ESRF Experimental Contribution Towards Working Group Introduction REVOL Jean-Luc

ESRF Experimental Contribution Towards Working Group Introduction Goal of this talk: Interdependence Longitudinal/transverse/single-bunch/multi-bunch based on ESRF observations Longitudinal Single Transverse Single Transverse Multi. Longitudinal Multi. Vertical Horizontal WG1 WG2 Longitudinal Single Transverse Single Transverse Multi. Longitudinal Multi. Vertical Horizontal 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

ESRF Experimental Contribution Towards Working Group Introduction Goal of this talk: Interdependence Longitudinal/transverse/single-bunch/multi-bunch based on ESRF observations IMPEDANCE MODELING WITH BEAM Resistive wall contribution to single bunch dynamics Broad Band contribution to multibunch dynamics Effect of partial filling on multibunch transverse instabilities Transverse feedbacks WG1 Transverse Single Transverse Single Longitudinal Single Longitudinal Single Vertical Vertical Horizontal Horizontal Vertical Vertical Horizontal Horizontal Transverse Multi. Transverse Multi. Longitudinal Multi. Longitudinal Multi. WG2 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Longitudinal /Single Bunch Bunch lengthening Recent measurements show oscillating values. Is it a sign of a Saw-tooth mechanism ? Stable turbulent regine above threshold 3.5 mA At zero current, discrepancy between theoretical bunch length and measured bunch length > 20 % 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Longitudinal /Single Bunch Measurement of energy spread based on two emittance measurements one in a dispersive section and one in a non dispersive section Microwave regime starts at 3.5 mA for the ESRF. From bunch lengthening and energy spread (simulated by tracking code) BBR model for the longitudinal impedance fres = 30 GHz, Rs = 42 k, Q=1 3.5 mA 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch / BBR Z (M  ) Transverse / Single Bunch / BBR Z ( M ) 20 R/Q ==> no effect of the resistive impedance ==> Positive real part compensates exactly the negative real part Simulation of the mode coupling with a BBR Model 10 Real (Z) f (GHz) -40 -20 20 40 -10 Imag (Z) Mode 0 and mode -1 interact with the imaginary positive part of the impedance ==> defocusing -20 Zero Chromaticity m=0 Detuning of Mode 0 = f (R/Q) Detuning of mode -1 = f( fr) Mode with larger m are much less detuned m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch / Mode merging Z (M  ) Transverse / Single Bunch / Mode merging Z ( M ) 20 R/Q Simulation of the mode coupling with a BBR Model 10 Real (Z) f (GHz) s -40 -20 20 40 -10 Imag (Z) Q Threshold Fr = 22 GHz R/Q= 13.5 M -20 0.8 mA IThreshold m=0 Fit of the model to the experimental data using Moses m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch impedance 30 Z ( M ) 20 R/Q Simulation of the mode coupling with a RW Model For the ESRF the modeling of the mode merging could be obtained with a Resistive Wall model only!!! 10 f (GHz) -40 -20 20 40 -10 Imag (Z) But this resistive wall model should be much larger than the theoretical one !!!! (Factor 10) -20 Zero Chromaticity m=0 m=-1 m=-1 m=-1 From mode merging, We cannot conclude on the broadband shape BBR + RW m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch 20 30 -20 -10 10 -40 -20 20 40 With the negative chromaticity Mode 0 is strongly unstable Thanks to the amplitude dependant tune spread 0.8 mA can be stored independently of the chromaticity f (GHz) Mode 0 shift with negative chromaticity v = -0.5 v = -0.8 v = 0 -40 -30 -20 -10 10 20 30 40 50 ESRF reduced chromaticity should be multiplied by 14.39 Maximum interaction at v = -0.6 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch 20 30 -20 -10 10 -40 -20 20 40 f (GHz) Mode 0 shift with negative chromaticity v = -0.5 v = -0.8 v = 0 @ Measurement of Growth rate in the negative chromaticity -40 -30 -20 -10 10 20 30 40 50 Maximum interaction at v = -0.6 Stabilization by amplitude dependent tune spread induces a saw-tooth effect 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch 20 30 -20 -10 10 -40 -20 20 40 f (GHz) Growing Time Mode 0 shift with negative chromaticity v v = -0.8 v = -0.5 v = 0 -40 -30 -20 -10 10 20 30 40 50 At high chromaticity ==> No contribution at low frequency of the unstable mode ==> need of a high frequency pick-up Maximum interaction at v = -0.6 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch 20 30 -20 -10 10 Probing the real part of the impedance at negative chromaticity ==> no effect coming from the RW -40 -20 20 40 f (GHz) Expected trend from simulation with a BBR Fr = 22 GHz R/Q= 13.5 M Growing Time Mode 0 shift with negative chromaticity v v = -0.8 v = -0.5 v = 0 The too long observed growing time (factor 4) is in favor or Reducing the strength of the BBR (R/Q) obtained from mode merging (compensated by a contribution from the RW) -40 -30 -20 -10 10 20 30 40 50 Maximum interaction at v = -0.6 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Single Bunch 20 30 -20 -10 10 An attempt was made to measure head-tail damping as a function of the single bunch current In the positive chromaticity regime -40 -20 20 40 f (GHz) Despite our efforts, no valid data could be obtained Many difficulties in practice due to : @ Chromatic modulation @ Amplitude dependent tune spread Mode 0 shift with negative chromaticity v = 0 v = -0.5 v = -0.8 -40 -30 -20 -10 10 20 30 40 50 Maximum interaction at v = -0.6 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 -5 -4 -3 -2 -1 1 2 3 4 5 -10 10 f (GHz) * frf Z ( M ) Resistive Wall Impedance 20 Transverse multibunch instabilities are associated to narrow band impedance 10 -40 -20 20 40 -10 Multibunch beam spectrum -20 Zero Chromaticity -1.4 -1.05 -0.7 -0.35 0.35 0.7 1.05 1.4 f (GHz) m=0 n = 991 1.75 -4 -3 -2 -1 1 2 3 4 * frf m=0 Coupled bunch number = 0…. 991 m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 Z ( M ) 20 -10 -5 5 10 -4 -3 -2 -1 1 2 3 4 * frf 10 -40 -20 20 40 -10 -20 Chromaticity 0.3 m=0 m=-1 -4 -3 -2 -1 1 2 3 4 * frf n = 991 Fs spacing between the two lines m=0 m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Observation in uniform filling Transverse / Multibunch 991 mode 0 mode -1 Red: Chi V =0.3 Blue: Chi V =0.35 Fs 20 dB Mode 991 Only two unstable modes 991 990 F0 2 30 db Chiv=0.3 Fo Uniform Filling 990 1 2 3 4 5 6 7 8 9 10 11 12 F0 F0 F0 F0 F0 F0 F0 F0 F0 F0 F0 F0 Chromaticity 0.3 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

