Journal of Hydraulic Engineering

Parallel solutions of linear systems arising from velocity–pressure coupling in implicit two-dimensional (2D) hydrodynamic models are usually difficult and inefficient. Using domain decomposition, a prediction–correction parallelization method is proposed to solve such systems in parallel. It is proposed as a special method for parallelizing simulations of free-surface flows in alluvial rivers. Rather than solving a large-scale global linear system over the whole domain, the method solves sub linear systems for subdomains in two steps, prediction and correction. For free-surface flows in alluvial rivers, the gravity wave propagation over subdomains is divided into internal and external parts, moving within a subdomain and across its boundaries, respectively. The external part is assumed to be well solved at the prediction step; the whole wave propagation is then solved at the correction step using updated boundary values and initial estimates. A theoretical analysis is conducted to derive the computational errors at the prediction and correction steps of this method, resulting in the condition for its application. The method is tested on five meshes whose numbers of elements are 12,800–819,200. The grid scale, which is equal to or smaller than a common scale of real applications, provides grid-independent results. The method performs well for problems of various computational granularities. In solving linear systems with the different meshes, sequential runs were 41–96 times slower than parallel runs using 64 subdomains and 64 working cores.

An understanding of the clogging of fine sediments within gravel substrates is advanced through use of dimensional analysis and macroanalysis of previously conducted clogging experiments in hydraulic flumes. Dimensional analysis via the Buckingham pi theorem is used to suggest that the dimensionless clogging depth can be potentially collapsed using the original and adjusted bed-to-grain ratios, i.e., ratio of substrate diameter to fine sediment diameter, substrate porosity, roughness Reynolds number, and Peclet number. Macroanalysis and statistical analysis is performed using 10 previously published studies that include a total of 146 different test conditions reporting noncohesive, fine sediment clogging, or deposition in porous gravel-beds with hydraulically rough turbulent open channel flow. Results suggest that the adjusted bed-to-grain ratio is a reliable predictor of the initiation of clogging with clogging occurring below 27 for the fine fluvial sediment clogging in a gravel bed substrate for hydraulically rough turbulent flow flumes considered. The original and adjusted bed-to-grain ratios show little influence on the depth of clogging once the lower threshold for bed filling is reached. Contrary to conventional wisdom, the bed-to-grain ratio is not used to predict the maximum depth of clogging. Rather, results suggest that the dimensionless clogging depth can be collapsed using the substrate porosity and roughness Reynolds number reflecting the impact of the pore water velocity distribution on the dispersion of fine sediment into the gravel substrate. The clogging depth results are used to estimate the clogging profile for fine sands in gravel substrate for the datasets.

Plastic pipes are widely used in pressurized water systems. Their viscoelastic behavior has to be taken into account in the design phase as well as in the interpretation of water-hammer signals for diagnostic purposes. The aim of the present paper is to experimentally investigate the transient waves’ transmission and reflection in branched networks of high density polyethylene (HDPE) pipes and to test the efficacy of a viscoelastic model for the interpretation of the results. Two laboratory Y-shaped systems were built in the Water Engineering Lab of the University of Cassino and Southern Lazio (Italy) and a set of water-hammer tests was conducted on them. A significant variation of transmission and reflection coefficients at the pipe junction is observed in comparison with the values suggested in classical textbooks. The strong agreement between the experimental pressure values and those obtained using the viscoelastic model suggests that the differences from the classical theory are due to the viscoelastic behavior of the material, since unsteady friction effects are negligible in the investigated scenarios.

A two-dimensional (2D) hydrodynamic model is developed for the Jing-Dongting river–lake system, with the computation domain ($\mathrm{3,900}{\mathrm{km}}^2$) covered by a channel-refined grid of 300,000 cells. The model provides good descriptions of the river–lake characteristics (strongly coupled, annular branches, varying flow regimes). In a simulation of a 1-year unsteady flow process, the mean absolute error in simulated water levels is 0.11–0.15 m, and the mean absolute relative error in simulated cross-section discharges is 4.3–8.5%, compared with field data. The accuracy of the 2D model is obviously better than those of existing one-dimensional (1D) and 1D–2D nested models. The prediction–correction parallelization (PCP) method is then tested by simulating the river–lake system using 64 subdomains. The complexities of real rivers are revealed to have almost no negative effects on the quasi-coupled solutions and accuracy of the PCP method. In tests, the mean absolute error in water level is found to be approximately 0.3 cm, and the mean absolute relative error in cross-section discharges is 0.23%, comparing the results of sequential and PCP simulations. In solving linear systems, sequential runs are 40.5 to 76.8 times slower than parallel runs using 64 cores. It takes 6.6 h to complete a simulation of a 1-year unsteady flow process in the river–lake system. The 2D model is then applied to the study of regulations of real river networks.

Laboratory experiments on single dense jets oriented at angles from 15° to 85° to the horizontal are reported. The major flow properties were measured by laser-induced fluorescence at the maximum rise height, impact point, and, for the first time, at the end of the near field. The impact point dilution was insensitive to nozzle angle over the range of about 45–65°. Because the additional mixing that occurred in the spreading layer beyond the impact point depended on nozzle angle, the near-field dilution was more sensitive to nozzle orientation and was highest for 60°, consistent with generally accepted design practice. Bottom boundary effects on dilution were also addressed. Along the jet centerline, time-average dilution first increased and then actually decreased in a thin layer up to the wall. The concentration increase near the bed is due to an increase in turbulent intermittency and accumulation of more saline fluid elements at the bed. The presence of this layer may explain wide discrepancies in reported dilutions near the boundary and may be environmentally important due to exposure of benthic organisms and sea grasses to high salinity. It may not persist, however, as it can be swept up by vortices that propagate radially away from the impact point. The vortices entrain ambient fluid leading to increased dilution, but they eventually collapse under their self-induced density stratification, marking the end of the near field.