CSR in DL C Biscari LNFINFN J Barranco

CSR in DL C. Biscari, LNF-INFN J. Barranco, P. Skowronski, CERN Granada, LCWS 11, 27 -09 -11

Coherent Synchrotron Radiation basics Particles in bunch radiate coherently (~ N 2) when passing in a dipole at frequencies ~ 1/sl -> Energy loss proportional to r-2 and independent from E Strong dependence on bunch length Effects on Bunch distribution • Position-dependent energy modulation along the bunch • Modulation of the slopes of particle trajectories + x-Energy correlations • Growth of the projected emittance in the bending plane The CSR wake has the shape of the derivative of the longitudinal density. • CSR accelerates the head and decelerates the tail • The more uniform longitudinal distribution, the less strong CSR effects • Shielding effects by vacuum chamber reduce CSR kicks

DBRC parameters DL DL Emittance [mm rad] < 150 Energy spread <1% Bunch length [mm] 1 -2 L [m] 143 Combination factor 2 Final bunch distance [cm] 30 Final average current [A] 8. 4 N of dipoles 22/24 Dipole r [ m] 6. 9 Angle [°] 21. 32/2. 37 Minimum bending radius along the system CR 1 < 150 <1% 1 -2 292 3 10 25. 18 8 20 CR 2 < 150 <1% 1 -2 434 4 2. 5 100. 24 9. 0 15 TA TA < 150 <1% 1 -2 1200 2. 5 100. 114 57 2

Comparison with existing facilities dealing with CSR effects LCLS - FLASH CLIC Drive Beam • E ~ 1 Ge. V(FLASH) 3 -15 Ge. V (LCLS) • sz ~ 10 -20 mm • Q ~ 0. 2 – 1 n. C E = 2. 4 Ge. V • One passage per bunch compressor Multi passages in isochronous rings sz = 1 -2 mm (x 100) Q = 8. 4 n. C (> x 10) From CLIC parameter list – 7. 7 n. C?

Simulations of CSR effects • ELEGANT (Borland) code has been used and results are presented – no shielding effect included Dipoles are cut into 100 slices and the CSR wake is computed from the longitudinal density at the end of each slice. This is used to modify the energy of each simulation particle. • BMAD (Sagan) code simulations including shielding are in progress – not presented here

Bunch distribution from Linac (from Avni Aksoy) - 23000 macroparticles Longitudinal phase space Different distribution shape From the Linac : sl = 1 mm sp = 0. 32 % After the stretcher: sl = 2 mm sp = 0. 32 % Transverse planes ~ gaussian distributions – Normalized emittances = 50 m m rad

Comparison between different DL isochronous cells • Significant emittance growth due to single particle dynamics (40% after non linear term optimisation) • csr effects (r = 4. 7): transverse emittance at DL output= decelerator acceptance with 2 mm bunch length • new DL cell: better non linear behaviour (< 20% increase with not yet sextupole optimisation) • more space for dipoles > longer bending radius (r = 6. 9)

DL optics and layout LDL /2 Total bending angle = 525 ° LDL = 146 m L dipoles = 43%

Bunch length dependence on R 56 (s) Bunch length along the DBRC is not constant: Follows the evolution of R 56 (s) Constant outside the dipoles Oscillations proportional to sp Sigmap = 0. 31 % Sigmal = 1 mm Sigmap = 0. 15 % Sigmal = 2 mm Sigmap = 0. 31 % Sigmal = 2 mm

sl = 1 mm Q = 8. 4 n. C • Energy loss: 0. 3 % • Energy spread : 0. 37 (+ 15%) • Large emittance growth > 100%

sl = 1 mm Q = 8. 4 n. C No csr With CSR Microbunching instability

sl = 2 mm Q = 8. 4 n. C • Energy loss: 0. 1 % • Energy spread : 0. 37 (+ 15%) • Large emittance growth > 100%

Transverse space sl = 2 mm Q = 8. 4 n. C NO CSR Emittance : With CSR initial 50 m after DL – no CSR : 60 m after DL + CSR 80 m < 150 m

Longitudinal phase space sl = 2 mm Q = 8. 4 n. C

sl = 2 mm Q = 8. 4 n. C Time- x No csr With CSR Microbunching instability

CSR along the DBRC Example of Transfer Line between CR 1 and CR 2

1 mm sl = 1 mm versus 2 mm in TL 2 (design in progress) 10% increase 2 mm 2% increase

Emittance growth due to non linear single particle dynamics (DL) (sextupole configuration to be optimised) Final emittance/Initial emittance From 50 m To 60 m (normalized) Issue : Preserving low emittance beams with high energy spread

Longitudinal phase space with CSR (DL) Bunch length Energy spread Total colour scale = 3% Good bunch length preservation

DL final emittance with csr From 50 m To 80 m (normalized) H Having a lower emittance from the Linac does not help in optimising the final emittance value V

Conclusions CSR effects in DL: • Beam distribution from the LINAC stretched to 2 mm shows some increase in energy spread and no degradation of the bunch length. • The horizontal emittance degradation is of the order of 50% (single particle + CSR), which is still a factor of 2 lower than the decelerator acceptance (final normalized emittance = 80 m m rad). CSR must be considered along the whole DBRC, and may impact in final design There is still room for improvement in DBRC design: • optimisation of the sextupole configuration is in the preliminary phase • adoption of the Chasman-Green isochronous cell in the whole system promises an overall improvement of the performances • Vacuum chamber will be designed in order to optimize the shielding effects, and further reduce the emittance degradation. Simulations of CSR effects including shielding are underway. Granada, LCWS 11, 27 -09 -11