Erosional behavior of deposited cohesive sediments

UFL/COEL-TR/071

EROSIONAL BEHAVIOR OF DEPOSITED COHESIVE
SEDIMENTS

by

T. M. Parchure

Dissertation

1984

EROSIONAL BEHAVIOR OF
DEPOSITED COHESIVE SEDIMENTS

BY

T.M.PARCHURE

A DISSERTATION PRESENTED TO THE
GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1984

I

REPORT DOCUMENTATION PAGE
1. Report No. 2. 3. Recipient's Accession No.

4. Title and Subtitle 5. Report Date
1984
Erosional Behavior of Deposited Cohesive Sediments
6.

7. Author(s) 8. Performing Organization Report No.
T. M. Parchure UFL/COEL-TR-071

9. Performing Organization Name and Address 10. Project/Task/lork Unit No.
Coastal and Oceanographic Engineering Department
University of Florida 11. Contract or Grant No.
336 Weil Hall R806684010
Gainesville, FL 32611 13. Type of Report
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency Final
and
Special Appropriation by
Florida State Legislature 14.
15. Supplementary Notes

16. Abstract

Erosional characteristics of fine cohesive sediments must be known for predicting
the fate of sorbed pollutants in the estuarine environment. The surficial layers of
fine sediment deposits in estuaries have a consolidation time typically ranging from a
few hours to a week. The flow-induced bed shear stress, Tb, and the bed shear
strength, Ts, are the two parameters that determine the erosional behavior of such
deposits under a constant temperature. Erosion rates were measured at different bed
shear stresses under four variable factors, namely, the type of sediment, the electro-
chemical composition of the eroding fluid, the consolidation period and the effect of
microorganisms. The effect of these factors on the rate of erosion is a result of the
corresponding influence of these factors on Ts. Under a constant Tb, a decrease in the
rate of erosion of deposited beds with time was observed. The reason for this behavior
is that erosion decreases as Ts increases with depth, rapidly at first and more slowly
in the lower layers. This increase in Ts is believed to be primarily due to a change
in the structure of the aggregates caused by consolidation. A three-zoned bed
structure has been proposed to characterize the depth-variation of Ts. Finally, an
expression for the rate of erosion has been proposed. The relationship between the
logarithm of erosion rate and the excess shear stress, Tb Ts, in this expression has
been explained on the basis of rate process theory involving the activation energy
concept.

Y

17. Originator's Key Words
Annular flume
Cohesive sediment
Erosion
Rate of erosion
Resuspension

19. U. S. Security Classif. of the Report

Unclassified

18. Availability Statement

20. U. S. Security Classif. of This Page
Unclassified

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...............................................

LIST OF TABLES................ ....... ......... .. ..............

LIST. OF F.IGURES.... ............-.................................

KEY TO SYMBOLS.................................................

ABSTRACT..... ................ ....... ....... .....................

CHAPTER

1 INTRODUCTION.. ..........................................

1.1 Importance of Cohesive Sediment Studies...........
1.2 Fine Sediment Erosion in the Estuarine Environment.
1.3 Objectives and Scope of the Present Study...........
1.4 Outline of Presentation.............................

2 FINE SEDIMENT PROPERTIES AND TRANSPORT PROCESSES.........

2.1 Fine Sediments ............................. .......
2.2 Composition and Structure of Clay Minerals.........
2.3 Clay-Water Interaction.................. ........
2.4 The Diffuse Double Layer...........................
2.5 Sediment Pollutant Relationship.................
2.6 Estuarial Fine Sediment Transport Processes.........

2.6.1 Source of Sediment...........................
2.6.2 Flocculation................................
2.6.3 Settling and Deposition.....................
2.6.4 Consolidation............................ ..
2.6.5 Erosion.....................................
2.6.6 Convective and Diffusive Transport...........

2.7 Mathematical Modeling..............................

3 PHYSICAL AND PHYSICO-CHEMICAL FACTORS
IN COHESIVE SEDIMENT EROSION..............................

3.1 Introduction......................................

Page

ii

vii

ix

xiii

xvii

1

1
3
7
8

10

10
12
13
14
16
20

20
20
24
27
28
28

29

35

35

3.2 Flow-rdlated Factors............................... 35

3.2.1 Bed Shear Stress........................... 35
3.2.2 Effect of Suspended Sediment on Turbulence.. 38

3.3 Sediment-related Factors.......................... 40

3.3.1 Soil Mechanical Indices.................... 40
3.3.2 Sediment Composition........................... 41
3.3.3 Organic Matter.............................. 41
3.3.4 Cation Exchange Capacity.................... 43
3.3.5 Dielectric Dispersion....................... 43

3.4 Fluid-related Factors............................. 44

3.4.1 Sodium Adsorption Ratio (SAR)............... 44
3.4.2 Combined Effect of SAR and CEC.............. 44
3.4.3 Temperature ............... 45
3.4.4 pH.......................................... 46
3.4.5 Pore Fluid.................................. 47

3.5 Bed-related Factors................................. 47

3.5.1 Bed Density and Shear Strength.............. 47
3.5.2 Water Content...... ......................... 48
3.5.3 Critical Bed Shear Stress................... 48
3.5.4 Characteristic Bed Shear Stress............. 52

4 PREVIOUS STUDIES ON EROSION............................. 54

4.1 Introduction...................................... 54
4.2 Soil Erosion............................ .... 55
4.3 Types of Beds in Erosion Studies................... 57
4.4 Procedures in Erosion studies.................... 59
4.5 Previous Studies at the University of Florida...... 60

4.5.1 Objective................... ........... ..... 60
4.5.2 Apparatus.................................. 61
4.5.3 Methodology............................... 66
4.5.4 Results of Studies under Series 1........... 68
4.5.5 Results of Studies under Series 2........... 69
4.5.6 Results of Studies under Series 3........... 70
4.5.7 Results of Studies under Series 4........... 72

4.6 Conclusions of Chapter 4........................... 77

5 EROSION MECHANISM AND BED STRUCTURE .................... 81

5.1 Introduction....................................... 81

5.2 Erosion Mechanism................................. 83

5.2.1 Concentration-time Relationships............ 83
5.2.2 Exchange of Sediment: Background Information 90
5.2.3 Investigation of Exchange of Sediment....... 95
5.2.4 Significance of Characteristic
Bed Shear Stress............................ 101

5.3 Bed Density Measurements........................... 111

5.3.1 Variation of Bed Density................. 111
5.3.2 Depth-averaged Bed Density ................. 112
5.3.3 Depth-variation of Bed Density.............. 114

5.4 Investigations for Shear Strength.................. 121

5.4.1 Measurement of.Shear Strength ............... 121
5.4.2 Minimum Shear Strength...................... 125
5.4.3 Estimation of the Maximum Shear Strength.... 128
5.4.4 Depth-variation of Shear Strength........... 134
5.4.5 Time-variation of Shear Strength............ 142
5.4.6 Depth-averaged Shear Strength
of Upper Zone.............................. 148
5.4.7 Depth-variation of Primary Particle Size.... 151

5.5 Conclusions of Chapter 5........................... 157

6 RATE OF EROSION......................................... 163

6.1 Introduction....................................... 163
6.2 Erosion Tests..................................... 164
6.3 Significance of the Empirical Parameters
E*i EF and a.................................. 167
6.4 The Floc Erosion Rate............................. 174
6.5 Factors. Influencing the Rate of Erosion............. 185
6.6 Erosional Properties of Kaolinite.................. 190
6.7 Review of the Rate Process Theory ................ 195

6.7.1 Background Information...................... 195
6.7.2 Application to Soil Creep................... 197
6.7.3 Application to Soil Erosion................. 199
6.7.4 Application to Cohesive
Sediment Bed Erosion....................... 200

6.8 Erosion Rate Expression......................... 203

6.8.1 Present Study............................... 203
6.8.2 Discussion................................ 213

6.9 Conclusions of Chapter 6........................... 220

I

7 LABORATORY MICROCOSM STUDY OF BIOLOGICAL FACTORS IN
COHESIVE SEDIMENT EROSION.............................. 222

7.1 Objective and Scope.............................. 222
7.2 Use of Microcosms................................ 222
7.3 Interaction Between Organisms & Sediments.......... 225
7.4 Effect of Microorganisms on Sediments.............. 226
7.5 Interaction between Microorganisms and Sediments... 228
7.6 Influence of Microorganisms on Soil
Cohesion and Erosion.............................. 231
7.7 Preliminary Study................................. 233

7.7.1 Preliminary Study 1......................... 233
7.7.2 Preliminary Study 2......................... 236

7.8 Rotating Channel Study...................... ... 238

7.8.1 Experimental Arrangement.................... 238
7.8.2 Experiment KN............................ 242
7.8.3 Experiment KM1.............................. 243
7.8.4 Experiment KM2 .............................. 248
7.8.5 Discussion ................................ 248
7.8.6 Conclusions of Chapter 7.................... 250

8 SUMMARY AND CONCLUSIONS............................... 253

8.1 Bed Shear Strength.............................. 253

8.1.1 Critical Shear Strength, .............. 253
8.1.2 Characteristic Shear Strength, ........ 255
8.1.3 Depth-variation of Shear Strength......... 256
8.1.4 Time-variation of Shear Strength.......... 257
8.1.5 Particle Sorting during Deposition........ 258

8.2 Bed Density........................ ................ 259

8.2.1 Depth-variation of Bed Density.... ........ 259
8.2.2 Time-variation of Bed Density ............. 260

8.3 Erosion Mechanism.. ........................... 261
8.4 Erosion Rate............. ....... ........ 262
8.5 Effect of Microorganisms......................... 264

APPENDIX

A. CHARACTERIZATION OF SEDIMENTS, FLUIDS, AND
NUTRIENTS USED FOR STUDY.............................. 266
B. DATA AND ANALYSIS OF EROSION TESTS................... 271
C. RHEOLOGICAL CONSIDERATIONS FOR BED STRUCTURE.......... 296
D. SELECTION OF CULTURE MEDIUM ................. ..... 301
E. LIST OF PREDOMINANT MICROORGANISMS COLONIZING
THE FINE SEDIMENT SUBSTRATE IN THE ROTATING CHANNEL... 303
REFERENCES ........ .. ............................ 304

