1. Low energy Beamstrahlung at CESR and the ILC Giovanni Bonvicini

2. With: M. Dubrovin, M. Billings, E. Wisniewski, several REU students over the years (E. Luckwald, N. Detgen, N. Powell, M. West) … and much help from many LEPP people I will discuss both visible (incoherent) and microwave (coherent) beamstrahlung (IB and CB)

3. Outline Why develop low energy beamstrahlung Phenomenology of IB Phenomenology of CB ILC detector concepts Current status of CESR IB monitor Feasibility of CESR CB observation Summary

4. What is beamstrahlung The radiation of the particles of one beam due to the bending force of the EM field of the other beam Many similarities with SR but Also some substantial differences due to very short “magnet” (L=  z /2√2),very strong magnet (3000T at the ILC). Short magnets produce a much broader angular distribution and have different coherence properties

5. Beam-beam collision (BBC) d.o.f. (gaussian approximation)

6. More realistic cartoon with asymmetric tails

7. BBC d.o.f. counting at the ILC 7 gaussian transverse d.o.f. 2 beam lengths At least 4 wake field parameters, and possibly 2 longitudinal currents well measured Beam energy spread not measurable by techniques described here but affected by properties of BBC Beam angle(s) and angular spread(s)?

8. Other possible BBC detectors Beam-beam deflection via BPMs. Limited to 2 quantities by Newton’s 3rd law. Semi- passive device. Gamma ray beamstrahlung monitor. Almost certainly a powerful device if it can be built with enough pixels, interferes with the beam dump (340kW) Pairs spectrometer (10 5 per BBC)

9. The rationale for developing CB and IB Very complex BBC phenomena in a large dimensional space Sensitivity to different variables than hard beamstrahlung, mainly through observation of polarization Simple, relatively inexpensive passive devices which can be located away from the beam line CB may provide imaging of the BBC CB so abundant (O(1kW)) so as to be a potential disruption for downstream sensors

10. IB power (stiff beams) CB largely leaves the spectrum unaffected and adds a factor N 1

11. Large angle incoherent power Wider angular distribution (compared to quadrupole SR) provides main background rejection CESR regime: exponent is about 10 ILC regime: exponent is very small

12. IB power dependence in CESR configuration

13. Some examples of IB pattern recognition

14. Coherence vs incoherence

15. Coherent beamstrahlung Coherent synchrotron radiation has been observed many times for very short beams Coherence condition is >  z (there is also a transverse coherence condition, negligible here) A similar situation arises when beams are separated - coherent beamstrahlung Coherent enhancement always proportional to N

16. Coherent enhancement at the ILC (dynamic beams, complete coherence)

19. CB coherent enhancement (vacuum, no angular divergence) C=P(CB)/P(IB) C(,  )=N exp(-(2  z / ) 2 ) (G. Bonvicini, unpublished) Angular effects reduce radiation by O ((  div /  rad ) 2 ) (not important at CESR, factor of 100 at the ILC). This gives a maximum CB power at the ILC in the neighborhood of 1kW

20. Beam pipe shielding Beam pipe effects are important for long magnets (Heifets, Mikhailichenko, SLAC-AP-083) However in the case of beamstrahlung the magnet is shorter than the beam. This needs to be computed. ILC CB not in doubt

21. Main low energy beamstrahlung observables Strong current dependence (N 3 and N 4 respectively) Strong  z dependence Observable dependence on beam-beam offset (very strong for CB) Correlated side-to-side radiation Strongly varying frequency spectrum which peaks at lower frequencies

22. ILC CB detector concept

23. ILC IB detector concept (1-2 mrad)

24. Large Angle Beamstrahlung Monitor Giovanni Bonvicini, Mikhail Dubrovin

25. ¼ Set-up principal scheme  Transverse view  Optic channel  Mirrors  PBS  Chromatic mirrors  PMT numeration

26. Azimuth angle dependence of radiated power Radiated power for horizontal and vertical polarizations  Two optic ports are reserved for each direction (E and W)

27. Set-up general view East side of CLEO Mirrors and optic port ~6m apart from I.P. Optic channel with wide band mirrors Installed ¼ detector Prilim. experiments, VIS and IR PMTs

