1. CLIC MDI Final Focusing Magnet Stabilisation Studies

2. Recent History Conventional Facility Design for NLC · Stanford Linear Accelerator Center, March 10 to 28, 2003 CARE/ELAN meeting @ CERN November 23 - 25 2005. CLIC07 Workshop, 16-18 October 2007 CLIC07 Workshop, Stabilisation day at CERN, March 18 Nanobeam 2008 (Novosibirsk, 27 May 2008 ) EUROTeV Scientific Workshop at Uppsala,August 2008 8 Oct. 20082Detlef Swoboda @ CLIC MDI working group

3. (Some) Literature Preliminary Results of the Ground Vibration Measurements at Potential Linear Collider Sites and Reference Places (2003 SLAC) Vibration stabilization for the final focus magnet of a future linear collider (November 2005) Status report on active stabilisation of a linear collider final focus quadrupole mock-up (2006) The stabilisation of final focus system (December 2007 Pramana) US Particle Accelerator School, January 22-26, 2007 in Houston, Texas Status of Mechanical Stabilization (2008 Uppsala) Vibration stabilization for a cantilever magnet prototype at the sub-nanometer scale (July 2008) Study of vibrations and stabilization at the sub-nanometre scale for clic final doublets (2008) CTF3 module and CLIC Final Doublet stabilization (??) 8 Oct. 20083Detlef Swoboda @ CLIC MDI working group

4. Global requirements Stability~ nm Spot sizeSome nm (40 x 1) Beam Measurement accuracy~ 1 % Q-pole strength accuracy ~ 10^-3 --> beam mismatch dominated by errors in measurement and not in magnets Tuning accuracy~ 10^-4 (0.1 mT) Alignment Tuning frequenciesvibration/ripple: t < 1/5 s stability/drift: t < 1 hr long-term stability: t < 1 week very long term stability: t < 1 month extremely long term stability: t < 1 year magnets can be constructed, supported, and monitored so as to meet alignment tolerances 8 Oct. 20084Detlef Swoboda @ CLIC MDI working group CLIC main parametersvalue Center-of-mass energy3 TeV Peak Luminosity7·1034 cm-2 s-1 Repetition rate50 Hz Beam pulse length200 ns Average current in pulse1 A Hor./vert. IP beam size bef. pinch53 / ~1 nm

5. 5 CLIC Linear Collider (~2019): Final doublets in cantilever 2m50 Detector Vertical beam size at the interaction point: 1nm Tolerance of vertical relative positioning between the two beams to ensure the collision with only 2% of luminosity loss: 1/10nm Interaction point Scope of FFS Below 5Hz: Beam position control with deflector magnets efficient Above 5Hz: Need to control relative motion between final doublets 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group

6. FF doublet (NLC ZDR) 8 Oct. 20086Detlef Swoboda @ CLIC MDI working group

7. 8 Oct. 20087 Final Focusing Use telescope optics to demagnify beam by factor M = f 1 /f 2 typically f 2 = L * f1f1 f 2 (=L * ) The final doublet FD requires magnets with very high quadrupole gradient in the range of ~250 Tesla/m  superconducting or permanent magnet technology. Detlef Swoboda @ CLIC MDI working group

8. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group8 Chromaticity correction Minimization of chromatic distortions: factors that influence the solutions to this problem: 1.a reduction in the momentum spread (not always feasible) would reduce the magnitude of the problem 2.The chromatic distortion of a FFS lattice is a function of the distance L*. The closer and stronger the lens the smaller is the distortion. 3.Sextupoles in combination with dipoles (provide dispersion) can be used to cancel chromaticity. Sextupoles introduced as pairs, separated by a –I transform do not generate second order geometric aberrations. However the dipoles introduce emittance growth and energy spread due to synchrotron radiation. Serious constraint. FF design  Balance between these competing effects

9. 8 Oct. 20089 Novel local chromaticity correction scheme P.Raimondi, A.Seryi, originally NLC FF and now adopted by all LC designs. Detlef Swoboda @ CLIC MDI working group

10. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group10 Elements of LC Final Focus System Summary In Linear Colliders, nanometer size beams are obtained by: Final Quadrupole Doublet telescopic system – FD Collateral effects: generate strong chromatic aberrations Sextupoles to correct FD chromatic aberrations – SEXT collateral effects: generate geometric aberrations Sextupoles located at beginning of -I transformer (or equivalent transform) then correct geometric aberrations Dipoles to supply dispersion for Sextupoles correction – BEND collateral effects: generate synchrotron radiation

