| Summary: | 博士 === 國立交通大學 === 土木工程學系 === 100 === To investigate the transport behavior of hyper-concentrated flow and as a result to provide as a reference for engineering planning of practical cases, a hydrostatic three-dimensional model for hyper-concentrated flow and sediment transport in alluvial channels was developed in this study. By following the vertical and horizontal splitting concept (VHS), the shallow water flow governing equations were split into two parts including the depth-averaged two-dimensional equations and velocity defect equation in vertical direction. The former one was transformed into orthogonal curvilinear coordinate system; the latter one was derived as the form of sigma coordinate. Incorporated with continuity equation, the three-dimensional velocity field can therefore be solved. Sediment transport governing equations include three-dimensional mass transport equation, active-layer continuity equation, and bed-layer continuity equation. The effects of hyper-concentrated flow were treated as follows: a quadratic rheological relation was used to reflect the characteristics of non-Newtonian fluid; a state function was used to reflect the influence of concentration to density; the empirical suspended- and bed-load formulae with hyper-concentrated flow effect were used for the sediment transport computation.
Sensitivity analysis was performed first to identify the weighting of parameters to be calibrated in the model. The influence extent induced by the parameters on water flow, suspended sediment concentration, and bed evolution, thereafter were examined and justified. To further investigate the characteristics of hyper-concentrated flow and sediment transport, several sets of experimental cases collected from the literatures were simulated. The case of dam-break wave propagation of non-Newtonian fluid was simulated to demonstrate the limit of traveling distance of hyper-concentrated flow. A channel bend flow experiment with 40% volume concentration of sediment was studied to investigate the effect of hyper-concentrated flow on super-elevation of water surface. A criterion of ratio of falling velocity and shear velocity was numerically examined and justified based on experimental data as a baseline for choosing the proper distribution type of turbulent diffusivity for simulation of suspended-load movement. At last, simple and concise regression relations for the increments of water depth, velocity, bed shear stress, and bed change caused by the hyper-concentrated flow were established for application of engineering planning and design.
|
|---|
