A Laboratory Facility for Flocculation-Related Experiments

school of ciVil engineering

research rePort r952march 2015

issn 1833-2781

a laBoratory facility for flocculation-related exPeriments

fiona h.m. tangfederico maggi

SCHOOL OF CIVIL ENGINEERING

A LABORATORY FACILITY FOR FLOCCULATION-RELATED EXPERIMENTS

RESEARCH REPORT R 952

FIONA H.M. TANG

FEDERICO MAGGI

MARCH 2015

ISSN 1833-2781


School of Civil Engineering The University of Sydney

Research Report R952

Copyright Notice

School of Civil Engineering, Research Report R952 A Laboratory Facility for Flocculation-Related Experiments Fiona H.M Tang Federico Maggi March 2015

ISSN 1833-2781

This publication may be redistributed freely in its entirety and in its original form without the consent of the copyright owner.

Use of material contained in this publication in any other published works must be appropriately referenced, and, if necessary, permission sought from the author.

Published by: School of Civil Engineering The University of Sydney Sydney NSW 2006 Australia

This report and other Research Reports published by the School of Civil Engineering are available at http://sydney.edu.au/civil

Page 2




ABSTRACT

This report describes the design and functions of a new experimental facility built in the School of Civil Engineering at the University of Sydney used for the investigation of flocculation-related processes. This facility was uniquely designed to replicate physical (hydrodynamic processes and sediment load), chemical (nutrients and contaminants) and biological (micro-organisms) processes in natural aqueous environment; hence, it allows for investigating the effects of these processes on the flocculation dynamics of suspended particle matter (SPM) through a fully controllable laboratory-based research. It consists of five major components, including a small- scale settling column, a turbulence generating system, a water quality mea-suring system, a μPIV system, and a micro-controlling system. Measurements, either imaging data of settling SPM or water quality readings, can be acquired automatically with any ar-bitrary scheduling. The innovation of this facility is the integration of physical, chemical and biological aquatic processes into one framework to explore the complexity of the interactions be-tween these processes and SPM dynamics. One of its major contributions to the advancement in sediment dynamics studies is the direct detection of possible repercussions the increased anthropogenic stresses has on the microbial population and the aggregation kinematics and statistics of suspended particles in aqueous ecosystem. Ultimately, this facility is expected to contribute to a comprehensive understanding of how all possible interactions in natural water bodies affect each other and consequently, how these interactions affect SPM flocculation and transport.

KEYWORDS

Aggregation Settling column μPIV Suspended particle matter Biological flocculation

Page 3


School of Civil Engineering The University of Sydney Research Report R952

Contents

Abstract 3

Keywords 3

1 Introduction 7

1.1 Background .................................................................................................................................. 7

1.2 Aim .............................................................................................................................................. 8

2 Design criteria 9

3 Facility components 10

3.1 Settling column ........................................................................................................................... 10

3.1.1 Flocculation section ....................................................................................................... 10 3.1.2 Diaphragm ....................................................................................................................... 11

3.1.3 Measuring section ......................................................................................................... 12 3.2 Turbulence generating system ................................................................................................. 12

3.2.1 Grid geometry .............................................................................................................. 12 3.2.2 Oscillation control ........................................................................................................ 13 3.2.3 Determination of turbulence shear rate ...................................................................... 14 3.2.4 Discussion ..................................................................................................................... 17

3.3 Water quality measuring system .............................................................................................. 18 3.3.1 Water quality meter ..................................................................................................... 18 3.3.2 Parameter calibration ................................................................................................... 18

3.4 μPIV system ............................................................................................................................. 203.4.1 Imaging system ............................................................................................................. 21 3.4.2 Illumination system ..................................................................................................... 21 3.4.3 Calibration of pixel size ................................................................................................ 23

3.5 Micro-controlling system .......................................................................................................... 24

4 Conclusion 26

5 Photographs of the facility 27

6 Concept drawings of the facility 31

Acknowledgment 40

Bibliography 41

Page 4



List of Figures

3.1 Schematic drawings of (a) the settling column, (b) the flocculation section, (c) the diaphragm, and (d) the measuring section. ........................................................................................................................................... 9

3.2 (a) the plan view and the cross-sectional view of a grid element, and (b) schematic drawing of the grid oscillation system. ..................................................................................... 11

3.3 Comparison of the turbulence shear rate G calculated based on equations proposed inHopfinger and Toly (1976) (Eq. (3.8)) and Matsunaga et al. (1999) (Eqs. (3.12)) for

the grid geometry described in Maggi (2005) against the experimentally measured G reported in Maggi (2005) at various grid frequency fg ..................................................... 14

3.4 The turbulence shear rate G determined based on Hopfinger and Toly (1976) approach(Eq. (3.8)) for the grid described in Chapter 3.2.1 as a function of the grid frequency

fg with different values of fitting coefficients C and m ........................................................ 153.5 (a) the turbidity CT of kaolinite suspensions at different kaolinite concentrations CK

over time and (b) the correlation between turbidity CT and kaolinite concentration

CK suspended in tap water with no addition of other substance. .......................................... 183.6 The CCD camera and the high magnification lens mounted on the height-adjustable

camera stand ............................................................................................................................. 19 3.7 The optical box for holding Cree LED and optical fibers. ..................................................... 20 3.8 (a) the optical fibers holder, and (b) the positioning stand for holding the optical

fibers holder. ................................................................................................................................ 20 3.9 Functions of (a) the horizontal and vertical sizes of field of view (FOV), and (b) the

pixel size Lpixel against magnification steps ranging between 0.58 and 7.0. ........................... 21

3.10 The interface built in Matlab2011b environment used for monitoring and controlling

all components of the facility ............................................................................................... 23

5.1 Photograph of the experimental facility ............................................................................... 25

5.2 Components of the settling column, including the flocculation section, the measuring section and the diaphragm. ...................................................................................................... 26

5.3 The multi-parameter water quality meter (TOA-DKK, WQC-24) equipped with Stan- dard and Ion modules ................................................................................................................ 26

5.4 Photographs of (a) an grid element, (b) evenly spaced grid elements connected through a stainless steel bar, and (c) the grid oscillation system. ........................................................ 27

5.5 Photograph of the CCD camera and the high magnification lens mounted on the height-adjustable camera stand ................................................................................................ 27

5.6 Photographs of (a) a Cree LED, (b) the optical box, and (c) the optical fibers holder fixed on the positioning stand. .................................................................................................. 28

5.7 Photographs of (a) the Arduino Uno board, (b) the motor shield, and (c) the screw

shield used in the micro-controlling system. ............................................................................ 28

Page 5



List of Tables

3.1 Table of parameters measured by the water quality meter (TOA-DKK, WQC-24) equipped with the Standard and Ion modules.

............................................................................................................................................. 1

7

3.2 Pixel sizes at magnification steps ranging from 0.58 to 7.0. ................................................... 22

3.3 Arduino Uno pin mapping. ...................................................................................................... 22

Page 6



Introduction

1.1 Background

The flocculation dynamics of suspended particle matter (SPM) in surface waters have been the main area of interest in both environmental and engineering contexts for the analysis and management of water quality, the monitoring of sediment and contaminant transport (e.g., Lick and Rapaka, 1996; Tye et al., 1996) and the prediction of sedimentation rate (e.g., van Leussen, 1988; Winterwerp, 1999).