28. On the top of set-up Input optics channel Radiation profile scanner Optics path extension volume

29. The ¼ detector Input channel Polarizing Beam Splitter Dichroic filters PMT’s assembly Cooling…

30. Sensitivity of R6095 vs R316-02

31. CESR beam pipe profile

32. Check for alignment @ 4.2GeV

33. 2D profile of radiation System is aligned Clear signal of radiation from dipoles and quads  Radiation from IP is covered by this geometry

34. Horizontal & vertical projections

35. Some late surprises ParameterCESR-C proposal CurrentIB change Curr.(mA)18580/600.15  z (mm) 1011.50.03  obs (mr) 1112-130.01- 0.001 (nm) 900-1000600-7000.01 Calc. err.N/A 10

36. PMT rate correlations with beam currents RED & VIS PMTs for this exp. are R6095 for visible light

37. Records selection For further analysis we exclude non- stable radiation periods at CESR currents re-fill In some cases we leave data for no- beam intervals

38. I(e+) vs I(e-) Depending on shift the 2D plot area of CESR currents might be different It can be used to search for correlations with observed PMT rate:  P1 – PMT dark noise  P2 – SR from dipoles and quads  P3 – SR reflected from masks  P4 – Luminosity term  P5 – Coherent beamstrahlung term

39. Fit to the rate for one of PMTs Rate vs record # Fit to the observed rate

40. Residuals In most cases everything is fine with data for vertical polarization Data can be used for all day in the same fit Pull distribution is consistent with statistics

41. Poor residuals for horizontal polarization Horizontal polarization is more sensitive to CESR tuning … Data should be split for “uniformly” performing periods to get fit consistent with statistics

42. Fitting strategies and results Term, HzFraction, % const I(e-) I(e+) I(e-) I(e+) I(e-) I(e+) I(e+) Separate fit for each PMT (no constrains for correlations) We find stable results for vertical polarization, use only PMT1,3 Fix / float coefficients for correlating terms Good quality fits are achieved in folowing cases:  Fit#1: w/o LUM and CBR terms (p4,p5 fixed to 0): fit returns ~10-30% fraction of I+ term that does not look right  Fit#2: Const, SR, and CBR terms are float (p3,p4 fixed to 0): CBR for RED ~7-16% (3-5 sigma significance) !!! CBR for VIS always consistent with 0 !!!  Fit#3: Const, SR, and LUM terms are float (p3,p5 fixed to 0): LUM term is 2 times larger than CBR in Fit#2 (arithmetic effect)

43. Discussion of results Fit#2 is the only indication on CBR so far But CBR term can be easily superimposed by the LUM term Const term is not a constant! It has ~1hour time constant for relaxation after illumination; depends on radiation background in the hall. Needs in special calibration. We may look at West-East and double-port correlations: need to install full scale setup We need to improve cooling of IR PMTs

44. The best try with IR: R316-02… Should be operating @ T=-40C Use cold gas N2 Cooling is not effective  Noise rate varies with temperature  No correlated signal have been observed

45. Acknowledgments Mike Billing John Sikora Stu Peck Mike Comfort Yulin Li Sasha Temnykh Richard Ehrlich Steven Gray Scott Chapman John Dobbins John Galander Valera Mdjidzade Georg Trout Margee Carrier et al. We appreciate many people who were involved or helped us to work on this project:

46. Summary ¼ of setup is installed on CESR and aligned Parts for the full scale set-up are produced at WSU and we are working on their installation Preliminary measurements show that system can observe VIS and RED radiation from IP/magnets region Experiencing technical problems with observation of IR… Need in better cooling system for IR PMTs Working on improvement of the cooling system for IR band

47. CB Observability at CESR Radiated power is propagating essentially in waveguide mode A short beam is still crucial. Observability at KEK-B (  z =6mm) appears very promising Waves will probably propagate in TM mode (M. Billings). TM cutoff is 0.82d and TM maximum power (for  z =10mm) is 2 pJ per BBC (1.7d and 2nJ for TE mode) Observation possible at two BPM stations, located at 0.68m and 3.6m from the IP respectively(M. Billings). One can look at both time and frequency domain Beam pipe bottleneck at SR mask a potential problem E. Wisniewski, S. Belomestnykh, M. Billings, computing the magnetic wake fields at the BPMs