multi: n = 991 I= 50 mA ==> single : m=0 Uniform filling Transverse / Multibunch Coupled bunch mode 991 threshold Mode -2 991 Mode -1 Mode 0 multi: n = 991 I= 50 mA ==> single : m=0 0.15 Vertical emittance blow up 20 dB Threshold In uniform filling, RW is a rather smooth instability, what is the criteria for the threshold ??? 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Black is the measured curve 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Resistive wall only, largely underestimates the threshold curve Fit of the threshold curve of mode 991 in uniform filling ==> simulation of the interaction of the beam spectrum with an impedance 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Resistive wall only, largely underestimates the threshold curve Fit of the threshold curve of mode 991 in uniform filling ==> simulation of the interaction of the beam spectrum with an impedance The addition of a broadband contribution helps to increase the threshold which gives more damping with the increased chromaticity The threshold at zero chromaticity is not sensitive to BBR.  b = 8 mm is deduced. Good agreement with beff = 7.3 mm from the evaluation of resistive wall components from the low gap ID vessels Which is also in favour of reducing the shunt impedance obtained with the mode merging The obtained combination of BBR+RW applied to mode-merging calculation gives a rather good reproduction Keeping fres to 22 GHz ==> R/Q = 6.8 M is needed for the fit 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 Z ( M ) 20 10 f (GHz) -40 -20 20 40 -10 Broadband impedance strongly helps stabilize the narrow band resistive wall impedance -20 Chromaticity 0.25 m=0 m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 Z ( M ) 20 10 No sign of transverse HOM induced by the cavity at ESRF. (Narrow band impedance ) What is the mechanism to cure transverse HOM? In 97 the installation of a new pair of cavities modified neither the SINGLE BUNCH dynamics nor the MULTIBUNCH TRANSVERSE dynamics f (GHz) -40 -20 20 40 -10 -20 Chromaticity 0.25 m=0 m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 Z ( M ) 20 10 10 5 f (GHz) -40 -20 20 40 n n + frf n - frf -10 -5 -20 -10 Chromaticity 0.25 Broadband impedance strongly helps stabilize the narrow band transverse HOM instabilities m=0 m=-1 m=-1 m=-1 m=-2 m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch 30 What is the status between impedance modeling and measurement with beam?? At ESRF a new modeling campaign is underway Z ( M ) 20 10 f (GHz) -40 -20 20 40 -10 Chromaticity is a very efficient tool to cure transverse instabilities. Nevertheless, the impact is strong on the LIFETIME mainly for single bunch. -20 Chromaticity 0.25 m=0 m=-1 m=-1 m=-1 m=-2 Other transverse impedance effects at ESRF ??? m=-2 f (GHz) -40 -20 20 40 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch What is the status between impedance modeling and measurement with beam?? At ESRF a new modeling campaign is underway 986 Uniform 124 mA Vertical Chih = 0.2 ChiV =0.19 Bump at 5 F0 991 990 Resistive wall Chromaticity is a very efficient tool to cure transverse instabilities. Nevertheless, the impact is strong on the LIFETIME mainly for single bunch. Other transverse impedance effects at ESRF ??? 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Experiments performed showed a STRONG STABILIZING EFFECT OF PARTIAL FILLING OF THE RING ON THE RESISTIVE WALL INSTABILITIES * and also of the “bump” Even a quasi uniform filling was more stable m = 0 and n =991 Resistive Wall effect “Bump” effect Partial Filling Threshold Is stabilization coming from longitudinal? In uniform, the use of the RF modulation for longitudinal Landau damping did not change the threshold 50 mA Multibunch 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Transverse / Multibunch Large incoherent betatron tune shifts observed are suspected of coming from an ASYMMETRY of resistive wall chamber cross sections. Horizontal Qs Is stabilisation due to an intra-beam tune spread arising from the current dependent tune shift resulting from the differently populated bunches ?? Vertical Quantification of this effect is on the way (R. Nagaoka et al). 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Brutal apparition of the instability at 0.08 Transverse / Multibunch Partial filling pattern 2*1/3 ChiV = 0.08 200 mA could be reached in 2*1/3 with a chromaticity of 0.08 instead of 0.4 (operation) and with a low emittance of 14 pm (at 0,07, the beam exploded) @ Feedback in 1/3 did not work. Brutal apparition of the instability at 0.08 ChiV = 0.08 Stabilization of all the lines by a 2 modes feedback (additional lines are induced by the filling pattern) Mode 991 and 990 Feedback ON Vertical emittance = 13 pm Brutal apparition of the instability at 0.03 (Feedback ON) ChiV = 0.03 2 F0 12 F0 I = 50 mA 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

