The expected photon flux incident on the shields from the calibration sources was calculated by taking into account the distance of the source, air absorption and the source physical properties reported in Table 2.1, i.e. activity, half life, lines branch ratios. By multiplying these fluxes for the slab geometrical area (1136 cm2) we obtain the expected counts/s at each energy. The on-axis detection efficiency at different energies in the GRBM and AC bands was estimated by dividing the on-axis counts measured in these bands by the incident photon rates. To evaluate the efficiency of each LS at energies lower than 60 keV we used the fluorescence lines provided by Ce139 (energies between 33 and 39 keV) and Sn113 (energies between 24 and 28 keV).
To derive the detection efficiency at energies > 200 keV it was mandatory, as discussed in previous section, to take into account the back-scattering component superimposed to the Compton spectrum associated with the given line.
As mentioned above, this component may be due to source photons back-scattered by the walls of the room in which the calibration tests where performed or by the satellite structures. It is very important to separate the contribution coming from the environment, that is only present in the calibration site, from that coming from the satellite, that there will be also in orbit, in order to achieve a reliable estimation of the expected in-flight efficiency at energies greater than about 200 keV. A good solution of the problem will be given by Monte Carlo (MC) simulation of the satellite/calibration room system, now in progress ([Rapisarda et al. 1997]), but so far we can also arrive to a first order estimate through a calibration analysis. The comparison of the GRBM detectors spectra with the NaI spectra and the ENEA calibrations sources spectra shows that most of the back-scattering contribution comes from the walls of the room. Indeed, the ratio between the photo-peak photons and the Compton photons is about 1 for the spectra measured by GRBM and those measured with the NaI detector in the place of the satellite (Fig. 2.28) . In addition, the source spectrum measured with the Germanium detector shows very low back-scattering component (Fig. 2.29). Thus, we corrected the >200 keV measured on-axis count rates by subtracting the counts due to back-scattered photons, modeling the high energy lines spectra Compton continuum in the way described in section 2.5.5.
In figures 3.5 and 3.6 we report the on-axis efficiencies in the GRBM and AC bands for LS1 and LS3, after the correction for the back-scattering. The efficiency at energies greater than 662 keV was obtained for extrapolation taking into account the CsI Compton cross sections.
 |
 |