48. CESR beam pipe profile

49. Possible plan for CESR Compare vacuum CB expectation vs ABCI- computed wake fields at BPMs If comparison is favorable, try a test with highest beam population, shortest beam

50. Future developments Design both detectors for the ILC Quadrupole radiation (A. Mikhailichenko) Detailed CB simulation Observables available in CB imaging Beam pipe effects and the possibility of waveguide CB at the ILC

51. Conclusions Some progress in IB. We have hopes that it will be seen this fall CB at the ILC will certainly be present in large but not threatening amounts. Potentially extremely useful for BBC imaging CB observation at present accelerators would be most useful If both these techniques develop, there is a tremendous amount of work to do

52. Spare slides

53. PMT and HV divider Hamamatsu PMTs VIS: R6095 IR: R316-02 HV divider: E990-07

54. Hamamatsu PMT R6095 28mm head on PMT for 300-650nm

55. Hamamatsu PMT R316-02 28mm head on PMT for Near IR Detection from 400-1200nm

56. Current version of N2 cooling IR photocathods are blown by the cold N2 Non-efficient cooling scheme, large termal flows through the metallic parts Working on improvement

57. Glan-Taylor Polarizers Specification Table Material Optical Grade Calcite Wavelength Range 400-1700nm Uncoated Clear Aperture 5mm or 10mm Housing Outer Dimensions12.7mm L x 9.5mm W x 9.5mm H (5mm CA) 19.0mm L x 16.0mm W x 16.0mm H (10mm CA) Length:Aperture Ratio1.5:1 (5mm CA) 1.4:1 (10mm CA) Beam Deviation <= 3 arc min Surface Quality20-10 on input/output faces 80-50 on escape window Wavefront Distortion <=1/4 wave at 632.8nm Field of View @ 632.8nm1.2° (FOV 1) limited by the critical angle of the prism hypotenuse 6.4° (FOV 2) limited by the critical angle of the input window Extinction Ratio 100,000:1 Damage Threshold10W/cm 2 CW @ 1064nm 20MW/cm 2 @ 1064nm, 1ns pulses, 10Hz Coating350-600nm: R ave < 1% per surface over entire range 600-1050nm: R ave < 1% per surface over entire range

58. Hot Mirrors Specification Table Substrate BOROFLOAT™ Dimension Tolerance ±0.5mm (sq.), ±0.25mm (dia.) Thickness 3.3mm

59. Cold Mirrors Specification Table Substrate BOROFLOAT™ Thickness 3.3mm Reflectance 90% visible light Transmission 80% IR waves Temperature Range 50°F to 450°F Dimensional Tolerances ±0.5mm (sq.), ±0.25mm (dia.)

60. 45° Reflective Dichroic Color Filters Specification Table Substrate BK7 Diameter Tolerance ±0.25mm Thickness12.5mm: 3mm ± 0.20mm 25.0mm: 4mm ± 0.20mm Clear Aperture 90% Surface Quality 60-40 Surface Accuracy 1/4 wave Parallelism <1 arc minute Angle of Incidence (AOI) 45° Coating Back Surface MgF 2 AR Coated Red615nm ± 15nm: R abs = 50% 640-750nm: R ave > 98% Blue510nm ± 15nm: R abs = 50% 400-475nm: R ave > 98% Green535nm ± 15nm: R max > 98% FWHM (R%): 65nm ± 15nm

61. Mirror coating Specification Table

62. Glass Specifications Specification Table

63. The best try with IR PMT R316-02 (next day, March 31, 2005)

64. Beam-beam vertical separation For beam-beam separation one would expect a “camel-back” dependence of CBR power vs separation distance

65. Beam-beam vertical separation (cont.) 2004/10/14 dedicated shift John Sikora & Mikhail Dubrovin VIS PMTs o’k but does not show CBR IR PMTs does not show any correlations with beam currents, presumably are not sensitive to IR