The determination of a representative particle impacting velocity is an important component in calculating solid particle erosion inside pipe geometry. Currently, most commercial computational fluid dynamics (CFD) codes allow the user to calculate particle trajectories using a Lagrangian approach. Additionally, the CFD codes calculate particle impact velocities with the pipe walls. However, these commercial CFD codes normally use a wall function to simulate the turbulent velocity field in the near-wall region. This wall-function velocity field near the wall can affect the small particle motion in the near-wall region. Furthermore, the CFD codes assume that particles have zero volume when particle impact information is being calculated. In this investigation, particle motions that are simulated using a commercially available CFD code are examined in the near-wall region. Calculated solid particle erosion patterns are compared with experimental data to investigate the accuracy of the models that are being used to calculate particle impacting velocities. While not considered in particle tracking routines in most CFD codes, the turbulent velocity profile in the near-wall region is taken into account in this investigation, and the effect on particle impact velocity is investigated. The simulation results show that the particle impact velocity is affected significantly when near-wall velocity profile is implemented. In addition, the effects of particle size are investigated in the near-wall region of a turbulent flow in a 90 deg sharp bend. A CFD code is modified to account for particle size effects in the near-wall region before and after the particle impact. It is found from the simulations that accounting for the rebound at the particle radius helps avoid nonphysical impacts and reduces the number of impacts by more than one order-of-magnitude for small particles (25 μm) due to turbulent velocity fluctuations. For large particles (256 μm), however, nonphysical impacts are not observed in the simulations. Solid particle erosion is predicted before and after introducing these modifications, and the results are compared with experimental data. It is shown that the near-wall modification and turbulent particle interactions significantly affect the simulation results. Modifications can significantly improve the current CFD-based solid particle erosion modeling.