BIOGRAPHICAL SKETCH ................................... 322

LIST OF TABLES

Page

2-1 Relationship of Valence and Atomic Number with Cation
Replaceability...................... ................. 17
2-2 Valence and Atomic Number of Common Toxic Metals.......... 17
5-1 Values of Characteristic Bed Shear Stress, Tch, for
Various Types of Beds.................. .................. 102
5-2 Characteristic Shear Strength, Tch and Characteristic
Depth zch Observed for beds in Series LM, KS and KT....... 108
5-3 Comparison of p/p Values for the Series LM, KS and KT..... 117
5-4 Comparison of Results of Normalized Bed Density at Bed
Surface................................................... 119
5-5 Comparison of measured and estimated Values of Bed
Density for Consolidation Time of 1.7 days................ 120
5-6 Values of Critical Shear Stress for Various Beds.......... 127
5-7 Estimated Maximum Values of Shear Strength for Beds
in Series LM, KS and KT................................... 129
5-8 Comparison of White River Sediment with Lake mud
Properties...................... ............. 139
5-9 Order of Aggregation versus Shear Strength for
White River Sediment (Krone, 1963)........................ 140
5-10 Depth-variation of Shear Strength of Lake Mud............. 140
5-11 Properties of surfacial Layer of Deposited Beds in
Series LM, KS and KT.................................... 150
5-12 Depth-variation of Primary Particle Size in Deposited
Beds of Kaolinite and Kaolinite plus Silt................. 155
6-1 Summary of Experimental Conditions for Erosion Tests......... 165
6-2 Qualitative Description of Depth-variation of Shear
Strength for a Deposited Bed ............................. 175
6-3 Depositional Properties of Kaolinite in Distilled
Water and Kaolinite in Tap Water Observed by
Mehta (1973)...................................... ........ 194
6-4 Settling Volume of Kaolinite in Water with Varying
Concentration of NaC1 observed by Hsi and Clifton (1962).. 195
6-5 Average Floc Erosion Rates c and Correlation
Coefficients for Sediment Beds used in the Present
Study..................... ........... .... ......... 210
6-6 Average Floc Erosion Rate Ef obtained by Reanylysis
of Data of Various Investigators......................... 214
6-7 Erosion Rate Expressions................................ 218
6-8 Nomenclature used in Table 6-7........................... 219
7-1 Culture Media in Preliminary Study 2...................... 237
7-2 Electrophoretic Mobility of Kaolinite Particles........... 247
7-3 Erosion Rates of the Beds KT, KN and KM ................... 249
7-4 Thickness of the Upper Layer for Beds KT, KN and KM....... 250

8-1 Properties of Surficial Layer of Deposited Beds............ 254
8-2 Comparison of p/p Values for Series LM, KS and KT......... 260
8-3 Values of 0 and f for the Series LM, KS, and KT.......... 263
f

viii

LIST OF FIGURES

Page
1-1 Variation of Heavy Metal Concentration with Sediment
Particle Size (Based on the Data of Salmons and
Mook, 1977) ....................... ... ................. 2
1-2 Schematic Representation of the Physical States
of Fine Sediments in Estuaries (Mehta et al., 1982)....... 4
3-1 Schematic Representation of Erosion Rate as a
Function of Bed Shear Stress.............................. 49
4-1 Schematic Viewof the Annular Flume Facility.............. 62
4-2 (a) Photograph of the Rotating Channel Facility........... 64
4-2 (b) Photograph of the Control Equipment.................. 65
4-3 Schematic Representation of Procedure used for
Bed Preparation and Erosion Tests in Experimental
Series 3 and 4 ........................................... 66
4-4 Relative Suspended Sediment Concentration against
Time, t, for Erosion of a Deposited Bed using Kaolinite
in Distilled Water at Tb = 0.207 N/m2. (Based on the
Data of Mehta and Partheniades, 1979) .................... 69
4-5 Suspended Sediment Concentration, C, against Time,
t, for Erosion of a Uniform Bed using Kaolinite
in Distilled Water at Tb = 0.413 N/m2 (Based on the
Data of Mehta and Partheniades, 1979)..................... 69
4-6 Suspended Sediment Concentration versus Time Observed
by Yeh (1979) for Erosion of Deposited Kaolinite
Beds using Distilled Water.......... .. ......... ....... 73
4-7 Variation of Suspended Sediment Concentration With
Time during a Test using Kaolinite in Salt Water
following the Experimental Methodology shown in
Figure 4-3 (Parchure, 1980)...................... ... 73
4-8 Apparatus for Measurement of Density as a Function
of Depth for Deposited Beds (Parchure, 1980).............. 74
4-9 Variation of Bed Density, p with Depth, z, in
Test corresponding to Figure 4-7.......................... 75
4-10 Variation of Bed Shear Strength, Ts with Depth z,
in Test corresponding to Figure 4-7....................... 75
4-11 Suspension Concentration versus Bed Shear Stress
for Different Flow Deposited Beds (Parchure, 1980)........ 76
4-12 Variation of Tch with Consolidation Time (Dixit, 1982).... 78
4-13 Variation of ln (E/6) with (T Ts)/T for a Test
using Kaolinite in Tap Water with T = 48 Hours
(Based on the data of Mehta and Partheniades, 1982)....... 78
5-1 Schematic Representation of Type I and Type II
Concentration-time Profiles and the Depth-variation
of Shear Strength for Each Type........................ 84
5-2 Examples of Concentration-time Profiles of type I

(Based on the Data of Partheniades, 1962; and
Fukuda, 1978) ......... ...... ......................... 85
5-3 Examples of Concentration-time Profiles of
Type I (Based on the Data of Lee, 1979;
Parchure, 1980;.Thorn and Parsons, 1980).................. 86
5-4 Examples of Concentration-time profiles of
Type II for Erosion of Uniform Beds (Based on the
Data of Partheniades, 1962; and Alizadeh, 1974)........... 87
5-5 Suspended Sediment Concentration in Flume during
Deposition (Based on the Data of Einstein and Krone, 1962) 94
5-6 Results of Erosion Tests conducted by Partheniades (1962). 94
5-7 Erosion of KT Series Bed (Tdc = 1 Day) with Tb=0.2 N/m2... 97
5-8 Suspension Concentration in the Channel as a Function
of Time, during and after Flushing of Suspended
Sediment .................................................. 97
5-9 Suspension Concentration in the Rotating Channel
during Flushing of Suspended Sediment..................... 99
5-10 CT as a Function of Tb for the Bed KTB1 with
(Arbn)= 0.2...................... ......................... 104
5-11 CT as a Function of Tb for the Bed KTB1 with
(ATb ). = 0.35............................................ 104
5-12 CT as a Function of Tb for the Bed KTB1 with
with (ATbn)* = 0.5................................ 105
5-13 CT as a Function of Tb for the Bed KTB1 with
(ATbn)* = 0.75........................................ ... 105
5-14 CT as a Function of Tb for the Beds KTBl and KTB8......... 107
5-15 Variation of Depth-averaged Bed Density as a Function
of Tdc for Series KT and KS............................... 113
5-16 Variation of Depth-averaged Bed Density as a Function
of Salinity for Series LM................... .......... 113
5-17 Effect of Initial Concentration of Suspension on the
Depth-averaged Bed Density for Various Values of Tdc...... 115
5-18 p/p as a Function of z/ho for Series LM ................... 115
5-19 p/P as a Function of z/ho for Series KS.................. 116
5-20 p/p as a Function of z/ho for Series KT.................. 116
5-21 Schematic Representation of a Typical Erosion Experiment.. 123
5-22 Correlation of Empirical Parameter a and Bed
Shear Strength Ts for Series LM.......................... 130
5-23 Correlation of Empirical Parameter a and Bed
Shear Strength Ts for Series KS.......................... 131
5-24 Correlation of Empirical Parameter a and Bed
Shear Strength Ts for Series KT.......................... 132
5-25 Bed Shear Strength T as a Function of
Depth for Series LM......................... .............. 135
5-26 Depth-variation of Bed Shear Strength for Series LM....... 136
5-27 Depth-variation of Bed Shear Strength for Series KS....... 136
5-28 Depth-variation of Bed Shear Strength for Series KT....... 137
5-29 Concentration-time Relationships for Erosion of Beds
KTB1 and KTB8 with a Bed Shear Stress of 0.6 N/m2....... 144
5-30 Erosion Rate as a Function of Time for Kaolinite
Beds in Tap Water with Tdc = 1 Day and 8 Days
under a Bed Shear Stress of 0.6 N/m ...................... 145
5-31 Time-variation of Bed Shear Strength from
Tdc = 1 Day to 8 Days..................................... 147

5-32 Average Bed Shear Strength over Zone 1 as a Function
of Tdc for Series KS and KT................... ....... .149
5-33 Average Bed Shear Strength over Zone 1 as a Function
of Salinity for Series LM ................................. 149
5-34 Particle Size Distribution at Various Depths for
a Deposited Kaolinite Bed in Tap Water.................... 156
5-35 Particle Size Distribution at Various Depths for
a Deposited Bed Containing Silt and Kaolinite in
Tap Water ......................... ..... ........... 156
5-36 Schematic Representation of the Proposed Three-zoned
Stracture for the Depth-variation of Shear Strength
of Deposited Beds......................................... 160
6-1 Schematic Representation of Empirical Parameters
e, and X...................................... 168
6-2 Schematic Representation of Empirical Parameters
ef and a.................................................. 168
6-3 Correlation of..c* -and .(-b Ts) for. Series LM............. 170
6-4 Correlation of Empirical Parameters X and a for
Series LM........................ ................... 171
6-5 Concentration-time Relationship for Erosion of Bed
in Series KT with Tdc = 1 Day and Tb = 0.6 N/2........... 177
6-6 Concentration-time Relationship for Erosion of
Bed in Series KT with Tc = 8 days andTb=0.2 N/m ......... 178
6-7 Erosion Rate as a Function of Time for the Bed
with Tdc = 1 Day in Series KT ............................. 179
6-8 Erosion Rate as a Function of Time for the Bed
with Tdc = 8 Days in Series KT................... ........... 180
6-9 Concentration-time Relationships for an Erosion
Experiment with Long Durations of Time Steps............... 181
6-10 Erosion Rate as a Function of Time for a Bed
in Series KS with Tdc = 1.7 Days...................... .... 182
6-11 Effect of Salinity on the Overall Erosion Rate
E60 for the Series LM....................................... 186
6-12 Influence of Salinity on the Erosion Rate of
Kaolinite with Tdc.= 3 Days............................... 186
6-13 Influence of Tdc on the Erosion Rate of
Kaolinite in Salt Water of 35 ppt Concentration .......... 188
6-14 Influence of Tdc on the Erosion Rate of
Kaolinite Beds in Tap Water............................... 188
6-15 Influence of the Type of Sediment and the
Salinity of Fluid on the Erosion Rate..................... 192
6-16 Salt and Non-salt Flocculation of Kaolinite Particles..... 192
6-17 Energy Change in Chemical Reaction...................... 196
6-18 Strain Rate as a function of 1/T for undisturbed
San Francisco Bay Mud (Based on the Data of Mitchell
et al., 1968) ...................... ..................... 198
6-19 Variation of Erosion Rate with inverse of
Absolute Temperature (Based on the Data of Kelly
and Gularte, 1981)....................................... 202
6-20 Effect of Temperature Change on the creep of
undisturbed San Francisco Bay Mud (Based on the
Data of Mitchell et al., 1968)............................ 202
6-21 Rate of Erosion for Series LM............................. 205
6-22 Rate of Erosion for Series KS............................. 206

