Analysis of branching zero-variance Monte Carlo games for radiation shielding problems

Inside the framework of the Monte Carlo variance reduction techniques lie the so called ”zero variance” techniques which, using the solution of the adjoint equation, can give estimates of the sought physical quantities with no error. This techniques are of considerable relevance when the event to be estimated is rare, like in most of the radiation shielding problems. Among this family of methods it is possible to distinguish the ”branchless” and ”branching” methods looking at how the multiplicative phenomena are treated. This work will analyze a series of zero variance branching methods and will compare them with other methods in order to assess their robustness and efficiency.