Fast neutron inelastic scattering is a promising non-destructive assay technique for various analytical applications. As an active neutron interrogation technique, its performance is a function of various different factors and parameters that require optimization. Monte Carlo simulation codes are indispensable for such tasks. However, the internal simulation routines implemented in such codes can rely on different physical models that can yield discrepancies in the simulation results. In this work we conduct an intercomparison of PHITS and Geant4 codes performance in application to fast neutron inelastic scattering simulations. The goal of this paper is twofold. First, we explain the differences in code configuration with respect to gamma and neutron transport, as well as internal simulation routines. Second, we conduct a performance assessment of the two codes using two different measurement configurations. One configuration consisted of a source of gamma rays in a broad energy range (100–9000 keV) and a CeBr3 detector. The other configuration consisted of a monoenergetic 2.5 MeV fast neutron source, Fe, Nd, Dy, B targets and a CeBr3 detector. Selected simulation configurations were chosen with a goal to compare the performance differences in neutron energy distribution, produced prompt gamma rays and energy deposition in CeBr3 detector between the two codes. Results of our study reveal a good coherence of both codes performance in the application of fast neutron inelastic scattering simulations. The simulation geometries and observed differences are described in detail.
