Modelling of a simple bunker problem with Monte carlo codes TRIPOLI 4.3 and MCNPX 2.4 to test the efficiency of the biasing methods
Titre de la revue : Radiation Protection Dosimetry
Volume : 115
N° : 1-4
Pagination : 306-309
Date de publication : 01/12/2005
The Monte Carlo codes are particularly used at IRSN to simulate the particle transport in complex geometries such as multi-element detectors, voxel phantoms and irradiation facilities. These calcula-tions without any optimisation could run over several CPU days. The biasing methods of TRIPOLI 4.3(1) or MCNPX 2.4(2) appear to be very powerful but they require a careful control in order to obtain reliable results. This is why IRSN users of these codes have developed a simple model, i.e. a bunker room, in order to test in terms of CPU time and control difficulty different variance reduction meth¬ods proposed by the codes. The geometry of the model is a square room in which there is a neutron isotropic source of UO2, which is typical of the sources simulated in engineering calculations to evaluate the protection shields of the installation facilities. The ceiling, floor and walls are made of concrete. The purpose of the simulation is to calcu-late the ambient dose equivalent rate outside the room at 20 cm from a wall. The presented results obtained with the two codes are compared with respect to CPU time.