1. The Linear Collider Alignment and Survey (LiCAS) Project Richard Bingham, Edward Botcherby, Paul Coe, John Green, Grzegorz Grzelak, Ankush Mitra, John Nixon, Armin Reichold University of Oxford Andreas Herty, Wolfgang Liebl, Johannes Prenting Applied Geodesy Group, DESY

2. 7 March 2003 LiCAS Project: UCL Seminar2 Contents  Introduction  Survey and Alignment of a Linear Collider  Survey Concept  LiCAS System Overview  Frequency Scanning Interferometry (FSI)  Straightness Monitors (SM)  Simulation of LiCAS performance  Summary

3. 7 March 2003 LiCAS Project: UCL Seminar3 Why do we need another collider ?  What’s wrong with the LHC ? It’s a high energy, high luminosity hadron collider Good as a discovery machine; eg: Higgs Hunting But hadron colliders are messy −Difficult to make precision measurements −Cannot determine quantum numbers of initial state NEED A LEPTON COLLIDER

4. 7 March 2003 LiCAS Project: UCL Seminar4 Physics with a (Linear) Lepton Collider  LHC: Can see 120 GeV Higgs  LC: Can see 120 GeV Higgs more clearly M H = 120 GeV, 310 4 pb -1 S/  B=3.6 (5.0 10 5 pb -1 )

5. 7 March 2003 LiCAS Project: UCL Seminar5 Why do we need a Linear Collider ?  Can’t we build a Super-LEP ? Synchrotron Radiation For 1% Synchrotron radiation loss LEP IISuper-LEP Energy180 GeV500 GeV  E / Rev 1.5 GeV5 GeV Radius4.3 km255.8 km Beam Energy Bend Radius

6. 7 March 2003 LiCAS Project: UCL Seminar6 LEP Synchrotron radiation loss sets the size of a Super-LEP Let’s try a Linear Particle Accelerator The Super-LEP

7. 7 March 2003 LiCAS Project: UCL Seminar7 Requirements for a Linear Collider  To study interesting physics, LC must be High Energy to create massive particles High Luminosity to create large numbers of particles  LC must have Large accelerating gradients VERY small beam cross-sections at IP: O(nm) You need to line-up your accelerator VERY precisely

8. 7 March 2003 LiCAS Project: UCL Seminar8 33km 200  m over 600m X-FEL Proposed Linear Collider: TESLA  Collider Length: 33km  Beam Energy: 500 GeV  Beam Luminosity: 10 34 cm -2 s -1  Beam Alignment at IP: O(nm)  Collider Alignment & Survey:

9. 7 March 2003 LiCAS Project: UCL Seminar9 Why is this hard ?  Temperature & pressure gradients inside collider tunnel affect open-air measurements A 600m line of sight can be bent by 4.5mm for 0.1 o C/m temperature gradient  Ground motion will misalign collider; so survey must be quick 200m over 600m Light gets bent by air refraction T

10. 7 March 2003 LiCAS Project: UCL Seminar10 Ground Motion: Effect on Luminosity 2s20s 1week Time to reset collider

11. 7 March 2003 LiCAS Project: UCL Seminar11 Extra Survey Constraints  Confined space (also used as emergency escape)  Collider has mixture of straight and curved sections  Electrically noisy environment

12. 7 March 2003 LiCAS Project: UCL Seminar12 When to Survey Accelerator  Tunnel Construction Check tunnel has stopped settling  Accelerator Installation Check component positions (& correct them)  Accelerator Maintenance If a component is replaced; the accelerator will be re-surveyed Each step has to achieve 200  m over 600m precision  Accelerator Diagnostics Check accelerator maintains alignment (& correct it) Find out what went wrong

13. 7 March 2003 LiCAS Project: UCL Seminar13 Traditional Accelerator Surveys  A team of surveyors using theodolites, laser trackers, etc Make precision measurements of accelerator site and accelerator A survey takes months to complete and requires a large team of people.  But this approach is not suited to LC because: Cannot achieve required accuracy Slow Manual Large space required

14. 7 March 2003 LiCAS Project: UCL Seminar14 Solutions: Hydrostatic Levelling Systems  Traditional method to measure vertical alignment  But water only follows local geoid…some parts of TESLA don’t  ….while NLC does not at all Measured Vertical Height NLC

15. 7 March 2003 LiCAS Project: UCL Seminar15 Other Solutions  Use a long stretched wire The wire will sag under gravity: Only good for horizontal alignment  Use a laser to align accelerator In open-air, it will be refracted by temperature gradients TESLA follows Earth’s geoid. So cannot be used for TESLA

16. 7 March 2003 LiCAS Project: UCL Seminar16 Survey Procedure  Two-step Survey procedure 1.Survey equidistant tunnel wall markers via multiple overlapping measurements: LiCAS Job 2.Measure collider components against wall makers:  Advantage: The same procedure is employed during tunnel construction, collider installation, operation and maintenance Accelerator wall Survey Train Accelerator

