An alternative method of biasing the collision kernel in Monte Carlo calculations

A simple alternative way of biasing the collision kernel for variance reduction is justified due to the great difficulties in implementing most techniques founding temperature in the general Monte Carlo programs available. These difficulties are related to programming necessities and the extensive data amounts usually required for the general methods. The random walk of particles is heavily driven by the probability distribution of the azimuthal scattering angle in each collision, even though it has a very simple form. In this work, a variance reduction scheme is devised to sample azimuthal scattering angle in each collision, even though it was a very simple form. In this work, a variance reduction scheme is devised to sample azimuthal scattering angles as a means to guide the scattering particle towards a preferred direction. There, the particles trajectories can be forced into a virtual pipeline drawn between the source and the point of interest. As an alternative method, it can be easily included in Monte Carlo programs and the results show a promising performance in the demonstration problem solved in this work. (author)