Jennifer G. Duan

The applications of the enhanced CCHE2D model to study the alluvial channel migration processes

This paper is to report a newly developed numerical-empirical model, the Enhanced CCHE2D (EnCCHE2D), and its application to simulating the alluvial channel migration phenomena. EnCCHE2D model is capable of predicting quasi-three-dimensional (3D) flow field and shear stress distribution on the bed, because a set of empirical functions of 3D flow characteristics formulated by results of a 3D model, CCHE3D, was integrated with CCHE2D, a depth-averaged hydrodynamic model, the predecessor of EnCCHE2D. The processes of sediment transport and meander migration were predicted based on these quasi-3D flow solutions. The advance or retreat of bank is calculated by considering not only the hydraulic erosion of bank surface and toe, but also the mass balance of sediment flux in the near-bank zone. As a result, the simulation of bank erosion, bar/pool formation and shifting, bank advance and retreat, channel widening and migration and meander evolution phenomena agree well with the available measurements of physical experiments.

Mean Flow and Turbulence around a Laboratory Spur Dike

The three-dimensional turbulent flow field around a spur dike in a plane fixed-bed laboratory open channel was studied experimentally using a microacoustic Doppler velocimeter. Mean and turbulence characteristics in all three spatial directions were evaluated at upstream and downstream cross sections near the dike. Results showed that the primary flow separated in both lateral and vertical directions. Two counter-rotating flow circulations, consisting of the lateral and vertical velocity components, originated at the dike section. Downstream of the dike, the circulation in the flow-separation zone is stronger than the one in the contracted primary flow zone. The maximum bed-shear stresses estimated using Reynolds stresses is about three times as large as the mean bed-shear stress of incoming flow.

Turbulent burst around experimental spur dike

Three-dimensional turbulent flow field was measured around an experimental spur dike by using a micro-acoustic Doppler velocimeter. Time fractions of turbulent burst events including outward interaction, ejection, inward interaction, and sweep were analyzed in each quadrant at the neighbor of the dike before and after the formation of scour hole. Over 80% of burst events near the bed have lower-order of magnitudes for both flat and scoured bed surfaces. Ejections and sweeps are prevalent before the local scour was initiated, and then outward interactions are dominant after the scour hole was formed. Conditional Reynolds stresses and high-order moments of turbulent velocities were analyzed along the thalweg. The magnitudes of u′–w′ pair were much larger than those of v′–w′ pair in the scour zone. Among four burst events, ejections and sweeps are the higher order events contributing to the Reynolds stresses. Since sediment entrainment and transport are closely associated with turbulent bursts near bed, the development of scour hole greatly depends on the higher order event near the bed.

Numerical Simulation of Unsteady Hyperconcentrated Sediment-Laden Flow in the Yellow River

The major obstacles to simulating flood flow in the Yellow River are its high sediment concentration, complex compound cross section, and rapid change in channel planform. This paper presents an improved one-dimensional numerical model that takes into account the effect of sediment concentration and bed change on mass and momentum conservation of flood flow in the Yellow River. The model is calibrated and then validated by simulating three individual flood events. Results show that an increase in sediment concentration leads to a reduction in flood wave celerity and peak discharge. The generalized likelihood uncertainty estimation (GLUE) method is used to evaluate the uncertainty of modeling results. A sensitivity index, analogous to the Nash-Sutcliffe efficiency factor, is adopted to quantify the sensitivity of calibration parameters. The modeling results are sensitive to the choice of Mannings roughness coefficient and the empirical recovery coefficient for suspended sediment transport at reaches of transitional channel planform. DOI: 10.1061/(ASCE)HY.1943-7900.0000599

Physically based simulation of dam breach development for Tangjiashan Quake Dam, China

Physically based modeling approach has been widely developed in recent years for simulation of dam failure process resulting from overtopping flow. Due to the lack of field data, there exist few applications to natural quake dams with complex erosion mechanisms. This paper presents a physically based simulation of the failure process of the Tangjiashan Quake Dam formed as a result of the “May 12, 2008” Wenchuan earthquake in China. The one-dimensional model adopted features as cost saving but enables capturing the main characteristics of the failure process, where selective sediment transport and gravitational slope collapse are accounted for. The simulated flow hydrograph and breach progression process are generally in good agreement with the observed data. Unsteadiness and non-uniformity are found to be substantial characteristics of breach progression during the failure process of natural quake dams. Sensitivity analysis showed that the Manning resistance coefficient and the erodibility coefficient in Osman and Thorne’s (J Hydraul Eng 114(2):134–150, 1988) model significantly influences the flow peak discharge but has less influence on its occurrence time, while the velocity lag coefficient associated with bed-load transport may affect the two breaching parameters substantially.

