Towards Depth-Averaged Modelling of the Decay of Granular Flows by Deposition

Abstract

Depth-averaged (DA) numerical models are often used to simulate the runout of flow-type landslides because of their practicality and computational efficiency. However, existing DA models generally overlook the effects of deposition, which governs the decay of the flow momentum and the runout distance. In this study, a new approach is proposed to model the effects of deposition. The flow-normal accelerations that drive the deposition process are simulated using a high-fidelity two-dimensional (2D) Smoothed Particle Hydrodynamics (SPH) model. Based on the 2D non-DA SPH results, a deposition model that considers the initial aspect ratio (i.e., height to length) of the debris and slope angle is developed and implemented into the DA context to provide a practical tool for engineering simulations. The new DA SPH model is evaluated against experimental results in the literature. By considering the effects of deposition, predictions of the runout and flow duration are improved by up to 38%. The new DA SPH model works particularly well for runout predictions of large initial aspect ratios of the debris, which has traditionally been the limitation of DA models. The new framework can be used to advance towards realistic deposition and entrainment modelling to improve the delineation of flow-type landslides in mountainous regions.