Neural Modeling of Square Surface Aerators

Applications of artificial neural networks in the field of aeration phenomena in surface aerators, which are not geometrically similar, are explored to predict reaeration rates under varying dynamic as well as geometric conditions. The primary network for prediction is a feed forward network with nonlinear elements. The network consists of an input layer, an output layer, a hidden layer, and the nonlinear transfer function in each processing element. The network requires supervised learning and the learning algorithm is the back-propagation. As back-propagation learning is affected by local minima, and to get over this aspect various other modifications have been suggested like Levenberg-Marquardt, quasi-Newton, conjugate-gradient, etc. The present study suggests that the Levenberg-Marquardt modification is a very efficient algorithm in comparison with others like quasi-Newton and conjugate-gradient. In the situations when the dimension of the input vector is large, and highly correlated, it is useful to reduce the dimension of the input vectors. An effective procedure for performing this operation is principal component analysis. The best prediction performance is achieved when the data are preprocessed using principal components analysis before they are fed to a back-propagated neural network, but at the cost of losing the physical significance of experimental data. The model thus developed can be used to predict the reaeration rate for different sizes of geometric elements (like rotor diameter, sizes of rotor, aerators’ geometry, water depth, etc.) under various dynamic conditions, i.e., the speed of the rotor.     

Chute Aerators. I: Air Transport Characteristics

Chutes with flow velocities in excess of some 20 to 30 m/s are usually prone to cavitation damage. Therefore, these flows are aerated using chute aerators. The current literature describes the aerator efficiency mainly in terms of the air entrainment coefficient as the ratio of the entrained air and the water discharges. However, this global coefficient neither specifies the air distribution nor its detrainment rate. The present investigation focuses on the flow structure and the air transport downstream of chute aerators. Systematic hydraulic model tests were conducted including a data analysis of the spatial air concentration distribution in both the near and the far aerator fields. Based on these, three flow zones were introduced, namely: (1) jet zone; (2) reattachment and spray zone; and (3) far-field zone. It was further found that aerators have primarily an effect on the average air concentration, whereas the increase of the bottom air concentration is typically small.     

Turbulence Structure of Hydraulic Jumps of Low Froude Numbers

Turbulence characteristics of hydraulic jumps with Froude numbers of 2.0, 2.5, and 3.32 are presented. A Micro Acoustic Doppler velocimeter was used to obtain measurements of the velocities, turbulence intensities, Reynolds stresses, and power spectra. The maximum turbulence intensities and Reynolds stress at any section were found to decrease rapidly from the toe of the jump towards downstream within the jump and then gradually level off in the transition region from the end of the jump to the friction dominated open channel flow downstream. The maximum turbulence kinetic energy at each section decreases linearly with the longitudinal distance within the jump and gradually levels off in the transition region. The Reynolds stress and turbulence intensities within the jump show some degree of similarity. The dissipative eddy size was estimated to vary from 0.04 mm within the jump to 0.15 mm at the end of the transition region. The dominant frequency is in the range from 0 to 4 Hz for both horizontal and vertical velocity components.     

Direct Equations for Hydraulic Jump Elements in Rectangular Horizontal Channel

Many problems in the design of stilling basins require a knowledge of various elements of a hydraulic jump with known values such as discharge intensity and the energy loss in the jump. Even for a simple case of a rectangular horizontal channel the solutions of equations involve tedious methods of trial and error. In this paper, direct explicit empirical equations for prejump and postjump depths and specific energies in a rectangular horizontal channel have been developed. These present forms of equations for hydraulic jump elements having very high accuracy and are applicable for a very wide range of values of discharge intensity and head loss without any limitations in comparison to other methods attempted so far.     

