Thorsten Hellert intra-bunch-train orbit distortion at FLASH FEL Seminar,

Transcription

1 Thorsten Hellert intra-bunch-train orbit distortion at FLASH FEL Seminar,

2 table of contents > motivation > analysis of multi-bunch data recorded from DAQ > RF dynamics > data modeling > plan for further investigation

3 motivation large bunch-trains have been studied in 9mA runs main focus on longitudinal stability > investigate transverse dynamics in bunch-train

4 motivation 4 orbit variations 4 x in mm y in mm s in m > data: user run with

5 motivation.3 relative orbit variations 4 x in mm y in mm s in m > data: user run with

6 analysis of multi-bunch data available in DAQ!! several data sets recorded from DAQ since 214 problems with server problems with BPMs problems with pulse jitter > only one data set available for investigation only 92 pulses recorded no HOM signals no GUN signals no TOROID signals no energy server

7 analysis of multi-bunch data available in DAQ 5.15 BPM3GUN.2 BPM9ACC1! bunch-to-bunch jitter y in mm y in mm BPM3DBC2.7 BPM5UND y in mm.2.15 y in mm number of bunches number of bunches > data: user run with

8 analysis of multi-bunch data available in DAQ 5.15 BPM3GUN.2 BPM9ACC1! bunch-to-bunch jitter y in mm y in mm.4 pulse-to-pulse jitter BPM3DBC2.7 BPM5UND y in mm.2.15 y in mm number of bunches number of bunches > data: user run with

9 analysis of multi-bunch data available in DAQ y in mm BPM3GUN y in mm BPM9ACC !! bunch-to-bunch jitter pulse-to-pulse jitter > stable over 1 2 pulses y in mm BPM3DBC number of bunches y in mm BPM5UND number of bunches averaged over 92 pulses 46 pulses single pulse > data: user run with

10 analysis of multi-bunch data available in DAQ in mm in mm BPM3GUN hor : 1.83mm ver : 5.28mm BPM3DBC2 hor :.94mm ver :.19mm number of bunches BPM9ACC1 hor : 1.6mm ver :.41mm BPM5UND1 hor :.81mm ver :.41mm number of bunches!! bunch-to-bunch jitter pulse-to-pulse jitter > stable over 1 2 pulses different intra-pulse patterns clearly seen > data: user run with

11 analysis of multi-bunch data available in DAQ x in mm y in mm s in m > orbit variations > 1kHz no iron magnets no vibrations GUN RF modules non closed dispersion wakefields, resonances (?) > difficulties unknown sources small number of BPMs insufficient model

12 analysis of multi-bunch data available in DAQ y in mm x in mm > first approach: model independent analysis of beam line s in m

13 analysis of multi-bunch data available in DAQ MIA! > model independent analsysis (MIA) find correlations in data matrix no model needed no physical statements # of bunch # of bpm x 11 x 12! x 1m x 21 x 22! x 2m " " # " x n1 x n2! x nm

14 analysis of multi-bunch data available in DAQ! > singular value decomposition (SVD) y y y y δ x δ δ y δ δ y δ x x x x value of interest: δ σ x 45% σ y 55% σ x' 1% σ y' 9% x σ total σ y

15 analysis of multi-bunch data available in DAQ! > model independent analsysis (MIA) find correlations in data matrix no model needed no physical statements # of bunch # of bpm x 11 x 12! x 1m x 21 x 22! x 2m " " # " x n1 x n2! x nm SVD U $% 1 n s 1 n V $% 1 n spatial strength timing

16 data modeling x BPM1 x BPM2 x BPM3 x BPM4 x BPM5 t t t t t s # of bpm # of bunch x 11 x 21 " x n1 SVD σ1 = 6 σ2 = 2 σ3 = 1 σ4 = σ5 = # of bunch

17 analysis of multi-bunch data available in DAQ σ 1 = 5.94 σ 2 =.56 σ 3 =.32 σ 1 = 5.94 σ 2 =.56 σ 4 =.24 σ 3 =.32 σ 5 =.21 σ 4 =.24 σ 5 =.21 σ 6 =.14 σ 6 =.14 σ 7 =.11 σ 7 =.11 σ 8 =.11 σ 8 =.11 σ 9 =.1 σ 1 = σ 9 =.1 # of BPM # of bunch σ 1 = # of BPM # of bunch

18 analysis of multi-bunch data available in DAQ σ 1 = 5.94 σ 2 =.56 σ 3 =.32 σ 4 =.24 σ 5 = time in mus s in m

19 analysis of multi-bunch data available in DAQ σ 1 = 8.4 σ 1 = 4.89 σ 2 =.45 σ 2 =.53 σ 3 =.3 σ 3 = time in mus s in m

20 analysis of multi-bunch data available in DAQ x in mm y in mm s in m > first approach: model independent analysis of beam line > orbit variations > 1kHz no iron magnets no vibrations GUN RF modules non closed dispersion wakefields, resonances (?)

