Test of an Internal Wire Target at the Hera Proton Ring
Total Page:16
File Type:pdf, Size:1020Kb
TEST OF AN INTERNAL WIRE TARGET AT THE HERA PROTON RING THOMAS LOHSE MPIftir Kemphysik ffeiddberg, 69Of?9, ccrmcng 1. INTRODUCTION Shortly after the HERA machine started its luminosity operation a working group was formed to study the possibility of exploiting the HERA proton beam for a high statistics B experiment able to discover CP violation in the B system. The efTort resulted in two detailed reports I,’ submitted to the DESY PRC. HERA has four intersction paints, two of which are occupied by the two large exper- iments HI and ZEUS, and one which is planned to be used by the HERMES experiment. The fourth interaction zone is currently available and suited to house a dedicated B de$ector. In principle, HERA has two independent rings and one could investigate the possibility of proton proton collisions (with asymmetric energies). Since this would, however, interfere with the routine e-p operation, we concentrated on the fixed target option in which protons from the beam halo intersct with massive internal targets, e.g., thin wires grouped around the beam center at distances of about 4 r.m.s. beam sizes. In the fixed target mode, the center of mass energy is only slightly above 40 GeV. This is quite close to the b&threshold and thus the B cross-section is tiny; about 10 to 20 nb are expected theoretically. The charm cross section- a dangerous background- is three orders of magnitude larger and the total inelastic cross section dominstes by 6 orders of magnitude. In order to get a precision of 0.05 in sin(2p), one thus has. to produce a huge number, lo”, of inelastic events, requiring 5 snowmass-years of data taking at interaction rates of about 39 MHz. There rates are in principle achievable at the HERA proton ring: With a nominal 2. 1013 protons stared and lifetimes around 50 to 100 hours, natural beam losses correspond to 50-100 MHz. We have to require, however, that the internal target is very efficient and able to absorb around 50% of all halo protons (which are about to leave the machine anyhow). Since the feasibility of an internal target is of primary importance, beam tests were set up at the HERA proton ring. The mein results of the 1992 tests are summuizcd in the following. a. EXPERIMENTAL SET-UP The choice for the target location, llgm upstream of the center of the HERA w-t hall, was dictated by practical necessities like finding II beam pipe section void of other 641 machine elements. Relevant parameters for optics, beam and target are summarized in PI Table 1. The optics are quite different from that for .the proposed final target srrsngcment; n cl most significantly the relatively large horizontal 0-f unction of 93 m exceeds the maximal value I, \I” of 50m assumed for e B experiment. In addition, other (position independent) parsmeters, the total current in the machine, the machine aperture, and the number of target wires, were smaller than in the design in reference’. Table 1: Parametera of the target wire and the proton beam at the ted location 1lSm 1ow nr upstream of the HERA west hall. II IDP L Proton Beep: Figure 1: Sketch of the expedmental set-up. Sl, S2: scintillatars; Sh: shower counters; Beta function p. =96 m Vl, V2: scintillators used as VETO counters; T: target. Alpha a* = -1.68 Emittsnce tr = 9. 10-O red m t, = 9. lo-’ red m Spatial dispersion d. = 0.82 m d. = 0 bunches. Data on the background detected by Si-pin-diodes in the vicinity of the collima- Angular dispersion d., = 24 mrad d.. = 0 tors and from ZEUS (scintillators C.5, proton-gated) and Hl (veto weI1) were continuously Beam size (r.m.s) cz = 0.9 mm or = 0.5 mm displayed in the control room and were partly available off-line. Number of bunches 10 (with 96 ns spacing) In total we performed 7 experiments with the wire close to the beam, corresponding Typical current 1.6 “A (= 2. 