Debris flows carry enormous momentum as they surge downslope and could cause loss of lives and damage to infrastructure. Structural countermeasures such as rigid and flexible barriers are commonly installed in mountainous regions to intercept debris flows. The current design approaches regard the debris flow as a uniform equivalent fluid without considering the solid-fluid interaction. Estimation of the debris flow impact, i.e., impact force and run-up height, remains semi-empirical. The objectives of this research are to reveal the mechanisms of debris-barrier interaction and solid-fluid interaction in the process of debris flow impact on rigid and flexible barriers. Findings of this study help to improve the current design practice.

Two research methodologies, namely the centri...[ Read more ]

Debris flows carry enormous momentum as they surge downslope and could cause loss of lives and damage to infrastructure. Structural countermeasures such as rigid and flexible barriers are commonly installed in mountainous regions to intercept debris flows. The current design approaches regard the debris flow as a uniform equivalent fluid without considering the solid-fluid interaction. Estimation of the debris flow impact, i.e., impact force and run-up height, remains semi-empirical. The objectives of this research are to reveal the mechanisms of debris-barrier interaction and solid-fluid interaction in the process of debris flow impact on rigid and flexible barriers. Findings of this study help to improve the current design practice.

Two research methodologies, namely the centrifuge and numerical modelling, were adopted. To investigate the debris flow impact mechanisms, a broad range of geophysical flows were modelled condiering continuous variation of solid fraction, i.e., stream flow (less than 10% solid fraction), hyperconcentrated flow (debris flood), debris flow, and dry debris avalanche. Effects of boulders were studied separately by modelling the dry glass beads and dry glass bead-sand mixtures. A newly developed model flexible barrier together with a rigid barrier was used to simulate impermeable barriers with contrast in the barrier stiffness. Numerical modelling was used to back-analyse the centrifuge test results and investigate the effects of barrier stiffness by parametric study. Results from physical and numerical modelling are compared with the theoretical impact load models’ prediction. Discrepancies are found and the current impact models are revised based on the findings from this research.

It is found that with an increase of solid fraction, an enlarging dead zone is observed at the base of barriers, denoting the contribution of grain frictional stress in the transition from run-up to pile-up impact mechanism. Due to the elongated interaction duration of pure fluid and two-phase flows, the peak loads on flexible barrier are reduced by 30% to 50% of the impact on rigid barrier. The boulder impact induced impulse load on rigid barrier could reach up to 6 times of the static load. Furthermore, due to reverse segregation of boulder-sand mixture, the boulders tend to migrate to the free surface and front of the flows without much reduction in velocity. With the combination of static load and boulder impulse load, the peak load of boulder-sand mixture impact could be even higher than that of pure boulder impact.

A new impact load model adopting a triangular load distribution is proposed. A triangular load distribution highlights the contribution of static deposition in the peak impact force and the efficiency of frictional energy dissipation for dense two-phase flows like debris flow. The modelled flow types are further quantified by the friction number N_fric to reflect the degree of solid-fluid interaction. The friction number N_fric and square of Froude number Fr² form a framework to consider the effects of solid-fluid interaction rather than relying on empirical coefficients adopted in conventional hydrodynamic approaches. The Fr²-N_fric framework covers a broad spectrum of flows for impact problems.

Based on the deduced dynamic pressure coefficient α of this research, α = 2.5 seems to be enough to cover the impact scenarios. The Hertz equation with load reduction factor of 0.1 seems not sufficient to predict the impulse load by boulder impact causing the localized damage of structures. The flexible barrier is a reasonable alternative to the rigid barrier, because it can effectively reduce the impulse loads induced by both the debris and boulders.