6-23 Rate of Erosion for Series KT............................. 207
6-24 Non-dimensional plot of Erosion Rates for Series
LM, KS and KT................................... .......... 208
6-25 Non-dimensional plot of Erosion Rates based on
reanalyzed Data of Various Investigators.................. 212
6-26 Correlation of Parameterf with Floc Erosion Rate 216
7-1 Microcosm Arrangement for Preliminary Studies............ 235
7-2 Photograph of Experimental Facility for
Laboratory Microcosm Study .............................. 240
7-3 Photograph of Equipment for Laboratory Microcosm Study.... 241
7-4 Results of Erosion Experiments for the Beds KT, KN and KM. 244
B-l Concentration-time Relationship for Erosion Test No. 1.... 272
B-2 Concentration-time Relationship for Erosion Test No. 2.... 273
B-3 Concentration-time Relationship for Erosion Test No. 3.... 274
B-4 Concentration-time Relationship for Erosion Test No. 4.... 275
B-5 Concentration-time Relationship for Erosion Test No. 5.... 276
B-6 Concentration-time Relationship for Erosion Test No. 6.... 277
B-7 Concentration-time Relationship for Erosion Test No. 7.... 278
B-8 Concentration-time Relationship for Erosion Test No. 8.... 279
B-9 Concentration-time Relationship for Erosion Test No. 9.... 280
B-10 Concentration-time Relationship for Erosion Test No.10.... 281
B-1 Concentration-time Relationship for Erosion Test No.ll.... 282
B-12 Concentration-time Relationship for Erosion Test No.12 ... 283
B-13 Concentration-time Relationship for Erosion Test No.13.... 284
B-14 Concentration-time Relationship for Erosion Test No.14.... 285
B-15 Erosion Rate versus Time for LM Series Test No. 1......... 286
B-16 Erosion Rate versus Time for LM Series Test No. 2......... 287
B-17 Erosion Rate versus Time for LM Series Test No. 3......... 288
B-18 Erosion Rate versus Time for LM Series Test No. 4......... 289
B-19 Erosion Rate versus Time for LM Series Test No. 5........ 290
B-20 Erosion Rate versus (Tb Ts)/Tsfor LM Series Test No. i.. 291
B-20 Erosion Rate versus (Tb Ts)/Fsfor EM Series Test No. 2.. 292
B-20 Erosion Rate versus (Lb 7s)/ sfor LM Series Test No. 3.. 293
B-20 Erosion Rate versus (-b 7s)/isfor LM Series Test No. 4.. 294
B-20 Erosion Rate versus (Tb Ts)/Tsfor LM Series Test No. 5.. 295
C-1. Relative Differential Viscosities of.White River
Sediment in Salt Water From Rotating Cylinder
Viscometer Measurements (Based on the Data of
Krone, 1963) ......... .................. ................ 298

xii

KEY TO SYMBOLS

A = empirical constant
A = cross-sectional area
Ac = a constant
Ae = empirical coefficient
A'= reactants in a chemical reaction
B = empirical constant
B = empirical coefficient
B' = activated molecules
C = depth-averaged suspended sediment concentration-
C, = sediment concentration near the bed
C = coefficient of consolidation
C = constant of integration
C = initial concentration of suspension
C = steady-state concentration
C (t) = instantaneous suspension concentration at time t
C, = the fraction of the deposited sediment under a
shear stress
C' = products of a chemical reaction
oC = temperature in degrees centigrade
c = macrosopic shear strength of clay
c = capacitance
De = rate of deposition
Dr = average diameter of eroded clay particle or cluster
Ds, Dy, D .= effective turbulent dispersion coefficients

diameter
= energy
= energy
EyI Ez =

in tne x, y, diuLrecLt ons relspec.t vely
of spherical particle
of reactants A'
of activated molecules B'
turbulent diffusion coefficients in the x, y, z
directions respectively

E' = experimental activation energy
AE = energy of activation
e = void ratio
ei = initial void ratio
F = frequency factor
F' = floc shear strength
G = specific gravity of sediment
g= gravitational acceleration
h = depth of flow
h = total thickness of bed
h = Planck's constant
h9 = thickness of overburden necessary to reduce the
aggregation of lower layer by one order
i = number of time step
KB = Boltzmann's constant

xiii

d
,
,
,

Kd = constant of proportionality
K = rate of reaction
K = hydrostatic pressure ratio
S= temperature in degrees Kelvin
L = hydrodynamic lift force
L = length of specimen
Ms= erosion rate constant
m = mass of sediment
m' = a constant
N = Avogadro's Number
n = number of successive aggregates
Pd = probability of deposition
P = probability of erosion
Qe = flow rate
Qo = average sediment discharge
Q = average water discharge
q = net sediment flux
R = universal gas constant
R = hydraulic radius
S = salinity of water in parts per thousand by weight
S = slope of the energy grade line
SQ=source/ sink term
T = absolute temperature in degrees Kelvin
T = duration of time step
T = duration of deposition
c = duration of self-weight consolidation
Tc= time factor = Cf/z
T11 = lateral compressive stress
t =time
t = elapsed time
te = time at which C, is equal to 50 percent
t ) = time required for the breaking of a particle due
0 to shear stress
t' = duration of time over which the terms u, v, w and C
are averaged in a general three-dimensional convective
diffusion equation
t" = period of growth of the laminar sublayer
u, v, w = velocity components in the x, y, z directions
respectively
ui = instaneous velocity of flow in the x direction in the
turbulent region near the sublayer
u', v', w' = instantaneous velocities in the x, y, z
directions respectively
u, v, w = time-mean velocities in the x, y, z directions
respectively
V = volume of water in the channel
V = volume of flow units
= terminal fall velocity of spherical particle
W = buoyant weight of floc
W = settling velocity of a floc or a particle
Ws' = effective settling velocity = P W
X= a parameter dependent on time and structure
Y = (zo z)/z

xiv

x, y, z =Cartesian coordinate directions with the z axis in the
direction of gravity, x axis along the longitudinal
axis of the water body and y axis along the lateral
axis of the water body
z = depth below the bed surface
Zch = characteristic depth below the bed surface
z = initial bed thickness
z = depth of erosion
z' = thickness of overburden
a = empirical coefficient
a = a constant
Be= empirical coefficient
Y = specific weight of sediment
Ys = specific weight of water
6116263= empirical coefficients
63\z) = empirical coefficient as a function of z
e = rate of.erosion
e = empirical coefficient
sd = dielectric constant
Ei = erosion rate for time step
Ef = average floc erosion rate
ef. = floc erosion rate for time step
Sf = floc erosion rate
ef = erosion rate at time t
e t = erosion rate at time t + At
e = dielectric constant of vacuum
El = erosion rate under excess bed shear stress
E2 = erosion rate under excess bed shear stress
.* = initial rate of erosion
C' = rate of strain
Asd = dielectric dispersion
rn = a constant
ro = a constant
8 = empirical coefficient
X = empirical coefficient
S= dynamic viscosity of fluid
p- = dynamic viscosity of liquid
's = dynamic viscosity of suspension
v = kinematic viscosity of fluid
S= small distance above the theoretical bed at which the
bottom boundary condition is effectively applicable
p = bed density at a depth z below bed surface
Pf = density of fluid
P1 = density of liquid
P(z) = density of spherical particle
p (z) =bed density as a function of depth z below the bed surface
p = depth-averaged bed density
GI = downward stress

bl ,Tb2 = bed shear stress for erosion
Tb = time-mean value of bed shear stress
Tdl Td2 = bed shear stress for forming flow-deposited bed
Tch = characteristic bed shear stress
Te = excess shear stress = Tb- Ts

Tel = excess shear stress under Tb1
Te2 = excess shear stress under T~b
Ti = instantaneous bed shear stress
Tm = bed shear stress for initial mixing
Tr = ratio of conseactive bed shear stresses
Ts = bed shear strength
Ts (z) = bed shear strength as a function of z
Tsf = floc shear strength
Tmax = maximum bed shear strength
To = time-averaged shear stress
= /Tb* = TbTbmn
(AT ) = normalized incremental bed shear stress
(ATr = normalized excess bed shear stress
= volume fraction of solids
= fraction of suspension volume occupied by flocs
f = volume fraction of primary aggregates
p = volume fraction of n aggregates
pna- incremental volume fraction
S= durny variable

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EROSIONAL BEHAVIOR OF DEPOSITED COHESIVE SEDIMENTS

By

T. M. Parchure

April 1984

Chairman: Dr. Wayne C. Huber
Co-chairman: Dr. A. J. Mehta
Major Deptartment: Environmental Engineering Sciences

Erosional characteristics of fine cohesive sediments must be known

for predicting the fate of sorbed pollutants in the estuarine

environment. The surficial layers of fine sediment deposits in estuaries

have a consolidation time typically ranging from a few hours to a week.

The flow-induced bed shear stress, Tb and the bed shear strength, T ,

are the two parameters that determine the erosional behavior of such

deposits under a.constant temperature. Erosion rates were measured at

different bed shear stresses under four variable factors, namely, the

type of sediment, the electro-chemical composition of the eroding fluid,

the consolidation period and the effect of microorganisms. The effect of

these factors on the rate of erosion is a result of the corresponding

influence of these factors on Ts. Under a constant Tb, a decrease in the

rate of erosion of deposited beds with time was observed. The reason for

this behavior is that erosion decreases as Ts increases with depth,

rapidly at first and more slowly in the lower layers. This increase in

Ts is believed to be primarily due to a change in the structure of the
5

xvii

aggregates caused by consolidation. A three-zoned bed structure has

been proposed to characterize the depth-variation of s. Finally, an

expression for the rate of erosion has been proposed. The relationship

between the logarithm of erosion rate and the excess shear stress,

Tb Ts, in this expression has been explained on the basis of rate

process theory involving the activation energy concept.

xviii

CHAPTER 1

INTRODUCTION

1.1 Importance of Cohesive Sediment Studies

Most estuaries and several reaches of shorelines have cohesive

sediment bottoms consisting primarily of clay, fine silt and varying

quantities of organic matter. With the rapid development of estuarine

and coastal harbors and the increasing demand for deeper and wider

navigation channels to accommodate larger vessels, the need to predict

the transport of cohesive sediments has assumed special importance in

the recent years. Sediment dredged from navigation channels and harbor

areas is discharged at disposal grounds which are located as close to

the sites of dredging as possible in order to reduce hauling time and

cost. In many cases, under sufficiently strong tidal currents, a portion

of the- sediment finds its way back to the areas previously dredged, and

thus there is an increase in the rate of maintenance dredging. The

present technology of dredging and disposal of sediments is expensive,

and the quantities of sediments to be handled in a single estuary may

amount to several million tons each year. Substantial savings in the

cost of maintenance dredging may be achieved by selecting the most

stable sites for disposal of dredged material if better techniques

become available for a precise prediction of the movement of the

sediment in the area of interest.

An important aspect of fine sediments is related to water quality.

Sediments are known to sorb a major fraction of nutrients and pollutants

in the aquatic environment. Pollutants bind preferentially to different

sizes of sediments and detrital particles (Odum et al., 1969). In

general, sediments provide a permanent sink for pollutants and nutrients

contained in the overlying water column. Due to their relatively large

specific surface area and high capacity to exchange ions, fine sediments

generally sorb a greater proportion of different chemical species than

coarse sediments. This can be seen from the data presented by Salomons

and Mook (1977), Figure 1-1. An increasing amount of Zn, Cr, Pb and Cu

is sorbed to sediments with an increasing percentage of fine sediments.

0.28 i

0.24- *Zn

020

U
-0.16-

I- /0
S0.12
Pb
z

z 0.08
0

0.04
Cu

00 I I I
O 20 40 60 80 100
PERCENTAGE OF SEDIMENT FINER
THAN 16 MICRONS

Figure 1-1: Variation of Heavy Metal Concentration with Sediment
Particle Size (Based on the Data of Salomons and Mook,
1977)

Fine sediments also sorb a large number of microorganisms (Zobell and

Feltham, 1942). Thus, in a typical aquatic environment, greater

quantities of chemical pollutants as well as microorganisms are found

sorbed to the sediments than those found in water. Distribution

coefficients of species between the sediment and water may reach 105

(Duursma and Gross, 1971; Duursma and Parsi, 1976). When the bottom

sediments are disturbed and brought into suspension, the bulk of the

pollutant load may be carried via the sediments rather than the water

(Kirby and Parker, 1973; Preston et al., 1972). The relationship between

fine sediments and pollutants is described further in section 2.5.

