Benchmarking the Particle Background in the LHC Experiments

The experiments for the Large Hadron Collider LHC at CERN have to work for 15 years in the presence of a very high particle background of photons in the energy range from 100\,keV to 10\,MeV and neutrons in the range from thermal energies ($\approx 0.025\,$eV) to 20\,MeV. \\ The background is so high that it becomes a major design criterion for the ATLAS ex\-peri\-ment, a general purpose experiment at LHC that will be operational in the year 2005. The exact level of this background is poorly known. At present an uncertainty factor of five has to be assumed to which the limited knowledge of the shower processes in the absorber material and the ensueing neutron and photon production is estimated to contribute with a factor 2.5. \\ So far, the background has been assessed only through extensive Monte Carlo evaluation with the particle transport code FLUKA. The lack of relevant measurements, which were not done up to now, are to a large extent responsible for this uncertainty. Hence it is essential to benchmark the background predictions with measurements in order to reduce the uncertainties resulting from the shower processes. This work describes in detail the benchmarking measurements and analysis of these backgrounds in an experimental arrangement that approaches rather closely the layout and shielding in the ATLAS detector. The absolute yield and energy of the particles ema\-nating from the final stages of the hadronic shower were measured using a Bi$_4$Ge$_3$O$_{12}$ detector. \\ In this study particular care was taken to guard against spurious effects, which could mask the measurements of the photon background. Typically we expect to measure a photon per $10^4$ incident hadrons which is equivalent to a reduction factor in energy of $\approx 10^8$. At first, calibration measurements with well known radioactive sources were carried out in order to evaluate the response to photons and neutrons of the used detector. The photon results show excellent agreement with the simulations, while the neutrons show some FLUKA specific discrepancies that are however well understood. The actual benchmarking task comprised measurements with different beam intensities and momenta, different positions and absorber thicknesses in order to reduce systematic effects and assess residual activities from other sources. \\ The careful analysis of the measurements including a detailed evaluation of the systematic uncertainties provides a good understanding of all effects due to residual activities, dead-time corrections and other rate effects of the set-up. Comparing the measurements with detailed FLUKA simulations shows that under all different measurement conditions the agreement is on the 20\,\% level. These studies also give answer to the nature of the particles emanating from the absorber. \\ Finally a method to obtain the measured photon rates and energies from the total measured and simulated numbers was developed. The comparison of the measurements with the FLUKA calculations can hence be used to reduce the uncertainties resulting from the shower processes, so that the background simulations can predict the ATLAS background with higher reliability.