Seminar 03/21/03 Experiences with the beam-beam effect at HERA mm Georg H.Hoffstaetter mm Cornell University (formerly DESY) m Georg.Hoffstaetter@DESY.de
HERA and its Pre-Accelerators Protons Electrons 20 keV Source Source 150 keV 750 keV RFQ Linac II 450 MeV H1 - 778 50 MeV Linac III Pia 450 MeV m 8 GeV DESY III DESY II 7 GeV 40 GeV PETRA PETRA 12 GeV HERMES 920 GeV HERA-p HERA-e 27.5 GeV HERA HERA-B Polarized DESY Electrons Protons PETRA ZEUS 6336 m long Georg.Hoffstaetter@DESY.de
HERA under Hamburg HERMES (7 GeV) HER H1 (318 ZEUS GeV) A HERA-B (42 GeV) PETR A Georg.Hoffstaetter@DESY.de
Superconducting HERA-p + HERA-e Georg.Hoffstaetter@DESY.de
Performance of HERA • Design luminosity had been surpassed • Then an Upgrade was needed • Beam separation by super-conducting magnets in the detectors • Focusing to ¼ of the old beam cross-section Georg.Hoffstaetter@DESY.de Georg.Hoffstaetter@DESY.de
(Courtesy F. Willeke) Specific Luminosity ( 1/cm 2 /s/mA 2 ) Specific Luminosity vs Proton Intensity Feb-03 3.00E+30 Luminosity Studies Luminosity Studies 120 Bunches 120 Bunches 120 Bunches Y2002 Goal 2.50E+30 L sp /cm -2 sec-1mA -2 I p I p < 70 < 70 mA mA 2.00E+30 design I p < 70 mA 1.50E+30 Y2002 Studies I e I e < 35 < 35 mA mA 1.00E+30 I e < 35 mA Feb03 Studies 5.00E+29 31 cm L peak L L peak <2.7 x 10 31 cm - peak <2.7 x 10 <2.7 x 10 31 cm - - 0.00E+00 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 I pb / mA 2 s s - Proton bunch current (mA) 2 -1 1 2 s -1 H1 February 03 Extrapolated Luminosity vs Bunch Luminosity extrapolation Luminosity ( 1/cm 2 /s ) Absolute H1 Luminosity Currents 8.00E+31 8.00E+31 7.00E+31 7.00E+31 Luminosity/cm -2 sec -1 L ( 1/cm 2 /s ) 6.00E+31 Feb03 Run 6.00E+31 5.00E+31 Feb2003 Oct02 Studies 5.00E+31 Y2002 Goal L / cm -2 sec -1 4.00E+31 Design Design 4.00E+31 3.00E+31 Y2002 Goal Y2002 Studies Series5 3.00E+31 Feb03 Studies 2.00E+31 Feb 03 Studies 1.00E+31 2.00E+31 0.00E+00 1.00E+31 0 5000 10000 0.00E+00 I pb X I e (mA 2 ) I pb / mA x I e / mA 0 0.05 0.1 0.15 0.2 0.25 I pb X I eb (mA 2 ) I eb / mA x I pb / mA Georg.Hoffstaetter@DESY.de
HERA III Polarized protons in HERA e-A in HERA • Deuteron acceleration: with same Linac • Ion Acceleration requires: - a new Linac - high energy e-cooling • Luminosity: • Polarimeters 7 10 31 = ⋅ = ⋅ ⋅ L L 1 1 • Flattening Snakes A p A A • Spin rotators • At least 4 Siberian Snakes Georg.Hoffstaetter@DESY.de Georg.Hoffstaetter@DESY.de
Parameters Georg.Hoffstaetter@DESY.de
Early experiences 0 . 5 h τ = • Beam sizes have to be matched p beam e to let the proton lifetime be long. • Beams have to meet head on to 10 h τ = about 0.1 sigma to avoid bad p beam electron lifetime. e • Proton and electron tunes have to 100 h τ = be controlled to about 0.002. p e beam • Tunes chosen to avoid resonances 50 h τ = Qx=0.293 Qy=0.297 p beam • e Crossing angles were avoided. Georg.Hoffstaetter@DESY.de
p lifetime drops with e current Proton lifetime (h) Electron bunch current ( � A) Georg.Hoffstaetter@DESY.de
Luminosity for different e currents L s ( 1/cm 2 /s 2 /mA 2 ) L s is independent of e-current time (min) Georg.Hoffstaetter@DESY.de
Higher p halo production for higher Ie Accumulated halo Newly produced halo HERA-B rates HERA-B rates p bunch number p bunch number Tail scraping at HERA-B HERA-B rates t(s) Georg.Hoffstaetter@DESY.de
Beam-Beam Force on e L s So far no reduction of L s by the bunch current I pp b No reduction of L s by the second experiment No reduction of L s by a larger � -funktionen Georg.Hoffstaetter@DESY.de
The evaluation of lumi scans l The Luminosity was initially too small: Lumiscan ( 60 ) ( 72 ) ° ° L L s s ( 60 ) ( 72 ) ° ° L L s s ( mm ) ( mm ) ∆ ∆ x x ( mm ) ( mm ) ∆ ∆ y y � ( ) ( ) ρ x ρ y Bunch has no product distribution: coupling A detailed analysis of lumi scans is ( 72 ) ° L l s only possible when the beam ( 72 ) ° beam kick is taken into account. L s For strong beam beam forces also l the changing � during the ramp ∆ x has to be considered. ∆ y Georg.Hoffstaetter@DESY.de
