Comparison of single bunch Comparison of single bunch simulations and measurements at the simulations and measurements at the Diamond Light Source Diamond Light Source R. Bartolini Diamond Light Source and John Adams Institute, University of Oxford Thanks to G. Cinque, I. Martin, J. Puntree, G. Rehm --- (WIP) TWIICE SOLEIL, 17th January 2014
23
Embed
Comparison of single bunch simulations and measurements at the Diamond Light Source
Comparison of single bunch simulations and measurements at the Diamond Light Source. R. Bartolini Diamond Light Source and John Adams Institute, University of Oxford Thanks to G. Cinque, I. Martin, J. Puntree, G. Rehm --- (WIP). TWIICE SOLEIL, 17th January 2014. Outline. Introduction - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Comparison of single bunch simulations and Comparison of single bunch simulations and measurements at the Diamond Light Sourcemeasurements at the Diamond Light Source
R. Bartolini
Diamond Light Sourceand
John Adams Institute, University of Oxford
Thanks to G. Cinque, I. Martin, J. Puntree, G. Rehm --- (WIP)
TWIICESOLEIL, 17th January 2014
OutlineOutline
• Introduction
• Review of measurements
• Single bunch tracking code – extension to sbtrack (RN)
• Comparison simulations-measurements
• Conclusions and open questions
TWIICESOLEIL, 17th January 2014
MotivationsMotivations
Diamond operates
low alpha for X-ray and THz users few weeks – incompatible with normal user operationand single bunch in camshaft modeseveral weeks per year – compatible with normal operation
There is interest in understanding and optimising the performance of the ring for these operating modes
THz users operate above the bursting threshold with negative alpha
THz detector development (P. Karataev, G. Rehm, et al.)
“Academic” understanding the dynamics above thresholds
TWIICESOLEIL, 17th January 2014
Low alpha latticeLow alpha lattice
TWIICESOLEIL, 17th January 2014
0 10 20 30 40 50 60 70 80 90
0
10
20
30
S (m)
Bet
a F
unct
ion
(m)
x
y
x
0 10 20 30 40 50 60 70 80 90
0
0.2
Dis
pers
ion
(m)
0 10 20 30 40 50 60 70 80 90
0
20
S (m)
Bet
a F
unct
ion
(m)
x
y
x
0 10 20 30 40 50 60 70 80 90
0
0.2
0.4
Dis
pers
ion
(m)
Parameter Standard User Lattice
Low Alpha Lattice
Emittance 2.7nm.rad 4.4nm.rad
α1 1.7×10-4 -1×10-5
α2 (with sext.) few×10-3 few×10-5
Natural bunch length (3MV) 10.0ps 2.4ps
Synchrotron frequency (3MV) 2608Hz 629Hz
DiagnosticsDiagnostics
TWIICESOLEIL, 17th January 2014
60-90 GHz 220-300 GHz
Sensitivity 28 V/W 1500 V/W 28 V/W 1500 V/W
Response Time <250 ps ~1 μs <250 ps ~1 μs
Measurement Bandwidth >4 GHz ~1 MHz >4 GHz ~1 MHz
Pre-amp input impedance 50 Ω 100 kΩ 50 Ω 100 kΩ
Pre-amp gain 60 dB 40 dB 60 dB 40 dB
60-90GHz SBD
220-300GHz SBD
•THz Schottky diode detectors from a dedicated dipole beamport
• B22 beamline (FTIR)• Streak camera• Orbit data in dispersive BPMs
B22 beamline (FTIR)B22 beamline (FTIR)
TWIICESOLEIL, 17th January 2014
CSR amplification factor for short pulse THz mode (above the bursting threshold)CSR gain > 1000 up to 100 cm–1
CSR gain ~200,000
CSR gain ~1000
Courtesy G. Cinque
Characteristics of THz emissionCharacteristics of THz emission
TWIICESOLEIL, 17th January 2014
105
106
100
101
102
bunc
h cu
rren
t (
A)
(fRF
frev
VRF
cos(s)/1/3)3/7
0
measured
shielded CSRfree-space CSR
α1 > 03
7
03
120
2 cos8
sRFrevRFth
thresh
Vff
cZI
856.043456.7 31
thBane, Cai, Stupakov, PRST-AB 13, 104402 (2010)
Wuestefeld et al., IPAC 2010, p. 2504 (2010)
Shielded CSR theory:
Cai, IPAC 2011, p. 3774, (2011)
Ries et al., IPAC 2012, p. 3030 (2012)
30 h
2
2
02.02
25.0exp385.034.05.0
thFrom VFP simulations:
From free-space CSR theory:
Instability thresholds and quadratic dependence with current in good agreement with theory:
Measurement of time structure of THz pulsesMeasurement of time structure of THz pulses
Includes RF nonlinear potential, chromaticity (and simplecticity)
TWIICESOLEIL, 17th January 2014
Equation of motion (II)Equation of motion (II)
Radiation damping and diffusion for electrons
Transverse
Longitudinal
RT2
'yT2
'y'yy
00'y1n
y
01n1n
RT
2T2
s
001n
s
01n1n
These terms guarantee that when tracking a distribution of macroparticles the equilibrium distribution is has the correct equilibrium emittances, beam sizes, divergences, bunch length and energy spread.