There is a wide consensus that SPM aggregation and breakup rates, its properties and characteristics (e.g., size distribution, density, porosity and internal architecture) and its settling rate and residence time are governed by numerous natural processes in aquatic ecosystem, such as, hydrodynamic processes (e.g., turbulence and differential settling Kranck , 1973; McCave, 1984), ionic interactions (e.g., Schofield and Samson, 1954; Tombacz and Szekeres , 2004), adsorption of chemicals and contaminants (e.g., Karickhoff et al., 1979; Ongley et al., 1981), presence of organic matter (e.g., fecal pellets, dead cells and metabolic products Droppo, 2001; Mietta et al., 2009) and colonisation of micro-organisms (e.g., Kiorboe, 2001; Grossart et al., 2006). Although effects of these processes on SPM flocculation have long been investigated, these investigations are often conducted indepen- dently in a labour-splitting system. For example, sedimentologists mainly focused on hydrodynamic effects on the physical attributes of SPM, which then affect SPM flocculation dynamics and sedimentation; whereas, chemical engineers and microbiologists investigated mainly the response of SPM flocculation toward a specific type of chemicals, contaminants or micro-organisms.

However, none of the above processes presents alone in natural ecosystem and, therefore, these processes may interact with one another to contribute an impact on SPM flocculation that is yet to be explored. For example, there is still lacking of a comprehensive understanding on how the change in chemical constitution of natural waters would impact the microbial communities, how hydrodynamic processes and increased sediment load can affect microbial colonization, how these interactions would affect SPM flocculation and which of these processes would have a dominating effect over the others. Hence, experiments focusing only on the effect of one controlling factor is not sufficient to answer the questions stated above. This, therefore, calls for a need to design an experimental system that can integrate the effects of all these interactions into one framework and that can allow the control of all physical, chemical and biological parameters at once.

Page 7



1.2 Aim


The aim of this report is to present an experimental facility that is able to incorporate many, if not all, possible physical, chemical and biological processes governing SPM dynamics in natural aqueous ecosystem into one framework. To this end, we design a small-scale settling column equipped with a turbulence generating system that creates isotropic and homogenous turbulence field at any desirable shear rate; and a water quality measuring system that keeps track of the changes of chemical concentrations and other water quality parameters in the settling column, which can then be used to infer chemical and biological interactions. The characteristics and dynamics of SPM are analysed based on imaging data acquired using an automated μPIV system. Through the use of a micro- controlling system in this facility, experiments can be carried out automatically at any arbitrary scheduling. The criteria upon which the design of this facility bases on are listed in Chapter 2, while, detailed descriptions of the design of each component of the facility is then presented in Chapter 3.

Page 8




Chapter 2

Design criteria

This experimental facility was designed to fulfil the following criteria:

1. This facility has to replicate the hydrodynamic (turbulence), sediment and nutrient characteristics found in natural water bodies.

2. This facility has to allow for the monitoring of chemical and biological processes in the controlvolume.

3. This facility has to be able to capture and fully preserve the detailed information of settlingSPM (e.g., size, shape, morphology, fractal characteristics, settling motion, etc.).

4. This facility has to allow for instantaneous, simultaneous and automatic monitor and controlof all parameters and measurements, including turbulence shear rate, sediment concentrations,chemical concentrations, SPM images and water quality readings.

Page 9


Research Report R952School of Civil Engineering The University of Sydney

Chapter 3

Facility components

To satisfy the design criteria stated in Chapter 2, this experimental facility was designed to consist of five major components: the settling column, the turbulence generating system, the water quality measuring system, the μPIV system, and the micro-controlling system (refer Chapter 5 for thephotographs of the full setup). The design of each of the above components is described in details in this chapter.

3.1 Settling column

A settling column (Figure 3.1), made of Perspex, consists of three main components: a flocculation section, a diaphragm and a measuring section. The three components are fastened together with screw connections while gaskets are used in between the connections to provide water proof seal. Both the flocculation and measuring sections are filled with water and are separated by the diaphragm. The test SPM suspension is present and mixed only in the flocculation section. During measurements, a slider is moved to allow the opening of a through hole on the diaphragm and SPM is allowed to flow from the flocculation section into the measuring section. In the measuring section, SPM flow is channeled into the camera field of view, where a μPIV system is used to acquire images of settlingSPM (described in Chapter 3.4).

3.1.1 Flocculation section

The flocculation section (Figure 3.1b) is the control volume where SPM suspension is tested; it has inner depth of 210 mm, inner width of 140 mm and height of 600 mm, with a total capacity of approximately 16 L. The flocculation section is partitioned into two compartments by a vertical

separator; the larger compartment has a square cross-section of 140 mm × 140 mm, where the test

SPM suspension is turbulently mixed; the other compartment has a size of 70 mm × 140 mm, andis used to host a water quality meter inserted vertically from the top of the flocculation section. A triangular slope was placed at the bottom of the water quality meter compartment to drive the

deposited SPM into the compartment with turbulence mixing. Several holes (1/8!! BSP thread)

equipped with elbow valves were made along the flocculation section to control overflow, to allow sampling of SPM at various depths, to drain and clean the column and to provide flexibility if insertion of other equipments is needed in future.

Page 10




Dia

phra

gmM

easu

ring

sec

tion

Floc

cula

tion

sect

ion

Overflow control

Sampling holes

(b)

3 V Motor

Fishing lines

Slider

Vertical separator

Slope to drive sediment into turbulence field

Flange for screw connections

Optical fibers inlets

(a)

Brass pin

(c)(d

)

Bubbles remover

Figure 3.1: Schematic drawings of (a) the settling column, (b) the flocculation section, (c) the diaphragm, and (d) the measuring section.

3.1.2 Diaphragm

The diaphragm (Figure 3.1c) separates the test SPM suspension in the flocculation section from the water volume in the measuring section, and allows only the flow of SPM into the measuring section during time at which image measurements are taken. Five sediment sampling holes of 5 mm diameter were made on the diaphragm to provide flexibility in choosing the position for SPM image acquisition. However, only one sampling hole is used at once and those not in use are temporary sealed. The opening and closing of the sampling hole are controlled using a slider connected to a two shaft 3 V DC motor through two fishing lines that run inside the flocculation section up to the top. To ensure the movement of the slider is not hindered by deposited SPM, a cleaning system, consisting of two immersible water pumps with pump outlets directed onto the slider, is used to remove deposited SPM from the slider.

Page 11




3.1.3 Measuring section The measuring section (Figure 3.1d) has a size of 210 mm × 140 mm × 270 mm. A lateral access(with a choice of 3 different positions) was implemented to insert optical fibers used in conjunction with the μPIV system (described in Chapter 3.4.2). These optical fibers are fastened to a holder (asin Figure 3.8 of Chapter 3.4.2) inside the measuring section. Attached to the holder, a SPM driver is used to guide the SPM flow coming from the diaphragm by gravitational settling toward the camera field of view along a two-wall separation chamber of 5 mm width. A pipe was placed at a corner inside the measuring section to remove bubbles trapped below the surface of the diaphragm. Several additional holes were also made for draining and cleaning of the column and for future need. Note that all holes implemented are of 1/8!! BSP thread.

