Numerical methods are increasingly being used to study granular flow phenomena, given the difficulty in extracting data from physical studies. Two common methods are the discrete element method (DEM), and a two fluid computational fluid dynamics (CFD) method. A companion paper (Sandlin and Abdel-Khalik, 2018) compared a DEM model and a CFD model of a granular flow through horizontal wire mesh screens, with experimental data. The aim of this study is to assess the sensitivity of both models to various input parameters. The results of this investigation will guide future modelers of particle heating receivers and other flowing particulate systems in selecting the appropriate modeling options and parameters to enhance the models ability to predict the actual particulate flow characteristics. For the DEM model, it was found that the granular material properties, especially the values for normal and rolling friction, had the largest impact on simulation results. For the CFD model, it was found that the constitutive relationships for frictional pressure, viscosity, and the treatment of physical boundary conditions had the largest impact on simulation outcomes. In addition, both numerical models exhibit a non-monotonic relationship between mass flux and the granular coefficient of restitution, and show reduced mass flux when using a simulation domain with offset wire meshes. The influence of other material properties and sub-modeling options is less pronounced. Methods of obtaining appropriate material properties and sub-modeling options are discussed.