A Preliminary Study on the Optimization of Source Angular Bias for Monte Carlo Simulation in a Large Radiation Generation Facility

In this study, a method for obtaining the parameter of the angular source bias was proposed for the large gamma radiation generation facility. A calculation program with the proposed method was developed for the automatic calculation of the source angular bias value from MCNP input. For the verification of the proposed method, a simple shielding problem was evaluated by MCNPX code. The results show that the radiation transport result with the proposed method give a good calculation efficiency over 6 times than that without the source angular bias. It is expected that the proposed method can contribute increasing the calculation efficiency and accuracy for the radiation shielding problem in using Monte Carlo method. In a large radiation generation facility, some restrictions of the gamma radiation shielding analysis using direct Monte Carlo simulation can be occurred due to the stochastic error of the calculation results in the large space. To increase the calculation efficiency, various variance reduction techniques are introduced for the large facility. The angular source biasing is one of the variance reduction techniques, which is that the true angular probability is biased with adjusting the source importance. However, the optimization of the angular biasing is not easy and some calculation error can be generated without the specific technique. In this study, a first collision method for the decision of the gamma source angular bias is proposed, and a radiation shielding problem in a large space was evaluated and verified with MCNP code calculations