Neutrons from p - p-> p 0n are generated uniformly distributed over the 30 neutron counters. The origins of the neutron momentum vectors correspond to the stopping distribution in the LH2-target of ±5 cm in beam direction and ±1 cm perpendicular to the beam. The momentum vector of the p 0 is pointing opposite to the direction of the generated neutron. The momentum vectors of the two photons are isotropically distributed in the rest frame of the p 0 and then transformed to the laboratory system. The detectors with tracks of a good event are shown in Fig. 6.11.
The simulation of the electromagnetic shower in the CsI-array and the properties of the individual crystals (number of photoelectrons per MeV and light collection non-uniformity) is done as described in section 5.8. For good events, the number of the neutron counter, the 26 ADC-values of the CsI-crystals and the 12 ADC-values of tag counters are written into a file. The simulated data are then analyzed with the program used for analysis of the experimental data, omitting temperature correction, pedestal subtraction and gain matching. The experimental data were taken with a trigger threshold of 5 MeV; the same trigger threshold was used for the simulated data. The veto condition in the simulation required none of the outer-crystal energy to exceed 2 MeV.
Nt = 160±13 (6.19)
events below a cut value of 55.5 MeV is found. The total number of events is
N = 2412±49, (6.20)
which leads to a low energy tail Rt of 70 MeV photons below a cut value of 55.5 MeV of
(6.6±0.6)·10-2
. (6.21)
This agrees with the result obtained from the simulated energy spectra of
(7.0±0.5)·10-2
. (6.22)
The quoted errors are pure statistical. The resulting values for Rt for cuts between 50.5 MeV and 60.5 MeV are displayed in the insert of Fig. 6.12.
The presented results should not be interpreted as values for corrections of the p b branching ratio. These corrections need to be calculated when the final crystal configuration in the detector is known and the properties of the individual crystals are determined. However, the results obtained are reflecting the present understanding of the crystal performance. The measured properties of the individual crystals, the experimental response to photons and positrons and the simulation produce consistent results.
However, it should be mentioned, that the large difference in the crystal response to a 70 MeV positron beam and to 70 MeV photons from the charge exchange reaction is not completely resolved. In this work, the main difference is attributed to the optical non-uniformity of the crystals, which needs further clarification. Other effects may be present: The positron data were taken within typically 2 hours whereas the 70 MeV photon data were acquired over several days. In the latter effects like gain matching, gain drifts (temperature and rate induced) were not controlled and monitored online but only corrected in the offline analysis, based on marginal statistics.
