TWO-PHASE DEBRIS-FLOW COMPUTATIONS THAT INCLUDE THE EVOLUTION OF DILATANCY AND PORE-FLUID PRESSURE

Pore-fluid pressure plays a crucial role in debris flows because it counteracts normal stresses at grain contacts and thereby reduces intergranular friction and enhances bulk flow mobility. Two-phase debris-flow models typically assume that pore-fluid pressure has both a hydrostatic component and a non-hydrostatic component that is established by initial conditions and dissipated diffusively in response to debris compaction driven by gravity. These models lack a key ingredient, however: explicit evolution of solid and fluid volume fractions coupled to changes in flow dynamics. This evolution is particularly important during the initial stages of debris-flow motion, when it is responsible for pore-pressure feedbacks that influence the balance of forces governing downslope acceleration. As a result of these feedbacks, a water-laden sediment mass can either creep stably or mobilize into a high-speed flow.

Here we summarize the rationale and predictions of a new, depth-averaged debris-flow model that accounts for coupled evolution of flow dynamics, solid and fluid volume fractions, and pore-fluid pressure by combining approaches previously used to model landslides, debris flows, submarine granular avalanches, and other dense granular flows. The model's structure is also consistent with a long-established tenet of critical-state soil mechanics: solid and fluid volume fractions evolve toward values that are equilibrated to the ambient state of effective stress and deformation. Dilatancy, pore-fluid pressure, and frictional resistance evolve as a consequence.

To emphasize physical concepts and minimize mathematical complexity, we focus on depth-averaged, one-dimensional motion of a two-dimensional debris flow descending a rigid, uniformly inclined, impermeable slope. Using finite-volume numerical methods well-suited for solving hyperbolic problems, we compare computational predictions of the behavior of such a flow to data from large-scale at experiments at the USGS debris-flow flume. Model predictions exhibit rapid evolution of pore-fluid pressure coupled to contraction (negative dilation) of loose debris during the first few seconds of motion, leading to positive feedback that enhances flow acceleration. At later times, motion is stabilized by relatively steady pore pressures that eventually decay to hydrostatic values.