17. 7 March 2003 LiCAS Project: UCL Seminar17 Survey Train  A survey train is used to perform the first step Mechanical concept developed by DESY Geodesy Group LiCAS provides an optical metrology for the train  Survey Train carries two systems Frequency Scanning Interferometry −Makes 1D Length Measurements Laser Straightness Monitors −Measures transverse displacements and rotations

18. 7 March 2003 LiCAS Project: UCL Seminar18  Each carriage measures the position of a reference marker in its own co-ordinates  Q: How to tie reference marker co-ordinates together Survey Train: External Measurements Carriage 1 Carriage 2 Marker 1 at (x1,y1)Marker 2 at (x2,y2) 1D FSI Length Measurements

19. 7 March 2003 LiCAS Project: UCL Seminar19  Use internal system to relative positions of carriages  Internal systems ties the external measurements together Survey Train: Internal Measurements Carriage 1 Carriage 2 (x c2,y c2 ) Marker 1 at (x1,y1)Marker 2 at (x2,y2) 1D FSI Length Measurements SM Measurements

20. 7 March 2003 LiCAS Project: UCL Seminar20 Survey Train: LiCAS Systems  An Optical metrology system for survey of a linear Collider Fast, automated high precision system Can operate in tight spaces Rails attach to tunnel wall z x y Vacuum tube 5m Internal FSI Lines SM Beam 0.5m External FSI Lines

21. 7 March 2003 LiCAS Project: UCL Seminar21 collider component Tunnel Wall Reconstructed tunnel shapes (relative co- ordinates) wall markersinternal FSI external FSISM beam LiCAS technology also applicable to second instrument ! Survey Implementation

22. 7 March 2003 LiCAS Project: UCL Seminar22 Frequency Scanning Interferometry  Interferometric length measurement technique  Require precision of 1  m over 5m  Originally developed for online alignment of the ATLAS SCT tracker Tunable Laser Reference Interferometer: L Measurement Interferometer: D Change of phase:  GLI Change of phase:  Ref time I Ref time I GLI (Grid Line Interferometer (GLI))

23. 7 March 2003 LiCAS Project: UCL Seminar23 FSI: Length Measurement  GLI  Ref

24. 7 March 2003 LiCAS Project: UCL Seminar24 FSI: Thermal Drift Cancellation  Thermal effects add subtle systematic errors to FSI −Nanometre movements can contribute micron errors (   Use two lasers tuning in opposite directions to cancel thermal drift

25. 7 March 2003 LiCAS Project: UCL Seminar25 FSI: Thermal Drift Cancellation  GLI  Ref   True Gradient Measured Gradient with Laser Tuning Up Measured Gradient with Laser Tuning Down

26. 7 March 2003 LiCAS Project: UCL Seminar26 FSI: 2-Laser Thermal Drift Cancellation

27. 7 March 2003 LiCAS Project: UCL Seminar27 FSI: ATLAS Implementation

28. 7 March 2003 LiCAS Project: UCL Seminar28 FSI: ATLAS Test Grid  6 simultaneous length measurements made between four corners of the square.  +7th interferometer to measure stage position.  Displacements of one corner of the square can then be reconstructed.

29. 7 March 2003 LiCAS Project: UCL Seminar29 FSI: ATLAS Resolution

30. 7 March 2003 LiCAS Project: UCL Seminar30 1m1m FSI: ATLAS Resolution  Stage is kept stationary  RMS 3D Scatter < 1  m

31. Retro Reflector ATLAS FSI System Laser 1 Laser 2 Reference Interferometer piezo detector C-Band Amplifier (1520-1570 nm) L-Band Amplifier (1572-1630 nm) Splitter Tree LiCAS FSI System 1m GLI Uncollimated Quill APD Collimated Quill 5m GLI ADC + AMPS RAMRAM To PC f1f1 f2f2 Amplitude Modulation @ f 1 Amplitude Modulation @ f 2 Detectors Demodulator @ f 1,  1 Demodulator @ f 2,  2 Demodulator @ f n,  n

32. 7 March 2003 LiCAS Project: UCL Seminar32 Erbium Doped Fibre Amplifiers  EDFA are optical power amplifiers Used to amplify low power tunable laser Standard equipment for Telecoms −but will it work for FSI ? Decay Signal ~1550nm Pump 980nm 4 I 15/2 4 I 11/2 4 I 13/2  Incoming Single Photon   Outgoing Photons fluorescence Wavelength / nm 1530 1610 Single Telecoms Channel

33. 7 March 2003 LiCAS Project: UCL Seminar33 Quill Collimation  Refractive  Reflective Quill end Retroreflector Collimation lens Retroreflector Reflective, off- axis paraboloid Quill

34. 7 March 2003 LiCAS Project: UCL Seminar34 Laser 1 M1 M2 Detector Laser 2 Demodulator @ f 1,  1 Demodulator @ f 2,  1 wavelength time 1 2 wavelength time 0 2 Volts time Volts time t0t0 t1t1 t0t0 t1t1 Amplitude Modulation @ f 1 Amplitude Modulation @ f 2 f1f1 f2f2 Two Laser AM Demodulation  Need 2 lasers for drift cancellation  Have both lasers present & use AM demodulation to electronically separate signals

35. 7 March 2003 LiCAS Project: UCL Seminar35 Volts Time 15% mod. Time Volts Amplitude Modulation on FSI fringe @ 40 & 80 kHz (now) 0.5 & 1MHz (later) FSI fringe stored as amplitude on Carrier (à la AM radio) Demodulation reproduces FSI Fringes High Pass Filter Two Laser AM Demodulation

36. 7 March 2003 LiCAS Project: UCL Seminar36 Results of Demodulation Demodulation of modulated laser does not effect interferometer signal Both signals have same frequency !!

