Performance of the GlueX Detector Systems

Transcription

1 Performance of the GlueX Detector Systems GlueX-doc-2775 Gluex Collaboration August 215 Abstract This document summarizes the status of calibration and performance of the GlueX detector as of summer Introduction The GlueX experiment took commissioning data in the fall of 214 and the spring of 215. Based on that data and extensive cosmic running, preliminary calibrations of the detector systems have been performed, and many of the systems have reached design specifications. This report presents a snapshot of the current status of the calibration and performance of the GlueX detector and the photon beam line. 2 Beam Line and Tagger Elements The Tagger Hodoscope The Tagger Microscope The Triplet Polarimeter The Pair Spectrometer 3 The GlueX Detector 1

2 Start Counter The quantities that measure the performance of the start counter are the timing resolution, σ t, the efficiency of detecting a charged track, ɛ, and the maximum rate at which the detector can run. Performance values are listed in Table 1. Figure 1 shows the timing resolution in element 29, where 3 ps has been achieved. Carrying out this fit for each of the 3 start counter elements yields the plot shown in Figure 2. σ t 35 ps 3 ps ɛ 95 % Maximum rate Table 1: The performance of the start counter. Entries /.1 ns Channel 29: t target ST - t RF tgt_rf_tdiff_chan_29 Entries Mean.215 RMS.3941 χ 2 / ndf 2128 / 11 Amp e+4 ± 1.46e+2 µ (ns).189 ±.1 σ (ns).2966 ± δt (ns) Figure 1: The time resolution of the start counter element 29 showing 3 ps resolution. Time Resolution (ns) Start Counter Time Resolution Performance Sector Number Figure 2: The time resolution of the start counter as a function of the start counter element number. The brown dashed line shows the design resolution, the blue points show the per element resolution. 2

3 Central Drift Chamber The quantities to measure performance of the CDC are the resolution on the distance from the wire, σ rφ, and the per wire efficiency, ɛ wire. Both of these are functions of the distance from the wire, and to some extent, the polar angle of the tracks. The resolution, σ rφ becomes worse close to the wire because the isochrones (regions of constant electron arrival times) get smaller. This in turn makes it less probable to have a primary cluster produced on the shortest time isochrone. The efficiency, ɛ wire, decreases close to the straw walls as due to the shorter track length in the active volume of the cell. The best measure of z resolution is the vertex resolution pairs of tracks detected in the CDC. Both z vert and xy vert are reported. σ rφ 15 µm 2 µm ɛ wire 95 % z vert 3 mm From two tracks. xy vert From two tracks. Maximum rate Table 2: The performance of the central drift chamber. σ(δd) [cm] Drift distance resolution for Ring Drift time [ns] Efficiency CDC Per Straw Efficiency Vs. DOCA Closest distance between track and wire [cm] Figure 3: The position resolution, σ rφ of the CDC as a function of drift time. Empty target z position counts/mm 4 χ 2 / ndf / p 35 ± 9.5 p ±.1 3 p2.373 ± p ± 1. p ±.1 p ± p ± z (cm) Figure 4: The z resolution at the target can be extracted from the empty target runs. The width of the thin target walls represent a measurement of z vert of 3 mm. 3

4 Forward Drift Chamber The performance parameters for the FDC were estimated using straight (no magnetic field) secondary tracks from the photon beam. In the FDC chambers we use the drift time to reconstruct the hit position in direction perpendicular to the wire (x), and the information from the strips of the two cathodes for the hit position along the wire (y). The wire resolution as function of the distance to the wire is shown in Fig.5(left). The gas mixture used in the chamber, 4/6 Ar/CO 2, is characterized with a big slope of the the time-to-distance function at small distances, resulting in deterioration of the resolution in that region. Such gas mixture was chosen to minimize the magnetic field corrections. The resolution at big distances to the wire is affected by the non-uniformity of the electric field there. The cathode strips register the avalanche produced very close to the wire, not the actual hit. Therefore, the cathode resolution (see Fig.5(right)) in x direction can be inferred simply from the reconstruction of the wire positions (blue points). On the other hand, the cathode resolution in y can be estimated by comparing the reconstructed avalanche position and the expected hit from the external tracking (red points). Note that due to the strip orientations w.r.t the wires, the x-resolution is expected to be about four times worse than the y-resolution. For the x-resolution, plotted as function of the charge, we see typical improvement of the resolution at higher charges. This is not the case for the y-resolution meaning there is a room for further improvement. Every chamber is capable of reconstructing a 3D hit position using both, wire and cathode information. The efficiency for such reconstruction is demonstrated in Fig.6. It is 95% except for that places with bad cathode channels (< 1%). σ x 2 µm 17 µm wire resolution for distances mm from wire σ y 2 µm 2 µm cathode resolution from track residuals in y direction ɛ 3Dhit 95 % Table 3: The performance of the Forward Drift Chamber. Parameters estimated with straight tracks without magnetic field. FDC Wire Resolution FDC Cathode Resolution microns Cathode Resolution, microns resolution of wire position reconstruction track residuals along wires distance to wire, mm charge, arbitrary units Figure 5: The FDC position resolutions: (left) from wires in x direction, σ wire, as function of distance to the wire and (right) from cathodes, σ cathode, in x (blue) and y (red) direction as function of the hit charge. Cathode x-resolution should be scaled down by factor of 3.86 to compare with y-resolution. 4