Rooster Tail Wave Hydraulics of Chutes

Supercritical flow below chute piers generates an aerated flow known as rooster tail. Their behavior varies along the spillway or chute for different Froude numbers. This paper aims to analyze the effect of various parameters including the Froude number, approach flow depth, chute pier width, and presence of a deflector ramp on the rooster tail geometry. Aerator systems, which generally include deflectors in chutes, are effective in the generation of rooster tails. The interaction between a deflector and supercritical flow of chutes causes a higher rooster tail. New equations define the rooster tail geometry in the presence of a deflector. The air concentration profile was measured and the effects of deflectors on the air entrainment were studied.     

Modes of Bed-Load Movement on a Smooth Rigid Surface

Sediment grains transported as supply-limited bed load on a rigid surface move either discretely or collectively as bed forms, with significantly different effective grain speeds and active storage volumes. The adopted mode has implications for sediment sorting and heavy mineral placer formation, dispersal of grain-associated pollutants, and accumulation and flushing of sediment deposits in unlined canals and sewers. The threshold condition between the two modes has been established for a smooth surface from flume experiments with different sediment types, flow conditions, and sediment supply rates. This is expressed in terms of a relationship between the sediment movability number, a dimensionless bed load parameter, and a grain shear Reynolds number.     

Aeration Characteristics of Ski Jump Jets

Scour downstream of ski jumps may be avoided by jet deflection to an area where the energy dissipation is accomplished. The main purpose of this experimental study was the analysis of the jet air entrainment downstream of a ski jump, both for pure water and preaerated approach flow conditions. A systematic variation of the Froude number and the flow depth in the approach flow channel resulted in a range of discharge characteristics, whereas the geometry of the ski jump was maintained for all tests. The air concentration profile was measured at different locations downstream from the ski jump to evaluate the: (1) jet air concentration distribution; (2) location of minimum air concentration along the mixture flow jet and development of the minimum and the cross-sectional average air concentrations; (3) jet trajectories; and (4) process of air entrainment characteristics and jet disintegration. The results demonstrate the significant effect of the approach flow Froude number, the approach flow depth, and of preaeration on jet disintegration.     

Horizontal Injection of Gas–Liquid Mixtures in a Water Tank

Experiments were carried out to investigate the behavior of horizontal gas–liquid injection in a water tank. Measurements of bubble properties and mean liquid flow structure were obtained. The turbulence in the liquid phase appears to help generating bubbles with relatively uniform diameters of 1–4?mm. Both bubble properties and mean liquid flow structure depended on the gas volume fraction and the densimetric Froude number at the nozzle exit. It was found that the bubbles strongly affected the trajectory of the water jet, which behaved similarly to single-phase buoyant jets. However, at gas volume fractions smaller than about 0.15, the water jet completely separated from the bubble core. Bubble slip velocity was also found to be higher than the terminal velocity for isolated bubbles reported in the literature. Dimensionless correlations were proposed to describe bubble characteristics and the trajectory of the bubble plumes and water jets as a function of the gas volume fraction and the densimetric Froude number. Finally, applications of the results for aeration/mixing purposes are presented.     

Sequent Depth Ratio of a B-Jump

A B-jump is defined as the jump having the toe section located on a positively sloping upstream channel and the roller end on a downstream horizontal channel. This jump often occurs in the stilling basins with a horizontal bottom and located downstream of a steep channel. For a B-jump, a completely theoretical approach is not sufficient to solve the momentum equation and to establish the sequent depth ratio. In this paper, by using the laboratory measurements carried out in this investigation, some available empirical relationships useful for estimating the sequent depth ratio are tested. Then, by using the Π theorem of the dimensional analysis and the incomplete self-similarity theory, a generalized functional relationship for estimating the sequent depth ratio for different types of jumps is deduced. The estimate of the coefficient appearing in this relationship is dependent on the particular type of jump. In conclusion, the analysis established that the sequent depth ratio for a B-jump depends on a parameter E accounting for the toe section position, the upstream Froude number F1, and the channel slope.     