3.2 Turbulence generating system

An isotropic (i.e., all averages relating to the properties of turbulence do not change under rotations or reflections of the coordinate system, Tennekes and Lumley , 1972) and homogenous (i.e., the mean velocity of the turbulence field is uniform in space, Tennekes and Lumley , 1972) turbulence is generated in the flocculation section by using an oscillating grid, which is a technique widely used in sediment related experiments (e.g., Wolanski et al., 1992; van Leussen, 1994; Liem et al., 1999; Gratiot et al., 2005; Maggi , 2005). The oscillating grid was designed to create a turbulence field of different intensities. The turbulence shear rate G induced by the oscillating grid was determined fromthe geometry, stroke and frequency of the grid using analytical and empirical equations proposed in previous literature (e.g., Hopfinger and Toly , 1976; Fernando and De Silva, 1993; Bache and Rasool , 1996; Matsunaga et al., 1999).

3.2.1 Grid geometry

Tgrhide eolsecmilelanttin(gFigurirde c3o.2nas)i,stms aodfeeoigfhptlahsotirci,zohnatsaal

siqzueaoref 1g2ri0d melmem×en1ts20pamramll.elTthoeegarcihd oeltehmer

e.nEt aisch

made of diamond-shaped bars with bar diameter d = 6 mm and has 16 units of square mesh with

mesh size M = 28.5 mm, resulting in a solidity (i.e., the ratio of the area of diamond-shaped bars tothe total grid area) of approximately 43.75 %. The grid elements are vertically and evenly displaced by 60 mm spacing between each other, and are connected through their centers with a stainless steel rod.

The grid element was made by plastic because plastic is very light in weight, non-reactive to chemicals, non-adhesive and has high resistance to oxidation and corrosion. It was designed as square so as to be symmetrical to the flocculation section (the compartment with square cross section of 140

mm × 140 mm, described in Chapter 3.1.1). In addition, tests reported in Maggi (2005) suggestedthat square grid created more isotropic and homogenous turbulence field in contrast to triangular

ginrdidu,cetdhuts

u,

rbjuuslteinfycien

.

gFtuhrethuesrme orfe,sqthuaeregrigdridwaisndeensisgu

nr

eindgtothae siiszoetr(oi.pei.c, i1ty20anmdmho×m1o2g0enmeimty) otfhatthe

gave 10 mm of clearance between the edge of the grid and the inner wall of the flocculation section. This is to provide sufficient space to introduce the slider mechanism on the diaphragm (described in Chapter 3.1.2) and yet to be small enough to minimise wall effect. The use of diamond-shaped bars enhances the turbulence field as a few studies observed that diamond-shaped prism created the strongest eddies and the highest vorticity compared to circular and square prisms (e.g., Tonui and Sumner , 2011; Ghozlani et al., 2012).

Page 12




Wheel

M = 28.5 mm d = 6 mm Slot to adjust the piston's position

Piston

120 mm

A A

Connector for the grid

Electrical spring interrupter

Tension springs

Metal plate to interrupt the springs

120 mm

A A

Cross section A-A (a)

Hollow metal tube for minimizing rotational and horizontal oscillations

(b)

Figure 3.2: (a) the plan view and the cross-sectional view of a grid element, and (b) schematic drawing of the grid oscillation system.

3.2.2 Oscillation control

The grid is driven through the vertical stainless steel rod by a 12 V DC motor, which is connected to a 140 mm diameter wheel and a piston of adjustable stroke S (Figure 3.2b) that transformed therotational motion of the motor into vertical motion perpendicular to the grid element plane. The adjustable stroke of the piston ranges from 10 mm to 60 mm, which corresponds to the minimum and the maximum stroke, respectively. Note that, the stroke S is defined here as the distance fromthe top to the bottom dead center of the piston, indicating the furthest possible travel of the grid in one direction.

The downward motion of the wheel is facilitated as a result of gravitational effect and, thus, the wheel moves faster downward than upward. This irregularity of the wheel motion is reduced with the aid of tension springs attached to the piston. A driver (a hollow rigid metal tube) is used to guide the stainless steel bar connecting the grid elements to minimise any undesired rotational and horizontal oscillations.

The frequency of the grid fg (i.e., the number of oscillations per second) is measured using anelectrical spring interrupter, which consists of two springs and a metal plate attached to the bottom end of the piston (Figure 3.2b). The two springs are connected to a 5 V electricity supply and a signal is detected when the metal plate touches the springs. The signal is then sent to the micro-controlling system (described in Chapter 3.5) to allow instantaneous, real-time monitoring and adjustment of the frequency.

Page 13




M

B

3.2.3 Determination of turbulence shear rate

The root mean square velocity u and the energy dissipation rate E of the turbulence induced by anoscillating grid can be correlated to the geometry, stroke and frequency of the grid (e.g., Hopfinger and Toly , 1976; Fernando and De Silva, 1993; Bache and Rasool , 1996; Matsunaga et al., 1999). In this section, the approaches proposed by Hopfinger and Toly (1976) and Matsunaga et al. (1999) were tested against the grid geometry and experimental data reported in Maggi (2005). The approach that best matches the experimental data in Maggi (2005) was used to determine the turbulence shear rate G induced by the oscillating grid in this study.

Hopfi nger and Toly (1976) approach Based on a number of experimental observations, Thompson and Turner (1975) and Hopfinger

and Toly (1976) found that at a fixed position z and a fixed stroke S, the turbulent fluctuatingvelocity u is proportional to the grid frequency fg, i.e., u ∝ fg

m (where m = 1 for square bars andm = 4/3 for round bars). From dimensional analysis, they proposed to include a physical quantitythat involves the unit of time, i.e., the kinetic viscosity υ, to account for the exponent m, and that,υ could be accounted for by the Reynolds number Re. By varying S around a fixed z, Hopfinger and Toly (1976) also observed a dependence of u on S and the mesh size M . Hence, at any position

z, u can be written as,

where Re = SfgM/υ.

u fgS

= F Re, S , (3.1)

To derived the spatial decay of homogeneous turbulence induced by the oscillating grid, the turbulence kinetic energy dissipation rate E can be expressed in terms of u as (e.g., Batchelor , 1953;

Tennekes and Lumley , 1972),

E A u3

= l , (3.2)

where l = βz (Thompson and Turner , 1975) is the integral length scale with A and β being constantparameters.

