Journal of Hydraulic Engineering

The results of an experimental study of a drill-drop manhole used in the City of Edmonton, Alberta, Canada, are presented. A drill-drop manhole is a vertical drop manhole connected to a deep trunk sewer with an upper manhole and a lower manhole. The upper manhole diameter is only two or three times that of the lower manhole. Further, the flow in the upper manhole is dominated by the momentum of the incoming flow and as a result, the weir and orifice outflow types occurring in a relatively larger upper reservoir do not occur in drill-drop manholes. Instead, four regimes were discovered: deflected jet, stable pool, filling and emptying, and full pipe flow regimes. Dimensionless head-discharge curves were developed for the upper manhole for all flow regimes, and an expression for the sum of the losses in the upper manhole, including the entrance loss to the outlet pipe was found from the experimental results. Effects of outlet diameter and its length, inlet diameters, and their elevations were studied.

Large-eddy simulation (LES) of a three-dimensional, turbulent free surface flow past a stream restoration structure with arbitrarily complex geometries is presented. The three-dimensional, incompressible, spatially filtered Navier-Stokes and continuity equations are solved in generalized curvilinear coordinates. For the solution of mixed air-water flows, the curvilinear immersed boundary (CURVIB)–level set method developed previously is used and extended to carry out LES. Complex solid geometries are handled by the sharp-interface CURVIB method, and the subgrid scale stress terms arising from the spatial filtering of the Navier-Stokes equations are closed by the dynamic Smagorinsky model. To demonstrate the potential of the CURVIB–LES–level set model for simulating real-life, turbulent free surface flows involving arbitrarily complex geometries, LES is carried out for the flow past a complex rock structure that is fully submerged in water in a laboratory flume. The simulations show that the method is able to predict the time-averaged value as well as the root-mean-square fluctuations of water surfaces with good accuracy. Moreover, it is seen that the free surface flow at a high Froude number causes a significant level of fluctuations of water surface elevation and velocity at the water surface.

Side weir flows induce complex three-dimensional (3D) structures on the flow field in the main channel. Extensive experiments at the laboratory scale were conducted in order to investigate the flow field for the particular case of a side weir with a zero height crest in both fixed bed and steady flow conditions. Detailed measurements of the flow surface, discharge, and velocity field were performed in the main channel upstream, alongside, and downstream of the side weir location. They allowed detection and analysis of the three-dimensional flow structures. Elaborations of the measurements show the implications of such three-dimensional nature of the flow on integral quantities such as Coriolis and Boussinesq coefficients as well as the specific energy and discharge coefficient that are important in the widely used one-dimensional (1D) models. Moreover, the results may help to better understand other free-surface problems similar to the side weir flow, such as lateral diversions, river bifurcations, and levee breaches.

An expression for the dispersion coefficient in a rectangular channel is derived to evaluate the importance of transverse and vertical variations in velocity for dispersion. The contribution of vertical variations to dispersion depends not on the ratio of the width $B$ and depth $H$ of the channel—as is usually assumed—but on the ratio of mixing times, $\tau =(H^2/D_z)/(B^2/D_y)$, where $D_y$ and $D_z$ are the transverse and vertical mixing coefficients, respectively. The analysis allows the role of vertical variations to be assessed quantitatively as a function of the time scale ratio and the shape of the velocity profile. The time scale ratio is estimated using data sets compiled by others and several empirical formulas for $D_y$. In almost all cases, vertical variations contribute a small amount to the overall dispersion. The results support the usual practice of considering only transverse variations in computing the dispersion coefficient, and the analysis provides an approach for including vertical variations in calculations of dispersion in cases in which $\tau$ is not small.

A single-layer velocity model is proposed in this study to predict the cross-sectional average velocity for open channel flows with submerged rigid cylinders. This model is preferable to previous two-layer models in that it does not involve evaluations of vegetation drag coefficient and hydrodynamic height of roughness elements. In the development of the single-layer model, the concept of the hydraulic radius is applied to determine the length scale of the entire cross-section affected by the submerged vegetation. The resulting friction factor scales largely with the ratio of stem height to flow depth with exponent $3/2$, and also decreases slightly with an increasing Reynolds number. Comparisons of the single-layer model with experimental data show that its performance is comparable to the previous two-layer models, and the model is also applicable to open channel flows subject to submerged flexible vegetation. However, the proposed model cannot be extended to two extreme conditions including flows over a smooth channel bed without vegetation and flows passing through emergent vegetation.