Flow Characteristics of Skimming Flows in Stepped Channels

Skimming flows in stepped channels are systematically investigated under a wide range of channel slopes (5.7°?θ?55°). The flow conditions of skimming flows are classified into two flow regimes, and the hydraulic conditions required to form a quasi-uniform flow are determined. An aerated flow depth of a skimming flow is estimated from the assumption that the residual energy at the end of a stepped channel coincides with the energy at the toe of the jump formed immediately downstream of the stepped channel. In a quasi-uniform flow region, the friction factor of skimming flows is represented by the relative step height and the channel slope. The friction factor for the channel slope of θ=19° appears to have a maximum. The residual energy of skimming flows is formulated for both nonuniform and quasi-uniform flow regions. Further, a hydraulic-design chart for a stepped channel is presented.     

Laminar Surface Water Flow over Vegetated Ground

This study examines low Reynolds number flow over vegetated sloping ground. Instead of using the traditional empirical formula related to resistance or frictional roughness, a new approach is presented by considering the flow inside the vegetation layer as porous media flow governed by Biot’s poroelastic theory as well as inside the soil layer. There is a discrepancy in the velocity distributions of flow over vegetative mantle area and bare area, respectively. Due to the change of the vegetation density, the effects of vegetation porosity and vegetation blockage on surface water flow are discussed.     

Influence of Selected Operating Parameters on Fungal Biomass Production in Corn-Ethanol Wastewater

Wastewater from a corn wet-milling ethanol plant was treated with Rhizopus microsporus mold in a continuous biofilm reactor (attached growth system). Plastic composite support tubes, composed of 50% (w/w) polypropylene and 50% (w/w) agricultural products were used as support media. The effects of operating pH (3.5, 4.0, and 4.5) and hydraulic retention times (HRTs) (5.0, 3.75, 2.5, and 1.25 h) on fungal growth, chemical oxygen demand (COD) removal and unwanted bacterial growth were evaluated under nonaseptic conditions. COD removal and biomass production were highest at pH 4.0 with lowest bacterial competition. Maximum COD removal of up to 80% was achieved at a 5.0 h HRT with a biomass yield of 0.44 g volatile suspended solids per g COD removed. A higher biomass yield was achieved at a shorter HRT of 2.5 h due to increased substrate availability; however, the biofilm was more sensitive to changes in wastewater composition. A HRT of 3.5–4 h was considered optimal in achieving organic removal and fungal biomass production. Significant loss of fungal biomass due to washout occurred at a 1.5 h HRT. Undesirable bacterial populations as a fraction of total biomass decreased with reducing HRT, excluding the 1.25 h HRT. Reductions in COD removal and biomass production were observed with decreases in aeration rate (1.0–0.25 L/min or 0.8–0.2 vvm (air volume per reactor working volume per minute). The recovered fungal biomass was found to contain protein of up to 40% (dry mass basis), which could serve as a source of high-value animal feed.     

Energy Model to Predict Suspended Load Deposition Induced by Woody Debris: Case Study

Large woody debris (LWD) has been used repeatedly to create aquatic habitats. This study attempts to quantify and predict geomorphological changes induced by LWD. Six cylindrical bundles of LWD were anchored in a stream in central Ohio, and bed elevations were monitored for up to seven months. A model was developed to predict deposition downstream of the LWD. Sediments accumulated immediately downstream of the LWD structures. The average accumulation depth 0.25 m downstream of the LWD was 0.10 m. At 1.25 m downstream, accumulation depth averaged 0.07 m. The model to predict the sediment deposition had R2 values of 0.87–0.77, respectively, at the two downstream locations. The most important terms in the model were the Froude number and bankfull depth. An advantage of this model was the use of easily measurable variables including average bankfull velocities, depth, and cross-sectional area. This fact will facilitate the use of the model in field settings. Suggestions for future improvements to the model include calibration/validation in different streams, inclusion of a temporal variable, and sediment characterization.     