1.2 Fine Sediment Erosion In the Estuarine Environment

Fine sediment processes occurring in the estuarine environment

consist of flocculation, settling and deposition of suspended sediment,

bed consolidation, resuspension or erosion, and horizontal transport in

suspension. These are noted in Chapter 2. Since currents in estuaries

are periodic and variable in magnitude, these processes occur in a

cyclic manner. Because of the complexity of erosion-deposition

phenomena, more than one interpretation is possible as far as any

schematic representation of these phenomena is concerned. One such

representation is shown in Figure 1-2 (Mehta et al., 1982). According to

this description, fine sediments can be considered to exist in four

states, namely, 1) a suspension in horizontal transport 2) a stationary

near-bed suspension of high density with no horizontal velocity but with

a downward velocity component; 3) a partially consolidated bed; and 4) a

settled or fully consolidated bed.

L

Suspension in Transport
I I-
Deposition Redispersion Resuspension Resuspension

Stationary Suspension
Partially
Consolidated Bed
Settled Bed

Figure 1-2: Schematic Representaion of the Physical States of Fine
Sediments in Estuaries (Mehta et al., 1982)

A stationary near-bed suspension with practically no mechanical strength
results from the settling of aggregates during transport, particularly
at times close to slack water. Under suitable conditions a skeletal
framework of aggregates develops in the lower part of the stationary
suspension which leads to the formation of a deposited bed. Due to
overburden and thixotropic rearrangement, the shear strength of such
partially consolidated deposited beds increases with time until it
attains a certain maximum value. This state of bed is referred to as the
settled bed which has a relatively low water content and void ratio, a
higher shear strength and a more stable structural configuration than
the partially consolidated bed.
Fine sediment erosion is a loose boundary phenomenon which occurs

at the bed-fluid interface and essentially involves removal of particles

5

from the bed when the flow-induced bed shear stress exceeds the

resistance to erosion characterized by the shear strength. Since a

stationary suspension has practically no mechanical strength, it is

re-entrained in the water column under a relatively small bed shear

stress. Entrainment of stationary suspension has been referred to as

redispersion, and it typically occurs shortly after current reversal

following slack water (Parker and Kirby, 1977). Entrainment from a

partially consolidated bed or from a settled bed may be referred to as

resuspension because it.involves entrainment of the sediment which was

in suspension before it became a part of the deposited bed. The terms

erosion and resuspension are used synonymously in the present study.

Redispersion of stationary suspension has not yet been investigated and

is a subject matter for further research. The settled bed has properties

similar to that of a compacted bed or a placed bed. Erosion of these

latter types of beds has been investigated to a considerable extent (see

Chapter 4). The resuspension behavior of partially consolidated

deposited beds is the subject of investigation for the present study.

Several factors influence the erosion characteristics of fine

sediment deposits. These factors, which are hydrodynamic, mineralogical,

and electro-chemical in nature, are described in Chapter 3. The present

knowledge of the influence of these factors on the properties of fine

sediments is not adequate to permit an accurate prediction of erosion

rates of different sediments as a function of several parameters

characterizing the above factors taken together. Under conditions

prevailing in nature, a variety of combinations of these hydrodynamic,

mineralogical, and electro-chemical interactions exist with various

degrees of predominance in terms of their effect on the erosion of the

1

sediment bed. As a first step towards understanding the effect of

different parameters on the rate of erosion, the effect of individual

parameters must be isolated and studied in the laboratory in detail.

The effect of each parameter could ultimately be expected to be

reflected in the properties of bed, particularly the shear strength of

the bed which in turn affects the rate of erosion under a given bed

shear stress. Hence one objective of the present study was to

investigate influence of selected important parameters on the shear

strength of the deposit. The other objective was to obtain an

appropriate expression relating the rate of erosion to the bed shear

strength.

Deposited beds of two types can be investigated in laboratory

studies. The first type is a statically deposited bed in which

deposition of suspended sediment takes place under static conditions of

no flow. The second type is a flow deposited bed in which a part of the

suspended sediment deposits under a relatively low flow velocity which

permits deposition to take place. In the present study, statically

deposited beds were primarily utilized. This is because the differences

between these two.types of beds are found.to be minor in comparison with

the effect of consolidation, which, for most practical purposes,

eliminates any difference that might exist between the properties of a

statically deposited and a flow-deposited bed, as far as bed erodibility

is concerned (Mehta et al., 1982). Preliminary studies on the erosion of

statically deposited beds were conducted by Krone (1962), Partheniades

(1962), Yeh (1979), Lee (1979), Fukuda and Lick (1980) and Thorn and

Parsons (1980). These studies have brought out the importance of the bed

shear stress, bed shear strength and various physico-chemical factors

that influence the rate of erosion. Thorn and Parsons (1980) made an

important observation that the shear strength of deposited beds varies

significantly with depth. However, they did not investigate this aspect

extensively. Furthermore, in all the studies listed above, sufficient

emphasis was not given to defining the pre-erosion stress history of the

bed, i.e. the manner in which the bed was formed. Finally, erosion rate

expressions proposed earlier did not appropriately relate the rate of

erosion to parameters characterizing bed resistance to erosion (Mehta et

al., 1982).

Another important factor which has not been so far investigated

systematically is the influence of microorganisms on the erodibility of

fine sediment deposits. Complex chemical and biological processes occur

at the sediment-fluid interface as described in Section 2.5. Also,

organisms have a variety of interactions with sediment as noted in

Chapter 7. Hence, physical and electro-chemical aspects related to the

influence of bed structure including the influence of microorganisms on

the erodibility of partially consolidated fine sediment deposits have

been specifically investigated under the present study along with

.certain other aspects.related to the elucidation of the basic erosion

mechanism as mentioned in the next section.

1.3 Objectives and Scope of the Present Study

The objectives of the present study were as follows:

1. to conduct laboratory experiments on the erosion of deposited

cohesive sediment beds,

2. to propose and verify a hypothesis concerning the bed structure in

terms of the variation of the bed shear strength with depth,

3. to study the influence of physical and electro-chemical factors

affecting the bed shear strength,

4. to study the influence of microorganisms on the erodibility of

cohesive sediment bed, and

5. to propose an erosion rate expression.

The main physical and electro-chemical factors selected for

investigating their influence on the shear strength (and hence on the

rate of erosion) were the bed shear stress, type of sediment, fluid

composition, and bed consolidation time.

The scope of the experiments was defined by the following factors:

1. All the erosion studies were conducted in a rotating annular flume

described in Chapter 4. The bed density was measured with the help

of a specially designed apparatus also described in Chapter 4.

Special facilities provided for conducting experiments involving

the effect of microorganisms are described in Chapter 7.

2.' The pore fluid and the eroding fluid were the same. This was

.achieved by equilibrating the sediment with the selected fluid.

3. Studies were limited primarily to statically deposited beds formed

in the annular flume.

4. Two sediments and three fluids were used for conducting

experiments. One was kaolinite and the other was a natural mud

from Lake Francis, Nebraska. The three fluids were 1) tap water, 2)

salt water with 1, 2, 5, 10 and 35 parts per thousand salinity and

3) reconstituted lake water. Characterization of the sediments and

the fluids is described in Appendix A.

5. The bed structure was varied mainly by changing the duration of

self-weight consolidation period.

6. The influence of microorganisms on the erodibility was examined in

experiments using kaolinite as the substrate.

1.4 Outline of Presentation

The study is presented in the following order. A description of the

sediment processes in a typical estuarine environment with particular

reference to fine sediments, and a brief description of some features of

mathematical modeling for fine sediment transport in estuaries is given

in Chapter 2. Chapter 3 describes the physical and electro-chemical

factors which influence the properties of cohesive sediments. A review

of previous erosion studies conducted at the University of Florida and

elsewhere is given in Chapter 4. Experiments conducted in order to

investigate the erosion mechanism and structure of deposited beds are

reported in Chapter 5. Experiments carried out to measure the erosion

rates of deposited beds are described in Chapter 6. Chapter 7 describes

a laboratory microcosm study to examine the influence of microorganisms

on the resistance to erosion of a kaolinite bed. Summary and conclusions

of the study are presented in Chapter 8.

CHAPTER 2

FINE SEDIMENT PROPERTIES AND TRANSPORT PROCESSES

2.1 Fine Sediments

Fine sediments are a product of weathering or hydro-thermal action

on the rock and other soil on the earth's surface. Mixtures of clays and

silts are usually called muds. The classification of fine-grained soil

either as a silt or a clay is not merely on the basis of particle size

but rather on the plasticity or non-plasticity of the material. Clays

are plastic over a wide range of water content from approximately 10 to

50 percent. They can be remolded or deformed without causing cracking,

breaking or change of volume, and retain the remolded shape. When dried,

clays have high resistance to crushing. On the other hand, silt has very

little or no plasticity and when dried, has little or no strength. The

;particles which.can get bonded to each other under suitable conditions

are called cohesive sediments.

The physical characteristics of cohesionless, non-clay soils are

determined mainly by particle size, shape, surface texture and size

distribution. The mineral composition of non-cohesive sediments

influences properties such as hardness, cleavage and resistance to

chemical attack. However, by and large the non-clay particles may be

treated as inert materials whose interactions are predominantly physical

in nature (Mitchell, 1976). The interaction of cohesive sediment

particles is electro-chemical in nature due to the presence of

electrical charge on their surfaces. Substitution of one ion for

another in the clay crystal lattice and imperfections at the edges lead

to negative charge on clay particles. The magnitude of charge varies

from 5 me milliequivalentt) to 150 me per 100 gram depending on the type

of clay. In general, all clays have a high specific surface area ranging
2
from 15 to 100 m per gram (Grim, 1968), as compared, for example, to

0.002 nr per gram for 1 mm diameter sand grains. The particles of

predominantly occurring clay minerals such as kaolinite,

montmorillonite, and illite are in the form of hexagonal flakes, whereas

those of attapulgite and halloysite are in the form of rods and tubes,

respectively.

The physico-chemical forces acting on cohesive particles could be

attractive or repulsive, and they result in floc formation under

suitable conditions when the net force is attractive. The repulsive

forces are caused by negatively charged particle faces, positively

charged cations adsorbed on particles, and osmotic pressure resulting

from the high concentration of cations near the surface of particles in

the pore water; The important attractive forces are London-van der Waals

forces which decay very rapidly with distance from the surface of the

clay particle. Other factors responsible for attraction are cation

bonds, water dipole linkage, hydrogen bond, and dipole-cation-dipole

linkage (Van Olphen, 1956). The London-van der Waals forces are a

property of the matter and hence are independent of the chemical

characteristics of the associated water. However, the electrical forces

depend strongly on the cation valence and the electrolyte concentration

in the ambient fluid. Both variables specifically affect the thickness

of the double layer (see Section 2.4). Because of very high specific

surface area, the surface forces are dominant in cohesive sediments as

against gravitational forces predominanting in non-interacting

particles. The average electro-chemical force exerted on a clay particle

is of the order of one million times the average weight of the particle

(Partheniades, 1962). The relevant properties of non-cohesive sediments

can be well described by a representative sediment size. However, the

transport properties of cohesive sediments must be described by

considering the floc as a settling unit which may be composed of a large

number of individual clay particles. The processes of floc formation,

settling and deposition are described in Section 2.6. Parameters

related to water quality such as pH, electrolyte concentration,

temperature etc. have an effect on the behavior of fine sediments.

