Geophysical flows, such as rock avalanches, debris flows and floods, are among the most destructive natural hazards, causing great fatalities and serious damage to key infrastructure worldwide every year. Commonly used mitigation measures include rigid and flexible resisting structures. When subjected to the impact of a geophysical flow, the resisting structure may interact complicatedly with the incoming flow and cause flow deceleration and redirection, force transmission, and energy dissipation. These processes are challenging for physical and numerical analysis, but are critical for understanding their physical mechanisms. This thesis presents a unified numerical framework based on coupled Computational Fluid Dynamics and Discrete Element Method (CFD/DEM) to examine key issu...[ Read more ]

Geophysical flows, such as rock avalanches, debris flows and floods, are among the most destructive natural hazards, causing great fatalities and serious damage to key infrastructure worldwide every year. Commonly used mitigation measures include rigid and flexible resisting structures. When subjected to the impact of a geophysical flow, the resisting structure may interact complicatedly with the incoming flow and cause flow deceleration and redirection, force transmission, and energy dissipation. These processes are challenging for physical and numerical analysis, but are critical for understanding their physical mechanisms. This thesis presents a unified numerical framework based on coupled Computational Fluid Dynamics and Discrete Element Method (CFD/DEM) to examine key issues on the interactions between a spectrum of geophysical flows and rigid and flexible resisting structures, in attempting to improve the fundamental understanding of the impact mechanisms and provide useful references for practical designs. Major new findings of this thesis are summarized as follows:

(i) When a multiphase geophysical flow hits a wall-like rigid obstacle, a metastable jammed zone called the Hydrodynamic Dead Zone (HDZ) may form whose characteristics are intriguingly complex but bear crucial scientific and engineering significance. The particle-liquid mixture impacting onto a rigid obstacle is modeled to quantitatively examine the progressive formation of HDZ. A modified granular temperature T_s accounting for the influences of inherent polydispersity and both translational and rotational motions of granular particles in geophysical flows is proposed to identify the distinctive features in the zonation of the HDZ. A conceptual source-sink model is further established to interpret the energy dissipation process in the unjammed-jammed transition in the HDZ, where the T_s serves as a function of either time or distance. Structural signatures of the time-dependent HDZ in the transition of the debris-structure interactions during the whole impact process and three key regimes, impact-up, roll-up, and heap-up in the flowing layer, are identified. These findings help us gain insights into the unjammed-jammed transitions and the underlying physics of the HDZ.

(ii) Two mechanisms, runup and pile-up, govern the impact behavior of geophysical flows against flexible barriers. Understanding of the two mechanisms helps for analyzing the debris-structure interactions, deriving adaptive analytical models, and guiding engineering designs. A total of 70 numerical tests of geophysical flows with varying fluid rheologies, pre-impact velocities (v₀= 1 m/s ~ 16 m/s), solid volume concentrations (ε_s= 0.1 ~ 1) and Froude-numbers (N_Fr= 0.43 ~ 6.89) have been conducted to examine the dynamics of flow-barrier interactions, regime quantification, impact load distribution and momentum reduction. Transitions between runup and pile-up mechanisms have been systematically identified by the peak-static load ratio δ and the momentum reduction ratio ζ for seven typical geophysical flows against a flexible barrier. Two transition diagrams are plotted to demonstrate the effects of ε_s, fluid rheology and N_Fr on the transition. A transit regime (0.5 < N_Fr <3.5, 0.45 < ζ < 0.55) is further defined by comparing the difference between the two criteria.

(iii)A ring net flexible barrier system consisting of ring elements, brake elements and supporting cables has been modeled considering the sliding behavior between interlocking rings and the ring-cable. Systematic tests of granular debris flow with varying flow/barrier height ratios h₀ / H_B and Froude numbers N_Fr have been conducted to examine their effects on the dynamics of the flow-barrier interactions, the evolutions of the load distribution and transmission, and the deflections and load-attenuation mechanism of the barrier. Numerical results have reasonably captured both experimental and field observations on the key debris-structure interactions, the build-up of the HDZ, the unique deformed-shape of the double-bulges of a barrier, and the rectangular deformed patterns of rings. The ratio h₀ / H_B is found strongly and positively correlated with the maximum values of the tensile forces sustained in cables, impact loads acting on the barrier, and the inflection and peak values of barrier deflections. The effect of the ratio h₀ / H_B appears to be more dominant than the N_Fr of flows under certain circumstances. A diagram is presented to show the relationships between the barrier deflection, impact load, and equivalent barrier stiffness, reveling the load-attenuation mechanism of the barrier.

(iv) To validate and calibrate key models and parameters for the modeling of the debris flow over a full-scale terrain with Erodible Beds (EBs), five groups of testing simulations, slurry, dry, mixture, dry with EBs and mixture with EBs, have been performed. Based on the back-analysis of the Yu Tung Road (YTR) debris flow event, the numerical results have been presented and compared to previous numerical studies and historical data in terms of the evolutions of the flow shapes and mobilities. The complicated four-way interactions among the terrain, erodible beds, slurry and boulders in a debris flow have been physically handled by the presented method. A criterion for estimating the reasonable ranges of key bond parameters has been proposed on the incipient motions of erodible particles. It is found that pure DEM simulations may experience difficulties in reproducing key features of debris flow over natural terrains while the slurry and mixture tests after calibrations can well match the historical data. Consideration of entrainment may help enhance the mobility of a mixture flow but impede the kinetics of a dry flow.