Multiscale study of basal effects in geophysical granular flow over rough topography

Abstract

Geophysical granular flow, one of the most destructive geohazards in the world, is a massive volume of granular earth materials flowing down natural terrains. Amid increasingly extreme weather, the risk of geophysical granular flow has been elevating in Hong Kong, leading to significant casualties and economic losses. A primary objective in risk assessment is to predict the flow velocity and runout distance, during which base topography plays a critical role in governing the overall flow dynamics and deposit morphology. Historical data indicated that variations in bed materials, such as glaciers and gravel, can result in more than a twofold difference in landslides run-up. However, a general physical model to characterize the flow-bed interactions over complex basal topography remains largely unresolved. Current research on landslide mobility primarily employs the framework of traditional friction models. However, these models lack a detailed characterization of base roughness and fail to fully account for the granular nature, which exists in a transitional state between fluid and solid. Consequently, their results often fail to accurately predict flow mobility and runout or rely heavily on back-analysis, which is limited to specific cases.

These research gaps present a significant challenge, necessitating the multiscale connection of granular micro-mechanics to continuum modelling. In response, this study proposes an integrated approach encompassing laboratory, microscopic, theoretical, and continuum investigation. Following this approach, we adopt a physically relevant setup of natural geophysical flows—specifically, granular column collapse.

First, laboratory tests with systematically controlled base roughness are conducted to obtain detailed flow measurements using particle imaging velocimetry. A general characterization of base roughness, denoted as Ra, is introduced to quantify the relative size and spatial distribution of base roughness.

Second, micromechanical simulations are performed on a calibrated setup using high-performance particle dynamics simulations. This analysis introduces new insights into scaling laws influenced by base roughness, where Ra effectively scales with runout distance and collapse duration, regardless of the flowing materials or initial conditions.

Subsequently, our theoretical investigation of granular micromechanics leads to the discovery of a mathematical model linking slip velocity with base roughness Ra. This general basal slip model successfully holds in all flow regimes, allowing for the quantitative consideration of base roughness in continuum modelling.

The implementation framework for the proposed boundary model in continuum simulations is then established, incorporating granular rheology and free-surface tracking technique. This improved continuum model shows a strong agreement with our micromechanical tests, whereas other existing boundary models fail to match the flow dynamics and final runout.

Future work will focus on expanding this physic-based method into a large-scale field-scale studies. It is expected that better predictions can be achieved in regarding of the flow velocity, runout distance, which is essential for mitigation engineering and risk assessments.