Application of surface-based bed load transport equations to a desert gravel-bed stream

Bed load measurements during low flow in a desert gravel-bed stream, LasVegasWash, were used to evaluate fractional bed load transport equations. Since the stream is an urban perennial channel that conveys primarily treated waste effluent, bed load transport is limited by the supply of transportable bed material. The Parker, Wilcock and Crowe, and Wu et al. surface-based fractional bed load transport equations were evaluated against the field data. The ratio between the calculated and measured bed load transport rate (C/O) as well as the data correlation coefficient with respect to each formula were calculated to judge the applicability of these equations to this field data set. Comparisons showed that both the Parker and Wu et al. equations satisfactorily predicted the transport rate, while theWilcock and Crowe equation has a larger scatter. The apparent success of Parker and Wu et al. equations can be attributed to two factors: a hiding function based on physical mechanism of the particle hiding and exposure, and the empirical coefficients calibrated using field data sources. The study results indicate that these two fractional bed load transport equations derived for perennial streams in wet regions are applicable to the field data collected at this arid region channel

Two-Dimensional Hydrodynamic Model for Surface-Flow Routing

AbstractA new surface flow routing algorithm based on numerical solutions of shallow water equations and the kinematic wave approximation (SWE-KWA) is proposed in this paper. The shallow water equations are discretized by the first-order Godunov-type finite-volume method. The stability analysis showed that the friction source term increased exponentially as flow depth became very small. This breaks the balance between the friction and the slope source terms. An approximate solution to the kinematic wave equation is introduced to restore this balance. This kinematic wave approximation makes it possible to apply the shallow water equations to both overland and channel flows. Test results show that this algorithm is accurate, robust, and stable for both very shallow overland and concentrated channel flows. The minimum allowable flow depth used in the tests is 10−10  m, two orders of magnitude smaller than the common rainfall excess rate (10−5 to 10−8  m/s). Because this algorithm applies to both the overland...

Celerity and Amplification of Supercritical Surface Waves

The amplification of supercritical waves in steep channels is examined analytically using a one-dimensional dynamic solution of the Saint-Venant equations. Existing methods were modified to describe the amplification of surface waves over a normalized channel length rather than over a single wavelength. The results are strikingly different, and a generalized graph shows that short waves amplify the most over a fixed channel length. The maximum amplification parameter over a normalized channel length is 0.53 when F=3.44 . Applications to the flood drainage channel F1 in Las Vegas indicate that the amplitude of waves shorter than 100 m would increase by 65% over a channel length of 543 m. These theoretical results await field verification. Supercritical waves could be dampened by increasing channel roughness to reduce the Froude number below 1.5.

Simulation of Unsteady Flow and Soil Erosion in Irrigation Furrows

AbstractThis study developed a one-dimensional numerical model for the simulation of unsteady flow and the resultant soil erosion in irrigation furrows. The model solves a modified version of the Saint-Venant equations that consider the loss of mass and momentum attributable to infiltration and sediment transport. The transport rate of fine sediment was predicted with a modified Laursen formula that treats the tractive shear stress as a function of both Reynolds number and the particle size. The modified Laursen formula was verified by using the erosion data measured in the field and in a laboratory flume. The model accurately predicted flow advance times and outflow hydrographs in comparison with data measured in irrigation furrows at Kimberly, Idaho. Sediment discharge predictions were less accurate.

Two-dimensional depth-averaged finite volume model for unsteady turbulent flow

A two-dimensional (2D) depth-averaged model is developed for simulating unsteady turbulent shallow-water flows with dry–wet fronts (e.g. dam-break flow). The model is based on 2D depth-averaged Reynolds-averaged Navier stokes equations coupled with the k−ϵ turbulence model. The high-resolution MUSCL scheme (monotone upstream-centred schemes for conservation laws) is implemented to minimize numerical diffusions. A novel augmented Harten–Lax–van Leer-contact Riemann solver is used to solve the governing equations simultaneously. A body-fitted mesh is generated by using the Cartesian cut-cell method to accommodate irregular boundaries. The model is tested against two laboratory experiments to examine whether or not turbulence model is essential for simulating unsteady turbulent flow. The results show that the addition of the k−ϵ turbulence model significantly improves the modelling results at places of strong turbulence activities. To further improve the results, a more accurate turbulence model for unsteady flow and dispersion terms in the momentum equations is needed.