Beam Instrumentation Challenges at the International Linear Collider

Beam Instrumentation Challenges at the International Linear Collider 2006 Beam Instrumentation Workshop Fermilab 03 May 2006 Peter Tenenbaum 1

ILC Schematic 03 May 2006 Peter Tenenbaum 2

How the ILC Works Every 200 msec: • Sources + DRs deliver 2820 bunches of 2 x 1010 over a period of ~900 μsec – – 307 nsec bunch spacing Normalized emittances 8 μm x 20 nm (400: 1 aspect ratio) RMS bunch length 6 -9 mm 5 Ge. V energy at DR exit • Bunches are collimated, spin-rotated, compressed, accelerated, focused, collided – – – RMS bunch length @ IP: ~300 μm RMS beam size @ IP: 650 nm x 5. 7 nm Normalized emittance @ IP: 10 μm x 40 nm Beam energy @ IP: 250 Ge. V Beam power @ IP: 11. 2 MW per beam 03 May 2006 Peter Tenenbaum 3

Challenges of the ILC • High-intensity collisions @ nanometer scale • Emittance generation and preservation • Luminosity, Energy, Polarization (LEP) measurements of the collision • Machine Protection 03 May 2006 Peter Tenenbaum 4

Beam Position Monitors • Main instrument for emittance control in the ILC – Damping Rings: ~500 x 3 – Ring to Main Linac: 318 x 2 – Main Linac: 314 x 2 – Beam Delivery System: ~300 total – Plus additional BPMs in injectors upstream of DRs, e- linac wiggler insertion (e+ production), etc etc 03 May 2006 Peter Tenenbaum 5

BPM Resolution RMS Beam size in Ring to Main Linac (RTML) – 1 -10 μm is typical Example of main linac emittance tuning – performance as a function of BPM resolution (single bunch of 2 x 1010) Figure courtesy K. Ranjan, FNAL Emittance performance and jitter detection/tracking both argue for resolution in the sub-micrometer regime for most BPMs 03 May 2006 Peter Tenenbaum 6

BPM Center Stability and XY Coupling Beam Aspect Ratio in RTML BPM Offset change of a few μm between global emittance tunings is probably OK 03 May 2006 XY coupling of BPM readings should be ~0. 1% Peter Tenenbaum 7

BPMs: Collisions Strong beam-beam interaction luminosity extremely sensitive to small offsets… Figures courtesy G. White (SLAC) and I. Reyzl (DESY) 03 May 2006 … But a small offset a large kick, which can be measured by BPMs with ~μm resolution Implies that IR BPMs, and some others, must have enough bandwidth to do bunch by bunch measurements, and have good behavior independent of fill pattern (μm level) Peter Tenenbaum 8

ILC BPMs – Additional Requirements • Environment – Many of the BPMs need to be installed near SC RF cavities – Some are in high-radiation environment of detector and/or dumpline • Fill Patterns – Most orbit tuning will be performed with single bunches – Luminosity is produced by bunch trains – Therefore, BPM offsets can’t vary much when operations switches from single bunches to trains • Behavior with varying intensity – BPMs need to operate with charge as low as ~ 1 x 108 (pilot bunch operation) • Reduced resolution okay at these charges (whew!) • How linear do the BPMs need to be over this operating range? How much tuning will be done with very low bunch intensities? • Near-IP BPMs – IR with 2 mrad crossing angle • Both beams pass through each IR BPM • Spent (ie, less-interesting) beam has large offset relative to incoming (ie, more interesting) beam – In general, near IP BPMs are very important! • They play a crucial role in keeping the beams in collision! 03 May 2006 Peter Tenenbaum 9

