Journal of Hydraulic Engineering

The hydraulic model study to determine the diffuser design for the Boston Wastewater Outfall is described in two papers. The outfall is tunneled with a diffuser consisting of multiport risers, and in this paper the tests leading to the final choice of the riser configuration are described. The study was done in a large density‐stratified towing tank. It was found that, for zero current speed with a deep thermocline, a thick, laterally homogeneous, submerged waste field was formed. For widely spaced risers, additional mixing in the plumes following collapse caused the minimum dilution to be higher than predicted by simple plume formulae. As the number of risers is decreased below about 50, the dilution decreases fairly rapidly, but increasing the number above about 50 caused the dilution to increase only slowly. The number of risers was chosen to be 55. This resulted in a savings of about $38,000,000 in construction costs compared to the preliminary design.

Hydraulic model tests were done on the Boston Harbor outfall diffuser with the number of risers and their spacing maintained constant at their final design values. It was found that eight ports per riser resulted in higher dilutions than 12, because with 12 ports the plumes merged and collapsed to a single rising column. Extended duration tests showed that the waste field was stable, that its thickness increased very slowly with time, and that dilution decreased very slowly with time. The flux‐average dilution was estimated from the movement of dye streaks to be only 1.1–1.2 times higher than the minimum dilution, a much smaller ratio than usually assumed. The mathematical model ULINE was found to be generally conservative in predicting dilution. Better predictions were obtained with a new mathematical model RSB, which well simulated the observed waste‐field characteristics for the final design. The effect of varying the number of risers was not so well predicted by RSB, however, due to the very large effective port spacing for this diffuser. Thus, hydraulic model tests are still needed for atypical diffuser designs, such as encountered here.

The problem of a small‐amplitude water wave propagating in a channel of constant depth with a horizontal poroelastic bed of infinite thickness is investigated. Using the theory of poroelasticity derived in Huang and Chwang (1990), three decoupled Helmholtz equations are applied to study the present problem. The general solution is obtained, and five physical parameters, $m$, $\epsilon$, $\Lambda$, $\Pi$, $\Psi$, for the porous media flow are defined. The first of these physical parameters represents the relative stiffness of solid and fluid, the second represents penetrability, while the other three reveal Mach numbers of two longitudinal waves and one transverse wave of the porous medium when penetrability is low, i.e., when $\Vert \epsilon \Vert \ll 1$. It is believed that these parameters will be especially useful for experimental modeling. We not only clarify the required constraints of Yamamoto et al. (1978) and Liu (1973) that are not explained in their articles, but also demonstrate successfully the generality of the present work.

Water losses in irrigation systems are a major problem in arid countries. The losses are primarily due to inefficient irrigation systems and uncontrolled water delivery. Improving the irrigation‐management systems is therefore necessary to conserve the available water. Numerical modeling appears as an important aid in solving this problem because it has been proven to be an excellent tool for designing and simulating flows in irrigation networks. However, shallow‐depth and dry‐bed situations, which are frequently encounted in irrigation canals, are classical problems in computational hydraulics, often causing computational difficulties. Extensive numerical experimentation, supported by analysis of the characteristic behavior of the governing equations, has led to development of new approach for avoiding the singularities of vanishingly small depths in canals. This new approach is based on suppression of the inertial terms in the equations when small depths appear imminent.

By applying the kinematic wave method to the analysis of overland flow through a rectangular plane, a functional relationship between the critical rainfall duration associated with the maximum peak discharge, and the characteristics of the rainfall intensity‐duration curve has been formulated. The traditional approach of equating the critical duration to the time of concentration is valid for planes with low infiltration loss. In this case, the full plane area contributes to the maximum peak discharge. However, for planes with high infiltration loss, the duration of the critical rainfall associated with the maximum peak discharge may be shorter than the time of concentration. Under this condition, only part of the plane contributes to the maximum peak discharge. A method for determining the critical rainfall duration and equations for determining the maximum peak discharge have also been developed.

The flow of water around a cylinder was calculated by a numerical model. The model used a finite‐volume method to solve the nontransient Navier‐Stokes equations for three dimensions on a general nonorthogonal grid. The $k‐\epsilon$ turbulence model was used to solve the Reynolds‐stress term. The flow field gave the shear stress on the boundaries, which was used to calculate sediment concentration for the bed elements. The diffusion/convection equation for the sediment concentration was solved and continuity for the sediment concentration gave changes in the bed elevation. Using the modified bed, a new flow field was calculated, and also a new sediment concentration. This sequence was repeated until the depth of the calculated scour hole was identical to the depth of the scour hole from a physical model study. The rest of the geometry of the resulting scour hole compared well with the measured scour hole from the experimental study. Also other flow characteristics given by the numerical model were found to be in agreement with observations from the experimental study.

The circular‐crested weir is used for flow measurement and has wide applications in hydraulic engineering, where it serves as an overflow structure. It can be used to control the water level in farm ponds and reservoirs. The discharge coefficient of such a weir is obtained experimentally as a function of the dimensionless total head of the approaching flow. It is shown that changing the upstream slope does not alter the weir‐discharge coefficient. However, increasing the downstream weir slope appears to increase the discharge coefficient. The weir crest pressure is determined as a function of the dimensionless head and its variation is related to the weir‐discharge coefficient. The experimental data indicate that the circular‐crested weir behaves like a sharp‐crested weir, when the dimensionless total head is extremely large.