In the recent years, the effect of hyper-concentrated flow has become a major factor for sedimentation problems in both rivers and reservoirs. In the previous studies, the correlations between concentration and the other three factors: flow velocity, flow depth and shear stress, were established as a guideline for the design flood protection facilities. A further detailed simulation on the variation of concentration in hyper-concentrated flow, to count on the stagnation of wave front and bed evolution, is needed further for planning with the consideration of stabilization of the channels.
This study presents the result of the first year in this 3-year research project. The rheological experiment, which aims at the movement of fine particles (<2mm) in Chen-Yu-Lan River, was completed to provide for validation of numerical parameters and the correlations between concentrations and the other two factors: Bingham yield stress and viscosity. The physical model test in this study used a smaller particle size other than that used in the previous study of 2008, in order to extend the formula of the sediment carrying capacity to cover wide range of particle size taken into account as a parameter considered. On the other hand, several empirical formulas were also selected and embedded in a sub-model for mobile-bed condition to perform the analyses on the capability of the numerical models with various experimental data. The results of numerical tests show the numerical model developed is capable of simulating both the hyper-concentrated flow and flow with low concentration without manual or artificial switching. In the analysis of sediment yield in Tung-Pu-Ray creek catchment, several approaches, including land sliding and surface erosion estimate method, sediment rating curves method, river bed evolution method, debris flow formula method, and flooding model method, were employed to estimate the inflow concentration for 10 storm events. Consequently, a flowchart was established and provided as a guide for engineers in choosing appropriate method in concentration estimation.