Measurements of Water Surface Profile and Velocity Field at a Circular Pier

Flow around a 60-mm-diameter pier on a smooth bed was measured in an open-channel flume. By varying the approach flow velocity and water depth, a wide range of subcritical flow conditions was produced. Water surface elevation was measured at 0, 90, and 180° to the approach flow direction near the surface of the pier; and three-dimensional velocity vector field around the pier was measured in two horizontal planes, one close to the bed and the other near the free surface. The velocity field measurements were obtained using a stereoscopic particle image velocimetry system. It was found that the change in water surface elevation around the pier was related to the Froude number and relative water depth. However, no direct relationship between the Froude number and the measured velocity fields was found. The approach flow conditions affected the pier flow field mainly behind the pier; the flow pattern was related to the pier Reynolds number. It was also found that the direction and magnitude of the ensemble-averaged velocity field was more dependent on the pier Reynolds number near the bed.     

Characteristics of Steady Horseshoe Vortex System near Junction of Square Cylinder and Base Plate

This paper presents an experimental investigation on the characteristics of a horseshoe vortex system near the juncture of a square cylinder and a horizontal base plate, using particle image velocimetry and flow visualization technique. Experiments were conducted for Reynolds numbers (based on the free stream velocity and the width of square cylinder) ranging from 2.0×102 to 6.0×103. The flow patterns are first classified into four major regimes: Steady horseshoe vortex system, periodic oscillation vortex system with small displacement, periodic breakaway vortex system, and irregular vortex system. The classifications can be demonstrated as a figure of Reynolds number versus the ratio of the height of square cylinder to undisturbed boundary layer thickness. The study then mainly focused on the characteristics of steady horseshoe vortex system (corresponding to Reynolds numbers ranging from 2.0×102 to 2.5×103). The nondimensional characteristics, including the horizontal and vertical distances from the primary vortex core to frontal face of the vertical square cylinder and bottom boundary of the base plate, respectively, the height of stagnation point at frontal face of the square cylinder, and the down-flow discharge as well as circulation of the primary vortex, all increase with increase of the ratio of the height of square cylinder to undisturbed boundary layer thickness. However, they all decrease with the increase of the aspect ratio (i.e., the height-to-width ratio) of the square cylinder. The study provides essential properties of a steady horseshoe vortex system and gives an insight for related engineering applications. It can be served as a basis for more complicated horseshoe vortex systems occurring at high Reynolds numbers.     

Cold Simulation of Metalloids Transport in Blast Furnaces

Metalloids normally get transferred at the interface of metal droplets passing through the slag system in the dropping zone and at the slag‐metal interface in the hearth zone in the lower region of a blast furnace. In these high temperature processes, the mass transport being the rate‐controlling factor, the viscosity of the slag system determines the kinetics of the refining reactions accompanied by mass and heat transfer at the metal droplets and slag interface. Slag systems generally possess random network structures comprising internal regions of weak ordering. The presence of these regions may result in non‐Newtonian behaviour of the slag. The rheological characteristics of a fluid relating to its network structure is expressed in terms of the indices consistency (k') and flow behaviour (n'). The extent of metalloids presence in hot metal is subjected to their residence time at the slag‐metal interface. The metal droplet descent through a surrounding fluid system has been studied and a co‐relation between drag Reynolds number and modified Reynolds number has been obtained. This correlation has been used to determine the drag velocity of a metal droplet falling through a slag system and the residence time distribution (RTD) of the metalloids at the slag‐metal interface in the lower region of the blast furnace.     

Effect of Organic Surface Load on Process Performance of Pilot-Scale Algae and Duckweed-Based Waste Stabilization Ponds

Removal efficiencies in pilot scale algae-based ponds (ABPs) and duckweed-based ponds (DBPs) were assessed during two periods of 4 months each. During Periods 1 and 2, the effect of low and high organic loading was studied. A linear correlation between ponds organic surface loading rates and the corresponding biochemical oxygen demand (BOD) removal rates was observed in both systems. For both periods, higher BOD and total suspended solids (TSS) removal efficiencies were found in DBPs compared to ABPs. Nitrogen removal rates (λr) in ABPs were linearly correlated with BOD surface loading rates (λs,BOD) and nitrogen loading rates (λs,N), while in DBPs, N removal rates were almost constant irrespective of λs,BOD or λs,N. Overall N removal rate in the algae system was significantly higher than that in duckweed system. Organic loading had no effect on total phosphorus removal efficiency in both systems. Higher P removal efficiency was achieved in the duckweed system than in the algae system. In ABPs as well as DBPs, fecal coliforms were better removed during low organic loading in comparison with high organic loading. During the two operational periods, higher fecal coliform removal efficiency in the algae system than in the duckweed system was observed.     