37. 7 March 2003 LiCAS Project: UCL Seminar37 Reference Interferometer Phase Extraction  Reference Interferometer is FSI’s “yard-stick” Must measure interferometer phase precisely  Uses standard technique of Phase-Stepping Step1: I(  true -1.5  ) Step2: I(  true -0.5  ) Step3: I(  true +0.5  ) Step4: I(  true +1.5  ) Carré Algorithm  true Reference Interferometer mirror moved in 4 equal sized steps

38. 7 March 2003 LiCAS Project: UCL Seminar38 Raw DataReconstructed Interferometer Signal Software Phase Extraction  Telecoms laser tunes linearly  Extract phase with software “phase-stepping”

39. 7 March 2003 LiCAS Project: UCL Seminar39 FSI: Extensions for LiCAS  Collimation optics for quill outputs  Move to Telecoms wavelength (1510nm – 1640nm) Telecoms fibres and equipment are cheaper Exploit cheap, high quality lasers −Reduce drift errors –x300 increase in continuous tuning range (0.24nm ®130nm) –x3000 increase in tuning rate (100 GHz/min ® 5THz/sec) −New features such as Amplitude Modulation (AM) Use Erbium Doped Fibre Amplifiers (EDFA) −Modular power distribution

40. 7 March 2003 LiCAS Project: UCL Seminar40 Straightness Monitors  Used to measure carriage transverse translations and rotations  Require 1  m precision over length of train z y Translation: Spots move same direction Rotation: Spots move opposite directions CCD Camera

41. 7 March 2003 LiCAS Project: UCL Seminar41 z y x y SM beams coming out of the screen Image of beam spots observed on CCD Camera SM: Rotations about Z  Use two parallel beams to measure rotation about z-axis

42. 7 March 2003 LiCAS Project: UCL Seminar42 SM: Splitter Configurations 1.Single Beam Splitter + End carriage retroreflector 2.Double Beam Splitter per carriage z y CCD 1 CCD 2 z y  Pro: Measurements independent of splitter angle  Con: Retroreflector introduces unknown transverse walk to all carriages  Pro: No retroreflector: No unknown walks  Con: The angle of each beam-splitter in each has to be determined; 12 extra calibration constants

43. 7 March 2003 LiCAS Project: UCL Seminar43 SM: Low Coherence Beams  Low coherence length diode lasers are used to avoid CCD interference Stray reflections off surfaces can interfere if coherent Beam- Splitters The two reflected rays can interfere if coherent CCD Chip CCD Glass Face-plate

44. 7 March 2003 LiCAS Project: UCL Seminar44 SM: Interference Rings  Laser with long coherence length.  Interference rings observed on CCD  Laser with low coherence length  No interference structure is observed Interference Rings Perfect Gaussian Beam

45. 7 March 2003 LiCAS Project: UCL Seminar45 SM: Demagnification Lenses  CCD cameras are ½’’ square.  A long collimated beam  large beam This can be larger than the CCD  Use of demagnification lenses increase dynamic range Lenses must be high quality to prevent beam distortion CCD

46. 7 March 2003 LiCAS Project: UCL Seminar46 SM: Results

47. 7 March 2003 LiCAS Project: UCL Seminar47 SM: Stability Results 0.125 pixels = 1  m

48. 7 March 2003 LiCAS Project: UCL Seminar48 SM: Extensions for LiCAS  Use of two parallel beam to measure rotations about z-axis  Two beam-splitter configurations are under investigation  Simple SM under test Low coherence length laser under test Demagnification lenses are being designed

49. 7 March 2003 LiCAS Project: UCL Seminar49 Without tilt metersWith 1-Axis tilt meters Simulations (of single car)  FSI resolution 1  m, SM resolution 1  m  Weak measurement of rotation around z-axis due to small separation between two beams on CCD  Tilt meters resolution 1  rad

50. 7 March 2003 LiCAS Project: UCL Seminar50 Simulations of Train over 600m Imperative each point is known precisely!!! Error on positions < 200  m after 600m

51. 7 March 2003 LiCAS Project: UCL Seminar51 Summary  Future linear colliders require precision survey and alignment  The LiCAS group is developing optical metrology techniques to address this in collaboration with DESY  Proposed solution is being developed for TESLA but can be applied to any collider  Preliminary results have been encouraging  LiCAS is now PPRP approved