Numerical Investigation of Granular Flow Dynamics Over Complex Topography: Quantifying the Efficacy of In-Path Engineered Structures to Control Granular Flow Mobility and Impact Velocity
AuthorJensen, Sara Eileen
AdvisorMcCoy, Scott W
Geological Sciences and Engineering
StatisticsView Usage Statistics
Rapid mass movement events pose a significant threat to people, infrastructure and property. These processes initiate suddenly, can reach speeds in excess of 10 m/s, and can travel distances greater than a kilometer beyond the steep terrain where they initiated. Rapid mass movements are especially dangerous when they initiate in close proximity to populated regions where they can cause a loss of life, damage property, and decrease land production. Due to the significant threats mass movements pose to populated areas, and the difficulties associated with source area stabilization, it is important to understand their mechanics such that accurate assessments of the hazards they pose to downslope communities can be made and effectively mitigated.Of key importance are accurate predictions of flow characteristics, such as runout distance and impact velocity, as these properties set the extent and severity of potential damage. Traditionally, empirical methods are used to relate path-averaged properties, such as total drop height or average slope, to these important flow runout characteristics. However, it is well known that work done by frictional processes that lead to flow resistance are dependent on the path taken and hence should depend on flow path shape. The following study investigates how the inclusion of flow path shape, as opposed to using path-average properties, effect the accuracy of flow runout predictions and to what degree the inclusion of engineered structures placed along the channel can decrease hazardous characteristics of the flow (e.g., runout distance, flow front impact velocity, and total system kinetic energy). A suite of numerical experiments, using the Discrete Element Method (DEM), were conducted in which dry gravity-driven granular flows were allowed to flow down flumes with check dams of varying heights and locations along the flow path. Confidence in the flow dynamics stimulated by the model was gained by first validating the model against a benchtop granular avalanche experiment conducted by Iverson et al. (2004). No model tuning was required because DEM input parameters were constrained using independently tested granular properties reported by Iverson. The validation study concluded that the DEM can accurately simulate the initiation of flow from a static state, the rapid granular flow down complex three-dimensional topography, and the resulting deposition patterns at the base of the slope. The addition of a sensitivity analysis that investigated particle stiffness, size, and shape highlighted the DEM’s dependency to particle diameter; where, smaller particle diameter simulations decreased particle runout distances. The DEM also had a slight dependency on particle shape; however, the sensitivity was notably higher for static particle conditions compared to dynamic particle conditions; however, the validation investigation demonstrated that a majority of the bulk dynamic properties of the flow were adequately represented by means of a constant directional torque rolling friction parameter. Lastly, DEM simulations were found to be insensitive to particle stiffness for Young’s Modulus values within the rigidity definition outlined by da Cruz et al. (2005). A reduction in stiffness values significantly reduced model runtimes allowing for subsequent investigations to increase model complexities. The results of the numerical experiments investigating the influence of check dam height and location along the flow path demonstrated that flow kinetic energy and impact velocity decreased with increasing check dam height and were sensitive to the proximity of the dam in relation to the base of the slope. Events that traversed a linear flume compared to those with a check dam, at full sediment capacity, encountered up to a 40% reduction of peak kinetic energy. Although runout distance had clear trends with increasing dam height and dam location, observed changes were comparatively small. These numerical experiments highlighted the importance of including the specifics of flow path shape, as opposed to just using path-average properties, when predicting runout characteristics. They also provided some first-order guidance for engineers if the situation allows for flexibility in check dam height or placement along the flow path. With increased knowledge of the capabilities that engineered structures have to decrease the hazardous nature of a flow, engineers can more successfully repurpose their design.