2.2 Composition and Structure of Clay Minerals

Clays are composed of extremely small (2 microns or smaller)

crystalline particles of one or more clay minerals. Clay minerals are

essentially hydrous aluminum silicates with magnesium or iron proxying

wholly or in part for the aluminum in some minerals. Clay mineral

composition refers to the identity and relative abundance of all the

clay mineral components. The most commonly occurring clay minerals are

kaolinite, montmorillonite (or smectite), illite, chlorite and

vermiculite. The presence of even a small amount (on the order of 5

percent) of certain clay minerals such as smectite may exert a

tremendous influence on the attributes of a clay material. Clays also

contain some non-clay minerals of which quartz, dolomite, calcite,

feldspar, gibbsite and pyrite are important examples.

The atomic structure of most clay minerals involves two structural

units. One is an octahedral unit, which may be dioctahedral (called

gibbsite) having the composition A2 (OH)6 or trioctahedral (called

brucite) with the composition Mg3(OH)6. The other is a silica

tetrahedral unit having the formula Si406 (OH)4. Different clay minerals

have a layered structure with different combinations of octahedral and

tetrahedral units. For instance, kaolinite has a 1:1 arrangement giving

the formula Si4A1440 (OH)8.

2.3 Clay Water Interaction

Water held by clays is grouped into two categories. The

low-temperature water can be driven off by heating the clay to about 100

C to 150 C. The nature of low-temperature water and factors that

control its characteristics are of great importance since they largely

determine the plastic, bonding, compaction, suspension and other

properties.of clay.minerals which in turn control their behavior under a

given flow field. The high-temperature water is the OH lattice water

which is lost at temperatures above 300 C.

Although a water molecule has a zero net charge, the centers of its

positive and negative charges do not coincide, thus resulting in its

dipole character. Due to this property, water gets bonded to clays

either in a mono-molecule thick layer or in a multi-layer structure. The

plasticity of soils is attributed to the water which is attached and

held by the clay particles. Experiments performed using non-polar liquid

in place of water do not indicate plasticity, and the clay particles act

in a manner similar to those of non-interacting particles (McCarthy

1977).

Einstein and Krone (1962) indicated that a sediment concentration

higher than 10,000 ppm results in giving non-Newtonian characteristics

to the suspending fluid. Rosenquist (1961) presented evidence that clay

particles are surrounded by a thin layer of crystalline water of very

high viscosity. Martin (1962) claimed that adsorbed water densities

increase rapidly for a distance less than 10 Angstroms, and the density

may exceed 1.2 grams/cm3. Thus, the interaction of clays and water is

complex in nature. The processes occurring at the molecular level define

the properties of clays which are conspicuously different from the

properties of non-interacting sediments.

2.4 The Diffuse Double Layer

In a dry clay, adsorbed cations are tightly held by the negatively

charged clay particle surfaces. Cations in excess of those needed to

neutralize the electro-negativity of the'clay particles and their

associated anions'are present as salt precipitates. When the clay is

placed in water, the precipitated salts go into solution. Because the

adsorbed cations on the surfaces of particles are present in a much

higher concentration, there is a tendency for them to diffuse away in

order to equalize the concentration. Their freedom to do so, however, is

restricted by the negative electric field originating at the particle

surfaces. The escaping tendency due to diffusion and the opposing

electrostatic attraction lead to a certain distribution adjacent to a

clay particle. A clay particle is thus surrounded on either side by a

diffused layer of counter-ions. This layer, known as the "diffuse double

layer," plays a dominant role in the mechanical properties of clays and

clay deposits. A decrease in the thickness of the double layer reduces

the electrical repulsion, which, in turn, causes a tendency towards

flocculation. Any change in the soil-water system which expands the

double layer tends to decrease the soil strength (Lambe, 1958). At any

concentration and cation valence, the double layer contains the same

positive charge, which should be equivalent to the total negative charge

of the particle surface. An increase of the concentration of cations in

the ambient water would reduce the diffusive tendency of the cations,

since this tendency strongly increases with concentration gradients.

Therefore, the cations will come to an equilibrium position under the

combined effect of electrostatic attraction and their diffusivity closer

to the particle surface, and the thickness of the double layer will thus

be depressed. Likewise, a substitution of a monovalent cation by a

divalent cation will result in a double layer containing half as many

cations as a double layer of a monovalent cation.

The study of.the effect of gross parameters such as the type of

electrolyte or. concentration of an electrolyte (e.g. salinity) on the

erosion rates of sediment beds often yields ambiguous results because

they fail to separate out the influence of the related basic parameters

which affect the double layer thickness. Lambe (1958) has pointed out

that according to the Gouy-Chapman theory the tendency towards

flocculation is caused by increasing a) electrolyte concentration, b)

ion valence, and c) temperature and by decreasing a)dielectric constant,

b) size of hydration ion, c) pH, and d) anion adsorption. Hence Lambe

has indicated that in general, a reduction in the shear strength of a

clay will result from the following factors:

1. reduction of electrolyte concentration,

2. cation exchange from high to low valence (e.g. Ca++ to Na+T,

3. adsorption of anions (e.g.phosphates),

4. exchange from a cation of small hydrated radius to cation of large

hydrated radius(e.g. Na+ to Li+),

5. increase of dielectric constant of pore fluid,

6. decrease of temperature,

7. increase in water content.

The fact that the influence of electrolytes occurs at the molecular

scale of individual particles is also substantiated by the observations

of Banin and Lahav (1968) who pointed out that the relative size of the

average montmorillonite particle depends on the type of adsorbed ion,

and the size tends to increase systematically with the position of the

ion. The complexity of interaction of clay minerals and the electrolyte

ions has been described well by Amorocho (1961). The influence of

electrochemical parameters is described in Chapter 3.

2.5 Sediment-Pollutant Relationship

It has been mentioned in Section 1.1 that fine sediments sorb a

major fraction of contaminants in the aquatic environment. The main

reason for the adsorption process is the high capacity of clays to

exchange their cations. The replaceability of exchangeable cations

depends upon the concentration of the cations, population of

exchangeable positions, nature of the anions and most importantly, the

nature of the cation. The higher the valence of the cation, the greater

is the replacing power and the more difficult it is to displace when

already present on the clay. With the same valence, the replacing power

increases qualitatively with the atomic number of the element as can be

seen from Table 2-1 (Grim, 1968). The valence and the atomic number of

some of the common toxic metals are given in Table 2-2. All the ions

listed in Table 2-2 have a relatively high valence as well as high

atomic numbers. Hence they readily replace cations of lower valence and

lower atomic number and continue to accumulate on the surfaces of

sediment.

Table 2-1

Relationship of Valence and Atomic Number
with Cation Replaceability

Cation replaceability Li < Na < K < Mg < Rb < Co

Valence 1 1 1 2 2,3 2,3

Atomic Number 3 11 20 12 37 27

Table 2-2

Valence and Atomic Number of Common Toxic Metals

Metal ion Cr Ni Cu Zn Ag Cd Hg Pb

Valence 2,3,6 2,3 1,2 2 1 2 1,2 2,4

Atomic Number 24 28 29 30 47 48 80 82

Fine cohesive sediments also adsorb a large number of

microorganisms. Zobell and Feltham (1942) estimated the number of

bacteria in the sediments in Mission Bay, California, to be on the order

of 10 million per cubic centimeter of mud on the average. The number

decreased from the surface with increasing depth of sediment below the

surface. In the uppermost 5 cm depth, the number of bacteria was found

to be from 172 million to 460 million per gram of dry mud. Maclean and

Smart (1978) showed that bacteria affect the structure and properties of

sediments. Clay particles and bacteria are attracted and attached to

each other because of the electrostatic charge on their surfaces

(Marshall, 1969). In addition, the bacteria have polysaccharide pads,

pili, microfibrils or anchoring appendages to facilitate attachment to

particulates. Production of extracellular acid polysaccharide by marine

bacteria has been reported by Corpe (1970).

In a study of the activity of bacteria attached to organo-mineral

particles, Romanenko (1971) concluded that the sediment may constitute

more than 80 percent of the total activity of microflora. Ivatin (1973)

studied the number of bacteria in the water of the Kuibyshev Reservoir,

U.S.S.R. The number varied from 1 million to 6.56 million bacteria per

ml of water and the total number of microflora was directly proportional

to the amount of organo-mineral particles in water. Zobell and Feltham

(1942) found that the mud contained more bacteria than the overlying

water. Whereas there were thousands of bacteria per cubic centimeter of

water, there were millions of bacteria in an equivalent volume of mud.

There is a continuous exchange of chemicals between sediment and

water through processes of sorption and desorption. Such an exchange

across the sediment-water interface is regulated by a variety of abiotic

and biotic mechanisms. Some of the examples of these mechanisms are 1)

processes associated with mineral-water equilibrium (Kramer, 1964); 2)

sorption processes, notably ion exchange; 3) redox interactions

dependent on oxygen supply (Ward and Brock, 1978); and 4) the activities

of microorganisms such as plankton, fungii and bacteria (Berner, 1977).

Abiotic factors control the partitioning and bioavailability of

pollutants within benthic ecosystems (Lyoma and Jenne, 1976). Such

complex biological and chemical processes may provide a stratified

structure to deposited beds in stagnant water. The oxidized layer at the

surface is yellowish in color and is relatively rich in oxygen, ferric

oxides, nitrates and nitrites. It supports the bulk of benthic organisms

such as.polychaete worms and bivalves in shallow water, flatworms and

copepods in deep water and a rich growth of aerobic bacteria throughout

(Smith, 1980). Below the oxidized layer is a greyish transition zone to

the reduced black layer characterized by lack of oxygen, iron in the

ferrous state, nitrogen in the form of ammonia and hydrogen sulfide. It

is inhabited by anaerobic bacteria.

Sediments provide a habitat for both macroorganisms and

microorganisms. The activity of macroorganisms essentially consists of

burrowing and ingestion of detritus. The effect of the activity of

burrowing organisms has been experimentally studied by Rhoads and Young

(1970) who showed that the burrowed bed surfaces were more erodible than

those in which there was no burrowing. Details of interactions between

sediments and organisms are given in Chapter 7.

L

2.6 Estuarial Fine Sediment Transport Processes

2.6.1 Source of Sediment

Sediment in estuaries may originate a) from rivers, creeks and

natural land drainages, b) from beaches and offshore deposits or c) from

sources such as waste discharges or reclamation and other construction

activities. River-borne sediments in estuaries are predominantly fine,

i.e. a mixture of clay, silt and fine sand. Sediment originating from

adjacent beaches.by littoral currents is typically sandy and is

transported inside the estuary by the combined action of waves and

tides. The source of organic matter in estuarial sediments is the

runoff from land and marsh areas and the biogenic processes taking place

in the estuary.

Sediment related processes which occur in a typical estuarine

environment are 1) flocculation 2) settling and deposition 3)

consolidation 4) resuspension or erosion and 5) convective and diffusive

transport.These processes are briefly described below.

2.6.2 Flocculation

It has been mentioned under Section 1.3 that the predominantly

occurring clay particles typically have the following properties: 1)

particle size less than 2 microns, 2) shape of platelette and 3) a large

specific surface with a negative electrical charge. Various repulsive

and attractive forces have also been noted in Section 2.1. Depending

upon the mineralogical characteristics of fine sediments and the

dissolved ions in water, the net effect of the physico-chemical

inter-particle forces can be either repulsive or attractive. In the

first case, particles tend to stay dispersed in suspension. On the other

hand, when the net force is attractive, particles are bonded together in

a process known as flocculation. This results in face to face or edge to

face bonds between the clay particles (Van Olphen, 1963). Both collision

and cohesion are essential for flocculation. The frequency of collision

is determined by the sediment concentration and the flow field.

