Twin bunches at FACETII Zhen Zhang Zhirong Huang

Twin bunches at FACET-II Zhen Zhang, Zhirong Huang, Ago Marinelli … FACET-II accelerator physics workshop Oct. 12, 2015

Outline q Twin bunches at the LCLS § 2 -color free-electron lasers q Wakefield effect in the compression of the twin bunches § Increase the time delay q Preliminary simulations of the twin bunches for the FACET-II § Larger charge, higher peak current and how to control the time delay q Summary Twin bunches at FACET-II 2

Twin bunches & Two bunches Twin bunches at FACET-II Born/accelerated in the same RF bucket Initial time delay ~ 10 ps Final time delay ~ 10 s-100 s fs Delay compression of ~200 Born/accelerated in different RF buckets Initial time delay ~ several ns Final time delay ~ several ns Almost no delay compression 3

A. Marinelli et al. , Nat. Commun. 6, 6369 (2015) Adjustable delay stage Photocathode Laser Pulse Double Pulse Bunch Compressor 1 Eph oton Bunch Compressor 2 Energy Linac Electron Gun 35 e. V Power Splitter Power Twin bunches at the LCLS time Few ps delay T. Maxwell time 4 Few fs delay ~1% energy separation ~35 f s 10 fs UNDULATOR

Wakefield effects Z. Zhang et al. , PRST-AB 18, 030702 (2015) q Wakefields in the accelerating structure complicates the compression of the twin bunches. Bunch profile Wake-induced energy loss Two effects: 1. Reduce the energy chirp over the two bunches 2. Reduce the “local” chirp of each single bunch If we fix the compression factor (or the final peak current), the time delay of the twin bunches will be 5

Twin bunches for FACET-II q The twin bunches at FACET-II require larger charge and higher peak current. Parameters LCLS FACET-II Charge 75 p. C + 75 p. C 1 n. C + 1 n. C Peak current ~4 k. A ~10 k. A Emittance ? Time delay 10 fs~130 fs >~100 fs Energy spread Small for selfseeding ? q The layout and parameters of the beamline in the Li. Track simulation. The numbers listed here are fixed and we vary the phases of L 1, L 2 and L 1 X, and the initial distribution of the two bunches. Gun L 1 X L 0 R =-48 mm L 2 -linac. R 56=-36 mm L 1 S 56 DL 1 Twin bunches at FACET-II 135 Me. V BC 1 335 Me. V BC 2 4. 5 Ge. V L 3 -linac 10 Ge. V 6

Parameters range in the simulations q For a specific initial distribution of the two bunches, we scan the phases of L 1 and L 2, the amplitude of L 1 X to vary the compression in the Li. Track simulation. Parameter Value Charge 1 n. C + 1 n. C Initial delay 6 ps -10 ps Profile of single bunch parabolic Scan step size Phase of L 1 X Amplitude of L 1 X 0 ~ 25 Me. V 0. 5 Me. V Phase of L 2 Phase of L 3 Twin bunches at FACET-II 7

Simulation results q If we require the final peak current >10 k. A, we can pick out the “good settings” in the scan results and show them in the histogram with energy separation and time delay. q Case 1: 1 n. C + 1 n. C, Initial time delay 6 ps, peak current 250 A Twin bunches at FACET-II 8

Simulation results (time delay 100 fs ~ 500 fs) q Case 1: 1 n. C + 1 n. C, Initial time delay 6 ps, peak current 250 A 9

Simulation results q Case 2: 1 n. C + 1 n. C, Initial time delay 8 ps, peak current 190 A When we scale the beam profile to 8 ps and 190 A, the final time delay ranges from 200 fs to 1. 2 ps. 10

Simulation results q Case 3: 1 n. C + 1 n. C, Initial time delay 10 ps, peak current 150 A When we scale the beam profile to 10 ps and 150 A, the final time delay ranges from 400 fs to 1. 4 ps. 11

Comments and summary q From the analysis in the reference (PRST-AB 18, 030702), wakefield effect will increase the final time delay of the twin bunches. q During the simulations for FACET-II, we find the wakefield effects are very strong and the final time delay will become large. It is easy to get >400 fs, even ~ps time delay. In order to cover the small delay (~100 fs), we have to optimize the initial distribution (let the two bunches initially have some overlap). q Considering the length of single bunch is ~100 fs (1 n. C, 10 k. A), the two bunches will overlap in the time domain if the delay is too small (<100 fs). q The simulation results show that the twin-bunch method can cover the range of the time delay from 100 fs to several ps. q The two bunches cannot have the same energy and also there is always large remaining energy chirp. This is because the wakefield at L 3 is very strong due to the high peak current (~10 k. A) and long accelerating tubes. 12

Questions and comments Thanks ! Twin bunches at FACET-II 13