LUMINOSITY AND BEAM MEASUREMENTS USED FOR PERFORMANCEThe vertical beam - PDF document

1 LUMINOSITY AND BEAM MEASUREMENTS USED FO
LUMINOSITY AND BEAM MEASUREMENTS USED FOR PERFORMANCEThe vertical beam-beam parameter in LEP reached0.083 in 1999. In order to achieve and maintain this highperformance a number of different observables are con-tinuously monitored and optimised. The beam sizes aremeasured using X-ray detectors and UV telescopes. Theluminosity is determined directly with tungsten-siliconcalorimeters and indirectly through an accurate measure-ment of the beam lifetime. The tune shift is measuredfrom the tune spectrum in collision. Beam-beam deflec-tion scans provide information about the beam sizes andseparation at the interaction points. The different meas-urements are shortly reviewed and their resolution andtime response is analysed. Their use for the optimisationof LEP is described.The magnitude of unavoidable machine imperfectionsand the efficiency of correction methods determine thevertical emittance in LEP. Ideally it can be reduced to asmall residual value, that is in practice limited by thebeam-beam interaction. The empirical fine-tuning of LEPinvolves more than a dozen methods/knobs, relying onfast and accurate performance measurements [1,2]. Ta-ble 1 summarises some relevant beam and performanceparameters for LEP operation.Table 1: LEP performance and basic beam parameters forhigh energy running in 1999 and 2000. Bunch current Beam energy Repetition frequency Number of bunches 4 per beam Horizontal emittance Vertical emittance Max. luminosity 2 BEAM DIAGNOSTICSThe BEUV telescopes detect the synchrotron light emit-ted by the beam in the near ultra-violet [3]. It provides areal time 2D image of the beam, integrating over 224turns and all bunches in the storage ring. Diffraction anddeformations of the mirror limit its absolute precision inemittance to ~ 0.25 nm. The BEUV image is used forhorizontal beam size measurement and for real-time ob-The BEXE detectors observe the synchrotron light in theX-ray range [4], providing a measurement of the localvertical beam size . A “turn average display” shows a25-turn average of every second, with a precision of~1 % for down to 300 m [4]. The vertical beam sizecan also be measured turn-by-turn for a selected bunch inthe machine. The BEXE is used extensively for emittanceoptimisation, however, its resolution becomes limited forCollision offset monitoringBeam-beam deflection scans are used to collision offset and the convoluted beam sizes at the IP[5]. A scan takes about 20 minutes per IP and the accu-racy in the convoluted beam size is about 10%. TheBBDS is used to optimise the beam-beam overlap and todetect local problems in the beam size, e.g. due to errorsin the IP beta functions.The LEP experiments measure local luminosity. Thenumbers are calibrated with the measured cross-sectionsand therefore quite accurate. However, the short-termresolution is limited with a time response of ~4 min.LEP luminosity monitors: The LEP luminosity monitorswhich consist of 16 Tungsten-Silicon calorimeters detectBhabba scattering events [6]. The high-energy resolutionis limited due to higher background rate and smaller cross 5 10 15 20 25 30 35 40 45 50 18:00 18:00 19:00 19:00 20:00 20:00 21:00 21:00 Lifetime [h]TimeElectrons Lifetime withoutLifetime during collisionFigure 1: Evolution of beam lifetime in LEP. Luminosity from lifetime: The beam current lifetime in collision is limited by beam-beam bremsstrahlung (lowangle Bhaba scattering) at the highest LEP energies [7].In particular there is no significant effect of beam tails.The average luminosity L can then be written as: 3021(h)(h)10(cms)(mA) ˜˜with i

2 being the bunch current and the lifetim
being the bunch current and the lifetime withoutcollisions (from Compton scattering on thermal photons,beam-gas scattering). Figure 1 shows the beam lifetimeduring a fill measured by the bunch current transformer(BCT). A fast online measurement of the average lumi-nosity, based on beam lifetime, was implemented for the1999 run [8] with a resolution of ~2% for a 30 s runningaverage. This performance results from the improvedsignal to noise ratio of 95-105 dB achieved in 1999 forindividual bunch current measurements made at a rate ofBeam-beam tune shift from mode: tune spectrum in collision exhibits peaks at the and modes. The tune difference between the two modesis roughly proportional to the vertical beam-beam pa- beam-beam pa-v (and thus to average luminosity). The ap-proximate relationship p v | 'Qv/4 is being used to calcu-cu-v per IP. A continuous measurement of the mode was implemented for the 1999 run, allowing a fastonline tracking of of v [9]. A measurement is availableevery 3 s with a resolution of ~2% in v. The method al-ways works reliably with bunch currents below 450 The tune locks can be lost for higher intensities, limitingthe range of usability for luminosity optimisation. 30 35 40 45 50 55 Lum. [1030 cm-2 s-1] 45 50 55 60 19:00 19:05 19:10 19:15 19:20 19:25 19:30 Lum. [1030 cm-2 s-1]TimeALEPH Figure 2: Example of luminosity tuning as observed withthe current lifetime (BCT) and the experiments (ALEPH,DELPHI and OPAL; L3 was off). 62 64 66 68 70 72 74 76 00:30 00:35 00:40 00:45 00:50 00:55 01:00 01:05 Lum. [1030 cm-2 s-1]TimeBCT Figure 3: Comparison of the average luminosity from theexperiments (EXPT) and the beam lifetime (BCT). 0.25 0.3 0.35 0.4 0.45 0.5 0.55 19:00 19:00 20:00 20:00 21:00 21:00 ey [nm] 0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.22 00:50 00:55 01:00 01:05 01:10 01:15 01:20 ey [nm]TimeBEXE Figure 4: Comparison of vertical emittance from theBEXE beam size measurement and the BCT luminosity.The LEP performance during physics fills is optimisedby reducing the vertical emittance with corrections ofcoupling, the tunes, the vertical orbit and dispersion.More than a dozen different methods/knobs are used tothat purpose. The optimum performance can only beachieved if small luminosity improvements are detected,.Figure 2 shows an example of luminosity optimisationwith vertical orbit and tune. The luminosity changes areeasily visible from the current lifetime (BCT). The ex-perimental luminosities show a large spread.The average luminosity from the four experimentsshows a significantly smaller spread and in Figure 3 it iscompared to the BCT luminosity for a fill at an energy of101 GeV. The luminosity improvements (indicated bydashed lines) are due to automatic corrections from thevertical orbit feedback. The two luminosity signals are inexcellent agreement and show both the beneficial effect of the orbit feedback. The response of the lifetime signalis, however, much faster (30 s compared to 4 min). Themeasurement in Figure 3 can be used to estimate the lu-minosity and vertical emittance drift due to uncontrolledchanges in the vertical orbit: 0.002 nm per minute.In order to calculate the vertical emittance from themeasured luminosity the design betatron functions andthe design horizontal emittance are assumed (taking intoaccount J). We can then convert the BCT lifetime intovertical emittance and use it to calibrate the BEXEmeasurement of vertical beam size in terms of emit-yyyVEHV ˜Fits to 1998 and 1999 data give C = 0.6/1.0 m and 292/283 m. The fit results are in agreement with theexpectation. The vertical emittance is