5 efficiency FDC 3D RECONSTRUCTION EFFICIENCY chamber 1 chamber 2 chamber 3 chamber 4 chamber 5 chamber X coordinate, mm Figure 6: The combined (wire and cathode) 3D hit reconstruction efficiency for the six chambers of third FDC package vs wire position. The dips in the efficiencies correspond to bad cathode channels, most numerous in this package. 5

6 Barrel Calorimeter The performance of the barrel calorimeter is characterized by its ability to measure the energy and position of electromagnetic showers, which improve with increasing energy deposition (E). The energy and position resolutions determine the experimental width of the π and η decays to two photons, which become benchmarks for our ability to reconstruct electromagnetic showers. An unfolding procedure is required to extract the energy resolution as a function of energy, which has not yet been completed. The position along the length of the barrel calorimeter (z) is determined by measuring the time difference between arrival times of the upstream and downstream hits. The position resolution can be determined for charged particles by comparing the calorimeter measurement to the value determined from tracking. π width, σ π 9 MeV MeV η width, σ η 28 MeV z-position resolution, σ z 1.1cm/ E Table 4: The performance of the barrel calorimeter. The width of the π depends on the minimum-energy cut placed on the daughter photons and leads to the range shown. counts/3 MeV BCAL π γγ reconstruction Spring Field on Runs E γ > 35 MeV M π = MeV σ π = 1.3 MeV σ/m = 7.7 % counts/3 MeV 3 BCAL π γγ reconstruction Spring Field on Runs E γ > 55 MeV M π = MeV σ π = 9.4 MeV σ/m = 7. % counts/3 MeV invariant mass (GeV) BCAL π γγ reconstruction Spring Field on Runs E γ > 75 MeV M π = MeV σ π = 9. MeV σ/m = 6.7 % counts/3 MeV invariant mass (GeV) BCAL π γγ reconstruction Spring Field on Runs E γ > 95 MeV M π = MeV 3 σ π = 8.7 MeV σ/m = 6.4 % invariant mass (GeV) invariant mass (GeV) Figure 7: The mass distribution of two photons reconstructed in the barrel calorimeter. The four plots show different minimum photon energy of the decay photons of the π. These range from 1.3 MeV for 35 MeV photon energy to 8.7 MeV when the π is reconstructed with photons with energies greater than 95 MeV. 6

7 Forward Calorimeter Figure 8 shows a plot of the π width versus the lowest energy of the two decay photons. In addition to the current measured resolution, the design resolution is also shown. The energy dependence of the design resolution is dominated by photostatistics, the amount of light from the shower making it into the PMT. For the actual data we seem to see no strong energy dependence on the resolution, which suggests we arent at the intrinsic resolution limit in the actual detector. π width, σ π 7 MeV 12 MeV for 1 GeV photons Table 5: The performance of the barrel calorimeter. FCAL Energy Resolution as of June 215 Mass Resolution [MeV] π Present Calibration Design Resolution E and E γ 1 [GeV] γ 2 Figure 8: The width of the reconstructed mass of th e π using photons in the FCAL as a function of the minimum energy of the two decay photons. The points with errors are the current resolution while the solid red line is the design resolution. 7

8 Time-of-flight Wall The time-of-flight wall is expected to have a per-paddle time resolution, σ t, of 1 ps. σ t 1 ps 96 ps πp separation at 2 GeV 9.55 σ πp separation at 3 GeV 4.65 σ πk separation at 1 GeV 9.35 σ πk separation at 1.5 GeV 3.65 σ Table 6: The performance parameters of the time-of-flight system. Mean Time Difference 1 MTdiff Entries Mean.175 β 1 + TOF q Entries RMS.1833 χ 2 / ndf 1431 / Prob Constant 9941 ± 58.1 Mean.4953 ±.637 Sigma.1362 ± time difference [ns] p (GeV/c) 1 Figure 9: (left) The time difference between the summed time in the two layers of time-of-flight paddles. The 136 ps sigma measurement corresponds to 96 ps time resolution per layer. (right) The β versus p plot for positive particles. Clear bands are seen for e +, π +, K + and p (top to bottom). 8

9 4 Data Acquisition and Trigger Data Acquisition System Level-1 Trigger 5 Summary 9