BPM Technology Choice • “Standard” ILC BPM: RF cavity Measurement of noise in RF-BPM triplet yields ~20 nm single-pulse resolution – Naturally high resolution – Naturally stable centering – Can be made relatively free of common-mode signal – Compatible with ultra-clean cryogenic RF systems – Tend to be “fussy” Some areas: striplines BPM Error [mm] • Figure courtesy M. Ross and G. White, SLAC – Large apertures – Directional – Not great on orbit stability Error signal from FFTB asymmetric BPM Triplet test. The signal corresponds to 2. 7 μm RMS center variation over a period of 1 week. • Though FFTB did pretty well! – Good for extraction lines and 2 mrad IR 03 May 2006 Peter Tenenbaum 10

Transverse Profile Monitors • Requirements: – Good resolution for small spots (a few μm vertical RMS) – Acceptable systematics for large xy aspect ratio (15 -25) – Capacity to endure high charge density of beam – “Turnkey” operation • No solid materials meet all these criteria • Liquid and gas targets have their own problems • For ILC: Photons! – AKA “Laser Wire” 03 May 2006 Survival of SLC wire scanners as a function of particle density. ILC beams are off the scale to the right by at least a factor of 10. Figure courtesy C. Field, SLAC. Peter Tenenbaum 11

• Laser Wires Used successfully at a number of locations – – – • SLC IP (45. 6 Ge. V) KEK-ATF (1. 28 Ge. V) PETRA (7 Ge. V) Challenges – – Signal strength and system design to extract signal Wide variation in beam energies variation in signal strength and form • • – Matching laser time structure to electron beam • • – Can we perform an entire “scan” in 1 train by “moving the light” within the train? Slow scans in 1 bunch / 5 Hz mode Large aspect ratio • – Injector: ~ 1 Ge. V RTML: 5 -15 Ge. V Linac: 15 -250 Ge. V BDS: 250 Ge. V Systematic overestimation of y beam size due to finite Rayleigh range Limited dynamic range • Images courtesy K. Balewski, M. Ross, H. Sakai Can’t reliably scan below ~1 μm RMS size 03 May 2006 Peter Tenenbaum 12

Optical Transition Radiation (OTR) Profile Monitors • Demonstrated capacity to measure 5 μm RMS sizes for > 1 Ge. V beam energy • Single-shot measurement of beam ellipse • Invasive to luminosity production • Single-bunch only – Susceptible to damage • Used in a few locations – Relatively large beams – “Diagnostic of last resort” – Measuring “streaked” beams (more on this next slide!) 03 May 2006 Image of a tilted 10 μm x 13 μm RMS spot at KEKATF via a beryllium OTR profile monitor. Image courtesy M. Ross. Peter Tenenbaum 13

Bunch Length Measurement • RMS bunch lengths of 150 -300 μm – Or 0. 5 -1. 0 psec • Use S-band dipole-mode cavity on zero-crossing – d. V/dz “streaks” beam in vertical – Image streaked beam on a downstream profile monitor • Typically OTR • Occasionally laser wire – Used at SPPS, TTF 2 – Will be used at LCLS, XFEL – Voltage determined by beam energy, RF wavelength, ratio of bunch length to vertical angular divergence • ILC needs are modest because of small vertical divergence 03 May 2006 TTF beam at 450 Me. V which has been vertically “streaked” by a 20 MV dipole-mode cavity at 2856 MHz. Image courtesy M. Huning. Peter Tenenbaum 14

Bunch Length Measurement (2) • ILC dipole mode cavities will be short (~0. 4 m) – Fill time ~150 nsec • Can fill in the inter-bunch interval – becomes a singlebunch device • Limits long-range wakefield buildup • Relatively power hungry • Other diagnostic uses besides σz – Ez correlation • Requires dispersion at readout point – Yz correlation • Requires horizontal streak 03 May 2006 Peter Tenenbaum 15

Bunch Length Measurement (3) • RF Detector method – Measures beam power in highfrequency band • “Poor man’s hardware FFT” aka “xylophone” – Non-invasive, cheap and simple – Only a relative measurement • Needs dipole-mode cavity as calibration 03 May 2006 Measurement of several RF diodes at the end of SLAC as a function of bunch compressor parameters (voltage and phase of the compressor RF). Figure courtesy M. Woods. Peter Tenenbaum 16