21 RF dynamics cavity 1 cavity 2! gradient in MeV/m - vector sum! Q1 Q2 E1 E2 beam loading effects detuning of cavities cavity misalignement > Δki Δkj Δxi,j < 1-3 mm/bunch time in µs many bunches required

22 data modeling! set up a theoretical model for ACC1: chamber model of cylinder-symmetric cavity analytical approximation insufficient for ACC1 1 ASTRA vs. chamber model R 11 R 21 > numerical model required to many free parameters for start-to-end tracking linearization needed R 12 R number of cavity

23 data modeling GUN BPM1GUN BPM3GUN ACC1 BPM9ACC1 ASTRA tracking,6 1,3 13,4 s in m free parameters: (x/y/kx/ky)acc Σi (x/y/kx/ky)cavi Δx, Δy V,iCAV Δxf, Δyf (x/y/kx/ky)beam.5.4 > Σ ak = Δkx, Δky φicav

24 data modeling GUN BPM1GUN BPM3GUN ACC1 BPM9ACC1 ASTRA tracking,6 1,3 13,4 s in m free parameters: ( model setup: parameterized lin. transfer matrix: ASTRA tracking > Σ Σ (

25 data modeling horizontal phase space vertical phase space 4 3 initial tracking transfer matrix 4 3 initial tracking transfer matrix x in mrad 1 y in mrad x in mm y in mm

26 data modeling GUN BPM1GUN BPM3GUN ACC1 BPM9ACC1 ASTRA tracking,6 1,3 13,4 s in m free parameters: ( Σ ( > Σ model setup: parameterized lin. transfer matrix: ASTRA tracking misalignments: coordinate system switches

27 data modeling χ lcav Δx β x = % ε = s χ Δx lcav Δx β x = - Δx - lcav / 2 tan(β)% ε = -β s

28 data modeling horizontal phase space vertical phase space initial tracking transfer matrix 2 2 x in mrad 1 1 y in mrad x in mm initial tracking transfer matrix y in mm > misalignments Δx = 14mm Δy = 9mm Δx = 1mrad Δy = -19mrad

29 data modeling GUN BPM1GUN BPM3GUN ACC1 BPM9ACC1 ASTRA tracking,6 1,3 13,4 s in m free parameters: ( Σ ( > Σ model setup: parameterized lin. transfer matrix: ASTRA tracking misalignments: coordinate system switches multi-bunch-interaction: RF-data

30 data modeling probe amplitudes forward amplitu gradient in MeV/m cavity 1 2 cavity 2 15cavity 3 cavity 4 1 cavity 5 5cavity 6 cavity 7 cavity probe phases 5 1 forward phase 3 5 phi in t in mus t in mus

31 data modeling GUN BPM1GUN BPM3GUN ACC1 BPM9ACC1 ASTRA tracking,6 1,3 13,4 s in m free parameters: ( Σ ( > Σ model setup: parameterized lin. transfer matrix: ASTRA tracking misalignments: coordinate system switches multi-bunch-interaction: RF-data > SVD-based fitting procedure for ak

32 data modeling influence of beam x beam kx module x module kx cav 1 x cav 1 kx cav 2 x cav 2 kx cav 3 x cav 3 kx cav 4 x cav 4 kx on horizontal orbit on horizontal kick on energy number of bunch mm per [mm or mrad] mrad per [mm or mrad] MeV per [mm or mrad]

33 data modeling y in mm y in mm x in mm x in mm number of iteration module d kx module d ky

34 data modeling y in mm y in mm x in mm x in mm phi in gradient in MeV/m probe amplitudes probe phases for f number of iteration t in mus 2 7 8

35 data modeling σ 1 = 8.4 σ 1 = 4.89 σ 2 =.45 σ 2 =.53 σ 3 =.3 σ 3 = time in mus s in m

36 data modeling y in mm y in mm x in mm x in mm number of iteration decreased: std(δx): 3µm -> 3µm std(δy): 57µm -> 21µm required parameters: Δkx,ACC1 = -1,15 mrad Δky,ACC1 = -1,3 mrad > remaining features module d kx module d ky

37 data modeling horizontal 9 1 vertical final orbit - measured - by model - measured - by model,4 mm > remaining features wakefields(?) coupler kicks(?) gradient in MeV/m 9 probe amplitudes 19 forward amplitudes probe phases forward phases

38 further investigations parasitical measurements cavity 1 cavity 2 gradient in MeV/m - vector sum > collect more data points time in µs user run with nbunch > 1 different set of gradients read more DAQ channels > GUN > TOROID > energy server find correlations > HOM-BPM (?) > impact on SASE

39 further investigations improve data modeling cavity 1 cavity 2 gradient in MeV/m - vector sum > improve model coupler kicks long range wakefields energy variation ACC39 time in µs

40 further investigations invasive measurements cavity 1 cavity 2 gradient in MeV/m - vector sum > more significant data detuning cavities > 1% > increase and vary Δk(t) time in µs

41 further investigations measurment by C BPM1TCOL forward power time reduced detuning! less forward power vector sum constant impact on vertical orbit no impact on horizontal orbit / energy > H.Schlarb: cavity misalignment in ACC2/3 (?)

42 further investigations invasive measurements cavity 1 cavity 2 forward power in MeV/m - mean forward power > more significant data detuning cavities > 1% > increase and vary Δk(t) manipulate forward power > isolate coupler-kicks time in µs

43 further investigations invasive measurements cavity 1 cavity 2 gradient in MeV/m - vector sum > more significant data detuning cavities > 1% > increase and vary Δk(t) manipulate forward power > isolate coupler-kicks dispersion measurement > isolate energy time in µs > thanks for your attention!