Oxygen Transfer in High-Speed Surface Aeration Tank for Wastewater Treatment: Full-Scale Test and Numerical Modeling

Oxygen transfer is one of the key processes in the bioreactor. Herein a computational fluid dynamics model for the oxygen transfer in high-speed surface aeration tank has been developed and validated through a full-scale aeration test. The test results indicate that the oxygen transfer mainly comes from the spray water in air and that the gas entrainment by the plunging of spray water and the surface reaeration in the aeration tank contribute little to the total oxygen transfer in high-speed surface aerator. A simple method was proposed to measure the oxygen transfer rate for high-speed surface aerator.     

Flow Characteristics around a Circular Cylinder Placed Horizontally above a Plane Boundary

Flow characteristics around a circular cylinder positioned near a plane boundary (on which laminar boundary layer flow develops in the absence of circular cylinder), are investigated for Reynolds numbers R ranging from 7.8×102 to 1.15×104. Particle image velocimetry and fiber laser Doppler velocimetry were used to measure the velocity fields and velocity time histories, respectively. Flow structures are particularly revealed using flow visualization technique at R = 7.8×102 for gap ratios G/D (where G is the net gap between the surface of circular cylinder and the plane boundary), varying from 0 to 4. Based on the experimental results, the variation of Strouhal number of shedding vortex (or eddy) with G/D, the mechanism of vortex shedding suppression, and the streamwise velocity profiles of the upper shear layers and gap flows for small G/D are all discussed. Although the regular, alternate vortex shedding is suppressed for G/D<0.5, the periodicity could be detected due to the vortex (or eddy) shedding from the upper shear layer of the circular cylinder. Gap flow switching randomly is found and first put forward to be the main reason of multipeak or broadband spectral characteristics of the shedding event at a certain small gap ratio. It is also found that the streamwise velocity profiles of the upper shear layer, where periodic shedding eddies originate, exhibit well-behaved similarity. In addition, a unique similarity of mean streamwise velocity profiles of the gap flows is demonstrated for G/D ? 0.3. For R<4×103, the S increases as G/D decreases to its maximum around G/D ? 0.5 and then decreases as G/D decreases. For R ≥ 4×103, although most of the previous studies indicate that the S is insensitive to G/D, the present study shows that S still increases as G/D decreases but the variations of S are in a small range (i.e., 0.18 ? S ? 0.22).     

Design Characteristics of Single Hub Paddle Wheel Aerator

Aeration experiments, based on dimensional analysis, were conducted in brick masonry tanks of dimensions 2.9×2.9×1.6?m3 and 5.9×2.9×1.6?m3 to study the design characteristics of a single hub paddle wheel aerator. A generalized equation has been developed to estimate the optimum volume of water to be aerated at a given paddle wheel diameter (d) and speed (N). A simulation equation with E [nondimensional number characterizing standard aeration efficiency (SAE)] and P* (nondimensional number characterizing power input per unit volume) provided a correlation better than that with E and F (Froude number) and E and R (Reynolds number) only. A simplified generalized equation correlating E and P* was developed which can be used to predict aeration efficiency for P* ? 6.56. Similarly, a simulation equation with Ne [(power number)(nondimensional number charcterizing power consumption)] and P* provided a correlation better than that with Ne and F and Ne and R only. The optimum dynamic condition at which maximum SAE is produced has also been presented. All the above results are valid subject to 1.47 ? P* ? 15.54.