11. STUDY OF SOME OPTIONS FOR THE CLIC FINAL FOCUSING QUADRUPOLE CLIC Note 506 M. Aleksa, S. Russenschuck 8 Oct. 200811Detlef Swoboda @ CLIC MDI working group

12. Permanent Magnets MaterialCharacteristics samarium cobalt (Sm2Co17)brittle neodymium iron boron (NdFeB) can lose strength under irradiation ultrahigh coercivity grades show very small remanence losses, <0.4%±0.1%, for absorbed doses up to 3 Mgy from 17 MeV electrons Ductile irradiation by 200 MeV protons does reduce the remanence considerably Samarium erbium cobalt (Sm x Er l-x Co)Stability ~ 10 -6 /hr ProsCons No pwr cablesAdjust. Range limitation No cryoDemagnetization No vibrationTemperature gradient High coercitivityRadiation tolerance 8 Oct. 200812Detlef Swoboda @ CLIC MDI working group

13. Permanent Quad Concepts A new style of permanent magnet multipole has been described. achieve linear strength and centerline tuning at the micron level by radially retracting the appropriate magnet(s). Magnet position accuracies are modest and should be easily achievable with standard linear encoders Steel Pole Pieces (Flux Return Steel Not Shown) Rotatable PM (Nd-Fe-B) Block to Adjust Field (+/  10%) PM (Strontium Ferrite) Section 8 Oct. 200813Detlef Swoboda @ CLIC MDI working group

14. The first prototype of “superstrong” Permanent Magnet Quad. Integrated strength GL=28.5T (29.7T by calc.) magnet size.  10cm Bore  1.4cm Field gradient is about 300T/m PHOTO Cut plane view Axial view PM Soft iron 8 Oct. 200814Detlef Swoboda @ CLIC MDI working group

15. Double Ring Structure –Adjustable PMQ- The double ring structure PMQ is split into inner ring and outer ring. Only the outer ring is rotated 90  around the beam axis to vary the focal strength. 8 Oct. 200815Detlef Swoboda @ CLIC MDI working group High gradient  heat load

16. Options for Various L* & X’ing Angles NBS (For Head on) MBS (For 7mrad) SBS (For 2mrad) LBS (2 nd prototype) (For 20mrad) NBS=No Beam Separation,SBS=Small Beam Separation MBS=Medium Beam Separation,LBS=Large Beam Separation Q-size  160mm  100mm Grad. With  20mm bore 212T/m (max) -23T/m (min) 175T/m (max) -77T/m (min) 208T/m (max) 26T/m (min) 186T/m (max) 31T/m (min) Grad. With  14mm bore 296T/m242T/m282T/m272T/m 8 Oct. 200816Detlef Swoboda @ CLIC MDI working group

17. SC Magnets ProsCons Ramping, adjust settingServices Low sensitivity to external fieldsQuench Temperature stabilityVibrations Cryostat Cross-section, inner radius High gradient SC back leg coil 8 Oct. 200817Detlef Swoboda @ CLIC MDI working group Coil dominated

18. General Environment 8 Oct. 2008 Detlef Swoboda @ CLIC MDI working group 18 The CLIC luminosity performance critically depends on the main linac quadrupole vertical stability (<1nm @ 1Hz) and the final doublet stability (<0.1nm @ 4 Hz) in noisy site 0.2nm 1.3nm 4 Hz Measurement of the quadrupole vibrations on active table in vertical direction compared to linac and Final Focus (FF) tolerances at 4Hz (in 2003) S.Redaelli’s PhD 2003 Linac FF

19. Cultural Noise 8 Oct. 2008 Detlef Swoboda @ CLIC MDI working group 19 At SLS at PSI, quad vs ground: some peaks correlate with beam jitter, some can be explained by He compressors or beam infrastructure (seen also in beam), but some peaks are due to girder resonances  Stable or damped support vital, but test in accelerator environment essential R.Assmann et al CERN-AB-2004-074

20. 8 Oct. 200820 B.Bkalov et al PHY REV Spec Topics- Accel and beams 1 031001 (1998) Tevatron integrated rms in deep tunnel: at 0.1 Hz, 0.3micron and falls off with increasing frequency, there are strong peaks correlated with the beam frequency (magnet distortion during running cycle) below 5 Hz and above 20 Hz, the equipment in tunnel is noisier than surface at night. Also very strong correlation with He liquefier plant. In LEP tunnel, above 4 Hz, vibration goes from 0.2 nm to 20nm when beam equipment is turned on.( V.E.Balakin et al CERN-SL-93-30-RFL (1993). Ramila Amirikas and DESY team presented (at LCWS 07) some site measurements Fermilab surface 32 nm at 1Hz, CERN 22 nm at 1Hz Fermilab tunnel 3nm LHC tunnel 2nm Did some calculations to remove the 1/f and to keep only cultural noise: Quiet sites below 10nm, medium below 30nm, and noisy sites above 50nm. Why test in an accelerator environment? Detlef Swoboda @ CLIC MDI working group