From the observations that u decayed with increasing z, Thompson and Turner (1975) proposed

that

du3 =B u 3

dz l , (3.3)

where B is a constant. By integrating Eq. (3.3) with l = βz, the corresponding u at a particular

distance z from the grid can be expressed as,

u = u0z −

3β

z0, (3.4)

where u0 is the root mean square velocity at z = z0 and z0 is the reference position. Hopfinger and Toly (1976) suggested that the reference position z0 can be taken at z = M with M as the mesh sizeof the grid. Experimental results in Hopfinger and Toly (1976) showed that both β and B depend on

the grid geometry and stroke S; however, the ratio B/β is always constant (i.e., B/β ≈ 3) provided

that the turbulence is homogenous. At z = z0, u0 can be derived based on Eq. (3.1), such that,

Page 14




1

u0 =CfgSRe(m−1)

1

S 2 , (3.5) M

where C = 0.25 is a dimensionless coefficient calculated from experimental fitting in Hopfinger and Toly (1976). By substituting Eq. (3.5) into Eq. (3.4), u can be expressed as,

3 1

u = CS 2 M 2 fg z−1Re(m−1). (3.6)

Following Eq. (3.2), E can therefore be written as,9 3 E = A u

3 3 3 (3m−3)

= C S 2 M 2 f g Re , (3.7) l β z4

where A = 1 (Tennekes and Lumley , 1972) and β = 0.1 (Noh and Fernando, 1993).

The turbulence shear rate G can be derived from E as,( E C3S9 M f 3Re(3m−3)

32 2 G = = g . (3.8) υ βz4 υ

Matsunaga et al. (1999) approach

In Matsunaga et al. (1999), the analytical solutions of E were derived based on the Standard k E model, which does not require the assumption of constant eddy viscosity in the vertical direction. This gives the equation of E as,

E 1 −17/2

= z + 1 , (3.9) 3 E! 1.82 z!

where z! = k! 2 E!−1 with k! and E! as the turbulence kinetic energy k and the turbulence dissipationrate E at z = 0, respectively. Both k! and E! were observed to be dependent on the grid geometryand frequency.

The dependence of k! and E! on fg , S and M was derived in Matsunaga et al. (1999) based onfittings of experimental data with the boundary conditions fixed at z = 0, such that,

for Re < 5500,

k! fg

2S2 = a1

E!

1

S 4 1

M Re 2 , (3.10a) S

and for Re ≥ 5500,fg

3S2 = b1 MRe, (3.10b)

k! fg

2S2 = a2

E!

1

S 4

M , (3.10c)

S

fg3S2 = b2 M

, (3.10d)

where a1 = 8.1 × 10−3, b1 = 8.2 × 10−5, a2 = 6.0 × 10−1 and b2 = 4.5 × 10−1 are curve fitting

dimensionless parameters reported in Matsunaga et al. (1999). Note that, Matsunaga et al. (1999)

defined Re as fg S2/υ.

Page 15




+ 1

G (

s−1 )

Following Eq. (3.9) and Eqs. (3.10), E can be derived as,

E = b1fg3S3M −1Re 1 z + 1 −17/2 for Re < 5500, (3.11a)

E = b2 fg 3S3M −1

1.82 z!

1 z −17/2

1.82 z!for

Re ≥ 5500. (3.11b)

Finally, the turbulence shear rate G can then be derived based on Eq. (3.8) and Eqs. (3.11) as,(

G = E = b1fg 3S3M −1Re 1 z + 1 −17/2 for Re < 5500, (3.12a)

(υ 1.82 z! υ E 1 z −17/2 1b2fg

3S3 M−1

G = υ

=+ 1

1 . 82 z! υ for Re ≥ 5500. (3.12b)

Comparison of the approaches against experimental da ta in Maggi (2005)

The turbulence shear rate G derived in Eq. (3.8) (Hopfinger and Toly (1976) approach) and Eqs.(3.12) (Matsunaga et al. (1999) approach) for the grid geometry described in Maggi (2005) (grid with diamond-shaped bars of d = 8 mm, square meshes of M = 75 mm and stroke S = 37.5 mm) isdepicted in Figure 3.3. The values of G determined using Hopfinger and Toly (1976) approach (Eq.(3.8)) showed a relatively good agreement to the experimental G measured in Maggi (2005) with aslight underestimation at fg < 0.75 Hz. On the other hand, the G derived using Matsunaga et al. (1999) approach (Eq. (3.12)) showed very poor agreement and underestimated the experimental G by approximately one order of magnitude. Hence, in this study, we adopted the approach proposed by Hopfinger and Toly (1976) to calculate the turbulence shear rate generated by the grid described in Chapter 3.2.1.

100

80

Hopfinger and Toly (1976) Matsunaga et al. (1999) Experimental data in Maggi (2005)

60

40

20

0 0 0.2 0.4 0.6 0.8 1

f (s−1) g

Figure 3.3: Comparison of the turbulence shear rate G calculated based on equations proposedin Hopfinger and Toly (1976) (Eq. (3.8)) and Matsunaga et al. (1999) (Eqs. (3.12)) for the grid

geometry described in Maggi (2005) against the experimentally measured G reported in Maggi (2005)

at various grid frequency fg .

Page 16




G (

s−1)

Determination of G for the grid in this study In Eq. 3.7, the experimental fitting coefficients C = 0.25 and m = 1 (Hopfinger and Toly , 1976)

were derived for grids of square bars and we acknowledge that these coefficients may change when different shapes of the grid bar are used. This could also be the explanation of the underestimation observed when comparing this approach to the experimental G measured in Maggi (2005) wherediamond-shaped grid bars were used. Hence, after fitting the experimental data reported in Maggi (2005), we obtained C = 3 and m = 0.67 for grid of diamond-shaped bars.

The functions of G (Figure 3.4) derived based on the coefficients suggested by Hopfinger and Toly (1976) (C = 0.25 and m = 1) and the adjusted coefficients (C = 3 and m = 0.67) were relatively

similar. In this study, we adopted the adjusted coefficients to determine G.

80 C = 3, m = 0.67 (calibrated here)

70 C = 0.25, m = 1 (Hopfinger and Toly, 1976)

60

50

40

30

20

10

0 0.2 0.4 0.6 0.8 1 1.2

fg (s−1 )

Figure 3.4: The turbulence shear rate G determined based on Hopfinger and Toly (1976) approach(Eq. (3.8)) for the grid described in Chapter 3.2.1 as a function of the grid frequency fg with differentvalues of fitting coefficients C and m.

3.2.4 Discussion

The determination of G based on Hopfinger and Toly (1976) approach is strongly dependent on theassumption that the turbulence induced by an oscillating grid is fully isotropic and homogenous. The energy dissipation rate E was calculated based on the assumption that the overall root mean squareof the fluctuating velocity equalled the horizontal root mean square fluctuating velocity (i.e., u = w,where u and w are the horizontal and vertical root mean square fluctuating velocity, respectively).In addition, by using the proposed linear relationship of the integral length scale l with position z (i.e., l = βz), the turbulence velocity derived in Eq. 3.6 is proportion to z−1 under the assumptionthat the eddy viscosity is constant in the vertical direction.

