / 2 H - - >> / 600 > Ni (0.9%) * Zn (0.6%) ** Ge (7.8%) <* 53 <22 82 Se (9.4%) <* 493 <23 86 Kr (7.3%) ** 505 > Zr (2.8%) <* 504 <28 24 Sn (5.6%) ** Xe (7.3%) ** Yb (2.7%) Os (4.%) Pt (7.2%) Hg (6.8%) Pb (52.4%) x4.3** Th (0%) U (99.3%) x

12 Beam-Loss Calculations Final step of BD design studies Simulations on the multi-processor computer Up to 500 randomly seeded accelerators with all types of errors and misalignments, typically 200 seeds Beam steering is applied Wide range of rf errors, thickness fluctuation and their combinations have been studied Number of tracked particles: Up to 6, typically 2 5 in each seed Total number of simulated particles 40 million, some cases up to 200 million. 2

13 The RIA Driver Linac ECR Injector Section I: Lowβ Stripper I Q = 28, kev/u 2 kev/u 2 MeV/u Section II: Medium β Stripper II Q = 72, 73, 74, 75, 76 U 238 Beam Section III: High β 400 MeV/u Q = 86, 87, 88, 89, MeV/u Baseline: About 200 beam line elements: ~ 400 rf resonators, 90 solenoids, 0 quads, 6 bending magnets, 3

14 805 MHz Elliptical-cell design: Losses in Watts/m Static /Dynamic err..5 % / 0.3 %.5 deg / 0.3 deg 2.0 % / 0.3 % 2.0 deg / 0.3 deg.0 % / 0.5 %.0 deg / 0.5 deg.5 % / 0.5 %.5 deg / 0.5 deg Lost power (W/m) Lost power (W/m) Lost power (W/m) Lost power (W/m) Distance (m) Distance (m) Distance (m) Distance (m) Misalignment errors are kept at their typical values. Stripper thickness fluctuation: % FWHM. Transverse correction applied Correction for RF static error applied Simulated: 50 seeds with 2E+5 particles. To keep the losses below the W/m limit, the static errors should be about (%, deg) and the dynamic errors about (0.5 %, 0.5 deg). 4

15 345 MHz Triple-Spoke design: Losses in Watts/m Static /Dynamic err. 3.0 % / 0.3 % 3.0 deg / 0.3 deg 4.0 % / 0.3 % 4.0 deg / 0.3 deg 3.0 % / 0.5 % 3.0 deg / 0.5 deg 4.0 % / 0.5 % 4.0 deg / 0.5 deg Lost power (W/m) Lost power (W/m) Lost power (W/m) Lost power (W/m) Distance (m) Distance (m) Distance (m) Distance (m) Same conditions as for the Baseline design except for RF static and dynamic err. Double the RF static & dynamic errors used for the Baseline design. No losses observed at the typical error values of (2%, 2 deg) static and (0.5%, 0.5 deg) dynamic Up to static errors of (4%, 4 deg) and dynamic errors of (0.5%, 0.5 deg) the losses are still below the W/m limit. The Triple-Spoke design is more tolerant of errors 5

16 Proton Driver Linac Structure Spoke cavities to 4 MeV Major Linac Sections Front end Squeezed ILC-style ILS-style 325 MHz 300 MHz 300 MHz

17 Stability Diagram (transverse motion) Unstable due to parametric resonance µ = T µ L 2 Linac operating tunes (black dots) cos(mu T ) Stable for all particles inside the separatrix = s π 2 ( βγ ) S f eemsin s 3 2 λ mc ϕ S Defocusing Factor 7