VALIDATION OF THE MCNP MONTE CARLO CODE FOR GERMANIUM DETECTOR GAMMA EFFICIENCY CALIBRATIONS

Canberra designs a variety of instruments that are required to accurately measure the radioactivity content of large andor complex sources. The Monte Carlo code MCNF was evaluated to determine if it would be suitable to predict the performance of these instruments. A series of experiments was designed, starting with simple source- &sector geometries and becoming increasingly more complex. For each geometry the full energy peak efficiency was computed with MCNP, and also measured experimentally. Multiple energies (nominally 100-1500 keV) were used for each experiment. The lessons learned from each experiment are described, and were suitably incorporated into later experiments. Computer run time cm be quite long, and therefore a geometrical biasing scheme was developed to make these computations more practical. The speed advantage can be a factor of 10 or greater. It is concluded that thk technique is very powerful and accurate when properly applied. For the geometries tested, efficiency crdibrations arc accurately predicted to within 5% for simple geometries, and up to 15% for complex geometries at low energies.