The turbulence induced by the oscillating grid used in this study, however, may not be ideally isotropic and homogenous. Previous studies observed that grids with solidity > 40% tended tocreate secondary flows, which affected the isotropicity and homogeneity of the induced turbulence (e.g., Corrsin, 1963; Hopfinger and Toly , 1976; Fernando and De Silva, 1993). Although the solidity of the grid used in this study is just slightly greater than 40%, the possibility of having secondary flows could not be eliminated. Furthermore, Fernando and De Silva (1993) observed the existence of large-scale secondary flow and much slower decay of turbulence velocity when adopting the grid with

Page 17




end condition that has parallel bars adjacent to the column wall (i.e., similar to the end condition of the grid used in our study). They suggested that secondary circulation could be minimised by cutting away the grid bars at the edges so as to create reflection symmetricity with respect to the column wall (Fernando and De Silva, 1993). The use of diamond-shaped bars in this study may also promote secondary circulation. The axial oscillation tests for circular, square and diamond-shaped prisms conducted in Tonui and Sumner (2011) suggested that the triangular-shaped afterbody of the diamond-shaped prism aided in the formation of large concentrations of secondary vortices. Moreover, wall effect may also affect the isotropicity and homogeneity of turbulence, even though in this study, the clearance between the grid and the column wall has been kept to the minimum. Due to the existence of secondary flow and wall effect, the assumption of constant eddy viscosity and that u = w may not hold fully. However, the secondary flow and wall effect in this study are mainlyconcentrated at the near wall and, therefore, the assumption of isotropic and homogenous will still hold for the turbulence field in the water volume inside the grid.

3.3 Water quality measuring system

3.3.1 Water quality meter

A multi-parameter water quality meter (TOA-DKK, WQC-24), equipped with sensors to measure up to 12 parameters, is used to measure the physical and chemical properties of the suspension tested in the flocculation section (Table 3.1).

The water quality meter has a height of 510 mm and a width of 110 mm, and is connected to a data logger with LCD digital display. The water quality meter has an internal memory capacity that can record a maximum of 3360 data and the data can be continuously and automatically measured and recorded even with the water quality meter disconnected from the data logger. For indefinite recording of data, the measured data can be communicated and saved instantaneously to a PC through a RS-232C cable connected to the data logger. The water quality meter is fully waterproof and has a submerging depth limit of 5 m (depending on the type of sensor mounted to the water quality meter).

3.3.2 Parameter calibration

Prior to experiments, each of the parameters listed in Table 3.1 is calibrated according to the procedures suggested in the ’’Hand-held Water Quality Meter WQC-24 Instruction Manual’’provided by the manufacturer of the water quality meter. Calibration curves may sometimes require for certain parameters (e.g., turbidity, Cl−, NH4

+ and NO3−).

Calibration curves for ion sensors Ion sensors are initially calibrated based on the 2-point calibration procedures stated in the

‘‘Hand-held Water Quality Meter WQC-24 Instruction Manual ’’. The lowest and the highest calibration points are chosen based on the sensor measuring range stated in Table 3.1 and the concentration range of the sample to be measured. To increase the accuracy of measurement, calibration curves can be obtained by increasing the number of calibration points.

Sodium chloride powder can be used to prepare standard solutions at various concentrations for the calibration of Cl− ion sensor, while, ammonium nitrate powder can be used for NH4

+ and NO3 −

ion sensors. Standard solutions are to be prepared with deionized or distilled water. The ion sensors are then submerged in the standard solutions of known concentrations and the measured readings

Page 18




Standard module parameters Indication range Precision Measuring method

pH 0.00 - 14.00 0.01 Glass electrode method Dissolved oxygen (DO) 0.00 - 20.00 mg/L 0.01 mg/L Galvanization diaphragm electrode method Electric conductivity (EC) 0.00 - 10.00 S/m 0.1 mS/m AC 4 electrodes method Salinity 0.0 - 40.0 % 0.1 % Conversion from EC value Total dissolved solid (TDS) 0.0 - 100.0 g/L 0.1 g/L Converted from EC value Sea water specific gravity 0.0 - 50.0 σt 0.1 σt Converted from EC value Water temperature -5.0 - 55.0 ◦C 0.1 ◦C Platinum thin film resistive element Turbidity 0.0 - 800.0 NTU 0.1 NTU 90◦ scattered light measurement method Depth 0.0 - 100.0 m 0.1 m Diaphragm pressure sensor type

Ion module parameters Measuring range Precision Measuring method

Chloride ion Cl−

Ammonium ion NH4+

Nitrate ion NO3−

1 - 35000 mg/L 0.09 - 1800 mg/L 0.62 - 62000 mg/L

0.01 mg/L 0.01 mg/L 0.01 mg/L

Selective ion membrane Selective ion membrane Selective ion membrane

Table 3.1: Table of parameters measured by the water quality meter (TOA-DKK, WQC-24) equipped with the Standard and Ion modules.

are recorded. The calibration curves can then be obtained by plotting the known concentrations of standard solutions against the measured concentrations.

Relating turbidity to suspension concentration Turbidity, a parameter measuring light scattering effect of suspended particles, is often used as

an indicator for the clarity or ‘‘cloudiness’’ of the water samples, hence, denoting the water quality. Some previous studies suggested the potential of using turbidity as a cost-effective estimation of the sediment concentration and proposed various possible correlations between turbidity and sediment concentration (e.g., Lawler and Brown, 1992; Gippel , 1995; Smith and Davies-Colley , 2002; Chanson et al., 2008). However, the correlations established between turbidity and sediment concentration were found to be site specific and were dependent on factors such as, the shape, size and composition of the particles, the sensitivity of the instruments, the presence of biological matter and the lighting condition (e.g., Gippel , 1995; Riley , 1998). Hence, in this study, turbidity was specifically calibrated and restricted to be used only as an estimation of the concentration of non-aggregated suspended kaolinite type Q38 (primary particle size (diameter) ranging between 0.6 μm to 38 μm).

The effect of flocculation on the turbidity of kaolinite suspension was examined through a small- scale pilot test. Kaolinite type Q38 was hydrated for 10 minutes prior to testing and was then added to a beaker with 250 ml of tap water. A magnetic stirrer was used to generate constant stir and to ensure no deposition at the bottom of the beaker. The turbidity readings were taken continuously for 2 hours. These procedures were repeated for three different kaolinite concentrations CK (i.e., 100 mg/L, 200 mg/L and 500 mg/L). On the side, a sample of 500 mg/L kaolinite suspension with added flocculant (i.e., Chitosan) was also tested to examine the effects of flocculant-induced particle aggregation on the turbidity. The turbidity CT readings over time for different kaolinite suspensionsare depicted in Figure 3.5a.

In Figure 3.5a, three general trends are observed: first, the turbidity CT of kaolinite suspensionswith no added flocculant were generally constant over time, especially at low concentrations; second, the suspension with added flocculant had an initial CT higher than that without flocculant; and,thirdly, in the presence of flocculant, CT decreased over time. In addition, we also note that the

Page 19




Tur

bidi

ty C

T(N

TU

)

Kao

linite

con

cent

ratio

n C

(m

g/L

) K

standard deviation increased with increasing CK .

Hence, from the results of these tests, we deduce that: (i) kaolinite suspension in tap water did not result in flocculation rate that was detectable by the turbidity sensor; (ii) the presence of other substances in the suspension could alter the turbidity values measured by the sensor; and (iii) the formation of flocs in the suspension decreased the turbidity. We then suggest that the correlation between turbidity and suspension concentration can only be established if the suspension is of the same particle size, shape and mineralogy without the presence of other substances and that the flocculation of the suspended particles is negligible.