10” protona) to L IotaI of 22 hours of data taking. In particular, the wire was several timea systematically Target Wire: moved towards the beam in small steps, while HERA was operated in normal luminosity Material Copper mode. In these situations, where the test experiment was running in a purely par&tic Diameter 100 pm mode, the main collimators were positioned at about 5 1.m.8. beam widths from the nominal proton beam position end were not moved. In two occasions during the power saving hours Interaction length 151 mm in November, when only protons had been stored in the machine, the collimators could be Radiation length 14.3 mm opened with the wire already positioned in the beam halo. The beam lifetimes were~typically around 100 h when the target ,u retracted. With- The target was mounted verticeIIy on e movable fork. It consisted of a 1OOpm thick out e-p collisions, the lifetime wea probably much larger, but no messurement beyond 100 h copper wire, followed by a second spare wire and finally a 100pm thick copper foil. Since was available. The presence of the target reduced the lifetime to typically 30-SOh, either the first wire was not damaged during the tests, the back-up targets were never used. The with or without e-p collisiona, and no distinct differences between these two aitustions were fork war driven by a stepping motor (normally used for collimators) with a step size of 3pm. observed. It was directly controlled via the collimator control panel from the HERA control room. The target area we.8 very close to the main proton collimators. These beam scrapers were The data presented here were taken from those running periods where the conditions situated about 6 m downstream of the target. Secondary collimators existed 214 m and 259 m were rather stable, i.e., neither jumps in the beam position nor violent besm losses in either downstream of the main collimators. the proton or the electron beam occured. Fig. 1 shows the whole experimental set-up. Downstream of the target (T) and e beam pipe section of increased diameter (ZOOmm), telescopes were placed above, below s. INTERACTION RATES IN WIRE SCANS and at both sidea of the beam pipe. Each of the four telescopes consisted of two plastic scintillators, a smaller one (Sl, 40 x 40mm’) positioned at 1.4” from the target, end e. In the parasitic experiments, the wire wan carefully moved towards the beam in emaIl larger one (SZ, 80 Y 80mm’) at 1.8 m, followed by II acintillator-lead shower counter (Sh) of steps of 30pm. An example is shown in Fig. 2e, which displays the horizontal wire position the same active ares end a depth of 18 radiation lengths. After the first measurement, the ea c function of time. At time 2OOOee first increase of the trigger rate (Fig. Zb) wee observed. upper telescope wea dismantled and its two scintillators were plsced 0.3m upstresm of the Simultaneously the percentage of triggers with the VETO counters fired (Fig. 2c) dropped target M “VETO” counters (Vl, VZ). Th 1s rearrangement sllowed us to determine to what from 90% to 35%, demanstrsting that iI& incrcaae in the trigger rate wea not due to beam extent the telescopes were triggered by beam related beckground and not by intersctions in beckground (which could have possibly been created by the disturbance by the wire). The the target. obvious interpretation ie that at ibis time the wire just moved out of the shadow of the During the whole running period HERA WM filled by a short train of 10 consecutive collimator and started to scrape away protons from the beam hdo. bunches (with e hooch spacing of vx 96ns). The arrival times st the test experiment were In the following hours the wire was moved another 14 timea towards the beam, cu GUI derived from e close-by besm pick-up. Only the bunch-crossing (BX) signal of the Iead- be seen in Fig. 2. After each step, the coincidence rate first increased sharply, accompanied ing bunch wes allowed to trigger the tclcscopea. This rulea out any pile-up from preceding by e corresponding drop in beam lifetime. Within minutes, the rate then gradually settled 642 5.4 c) Yeto rate 17.1 5.2 1 ij ‘3-l 1 5 *.a a) wire pas. Imml z/u +.5 zoo0 4000 6000 8000 10000 0 60 80 60 80 c b\ coincidence rote fkHZI , I TfJC channel TDC channel E°Cfl Figure 3: Time spectra of events, where one (hatched histograms) or both (open histograms) VETO counters had fired. One TDC channel corresponds to 0.6 ns. The spectra are shown for the wire retracted or fully moved in. 0I’ 2000 A,000 6000 BhoO 10000 0 zoo0 moo 6000 BOO0 10000 time Is1 time Is1 background events set both VETO taunters at once in almost all CBS~S, When the target Figure 2: Wire scan information venzu~ time: a) Ho&x& wire position; b) ORed coinci- is moved in, however, x 80% of the events set only one counter, showing the presence of dence rate using three scintillator telescopes; c) percentage of triggers with VETO counters a new low multiplicity component in the counters. in addition, in these events where only fired; d) background rate measured by the proton gated ZEUS s&tiUator C5. one VETO counter is set, the VETO signal is delayed by zz 3 ns with respect to the beam background.