Automatic Importance Biasing in FLUKA: Application Scenarios

European Organization for Nuclear Research (CERN), Geneva, Switzerland

^* e-mail: francisco.ogallar.ruiz@cern.ch

Published online: 15 October 2024

Abstract

In the context of Monte Carlo simulation, importance biasing is a well-established variance reduction technique. It is typically employed to compensate for particle attenuation and is commonly used in FLUKA simulations. However, implementing importance biasing in a simulation can be a rather time-consuming exercise, especially for complex geometries as it requires the segmentation of solids. This document demonstrates the effectiveness of a newly developed tool designed to automate the implementation of problem specific importance biasing schemes in FLUKA simulations, with a focus on user-friendliness. By minimizing user intervention, the tool significantly reduces the amount of time required to implement the biasing scheme. Additionally, it reduces the risk of potential errors by eliminating the need of geometry modifications and manual attribution of importances. The document briefly outlines the tool and focuses on several examples to illustrate its usage and scope. It is important to mention that it does not aim for fully replacing the traditional region-based FLUKA importance biasing. In the hands of an expert the latter might indeed achieve faster convergence, depending on the problem. Yet, the automated solution’s overall benefits become apparent when considering the time investment required for manual implementation and fine-tuning of importance factors.