Since the results in Figure 3.5a denote the flocculation of kaolinite suspension in tap water is negligible, the turbidity was then calibrated against known kaolinite concentrations. The correlation

between turbidity CT and kaolinite concentration CK can be derived as in Figure 3.5b, such that,

CK = 0.95CT . (3.13)

This correlation, having correlation coefficient R= 1.00 and normalized root mean square errorNRMSE = 6.9 %, is used in experiments to estimate the concentration of kaolinite suspension tested in the flocculation section.

700

600

500

C = 500 mg/L− Flocculant K

800

700

600

400

300

200

100

C = 500 mg/L K

C = 200 mg/L K

C = 100 mg/L K

500

400

300

200

100

C = 0.95 C K T

00 50 100 150 200

Time (minute) (a)

0 0 200 400 600 800

Turbidity C (NTU) T (b)

Figure 3.5: (a) the turbidity CT of kaolinite suspensions at different kaolinite concentrations CK overtime and (b) the correlation between turbidity CT and kaolinite concentration CK suspended in tapwater with no addition of other substance.

3.4 μPIV system

The measurement of the floc geometrical characteristics is conducted by using a micro particle image velocimetry (μPIV) system that has long been used in most areas of experimental fluid mechanicsas a promising tool to detect single particle in space as well as in time, to track the flow motion and to determine the particle distribution within regions ranging from several millimeters down to a few micrometers (e.g., Grant , 1997; Santiago et al., 1998; Chakraborti et al., 2000; Maggi , 2005; Lindken et al., 2009). The μPIV system, which consists of an imaging system and an illumination system,enables the tracking of the characteristics and motion of settling SPM in an untouched environment, hence, preserving the detailed information of the SPM.

Page 20




3.4.1 Imaging system

A digital charge-coupled device (CCD) camera (Prosilica GC2450) and a high magnification lens (Navitar 12X Body Tube) are used to acquire images of settling SPM. The CCD camera and the magnification lens are mounted on a height-adjustable camera stand equipped with a cooling system to prevent over heating of the camera (Figure 3.6).

The CCD camera has a size of 2448 × 2050 pixel, 8-bit grayscale depth with a frame rate of 15

Hz at full size. The magnification lens is equipped with a continuous zoom that enables variation in

the field of view (FOV) over a wide range of magnification steps, ranging from 0.58 to 7.0. The size of the FOV and, hence, the size of a pixel at various magnification steps was determined through calibration against printed grids of known sizes (described in Chapter 3.4.3). The depth of field

(DOF) of the lens was also qualitatively determined to be approximately ± 5 mm from the focaldistance.

The CCD camera is connected to a host computer and SPM images are acquired using Image Acquisition Toolbox in Matlab. This toolbox allows the adjustment of camera properties (e.g., gain

and shutter speed) and enables automatic image acquisition at specified time.

Magnification lens Fan for cooling Aluminium plates for heat dissipation

CCD camera

Height adjustable stand

Figure 3.6: The CCD camera and the high magnification lens mounted on the height-adjustable camera stand.

3.4.2 Illumination system

A Cree LED (cool white colour) of 3.7 W, 400 lumens is used to illuminate the settling SPM. Light from the Cree LED is transported and shined directly onto the settling SPM through optical fibers that are inserted into the measuring section. The use of optical fibers enables the light to be shined directly onto the region where the measurements are to be taken without the need for the light to travel through the Perspex wall of the measuring section, hence, minimising the attenuation of the light beam.

An optical box (Figure 3.7), made of aluminium plates, was designed to hold the Cree LED and optical fibers in place. As the Cree LED radiates a great amount of heat, aluminium plates and a fan are used to dissipate the heat and control the temperature. Four optical fibers are inserted into an aluminium tube and are placed as close as possible to the bulb of the Cree LED so as to capture the maximum light intensity.

The other ends of the optical fibers are inserted into the measuring section and are fixed onto a holder (Figure 3.8a) attached to a positioning stand (Figure 3.8b). The optical fibers are tightly

fastened between two plastic plates with one of the plates having round grooves cut at different angles

Page 21




to concentrate the light beams from the optical fibers into a point where the region of measurement is. A mask of less than 1 mm opening is attached to the plastic plates to create a thin vertical light sheet. The SPM driver is attached to the optical fibers holder approximately 5 mm from the mask to channel the flow of settling SPM (as described in Chapter 3.1.3).

(FigTuhree 3o.p8tbic)aml fiabde osfhaol9d0ermismhe×ld2i8n0thmemmveearstuicrianlgplsaetcetiaonndbay 1a4tt0acmhimng×it 1o1n0tomampobsaitsieonpilantge.stTanhdeoptical fibers holder is positioned in such a way that the center of the 5 mm gap between the mask and the SPM driver coincides with the center of the hole on the diaphragm. This is to allow SPM to flow within the plane of measurement. The vertical position of the optical fibers holder can be adjusted with a slot on the positioning stand.

6 Aluminium plate

2 Aluminium tube for optical fibers 2 3 Support pin for aluminium tube

5 4 Cree LED

7 4

3

Side view

5 Support plate for Cree LED 6 Fan for cooling 7 Inlet of optical fibers

4 5

6 6 3

2 4

5 7 3 2

7

Perspective view Top view

Figure 3.7: The optical box for holding Cree LED and optical fibers.

2 and 3 7

6

I

6

5

I Round grooves for placing optical fiber

2 Plastic plate with grooving 3 Plastic cover

4 4 SPM driver for channeling flow stream

5 Screw hole for the attachment of the optical fibers holder onto the positioning stand

6 Screw hole for the attachment of 2 and 3

7 Mask for creating light sheet

6 2 3

(a)

Optical fibers holder 2 Screw and nut 3 Slot for the adjustment of the vertical

position of the optical fibers holder 4 Perspex plate

4 5 Perspex base 6 Plastic plate for supporting the optical

fibers holder

5

(b)

Figure 3.8: (a) the optical fibers holder, and (b) the positioning stand for holding the optical fibers holder.

Page 22




FO

V s

ize

(mm

)

Pixe

l siz

e L

(µ

m2 )

pixe

l

3.4.3 Calibration of pixel size

oTfhveieswize(FoOf aVp) iwxeitlhatprvianrtioedusgmridasgnoiffikcnatoiwonn sstiezpes,wrans gdientgerfmroimned2 bmymm×eas2urminmg toe0s.2iz5e mofmth×e fi0e.2ld5mm. The horizontal and vertical sizes of a pixel, Xpixel and Ypixel, respectively, can be determinedas,

Xpixel =

Ypixel =

XF OV

Nx

YF OV

Ny

, (3.14a)

, (3.14b)

where Nx = 2448 and Ny = 2050 are the number of pixels within the horizontal XF OV and thevertical YF OV field of view, respectively. The size of a pixel Lpixel is then taken as the averagebetween Xpixel and Ypixel.