21. State of the art inertial sensors 8 Oct. 2008 Detlef Swoboda @ CLIC MDI working group 21 NI PCI-6052 Multifunction DAQ Fast cardLow noise card  Compatible Matlab/Simulink (Softwares used for the algorithm) nm stabilisation equipment exists

22. 22 Isolator: Elastomer: Passive isolation 1 geophone / 1 vertical actuator 2 geophones / 2 horizontal actuators Honeycomb table Controller : Control actuators from geophone data Isolator Active isolation Active isolation from the ground: commercial system Presentation of the STACIS commercial system 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group

23. The prototype 8 Oct. 2008 Detlef Swoboda @ CLIC MDI working group 23 Active compensation 2.5 m long The large prototype and its instrumentation : Actuators used for the active control of vibration :  Force = 19.3 N  Maximal displacement = 27,8 μm  Resolution = 0,28 nm - A stacking of PZT patches -

24. 24 Vibration study of a cantilever beam at high frequencies Ground motion: decreases with frequency Team of DESY Waves on coasts Cultural noiseEffect of the earth Studies focused until now on highest motions (below 300Hz) Vibration study of a cantilever beam for f>300Hz  Amplification at resonances  Impact of acoustic noise 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group

25. 8 Oct. 200825 integrated displacement RMS (with active table ON) Tests with the large prototype: quiet room 1 nm Actuator electronic noise at 50 Hz Detlef Swoboda @ CLIC MDI working group

26. 8 Oct. 200826 Results : integrated displacement RMS Tests with the large prototype Active control CIM Detlef Swoboda @ CLIC MDI working group

27. 27 Conclusion and future prospects Vibration sensors and instrumentation  Ground motion measurements from low to high frequencies (0.1Hz  2000Hz)  Measurement chain found for active rejection of CLIC final doublets vibrations (1/10nm for f>5Hz)  Collaboration with PMD Scientific company to test new electrochemical sensors tending toward the final specification of CLIC  Test of small capacitive sensors with 0.1nm resolution (P75211C of PI) Vibration study of a canteliver beam at high frequencies (>300Hz)  High impact of acoustic noise up to at least 1000Hz for CLIC FD  Measurements to perform on canteliver magnets in an operating accelerator site 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group

28. 28 Active stabilization of a canteliver beam down to the sub- nanometre level above 5Hz  Feasibility of active isolation from the ground proven  Active rejection feasibility of resonances proven  On-going study: multi-sensors multi-actuators system in order to stabilize the beam all along its length  Stabilization to do on a more complex structure closer to the FD design Conclusion and future prospects Simulations give us information about optimal location of sensors and actuators and their number Simulations will allow us to follow the evolution of final doublets design 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group

29. Beam delivery system Crab XP 1 PM SC RM Magnet Technology IP concept 8 Oct. 200829Detlef Swoboda @ CLIC MDI working group

30. Summary (1) Vibration & stabilization – Several studies and R&D Passive damping & active compensation (table) Modeling & active compensation (cantilever support) – Commercial equipment for controlled environment like IC production in accelerator noise > 10 x. – Suspension vs. support? FF Quad magnet technology – High gradient ( N x 100 T/m) requires permanent/SC technology – Combination of both types? 8 Oct. 200830Detlef Swoboda @ CLIC MDI working group

31. Summary (2) IP layout – Push-pull vs. 2 nd IP? Need to define strategy, resources, timescale. 8 Oct. 200831Detlef Swoboda @ CLIC MDI working group

32. Reserve 8 Oct. 200832Detlef Swoboda @ CLIC MDI working group

33. Definition of “Switched-On” Length 16 series of outer ring switching No. 8 cm 4 cm 2 cm 1 cm “Switched- on” length 1 ON 15cm 2 ON OFF 14cm 3 ON OFFON 13cm 4 ON OFF 12cm 5 ONOFFON 11cm 6 ONOFFONOFF 10cm 7 ONOFF ON 9cm 8 ONOFF 8cm 9 OFFON 7cm 10 OFFON OFF 6cm 11 OFFONOFFON 5cm 12 OFFONOFF 4cm 13 OFF ON 3cm 14 OFF ONOFF 2cm 15 OFF ON 1cm 16 OFF 0cm ON Cut Plane View 8 Oct. 200833Detlef Swoboda @ CLIC MDI working group

