A Practical Look at Monte Carlo Variance Reduction Methods in Radiation Shielding

With the advent of inexpensive computing power over the past two decades, applications of Monte Carlo radiation transport techniques have proliferated dramatically. At Los Alamos, the Monte Carlo codes MCNP5 [1] and MCNPX [2] are used routinely on personal computer platforms for radiation shielding analysis and dosimetry calculations. These codes are attractive for such applications because of their ability to accommodate complex 3-D geometries, inclusion of flexible physics models that provide coupled electron-photon and neutron-photon transport, and the availability of extensive continuous-energy cross section libraries derived from evaluated nuclear data files. It should be noted that these codes are general purpose in nature, and are therefore not optimized for any particular application. It is left up to the user to select appropriate variance reduction (VR) tools from a rich palette of such tools bundled with these codes. An excellent overview of classic MCNP VR techniques was given by Booth [3]. Recent developments such as the weight window generator and automated VR using deterministically generated importance functions have been described by several authors [4,5,6,7,8,9]. Both the AMCNP patch to the MCNP code [7] and the stand-alone code ADVANTAG [6] appear to greatly improve the computational efficiency for deep penetration problems. The efficiency of a Monte Carlo simulation may be quantified using the Figure of Merit (FOM), which is defined by the MCNP code developers according to Equation 1, where R is the tally relative error and T is the computing time.