Panels in Figure 3.9 show that the size of FOV and the size of a pixel Lpixel decreased exponentially

with increasing magnification step. At the lowest magnification step (i.e., 0.58), a pixel has a size of

4.435 μm2, whereas, at the highest magnification step (i.e., 7.0), the size of a pixel is approximately0.375 μm2. In addition, the ratio of Xpixel to Ypixel (Table 3.2) is relatively close to 1, signifying that

the pixel is approximately a square. The horizontal Xpixel, vertical Ypixel and average Lpixel pixel

sizes at each magnification step are tabulated in Table 3.2. 12 5

X

10 Y FOV

FOV 4

8 3

6

2 4

2 1

0 0 1 2 3 4 5 6 7

Magnification step (a)

0 0 1 2 3 4 5 6 7

Magnification step (b)

Figure 3.9: Functions of (a) the horizontal and vertical sizes of field of view (FOV), and (b) the pixel size Lpixel against magnification steps ranging between 0.58 and 7.0.

Page 23




Magnification steps Xpixel (μm) Ypixel (μm) Lpixel (μm2) XpixelY pixel

0.58 4.40 4.47 4.435 0.984 1.00 2.59 2.62 2.605 0.989 1.50 1.82 1.84 1.830 0.989 2.00 1.28 1.30 1.290 0.985 2.50 1.05 1.06 1.055 0.991 3.00 0.84 0.85 0.845 0.988 3.50 0.72 0.73 0.725 0.986 4.00 0.64 0.65 0.645 0.985 4.50 0.57 0.58 0.575 0.983 5.00 0.51 0.52 0.515 0.981 5.50 0.46 0.47 0.465 0.978 6.00 0.42 0.43 0.425 0.977 6.50 0.40 0.41 0.405 0.976 7.00 0.37 0.38 0.375 0.974

Table 3.2: Pixel sizes at magnification steps ranging from 0.58 to 7.0.

3.5 Micro-controlling system

A micro-controlling system consisting of an Arduino Uno, two motor shields and a screw shield are connected to a PC and are used to automate each component of the facility, including the regulation of the oscillating grid frequency, the control over the illumination system, the slider, the water pumps, and the image acquisition. A map of the pins on Arduino Uno board used to control each of these components is depicted in Table 3.3. All operations can be scheduled and carried out automatically under the supervision of a script coded in the Matlab2011b environment (Figure 3.10). All measurements, including the grid frequency and water quality parameters, are saved in a text file.

Digital pin Taken by Analog Pin Taken by

0 - A0 Grid motor (current measurement) 1 - A1 - 2 Electrical spring interrupter A2 - 3* Grid motor(speed control) A3 - 4 Slider motor (direction control) A4 - 5* Slider motor (speed control) A5 - 6* Water pumps 7 - 8 - 9* Grid motor (brake control) 10* CCD camera 11* Cree LED 12 Grid motor (direction control) 13* -

The symbol ∗ indicates the pin can be either used as digital or analog.

Table 3.3: Arduino Uno pin mapping.

Page 24




Automatic control based on scheduling

Control of grid frequency

Control of slider system

Control of illumination system

Real-time monitoring of water quality parameters and grid frequency

Control of imaging system

Control of slider cleaning system

Figure 3.10: The interface built in Matlab2011b environment used for monitoring and controlling all components of the facility.

Page 25

Chapter



Conclusion

This report describes in details every component of a newly built laboratory facility used for investigating flocculation processes of suspended particles. The design of this facility aimed at replicating and integrating the hydrodynamic, chemical and biological processes of natural aquatic ecosystem into one laboratory-based experiment. This aim was met by adopting the use of an oscillating grid to generate isotropic and homogeneous turbulence and a water quality meter to control and monitor the chemical and biological processes in the control volume. The geometrical characteristics of suspended particles are fully preserved by using a μPIV system to acquire images of settling particles.With the use of a micro-controlling system, all parameters and measurements can be monitored and controlled simultaneously, instantaneously and automatically. This facility is expected to contribute to the discovering of the fundamental link that bridges the understanding of physical, chemical and biological interactions in waters and the flocculation dynamics of suspended particles.

Page 26

Chapter



Photographs of the facility

Turbulence generating system

Water quality measuring system

Settling column

!JPTV system

Micro-controlling system

Figure 5.1: Photograph of the experimental facility.

Page 27


separator

Vertical Motor that controls the opening and closing of the slider

Fishing line Slider

Diaphragm

Slope

Optical fibers inlets

Bubbles remover

Flocculation section Measuring section

Figure 5.2: Components of the settling column, including the flocculation section, the measuring section and the diaphragm.

Figure 5.3: The multi-parameter water quality meter (TOA-DKK, WQC-24) equipped with Standard and Ion modules.

Page 28


(a)

Wheel

Piston

Tension spring

Electrical spring interrupter

Hollow tube for minimizing horizontal and rotational oscillations

(b) (c)

Figure 5.4: Photographs of (a) an grid element, (b) evenly spaced grid elements connected through a stainless steel bar, and (c) the grid oscillation system.

Magnification lens

Fan for cooling

Aluminium plates for heat dissipation

CCD camera

Height adjustable stand

Figure 5.5: Photograph of the CCD camera and the high magnification lens mounted on the height- adjustable camera stand.

Page 29


(a) (b)

(c)

Figure 5.6: Photographs of (a) a Cree LED, (b) the optical box, and (c) the optical fibers holder fixed on the positioning stand.

(a) Arduino Uno (b) Motor shield (c) Screw shield

Figure 5.7: Photographs of (a) the Arduino Uno board, (b) the motor shield, and (c) the screw shield used in the micro-controlling system.

Page 30



Chapter 6

Concept drawings of the facility

Page 31



.

fLoc;GULATo

I

/

I rn 16

/

·I" [fj -It .... " f:w Iff

•Dt l

"' .... Tl,.. 'l"r*"- !&J,.....

I

A H.OS /

I±= · 1+2 Dl I z.

' ' '

)

Page 32



/

;i qub. 1- · ---1 l·

7 -I- 2•

,, ---.....

'

-;;?'....

tv

f I

/ ,

'>

.......

I ·" !c-.

Page 33



!- I

Page 34



- -

Page 35



/

----

/

jv--.\k

---il - l- .t-;1Je9 l \

f--

I (Z) \1. 2> b: 1t ct-U

r-1---+-1

-

-

Page 36



kl )

'

i,,..l' l

I l1)

I'--,

z,.

;::::=:

_.J

J / - {

I I "'-..

I .,r-· A l .(ip I·-

•-]· -:-- 4>- t;.l!.

1-.>ho-

,/'\/

[1jj

,.,,.,_

1---'

" '-

Page 37


School of Civil Engineering The University of Sydney Research Report R952 Page 38



u

"' c

.J (!)

4

.!3

(U \)

:)

d. e ,J

.!

18, t

.Q .........

<..._t:' l

<:::/.

lL

clo( "I"V- 4 ..J ()

:r .._j "'-

c:e:.. ft.

"" 1 -1--

i>

Page 39

A LABORATORY FACILITY FORFLOCCULATION-RELATED EXPERIMENTS


Acknowledgement

The authors are grateful to the technical support team in the School of Civil Engineering at the University of Sydney, in particulary, Mr. Garry Towell, Mr. Theo Gresley-Daines, Mr. Ross Baker and Mr. Henry Black, for their assistance in constructing this experimental facility.