34. Seismic Measurements 8 Oct. 200834Detlef Swoboda @ CLIC MDI working group

35. Ground motion @ CERN 2002 - 2003 8 Oct. 200835Detlef Swoboda @ CLIC MDI working group

36. 8 Oct. 200836 0.1 nm at 4Hz test No dedicated site in CTF3 Initial tests in Annecy CLEX with accelerator environment Mechanical measurement lab (not a stable zone) Work with survey to find a base-plate site (where alignment tests are made?) LHC tunnel had been suggested, but difficult to access (soon…?) Material: “old” CERN/CLIC table (currently in Annecy) or buy a new one with more accessible control Sensor development Install final focus magnet mock-up on test support Use of different sensors  compare with laser interferometer: reference on floor and measurement on mock-up Detlef Swoboda @ CLIC MDI working group

37. 8 Oct. 200837 Association of active and passive isolation :  Not heavy enough to use industrial products, but it is possible with a larger prototype. The small and elementary mock-up Active isolation The passive layer : Require active isolation Δf Passive isolation is efficient Resonant frequency of the rubber Detlef Swoboda @ CLIC MDI working group

38. 8 Oct. 200838 FD stability Things we don’t know: What is the FD configuration? Saclay? Is it normal or superconducting? (M.Aleksa’s work: Sm 2 Co 17 ) How close to detector? MDI issues=> free-fixed or fixed-fixed configuration? Simulations for different configurations: Free, free-fixed… 1 support, multi-support… Detlef Swoboda @ CLIC MDI working group

39. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group39 Telescopic system In practice, to achieve a telescopic system in both planes we need at least two quadrupoles to simulate each lens of the telescope, and the magnification may be different in each plane.

40. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group40 Final focus chromaticity Strong FD lens has high degree of chromatic aberrations Typically L*~4m,  ~0.01,  ~0.1mm  If uncorrected chromatic aberration of FD would completely dominate the IP spot size! Need compensation scheme. using FD

41. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group41 Chromatic corrections concepts particle with the same input coordinate but a different momentum p1 see the quadrupoles with strenghts than p0. to compensate for this chromatic difference a lattice can be designed where particles of greater momentum encounter an extra quadrupolar field to compensate for the increased momentum. This is achieved by the introduction sextupoles and dipoles into the lattice structures.

42. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group42 Chromatic corrections The magnetic induction of a quadrupole is a linear function of the variable x, y. A particle with momentum p will be affected differently than a particle with momentum p0. The corresponding strenghts of the quadrupole the focal strenght of the quadrupole decreases as the momentum increases. Chromatic properties of a sextupole may be interpreted similarly. Chromatic effects occur because particles with different momenta respond differently to a given magnetic field.

43. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group43 Chromatic corrections concepts This lattice has the potential of chromatic corrections. While a particle p0 follows the central trajectory, the particle p1 with   0 will follow the trajectory defined by the function  d x (s). The function is nonzero after the first dipole. At position 1, p1 encountered slightly different quadrupolar strengths than p0. Let’s arrange a sextupole at position 1, which is not affecting p0. Particle p1 will experience a gradient proportional to its displacement, therefore proportional to  B1QF B2

44. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group44 If proper sextupole strength is chosen, the extra gradient exactly compensates the difference in gradient experienced by particles with different momenta in the preceding quadrupole. However, in this process the sextupoles will in general introduce geometric distortions. A procedure to eliminate chromatic aberrations without introducing second-order geometric aberrations is the following. Chromatic corrections concepts

45. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group45 IP beta bandwidth SLC measured beam size at IP with momentum

46. 8 Oct. 2008Detlef Swoboda @ CLIC MDI working group46 Interlaced pairs Ideally, from second-order geometric aberrations point of view, is to assemble –I transformers that do not interfere between x, y planes (separated in space). This often requires prohibitively long and expensive sections. Consider interlaced sextupole pairs. A particle arrive at first sextupole S1 with displacement x1. As it gets to the first sextupole of pair S2, its motion is perturbed and particle reaches second sextupole S1 with a displacement not equal to –x1. Not exact cancellation from the second sextupole S1. However, since the disturbance by sextupole S2 is of order two, the uncorrected geometric aberrations of the pair S1 are then of order three and four  fine.