Inter-particle collision may be caused by one or more of the following

three mechanisms: 1) Brownian motion resulting from the bombardment of

suspended.sediment particles by the thermally activated water molecules,

2) velocity gradients causing collisions of particles spaced less than

the sum of their radii in the direction of the gradient and 3)

differential settling velocities of suspended particles resulting in

their collision. Results based on laboratory experiments show that the

settling velocities of flocs can be up to four orders of magnitudes

greater than the settling velocities of primary particles (Bellessort

1973). The presence of organic matter enhances flocculation. Factors

which affect properties of elementary particles influence the

flocculation process as described below.

Generally, the flocculation tendency increases with salinity.

Concentration of salts in ordinary river water is often sufficient to

flocculate most clay minerals. The effect of cation concentration on

flocculation is far more important than the valence of the cations. High

salinity is, however, not an essential factor for higher degrees of

flocculation. In fact, certain clay minerals such as kaolinite

flocculate more readily in distilled water than in salt water. A high

concentration of suspended sediment increases the frequency of

inter-particle collision and hence enhances flocculation. High pH

contributes to dispersion whereas low pH enhances flocculation. The

usual pH range experienced in most rivers and estuaries is not very far

from neutral and hence the influence of pH variation on flocculation is

not likely to be significant for practical purposes. The tendency

towards flocculation increases with temperature (Lambe 1958). Available

data indicate that a decrease of temperature through a maximum possible

range encountered in estuaries will decrease settling rates by a factor

of about 2 (Committee on Tidal Hydraulics, 1960).

Velocity gradients induced in the liquid, even by gentle stirring,

cause a relative motion of particles promoting flocculation. Such

velocity gradient-controlled flocculation is called orthokinetic

flocculation. While considering Brownian motion, it is useful to

consider one particle, called the collector particle, as remaining

stationary. Since particles may be considered to become attached to a

collector particle, a concentration gradient is formed radially outwards

from the collector particle. This diffusion-controlled flocculation is

called perikinetic flocculation. Temperature and viscosity effects are

significant under perikinetic flocculation, and the rate of flocculation

is independent-of particle size. In the benthic boundary layer,

orthokinetic flocculation has a much greater influence on the frequency

of collision between particles than perikinetic flocculation (Williams,

1980). The processes of cohesion and collision have been described in

detail by Kruyt (1952), Einstein and Krone (1962), Krone (1962), and

Hunt (1980). Zeichner and Schowalter (1979) shoed that hydrodynanic

effects can significantly alter the criteria developed for stability of

dispersions to Brownian coagulation. Turbulent disruption of flocs is

discussed by Thomas (1964). Papers on topics such as surface chemistry

of colloids, rate theories, hydrodynamic aspects, effect of inorganic

salts as flocculants and experimental methods for fundamental and

theological measurements have been presented by Akers and Gregory

(1977).

In general, the following three stages of flocculated clay

suspensions may be considered: 1) A small cluster of primary clay

particles, referred to as a "floc," 2) Floc aggregates formed by flocs

which may join together, and 3) An extensive aggregate network formed by

floc aggregates- ELocs of primary particles are designated as zero

order aggregates. These zero order aggregates are tight, near sphericl

units consisting of perhaps as many as a million elementary particles.

Aggregates of zero order aggregates are termed as first order

aggregates. Aggregates of the first order aggregates are called second

order aggregates and so on. A procedure for determining the order of

aggregation is given by Krone (1976). A suspended floc can be of any

order, and the size of a suspended floc is independent of its order.

Each succeeding order consists of aggregates of lower density and lower

shear strength (Krone, 1963). High rates of shear break the floc and

redisperse them. The rate of internal shear necessary to break the floc

has to be greater than the floc shear strength which depends upon the

properties of the sediment and the suspending fluid. For a given floc

shear strength, the floc size is limited by internal shearing. Since

internal shear can promote floc growth and also limit its size, flocs in

suspension approach a maximum equilibrium size under a sustained

condition of internal shearing (Krone, 1962). The lower the shear

stress, the larger is the floc diameter and vice versa. The floc shear

strength, although a function of the type of sediment and water quality,

decreases with increasing order of aggregation. Krone (1962) has

reported shear strength of 2.2 N/mr for the first order, 0.39 N/m for

the second order, and.0.14 N/m2 for the third order aggregates for San

Francisco Bay mud.

Krone(1962) noted that flocs grow rapidly when the shear stress is

less than 0.006 N/m2. With shear stress greater than this value, the

equilibrium floc size is reduced. For instance, the floc size reduces

from 90 microns to 30 microns when shear stress is increased from 0.02

to 0.06 N/m2.

2.6.3 Settling and Deposition

Deposition of sediment from suspension may take place either under

quiescent conditions or under a small flow velocity. The depositional

behavior of flocs is controlled primarily by the dynamics of the

interaction between two kinds of stochastic processes occurring just

above the sediment bed (Mehta and Partheniades 1973). One of these is

related to the kinetics of inter-floc collisions and the other to the

probability that a floc of a given size and shear strength may deposit

on the bed itself. The inter-floc collision frequency is highest in the

near-bed region of the flow, because the prevailing shear stresses are

also of the highest magnitude in this zone. Under the growth-promoting

influence of large number of collisions and local shear, a floc attains

an optimal size. When a settling floc approaches close enough to the bed

surface, it is subjected to the highest drag and hydrodynamic lift

forces. These two forces at a given location and instant of time are

proportional to each other under a fully turbulent flow, and their

variation with time about their mean values is of a stochastic nature.

Therefore, if the shear strength of floc is of a sufficient magnitude to

be able to withstand the high shear stresses acting on it, it will

ultimately touch the bed and become part of it. This process is termed

as deposition. If, however, the floc is composed of units held together

by relatively weak bonds, the settling floc will be ruptured by the

prevailing shear stress and the resulting smaller units will be

re-entrained into suspension.

Applying the continuity equation for the total amount of sediment

in a closed system, the rate of change of suspended sediment

concentration may be expressed as

dC Pd W C
dt h (2-1)

where, Pd is the probability of a particle or a floc sticking to the

bed, C is the suspended sediment concentration, Ws is the settling

velocity of the floc and h is the total depth of flow. The product

(PdWs) defines the effective settling velocity W's. The probability of

floc deposition and the effective settling velocity have been discussed

by Mehta and Partheniades (1973).

Krone (1962) conducted deposition experiments using mud from San

Francisco Bay. He pointed out that for sediment concentrations less than

300 ppm, the decrease in sediment concentration in suspension is

exponential with time. With sediment concentrations less than 300 ppm,

the flocs settle more or less independently without significant mutual

interference. Mehta and Partheniades (1973) noted that independent

settling of particles can occur for sediment concentrations up to 700

ppm. For concentrations between 300 ppm and 10,000 ppm, there is an

increasing amount of interference due to increased inter-particle

collisions, with the result that larger flocs are formed, causing a

faster rate of deposition. For concentrations greater than 10,000 ppm,

a continuous network of sediment particles is formed and its downward

movement is achieved by water moving upward through the sediment voids.

Let the initial concentrations of suspension be denoted by C1 (less

than 700 ppm) and C2 (greater than 10,000 ppm). The three ranges of

concentration may be denoted as: 1) C < C1, 2) C1 < C < C2 and 3) C >

C2. The rate of deposition equation given under Equation (2-1) may be

modified to make it applicable for all the three ranges and may be

written as

W C
dC s
dt = (2-2)

where Ws is defined as the effective median settling velocity for a

given system. Hayter and Mehta (1982) have given expressions for

determining W's for the three ranges of initial concentration.

A state of complete dispersal of suspended sediment is rare under

natural circumstances since a small concentration of electrolyte, on the

order of less than one part per thousand by weight is sufficient to

cause inter-particle attraction (Krone, 1962; Partheniades and Kennedy,

1966). In an estuarine situation, the majority of the suspended load may

deposit during slack water, which may result in a substantial reduction

in suspended sediment concentration. Under suitable flow conditions,

flocs deposit and form a bed. In a typical estuarine situation, a

measurable reduction in the suspended sediment concentration generally

occurs during slack water, and sometimes the majority of the sediment

load deposits.

2.6.4 Consolidation

The term consolidation is used to refer to that portion of the

compressibility of a soil that is essentially inelastic. For

non-cohesive sediments, the pore water and soil grains are relatively

incompressible. Hence the volume change is the result of expulsion of

water from the interstices between soil grains. In cohesive sediments,

the unit which participates in bed formation is a floc which may have a

zero or higher order of aggregation. Once the flocs are deposited, the

order or aggregation decreases due to overburden pressure. The aggregate

networks collapse into lower order aggregates. Krone (1963) found that

the order of aggregation changed by one with an overburden of about 2 to

3 cm. Deformation of floc structure continues to take place with

increasing time of consolidation, which results in the reduction of the

pore space due to expulsion of pore fluid and plastic flow.

The self- weight consolidation of a deposited bed occurs in two

stages, namely, primary consolidation and secondary consolidation.

During primary consolidation, sediment is supported by pore water

pressure. As pore water.escapes through a multitude of channels between

individual particles, the water content of the bed as well as the

effective stress decreases with time. At the end of primary

consolidation, the pore pressure and the effective stress are both equal

to zero, and the network of aggregates is supported against overburden

by the inter-particle bonds. The secondary consolidation process

involves re-orientation of clay particles and the adsorbed water

molecules. The rate of change in the bed thickness is extremely small

during secondary consolidation, but the bed shear strength may continue

to increase. Although different clays and the pore waters associated

with them have different properties of consolidation, the order of

magnitude of time for the primary consolidation may be in terms of days,

whereas the secondary consolidation may continue over a duration of a

year or even longer. Consolidation results in a significant change in

the bed density as well as the shear strength of deposited beds (see

Chapter 5).

2.6.5 Erosion

Erosion of cohesive sediment beds has been described under Section

2.6.5. Two modes of erosion can be identified. First there is mass

erosion, in which the plane of failure is somewhere beneath the bed

surface, at a level at which the flow-induced bed shear stress is equal

to the bulk shear strength. In this mode, relatively large portions of

bed are ripped off and entrained in the flow. The second is surface

erosion in which the inter-particle bonds must be broken in order to

detach them from the bed and entrain them into the suspending fluid.

Surface erosion takes place due to the relative instability of the

particles in relation to the remainder of the bed. This instability can

arise due to orientation.(edge to face orientation instead of face to

face) or due to insufficient bonding between the particles. Hence all

factors that influence inter-particle forces in turn have an effect on

the resuspension properties of cohesive sediments.

2.6.6 Convective and Diffusive Transport

Once eroded from the bed, fine sediments are transported entirely

as suspended load (not as bed load). Such transport is the result of

three processes: 1) convection (or advection): the sediment is

transported approximately at the speed of the local mean flow; 2)

turbulent diffusion: driven by spatial gradients of suspended sediment

concentration, the material is diffused laterally across the width of

the flow channel, vertically over the depth of flow and longitudinally

in the direction of the transport; and 3) longitudinal dispersion: the

spatially averaged suspended sediment load is dispersed in the flow

direction by spatial velocity gradients (Ippen, 1966). Thus, the spatial

distribution of suspended sediments is dependent upon the turbulent flow

regime. The convective diffusion equation for suspended sediment

transport is discussed in Section 2.7.