Page 40

A LABORATORY FACILITY FORFLOCCULATION-RELATED EXPERIMENTS


Bibliography

Bache D.H. and E. Rasool, (1996). Measurement of the rate of energy dissipation around an oscillating grid by an energy balance approach. The Chemical Engineering Journal and The Biochemical Engineering Journal, 63(2), 105-115.

Batchelor G.K., (1953). The theory of homogeneous turbulence. Cambridge University.

Chakraborti R.K., J.F. Atkinson and J.E. Van Benschoten, (2000). Characterization of alum floc by image analysis. Environmental science and technology, 34(18), 3969-3976.

Chanson H., Takeuchi M. and Trevethan M., (2008). Using turbidity and acoustic backscatter intensity as surrogate measures of suspended sediment concentration in a small subtropical estuary. Journal of environmental management, 88(4), 1406-1416.

Corrsin S., (1963). Turbulence: experimental methods. Handbuch der Physik, 8(2), 524-533.

Droppo I. G., (2001). Rethinking what constitutes suspended sediment. Hydrol. Process., 15, 1551-

1564. Fernando H.J.S. and I.P.D. De Silva, (1993). Note on secondary flows in oscillatinggrid, mixingbox

experiments. Physics of Fluids A: Fluid Dynamics (1989-1993), 5(7), 1849-1851.

Ghozlani B., Z. Hafsia, and K. Maalel, (2012). Numerical study of flow around an oscillating diamond prism and circular cylinder at low Keulegan-Carpenter number. Journal of Hydrodynamics, Ser. B, 24(5), 767-775.

Gippel C.J., (1995). Potential of turbidity monitoring for measuring the transport of suspended solids in streams. Hydrological processes, 9(1), 83-97.

Grant I., (1997). Particle image velocimetry: a review. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 211(1), 55-76.

Gratiot N., H. Michallet, and M. Mory, (2005). On the determination of the settling flux of cohesive sediments in a turbulent fluid. Journal of Geophysical Research: Oceans, 110(C6).

Grossart H. P., T. Kiorboe, K. W. Tang, M. Allgaier, E. M. Yam and H. Ploug, (2006). Interactions between marine snow and heterotrophic bacteria: aggregate formation and microbial dynamics. Aquatic microbial ecology, 42, 19-26.

Hopfinger E.J. and J.A. Toly, (1976). Spatially decaying turbulence and its relation to mixing across density interfaces. Journal of Fluid Mechanics, 78(01), 155-175.

Karickhoff S. W., D. S. Brown, and T. A. Scott, (1979). Sorption of hydrophobic pollutants on natural sediments. Water research, 13(3), 241-248.

Page 41



Kiorboe T., (2001). Formation and fate of marine snow: small-scale processes with large-scale im- plications. Scientia Marina, 65(2), 57-71.

Kranck K., (1973). Flocculation of suspended sediment in the sea. Nature, 246, 348-350.

Lawler D.M. and Brown R.M., (1992). A simple and inexpensive turbidity meter for the estimation of suspended sediment concentrations. Hydrological processes, 6(2), 159-168.

Lick W. and V. Rapaka, (1996). A quantitative analysis of the dynamics of the sorption of hydrophobic organic chemicals to suspended sediments. Environmental Toxicology and Chemistry, 15(7), 1038-1048.

Liem L.E., D.W. Smith, and S.J. Stanley, (1999). Turbulent velocity in flocculation by means of grids. Journal of Environmental Engineering, 125(3), 224-233.

Lindken R., M. Rossi, S. Grobe, and J. Westerweel, (2009). Micro-particle image velocimetry (PIV): recent developments, applications, and guidelines. Lab on a Chip, 9(17), 2551-2567.

Maggi F., (2005). Flocculation dynamics of cohesive sediments. Ph.D. Thesis, Delft Universitiy of Technology, The Netherlands.

Matsunaga N., Y. Sugihara, T. Komatsu and A. Masuda, (1999). Quantitative properties of oscillating-grid turbulence in a homogeneous fluid. Fluid dynamics research, 25(3), 147-165.

McCave I.N., (1984). Size spectra and aggregation of suspended particles in the deep ocean. Deep Sea Research Part A. Oceanographic Research Papers, 31(4), 329-352.

Mietta F., C. Chassagne, A.J. Manning and J.C. Winterwerp, (2009). Influence of shear rate, organic matter content, pH and salinity on mud flocculation. Ocean Dynamics, 59(5), 751-763.

Noh Y. and H.J. Fernando, (1993). The role of molecular diffusion in the deepening of the mixed layer. Dynamics of atmospheres and oceans, 17(2), 187-215.

Ongley E.D., M.C. Bynoe and J.B. Percival, (1981). Physical and geochemical characteristics of suspended solids, Wilton Creek, Ontario. Canadian Journal of Earth Sciences, 18(8), 1365-1379.

Riley S.J., (1998). The sediment concentration-turbidity relation: its value in monitoring at Ranger Uranium Mine, Northern Territory, Australia. Catena, 32(1), 1-14.

Santiago J.G., S.T. Wereley, C.D. Meinhart, D.J. Beebe and R.J. Adrian, (1998). A particle image velocimetry system for microfluidics. Experiments in fluids, 25(4), 316-319.

Schofield R.K. and H.R. Samson, (1954). Flocculation of kaolinite due to the attraction of oppositely charged crystal faces. Discuss. Faraday Soc., 18, 135-145.

Smith D.G. and Davies-Colley R.J., (2002). If visual water clarity is the issue, then why not measure it. In National Water Quality Monitoring Council National Monitoring Conference 2002, Madison, Wisconsin.

Tennekes H. and J.L. Lumley, (1972). A first course in turbulence. MIT press.

Thompson S.M. and J.S. Turner, (1975). Mixing across an interface due to turbulence generated by an oscillating grid. Journal of Fluid Mechanics, 67(02), 349-368.

Page 42



Tombacz E. and M. Szekeres, (2004). Colloidal behavior of aqueous montmorillonite suspensions: the specific role of pH in the presence of indifferent electrolytes. Applied Clay Science, 27(1), 75-94.

Tonui N. and D. Sumner, (2011). Flow around impulsively started square prisms. Journal of Fluids and Structures, 27(1), 62-75.

Tye R., R. Jepsen and W. Lick, (1996). Effects of colloids, flocculation, particle size and organic matter on the adsorption of hexachlorobenzene to sediments. Environmental Toxicology and Chemistry, 15(5),643-651.

van Leussen W., (1988). Aggregation of particles, settling velocity of mud flocs a review. In Physical processes in estuaries, (pp. 347-403). Springer Berlin Heidelberg.

van Leussen W., (1994), Estuarine Macroflocs, Ph.D. Thesis, University of Utrecht, The Netherlands.

Winterwerp J.C., (1999). On the flocculation and settling velocity of estuarine mud. Continental shelf research, 22(9), 1339-1360.

Wolanski E., R.J. Gibbs, Y. Mazda, A. Mehta and B. King, (1992). The role of turbulence in the settling of mud flocs. Journal of Coastal Research, 35-46.

Page 43

Documents

A Laboratory Facility for Flocculation-Related Experiments