3 shown in Figure 4for data from 1998 (to
shown in Figure 4for data from 1998 (top) and 1999 (bottom). The emit-tance is calculated from the BCT luminosity and from theBEXE measurement of vertical beam size. As the BEXEdata is calibrated with the BCT data and Equation 2, weexpect a general agreement. We do indeed find an excel-lent agreement for the larger emittance in 1998. The twosignals track very well, even for short-term variations.The data shown for 1999 (Figure 4, bottom) corre-sponds to the data shown in Figure 3, where the BCT andthe experimental data are in good agreement. For thesmaller 1999 emittance the BEXE beam size measure-ment cannot resolve the short-term variation as seen bythe BCT lifetime and the experiments. The limitation inresolution becomes understandable if the local beam sizem) is compared to the constant term in Equa-tion 2 (~300 m). In addition, the relationship betweenlocal beam size and emittance may be perturbed, for ex-ample by beta beating. 12.513.514.511:02:2411:09:3611:16:4811:24:0011:31:12 TimeLuminosity from BCT lifetimeBB tune shift (0- mode)mode and luminosity from the BCT versus time.A measurement of the vertical beam-beam tune shift modes in the tune spectrum is shown in Fig-ure 5. The signal shows a fast response to change in lu-minosity. The time resolution is only 3 s, compared to30 s for the BCT signal. The relative resolution in lumi-nosity is about 2% for both. The measurement of the modes provides a very fast and accurate tuning signal.However, the measurement is not fully operational, as itrequires manual set-up at the beginning of each fill. Inaddition, the tune locks can be lost with a more noisy tunespectrum for bunch intensities above 450 4 CONCLUSIONA number of dedicated beam diagnostic devices aroundthe LEP ring provide fast online tuning signals for theImaging telescopes (BEUV) look at synchrotron lightin the UV range and provide a real-time 2D image of theOptimisation of the vertical emittance during high-energy physics fills relies mainly on the vertical beamsize measurement from X-ray synchrotron light (BEXE),the luminosity provided by the experiments and the lumi-nosity estimates from the current lifetime (BCT) and the modes. The two later signals were newly imple-mented for the 1999 run. The time response is slow forthe experiments (~4 min), faster for BEXE (1 s),BCT (30 s), and modes (3 s). The BEXE resolutionfor the smallest vertical emittances (0.25 nm) becomesinsufficient for fine-tuning. BCT and modes providesufficient resolution ( 2%). Especially the use ofthe BCT measurement was essential for achieving thehigh performance in 1999 with a vertical beam-beamparameter of up to 0.083.[1]G. Arduini et al, “LEP Operation and Performancewith 100 GeV Colliding Beams”. These proceedings.[2]S. Myers, “Performance related measurements onLEP”. CERN-SL-99-002 DI.[3]R. Jung, “Precision Emittance Measurements in LEPwith Imaging Telescopes, Comparison with WireScanner and X-ray Detector Measurements”. CERN-SL-95-63 BI.[4]R. Jones et al, “Real Time Display of the VerticalBeam Sizes in LEP Using the BEXE X-Ray Detectorand Fast VME Based Computers”. CERN-SL-99-056-BI.[5]M. Placidi and J. Wenninger, “Interaction RegionDiagnostics in e Colliders”. Proc. DIPAC97.CERN-SL-98-13-BI.[6]E. Bravin et al, “Luminosity measurements at LEP”.CERN-SL-97-072-BI.[7] H. Burkhardt et al, “Beam Lifetimes in LEP”. Proc.EPAC 1994.[8]R. Assmann, “Luminosity from Lifetime”. Proc. SLBeam Instrumentation Day 1998.[9]M. Albert et al, “Zero-Pi modes in LEP”. To be pub-lished