2.7 Mathematical Modeling

The general three-dimensional convective diffusion equation for

suspended sediment transport in an incompressible, fully developed

turbulent flow field is given as follows (Partheniades, 1964; Hayter,

1983):

9c ac ac c aC
+ Iu +v +w + (WC) (E )
3t ax y 9z 9z s ax x

(y ( C E C
(E 3) -) (E -L) = S (2-3)

where,

u, v, w = velocity components in x, y and z directions,

respectively,

C = concentration of sediment in suspension,

x, y, z = Cartesian coordinate directions with the z axis in

the direction of gravity, x axis along the

longitudinal axis of the water body and y axis along

the lateral axis of the water body,

Ex,Ey,Ez= turbulent diffusion coefficients in x, y, and z

directions respectively,

Ws= terminal settling velocity of the suspended sediment

particles or aggregates,

S = source and/or sink term.

The terms u, v, w and C used in Equation (2-3) are time-averaged. The

time-averaging is done over a duration t' satisfying the following

criteria: 1.) t' is large enough so that the. time-averages of the

fluctuating components, i.e. u', v', w' and C', are approximately zero

over t', and 2) t' is small enough so that the variations of

time-averaged terms change with respect to a larger time scale

(MacArthur, 1979).

The two-dimensional case is often used to approximate conditions in

open channel flows; hence, it is of particular interest to estuarine

sediment problems. Two-dimensional vertical (i.e. laterally

integrated) fine sediment transport models have been developed by

O'Connor (1971) and Ariathurai et al. (1977) among others.

Two-dimensional horizontal (i.e. vertically integrated) models have

been developed by April and Brett (1975) and Ariathurai et al. (1977).

None of the models mentioned above take into account all the parameters

related to fine sediment transport. However, they provide a simplified

description of the essential processes involved. Integrating Equation

(2-3) over the width of flow and assuming that l)coefficients Ex and Ez

are are independent of x and z respectively, and 2) the concentration of

sediment is small and has no effect either on the diffusion coefficients

or on their spatial gradients, the laterally integrated 2-D form of the

equation is expressed as

3C @CC zC 8 C 8C
+ U + W (D:) + (Dz)
7t 3x az @x xZx Bz zz

S(WC) + S
Dz s

(2-4)

where D and D are the effective turbulent dispersion coefficients
x z
accounting for both the processes of turbulent diffusion and dispersion

in the x and z directions respectively.

In order.to solve Equation (2-3) for the suspended sediment

concentration at specified points in the flow field, the following

initial and boundary conditions are used:

1. The initial condition states that at time t=0, the distribution

of sediment concentration is a known function of space, i.e.

C(x,z) I t-O = Co(x,z)

(2-5)

2. At all the solid boundaries, other than the sediment bed, the

transport flux normal to the boundaries is zero.

3. At each external water boundary, the width-averaged sediment

concentration is a known function of time.

4. The free surface boundary condition states that there is no net

rate of transport in the z direction across the free surface.

This condition may be expressed as

(t) I = D (t)I z=h
s C(t) I z=h z DZ z=h

(2-6)

where, Ws = terminal settling velocity of the flocs, h = depth

of flow, and C(t) is the instantaneous concentration of

suspended sediment at time t.

5. The bottom boundary condition states that the eroded sediment

material is transported in the vertical direction away from the

bed by turbulent diffusion and that the deposited sediments

become part of the bed. Accumulation and diffusion of sediment

particles may be assumed to take place within a narrow zone in

the immediate neighborhood of the bed. Within that zone, the

sediment enters and leaves in the z direction by the following

physical processes: 1) erosion of bed at a rate e, 2)

deposition of sediment at a rate WsCb, where Cb is the sediment

concentration near the bed, and 3) turbulent diffusion from the
C
bed to the main flow at a rate equal to (-E-- ). The bottom

boundary condition may accordingly be stated as follows for

four different cases:

Case 1: Constant erosion rate without deposition

aC
S+ WCb = E at z = 5 (2-7)
Szat (27)

Case 2: Deposition without erosion

9C
(1 Pd) = Ez at z = (2-8)

Case 3: Simultaneous erosion and deposition

aC
E + (1 Pd) WC = Ez at z = (2-9)

Case 4: Equilibrium condition without net erosion or net

deposition

aC
WSC = Ez at z = 5 (2-10)

where Pd is the probability of deposition, E is a small distance

above the theoretical bed (z=0) at which the bottom boundary

condition is effectively applicable. Since Dz=0 at z=0, diffusive

flux will not occur at z=0 and the boundary condition will not

apply.

The initial condition and the boundary conditions stated above

cannot, however, be readily used to solve the convective diffusion

transport equation unless the values of Ws, C and dC/dz at z = h and at

z = are known as well as the value of probability of deposition Pd is

known. The bottom boundary condition is especially difficult to

determine because the structure of turbulence and the exact nature of

the interaction of flocs and the flow just above the bed surface are the

unknown factors.

Assuming that the settling velocity remains constant, the

depth-averaged two-dimensional form of Equation (2-3) is expressed as

ac+ u + v (D -L) + (D -L) + S (2-11)
t x y x ax y y +S (2-11)

The source/sink term can be expressed as

C C+ tC
S = + Dr-= l e+ d (2-12)
a =t e at d

where -r I is the rate of sediment addition to the suspension due to
e 9C
erosion from the bed, i.e. source; and -7 I is the rate of sediment
d
removal from the suspension due to deposition, i.e. sink. Reference may

be made to Hayter and Mehta (1982) and Hayter (1983) for a further

discussion on mathematical modeling.

In order to incorporate the bottom boundary condition for a

laterally integrated model or to include the source term for a

34

depth-integrated model, the rate of erosion of cohesive sediment bed

under different flow conditions for various fluid and bed properties

must be known. This involves aspects related to the structure of

turbulence, effect of sediment concentration on the turbulence, shear

resisting forces of the sediment bed, effect of electrolyte, pH and

temperature on the properties of bed, type and amount of clay minerals

and their cation exchange capacities, and so on. These parameters are

described in Chapter 3.

CHAPTER 3

PHYSICAL AND PHYSICO-CHEMICAL FACTORS IN
COHESIVE SEDIMENT EROSION

3.1 Introduction

The principal factors controlling the erosion of saturated cohesive

sediment beds may be broadly classified under four groups, namely 1)

flow-related factors 2) sediment-related factors 3) fluid-related

factors and 4) bed-related factors. Flow-related factors include the bed

shear stress and the influence of suspended sediment on turbulence.

Sediment-related factors include sediment composition and various

indices used to properly characterize the sediment. Fluid-related

factors include temperature, pH, salinity and characterization indices

to describe the fluid properties. Bed-related factors include bed

density, water content and bed shear strength. These factors have been

briefly described in this chapter.

3.2 Flow related Factors

3.2.1 The Bed Shear Stress

The hydrodynamic force on a sediment bed depends upon the

turbulence structure near the bed. It has been assumed that at the

boundary, the entire shear is transmitted by the viscosity of fluid and

that a thin layer within which the flow is laminar exists near the

35

boundary whereas outside of it the flow is turbulent. The shear stress

is governed by the thickness of this laminar sublayer. Einstein and

Huon-Li (1958) proposed a theory of periodically forming and

disintegrating laminar sublayer. It was assumed that at the beginning of

the formation of the sublayer, the turbulent flow continues all the way

to the boundary. A very high shear stress thus results, which generally

slows down with distance from the boundary, and a laminar sublayer of

increasing thickness is built up. This layer breaks down instantaneously

and completely as soon as its thickness becomes large enough to make it

unstable.

Because of the random nature of turbulent motion, its properties

need to be described by means of statistical parameters. The

instantaneous velocities in the x, y, z, Cartesian coordinates are

represented by

u. = u + u', v. = v + v' and w. = w + w' (3-1)
1 1 1
The mean velocity u is given by

t'
u = -T / u dt (3-2)
o

where, t' is a long time in comparison to the time scale of turbulence.

Similar expressions may be given for v and w Because the random

fluctuations of velocities are both negative and positive with respect

to the mean, the mean of u' is given by

t'
u' = I u' dt = 0 (3-3)
o

The statistical parameters include the root-mean -square (rms) of

fluctuations, which is called the average intensity of turbulence and is

given by

t'
rms = / U= (- / u2 dt) (3-4)
o

From hydrodynamic considerations, it can be shown that for a

two-dimensional case of a fully developed turbulent flow, the effective

instantaneous shear stress is given by

T. = pf u' v' (3-5)

where pf is the density of fluid.The negative sign is a consequence of

u' being negative for a positive v' and a positive u' for a negative v'.

The turbulent fluctuations are-of the order of 10 percent of average

velocity.

Instead of expressing the shear stress in terms of turbulent

fluctuations, it can be expressed in terms of the instantaneous velocity

u. by the expression
1
U.
1
Ti = 5 (3-6)

where u. is the instantaneous velocity in the turbulent region near the
1
sublayer, t is time, p and v are respectively the dynamic and kinematic

viscosities of the fluid. Equation (3-6) is valid for 0 < t < t" where

t" is the period of growth of the laminar sublayer. The instantaneous

bed shear varies from a minimum of p (ui /v7 W) to a theoretical

maximum of infinity. The instantaneous velocity of flow in the turbulent

region, u., varies randomly and therefore T. also varies randomly with
1
time and space coordinates. Einstein and El-Samni concluded from the

results of their experiment that the fluctuations resulting from

turbulence of the lift acting on a rough bed are distributed according

to the Gaussian probability density law. Partheniades (1965) accepted

this conclusion but assumed that for purposes of further analysis, the

L

time-distribution of the instantaneous shear stress,Ti is also normal,

having a mean value Tb and a standard deviation Tb To/ where To is a

numerical constant. Einstein and El-Samni (1949) found n0 to be 2.75.

Christensen (1965), however noted that the velocity, rather than the

lift (and shear stress) is distributed according to the normal law.

While the instantaneous shear stress fluctuates with both space and

time, it is convenient to use the average value Tb. For instance, the

magnitude of the average bed shear stress for an open channel flow is

given by the expression

Tb = Rb Se (3-7)

where w is the unit weight of water, Rb is the hydraulic radius

corresponding to the bottom, and S is the slope of the energy grade

line. The bed shear stress denoted by Tb (with units of N/m2) in the

present study refers to the magnitude of bed shear stress averaged with

respect to time and space coordinates. The procedure followed for

measuring Tb in the annular flume used in the present study has been

described in detail by Mehta and Partheniades (1973).

3.2.2 Effect of Suspended Sediment on Turbulence

The effect of suspended sediment on the structure of turbulence is

not yet completely understood. Vanoni (1946) and McLaughlin (1961)

showed that certain flow characteristics such as the Karman constant,

turbulence structure and turbulence intensity depend on the

concentration of the suspended material. Yalin and Findlayson (1972)

have shown that suspended sediment may influence the value of Karman

constant. However, the data indicate that this effect is negligible for

concentrations less than 3000 mg/l. Partheniades (1962) studied the

1

effect of concentration of suspended sediment on the erosion rate. He

found that the erosion rates at the same velocity but at concentrations

varying from 2 g/l to 4 g/l were the same. Hence he concluded that the

turbulence structure near the wall which controls the erosion is not

materially affected by the concentration of fines.

