A numerical study on the effects of particle size distribution on run-out distance of granular flow

The discrete element method is a powerful numerical tool widely employed in granular flow simulations. To reduce computational costs, either monodisperse granular models or models with a limited number of particle sizes are used to estimate the risk of sediment-related disasters. This highlights the need for a thorough understanding on the effects of particle size distribution on the accuracy of granular simulations. In this study, a series of granular flow simulations was conducted with different particle size distributions. The results clearly indicate that the run-out distance strongly depends on the number of particle sizes and is underestimated in the conditions employed by the monodisperse granular model. The longest run-out distance was observed in case that the bidisperse granular model was employed, whereas run-out distance tends to a constant value when polydisperse granular model was employed. Because particle size segregation was observed during the flow, a series of vibration simulations was also conducted to estimate the degree of the size segregation. It was shown that each granular model has an inherent degree of size segregation. From the results of the granular flow simulations and vibration simulations, it was confirmed that the inherent degree of the size segregation clearly affects the run-out distance.