To get low emittance. 200 mA in uniform at V = 0.1 instead of 0.5 Transverse / Multibunch 200 mA in uniform at V = 0.1 instead of 0.5 200 mA could be reached in 2*1/3 with a chromaticity of 0.08 instead of 0.4 (operation) and with a low emittance of 14 pm (at 0,07, the beam exploded) @ Feedback in 1/3 did not work. But still a lot of lines to stabilize to reach the low emittance !! Coming from the 5 f0 bump?? How many modes should we feedback ?? To get low emittance. Is it achievable with a bunch by bunch feedback?? 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

To get low emittance. Single bunch transverse feedback Transverse / Single bunch Transverse / Multibunch Single bunch transverse feedback is working on mode merging at ESRF Feedback allows to increase the current by a factor 5 !! ==> 0.7 * 5 = 3.5 mA !!!! With Empirical setting of the feedback parameters (most probably resistive) 200 mA could be reached in 2*1/3 with a chromaticity of 0.08 instead of 0.4 (operation) and with a low emittance of 14 pm (at 0,07, the beam exploded) @ Feedback in 1/3 did not work. But in single bunch, this performance is not competitive with the chromaticity 17 mA at 0.9 (and bunch lengthening) How many modes should we feedback ?? To get low emittance. Is it achievable with a bunch by bunch feedback?? 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Feedback in single bunch at higher chromaticity mode m At 3 mA: Without feedback, strong instability at v = 0.3 which could be damped by the feedback 6 mA: Could be reached with feedback, at v = 0.3 (instead of 0.5 without) Then we get two strong unstable modes which could not be damped independently (empirical setting of the phase to optimize the damping of both) 3 mA 134 kHz mode m mode m-1 Feedback is less and less efficient with increased chromaticity and the emittance is strongly affected A feedback acting independently on two single bunch modes is under design 6 mA 132 kHz 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc

Conclusion WG1 WG2 Why is the ESRF longitudinal turbulent regime stable? What are the experimental methods to probe the machine impedance? What is the comparison between experimental model and simulation? How far, RW should be considered in single bunch ? How far, BBR should be considered in multibunch? What is the multibunch transverse stabilizing mechanism in partial filling? What is the single bunch dynamics at high chromaticity? Could we use an harmonic cavity to increase the single bunch intensity threshold? What is the limitation of the transverse feedback in single bunch? How many modes should we consider for a transverse feedback in multibunch? Is transverse feedback compatible with very low emittances ? And much more questions will be asked during the working group discussions !!!! 10/2/2019 Beam Instability Workshop; Monday,13 March 2000 ; Revol Jean-Luc