Self Polarization of the Electron Beam Each 10 10 -th photon flips the spin of the electron In HERA every 38.5 minutes In HERA every 16.2 hours Ideal ring: equilibrium polarization 92.38% HERA: routine operation with 60-65% polarization Georg.Hoffstaetter@DESY.de
First longitudinal lepton polarization VEPP 1970 80% 0.65 GeV ACO 1070 90% 0.53 GeV VEPP-2M 1974 90% 0.65 GeV VEPP-3 1976 80% 2.0 GeV SPEAR 1975 90% 3.7 GeV VEPP-4 1982 60% 5.0 GeV CESR 1883 30% 5.0 GeV DORIS 1983 80% 5.0 GeV PETRA 1982 70% 16.5 GeV LEP 1993 57% 47 GeV HERA 1994 70% 27.5 GeV (longitudinal) Georg.Hoffstaetter@DESY.de
c t r o n P o l a r i z a t i o n L o n g i t u d i n a l E l e 80 transverse polarization [ % ] longitudinal polarization [ % ] 60 40 20 2.5 0.5 1 1.5 2 time [ h ] Georg.Hoffstaetter@DESY.de
���������������������������������� (courtesy M. Minty) ����� ����������������������������������� ��� ������� ��� ������������� ��������������������������� ����������� ������������������������������������������������������������ ��������������������������������������������������������
First polarization at H1 and Zeus 51% 3 Rotator Polarization Studies with Harmonic Bumps 26.-27. February 2003 40% 30% 20% 10% 51% polarization with e/p collisions was possible with Specific luminosities close to the design: �������������������������������� ����������������������� ���������� Georg.Hoffstaetter@DESY.de Georg.Hoffstaetter@DESY.de
Second e-fills have more polarization Ip pol Ie 0 1 2 3 4 5 6 7 8 9 Time(days) Explanation: The first fill and the refilling procedure have increased the proton emittances and decreased the beam beam force that acts on spins. Georg.Hoffstaetter@DESY.de
Runs with more lumi have less pol. March 1999 Polarization L Explanation: Runs with more initiall lumi (that is at the time of maximum lumi in this run) have a higher beam beam force than runs with lower initial lumi, given that the initial electron current is about the same from run to run. Georg.Hoffstaetter@DESY.de
Simulation by spin/radiation tracking 0 . 04 ξ = ey Explanation: The achievable polarization is the maximum of a dens resonance structure. This makes quantitative predictions hard, but a dependence on the beam beam tune shift is clearly visible. Georg.Hoffstaetter@DESY.de Georg.Hoffstaetter@DESY.de
Where are the Beam-Beam Limits? ( ) s ε e L , y calc L s , measured 2 ∆ ( m ) Q β e ( ) s ε y e y L , y measured 2 ∆ Q L s , x measured ε e ( m ) ( m ) β β e e y , measured y y ε e x , measured ( m ) β e y Upgrade and Ip=140mA: emittance starts to grow Georg.Hoffstaetter@DESY.de
Simulation of large beam beam forces Measured lumi Expected lumi for measured emit. Simulation ( ) s ε e L , y measured L s , measured ( m ) β e y Georg.Hoffstaetter@DESY.de
Dipole modes of Gaussian bunches • N Beam beam tune shift for one particle in the r ξ = β ppb e 2 ( ) ex ex πγ σ σ + σ beam beam field of a Gaussian bunch: e px px py ( ) σ σ + σ • Shift in the dipole modes oscillation px px py ∆ = ξ Q Frequency of a Gaussian bunch: ex ex ( ) Σ Σ + Σ px px py Assumption: the bunches remain Gaussian This approximation is justified for a stiff beam hitting a much less stiff beam when the first beam creates a small beam beam kick. Georg.Hoffstaetter@DESY.de
Simulated coherent modes 0 . 009 ∆ ν m = 0 . 041 0 . 027 ξ = = dQ Why? * = ex ex ex + 4 . 0 m β how ? 0 . 272 0 . 082 ξ = = 0 . 013 ey dQ ∆ ν m = ey ey ey (From work with Jack Shi, KU) ν ν px py / f / f f x f x 0 0 0 . 013 0 . 003 ∆ ν sim = ∆ ν sim = ey ex ν ν ν − ξ ex ey ν − ξ ey ey ex ex / f / f f x f x Georg.Hoffstaetter@DESY.de 0 0
Beam Beam experiments of Feb. 2003 Unexplained lumi change over each bunch train: Higher p current Lower specific luminosity Georg.Hoffstaetter@DESY.de
Beam Beam Tune shifts � ex (comp.,meas.) ) A m ( p I � Bunch number � ey (comp.,meas.) ) A m ( p I � Georg.Hoffstaetter@DESY.de
Recommend
More recommend