TWIICESOLEIL, 17th January 2014
Equation of motion (III)Equation of motion (III)
Wakefields are generated via the interaction of the beam with the vacuum pipe, by synchrotron radiation and by direct space charge
Transverse
Longitudinal
iz
||0
1n1n )'zz(W)'z('dzC
Nr
iz
p0
1n1n )'zz(W)'z(D)'z('dzC
Nr'y'y
The effect of the wakefield on the single particle motion is lumped in a kick added to the one turn map (both transverse and longitudinal)The kick is computed binning the longitudinal distribution of the electrons and computing W and W|| from analytical formula or numerical codes
TWIICESOLEIL, 17th January 2014
Wakefields implementedWakefields implemented
- CSR impedance with shielding
Simulating the THz pulses as detected by the SBDSimulating the THz pulses as detected by the SBD
The code is capable of
• reproducing the THz pulse as detected by the SBD within the BW of the detector• producing the spectrogrammes (FT of the time signals) vs current as measured
It has been used to replicate
• bunch lengthening curve• centre of charge shift• THz spectrum
mainly by using
• CSR wake• BBR impedance fit (single resonance, R,L,C)• purely inductive impedance L
TWIICESOLEIL, 17th January 2014
α1=-1.0x10-5 / VRF = 3.4 MV
The CSR wake was defined using the machine parameter for bending radius and height of the pipe – no fit
TWIICESOLEIL, 17th January 2014
CSR only• generates a fine structure in the THz spectrum• missing transition with current in THz spectrum• bunch lengthening off
frequency (Hz)
bunc
h cu
rren
t (m
icro
Am
ps)
0 2000 4000 6000 8000 10000 12000
0
10
20
30
40
50
60
70
80
90
100
measurements
simulations
α1=-1.0x10-5 / VRF = 3.4 MV
Adding a BBI with parameter fitted to achieve the best agreement with the data:e.g. Q = 2; r = 310 GHz; Rs = 41 kOhm
TWIICESOLEIL, 17th January 2014
CSR + BBI• generates a fine structure in the THz spectrum• transition with current in THz spectrum present albeit at higher current• bunch lengthening well reproduced
simulations
frequency (Hz)
bunc
h cu
rren
t (m
icro
Am
ps)
0 2000 4000 6000 8000 10000 12000
0
10
20
30
40
50
60
70
80
90
100
measurements
α1=-1.0x10-5 / VRF = 3.4MV
Adding a BBI with parameter fitted to achieve the best agreement with the data:e.g. Q = 1; r = 260 GHz; Rs = 36 kOhm
TWIICESOLEIL, 17th January 2014
CSR + BBI• generates a fine structure in the THz spectrum• transition with current in THz spectrum present albeit at higher current• bunch lengthening well reproduced
simulations
frequency (Hz)
bunc
h cu
rren
t (m
icro
Am
ps)
0 2000 4000 6000 8000 10000 12000
0
10
20
30
40
50
60
70
80
90
100
measurements
frequency (Hz)
bunc
h cu
rren
t (m
icro
Am
ps)
0 2000 4000 6000 8000 10000 12000
0
10
20
30
40
50
60
70
80
90
100
Same CSR+BBI, different machine conditionssimulationsMeasurements = -1.410–5 V = 4 MV
frequency (Hz)
bunc
h cu
rren
t (m
icro
Am
ps)
0 2000 4000 6000 8000 10000 12000
0
10
20
30
40
50
60
70
80
90
100
Measurements = 1.010–5 V = 3.4 MV
simulations
2*s= 2 * 677.94 Hz
Bunch current = 50 Amp nturrn = 100,000 / analysed last 80,000 turns
Time signals from SBD
Simulations
spectrum of SBD signal
Measurements = -1.010–5 V = 3.4 MV
No evidence of microbunching – bursting entirely described by the form factorcurrent = 50 microAmps in conditions of stable bursting
Longitudinal dynamics above threshold for negative alpha
Simulations = -1.010–5 V = 3.4 MV 50 uA - bursting
s= 677.94 Hz
FFT of horizontal motion
Time signals from SBD and orbit dataSimulations = -1.010–5 V = 3.4
Conclusions and ongoing workConclusions and ongoing work
Numerical tracking seems to be capable of reproducing the phenomenology of the single bunch dynamics above threshold and the THz emission
Qualitative agreement even with a simple BBR resonator but the results are very sensitive to the details of the impedance model used
Investigation of dynamics above the bursting threshold shows
• numerical evidence that bursting with negative alpha is not due to microbunching• numerical evidence (TBC) that different bursting modes (and transition between bursting mode within the same machine conditions) might be generated by different wakefields becoming unstable
• Code development: real lattice tracking, more macroparticles, ….