Luminosity Monitor • • ILC operates in regime of intense beam-beam interaction Studies have shown that for realistic bunch shapes, optimum luminosity does not occur at the point where the beam-beam deflections are zeroed Only way to optimize is to zero deflections, then vary beam offset to maximize luminosity Has to be done within 1 train – Optimum varies from train to train • Requires a luminosity monitor which can provide useful bunch by bunch information for dither feedback controller 03 May 2006 Simulation of optimizing luminosity as a function of collision offset and angle within 1 train. Necessary because the max lumi, head-on collisions, and zero deflection all occur at different points in the parameter space! Peter Tenenbaum Figures courtesy G. White, SLAC 17

L E P uminosity nergy olarization • Beam polarization crucial part of ILC’s power as a physics instrument • Precision measurement of energy similarly crucial • Strong beam-beam results in a long “tail” on luminosity spectrum, depolarization during collision • Precision measurement of luminosity (at collision), energy (pre- and post), polarization (pre- and post) necessary to understand luminosity spectrum 03 May 2006 Measurements Image courtesy International Linear Collider Technical Review Committee Peter Tenenbaum 18

Machine Protection • Damage from beam impact at normal incidence 2 1010 8 x 6 mm – Niobium: threshold is around 5 x 1014 e-/cm 2 – Copper is about the same (collimators) – Titanium can take ~15 x higher density • At glancing incidence – High-z materials: about the same as normal incidence – Low-z: factor of a few higher density can be tolerated • For β ~ 100 m, 2 e 10 e-/bunch, single bunch density: • ~8 x 1013 @ 5 Ge. V • ~4 x 1015 @ 250 Ge. V – Even at low energy, a few lowemittance bunches will damage anything they hit! 03 May 2006 Front Back Electron micrograph of 1. 4 mm thick Cu target with “silhouette of passage” of 30 Ge. V electron beam. Image courtesy D. Mc. Cormick, SLAC. Peter Tenenbaum 19

Machine Protection (2) • A significant amount of ILC commissioning will need to be done with low-density “pilot bunches” – Lower charge and/or larger emittance – Instrumentation must be adequately responsive for low-intensity tuning • Current baseline design includes a pilot bunch on every accelerator cycle – ~ 10 μsec ahead of luminosity bunches – Additional protection for some areas, including detector • Need dedicated MPS-linked beam instruments which detect the pilot bunch, or radiation generated when it hits something, reliably determine that it did/didn’t made it to the dump • • Need to monitor trajectories of main beam bunches as well – something could go wrong between pilot bunch and end of lumi bunches System from MPS instruments to MPS response devices (abort kickers, DR extraction kicker) needs very low latency – travel times for signals will consume most of the pilot bunch “head start” 03 May 2006 Peter Tenenbaum 20

Conclusions • • ILC will break new ground in terms of beam quality requirements… …which implies breaking new ground in terms of both quality and quantity of beam instruments – Can’t make a new accelerator, with innovative requirements, entirely out of COTS instruments! • • • Heavy reliance on high-precision, stable BPMs for emittance tuning, feedback, feedforward Transverse emittance measurement based on substantial number of laser wire scanners Two bunch-length monitoring technologies: dipole mode cavities and RF spectrum analysis Fast luminosity monitor required for collision orbit tuning on every accelerator cycle Energy and polarization measurements needed to decolvolve luminosity spectrum Machine protection will require some beam instrumentation designed solely for this purpose 03 May 2006 Peter Tenenbaum 21

Acknowledgements There are many people who have worked for a decade and more on linear collider beam instrumentation, either directly (hardware prototyping and experiments) or indirectly (studies of requirements and usage of the instrumentation suite). There are too many people in this category to list, but they know who they are and I thank them. 03 May 2006 Peter Tenenbaum 22