Space weathering can be defined as the gradual changes experienced by the surfaces of airless planetary bodies due to exposure to the vacuum of space, radiation, and micrometeorite bombardment. Characteristic visible and near infrared (VNIR) spectral changes due to space weathering include a decrease in albedo and a general “reddening” of spectra (increasing reflectance with increasing wavelength). Apollo-returned lunar soils contain grains with amorphous rims with submicroscopic metallic iron (smFe0) particles dispersed throughout. Optically absorbing particles of this scale (10s of nm) have disproportionately large optical effects compared to their size. The size of the smFe0 particles determines the amount of reddening observed -- smaller particles are associated with redder spectra in the VNIR than larger particles. A transition occurs at a particle size of about 30-50 nm, above which the spectra darken without reddening. Previous modeling work has failed to robustly reproduce this transition at the observed iron particle size, instead requiring larger iron particles. This work looks at modeled space weathered particles using the Multiple Sphere T-Matrix Model (MSTM). These simulated particles are designed to be more geometrically realistic than previous efforts involving single spheres or coated single spheres. We construct particles as spheres of olivine (referred to as the host) with a thin, even rim of amorphous silica glass surrounding it. Within the rim but outside the host grain, we randomly populate a large number of monodisperse smFe0 particles. We then construct particles with different smFe0 sizes and abundances. The scattering properties of these particles are calculated using MSTM for comparison to existing remote sensing and laboratory datasets. For comparison, we show several cases that have been simplified to a coated-sphere using effective medium approximations (EMA). Some previous studies have relied on the use of EMA to simplify these geometrically complex and heterogeneous particles to single homogeneous spheres, or homogeneous spheres with a homogeneous coating that can be more easily modeled. We show that the use of EMAs results in the loss significant information about the scattering properties of these particles and should be avoided. This result agrees with work done by Mishchenko et al. (2016) concerning the applicability of EMAs to heterogeneous atmospheric aerosol particles.