Gust (1976) conducted laboratory hot-wire anemometer measurements

in dilute clay suspensions using sea water. For fully developed

turbulent flow in an open channel with a smooth mud bottom, mean

velocity profiles were measured for non-eroding and eroding flow rates

and compared with Newtonian flows under the same experimental

conditions. Beneath the Newtonian core, a viscous sublayer was found to

exist whose thickness was enhanced by a factor of 2 to 5 due to

suspended sediment. The friction velocity determined by the gradient

method in the viscous sublayer was reduced by as much as 40 percent.

These measurements indicate that turbulent drag reduction occurs at

non-eroding as well as eroding velocities. Agglomeration of clay-mineral

particles was suggested as a possible explanation of this phenomenon.

In an experiment with attapulgite, a clay mineral, Bogue and

Metzner (1963) found no drag reduction under turbulent flow in a pipe.

Yalin and Metzner (1972) did not find drag reduction in river flow. Gust

(1976) used clay minerals consisting of a mixture of illite, chlorite

and kaolinite which did demonstrate the occurrence of drag reduction.

Hence he proposed that not all types of aggregates cause drag reduction,

and only those that can undergo shape changes or disruption under the

influence of shear can actually alter the shape of the boundary layer.

This agglomeration is also a function of the ionic strength of the

medium. Experiments conducted by Gust using sea water represent the

limiting case of large ionic strength.

3.3 Sediment-related Factors

3.3.1 Soil Mechanical Indices

Research on cohesive sediments began with a background of soil

mechanics and hence attention was first concentrated on conventional

parameters for estimating the shear strength of clays. Dunn (1959)

suggested an expression correlating the critical shear strength for

erosion to the vane shear strength and plasticity index as the factors

which characterize the properties of bed.- Masch et al. (1965)

conducted studies on remolded cohesive sediments using a rotating

cylinder apparatus and found that the critical shear stress was

correlated to the water content and the vane shear strength. Flume

studies of natural soils conducted by Lyle and Smerdon (1965) showed

that the critical tractive force correlated to the void ratio and

plasticity index. Grissinger (1966) found that the erosion rate of soil

decreased with increasing clay content and decreasing void ratio.

However, after reviewing all the available literature, Paaswell (1973)

concluded that the generally used soil classification indices have not

proved useful as erosion predictors..

Partheniades (1962) concluded that the bed shear strength as

measured by a direct-shear-apparatus has no relationship with the soil's

resistance to erosion, which is essentially governed by the strength of

inter-particle and inter-aggregate bonds between the deposited sediment

material.

3.3.2 Sediment Composition

Christensen and Das (1973) studied the erosion of compacted beds

made of kaolinite, grundite and Ottawa sand. It was concluded that the

erosion rates were highly dependent on soil composition i.e. the clay

content as well as the clay type. It appears reasonable that clay

content and clay type would affect the shear strength because the

particle size and surface area are determined by the mineral

composition, and cohesion between fine particles is a function of

electro-chemical bonds.which. in turn are related to particle size and

surface area. However, Thorn and Parsons (1980) who studied erosion of

natural mud from three different locations, each having a different clay

mineral composition and organic contents, found that there were

similarities in their physical properties and behavior and that their

quantitative erosion characteristics were not as dissimilar as might

have been expected.

3.3.3 Organic Matter

Organic matter may by present in sediments in particulate,

colloidal or molecular form. Colloidal organic matter is negatively

charged and most molecular-sized organic materials are strongly polar

(Baver, 1956). Hence their interaction with clay minerals is of an

electro-chemical nature. A relatively small percentage by weight of

organic matter can have a considerable influence on soil

characteristics. A summary of the knowledge compiled by the Committee on

Tidal Hydraulics (1960) revealed inconclusive results. It appears that

organic matter has two mutually opposing effects on settling 1) it

promotes flocculation by providing an additional bridge or link between

inorganic particles, and 2) it slows settling rates because the lower

density of organic matter reduces the overall density of flocs.

Kandiah (1974) found that organic matter strengthens soil

aggregates against slaking. Studies on the erodibility of 30 percent

illitic soil showed that the critical shear stress for erosion increased

from 1.7 N/m2 to 4.0 N/m2 when the organic matter content was increased

from 0 to about 4 percent.

The presence of organic molecules can be detected by X-ray

diffraction.tehnique. However, it is frequently impossible to go further

and identify the organic compounds present in small amounts or those

with extremely fine size, either discrete or adsorbed. Infrared

absorption and diffraction procedures in recent years have provided

considerable information on the kind of organic molecules that may be

adsorbed, the way in which they are tied to the clay-mineral surfaces

and the manner in which they are oriented on the clay-mineral surfaces

(Grim, 1968).

33.34 Cation Exchange Capacity (CEC)

Clay minerals-have the property of sorbing certain cations and

anions and retaining these in an exchangeable state, i.e. these ions are

exchangeable for other cations and anions when present in water

solution. The CEC has a range varying from 5-10 meq/100g for kaolinite

to 80-120 meq/100g for montmorillonite. Kandiah (1974) showed that the

effect of CEC on the critical shear stress for erosion, Tcr (see Section

3.5.3), must be considered in conjunction with the sodium adsorption

ratio, SAR (see Section 3.4.1). When the SAR is high,T decreases
cr
with increasing CEC due to face to face repulsion whereas with low SAR,

cr increases with increasing CEC due to face to face attraction.
cr

3.3.5 Dielectric Dispersion

The dielectric constant is a measure of the ability of the clay to

store electrical potential energy under the influence of an electric

field. The dielectric constant, ed, for a soil sample is defined as

cL
es
Sd (3-8)
vc

where ce =capacitance, Ls = length of specimen, Ac = cross-sectional

area, eV dielectricc constant of vacuum (= 8.85 x 10-14 farad/cm). The

dielectric constant of a dry silicate material is 4, and of water about

80.

Dielectric dispersion, Asd has been defined by Alizadeh (1974) as

the total decrease in the measured dielectric constant caused by a

change in the type and amount of clay. Other factors such as pore fluid

composition, water content, particle orientation etc. have a secondary

effect. The dielectric dispersion has been used as a quantitative index

for soil characterization. Each clay appears to have a characteristic

value of this parameter (Alizadeh, 1974; Arulanandan et al., 1973).

Measurements have shown that 10 percent kaolinite and 21 percent water

content gives Asd = 7.5, whereas 60 percent kaolinite and 30 percent

water content gives Asd = 18.

Alizadah (1974) showed that the critical shear stress for erosion

increased with an increasing magnitude of AEd when SAR had low values

(less than 15). For SAR values greater than 20, Tcr showed a very small

variation as a function Asd'
d

3.4 Fluid-related Factors

3.4.1 Sodium Adsorption Ratio (SAR)

The Sodium Adsorption Ratio (SAR) is defined by the following

expression:

Na+
SAR =
[ + (Ca+ + + Mg ) (3-9)

where the cation concentrations are in milliequivalents per liter. The

SAR represents the relative abundance of Na ions with respect to the

Ca+ and Mg ions. Low SAR produces inter-particle attraction and

consequently flocculation. High SAR causes particle repulsion. Alizadeh

(1974) studied the effect of varying SAR for the same electrolyte

concentration and found that for electrolyte concentrations ranging from

0.005 N to 0.25 N, the critical shear stress for erosion, Tcr, decreased

for increasing SAR values. For instance with an electrolyte

concentration of 0.125 N, the T for montmorillonite decreased from 2.0
cr
2 2
N/m2 at SAR = 10 to a value of 0.1 N/n at SAR = 80.

3.4.2 Combined Effect of SAR and CEC

Ariathurai and Arulanandan (1978) have reported the effect of

electro-chemical parameters on the rate of erosion of compacted beds.

Their rate of erosion expression (see Table 6-7) contains an erosion rate

constant M which is defined as the increase in the rate of erosion for

an increase in the interface fluid shear by an amount equal to the

critical shear stress of that soil. The effects of SAR and CEC were

pointed out as follows: a) With low SAR values M reduces with

increasing CEC. With high SAR values, M reduces drastically with

increasing CEC. With an increase in CEC greater than 10, the value of M

drops at a considerably reduced rate. b) The value of M increases with

increasing concentration of electrolyte in the low concentration range.

It then levels off indicating that the maximum degree of flocculation

was reached. c) Increasing SAR reduces the value of M, rapidly at first

and then the reduction is gradual. An ultimate structure seems to be

reached at SAR greater than about 30.

3.4.3 Temperature

Kandiah (1974) studied the effect of temperature on the erosion of

a remolded illitic soil and expressed the critical shear stress for

erosion T cr, as a function of temperature as follows:

-15 o
T = 1.8 x 1015 ex(4100/T ) (3-10)
cr

where T =absolute temperature in degrees Kelvin.

The effect of water temperature on the rate of erosion has also

been studied by Grissinger (1966), Christensen and Das (1973), Raudkivi

and Hutchison (1974) and Gularte (1978). It was observed that the

erosion rate increased with increasing temperature because the critical

shear stress was reduced due to a reduction in the inter-particle forces

of attraction. Campanella and Mitchell (1968) observed that an increase

in temperature resulted in a significant pore pressure increase and a

significant permanent decrease in the volume of saturated soils. The

range of temperature for these tests was from 40 C to 60 C.

Raudkivi and Hutchison (1974) concluded that although temperature

has an influence on the erosion rate of cohesive sediments, the effect

is insignificant with increasing salinity of eroding water and with

decreasing particle size because both factors increase the

inter-particle cohesion. Mitchell (1969) pointed out that while it is

acceptable to use diffuse double layer theory to explain the effect of

salinity, temperature variations should have little effect on the

inter-particle double layer and that it is more probable that the

inter-particle contact structure itself is weakened because of thermal

energy of constituent atoms.

3.4.4 pH

Kandiah (1974) found that the erosion rate increased with

increasing pH for the same bed shear stress. For instance, the rate

increased from 1.2 x 10-2 g/an2-min to 3.3 g/cm2-min at a bed shear

stress of 3.0 N/m2, when the pH was increased from 4.4 to 10.3. The

soil used for these experiments contained 30 percent kaolinite and 70

percent Yolo loam silt, which was a natural soil with a median diameter

of 0.07 nm and a CEC of 10.4 meq/100 gram. The pH values selected were

4.4, 6.9 and 10.3, and the shear stress ranged from 0.5 N/m2 to 4.5

N/m2

Liou (1970) used compacted beds to study the erosion of cohesive

sediments. He used a static load of 0.87 Kg/ca2 for 2 hours using a

hydraulic press. He showed that the critical shear stress for the

erosion of such beds decreased from 2.73 N/m2 to 0.19 N/m2 when the pH

was increased from 5.6 to 8.2.

3.4.5 Pore Fluid

The physico-chemical effects influence the magnitude of

inter-partical stresses. These stresses in turn control the soil

behavior in shear and in compaction. Surgunam (1973) and Arulanandan

(1975) used pore fluids and eroding fluids having different chemical

properties and studied their effect on the erodibility of clays. It may

be noted that the erodibility of clay beds having the same pore fluid

and eroding fluid has yet to be fully understood. Furthermore, under

estuarine.conditions, it is reasonable to consider the two fluids to be

the same as far as the top, active layer of the sediment bed is

concerned. Hence the aspects related to the difference in the properties

of pore fluid and the eroding fluid were not included in the scope of

the present study.

3.5 Bed-related Factors

3.5.1 Bed Density and Shear Strength

Deposited beds.have a vertical variation in bed density as well as

shear strength. The lower layers have a relatively longer consolidation

time and also a lower order of aggregation which results in their

attaining a higher density and a higher shear strength than the sediment

in the upper layers. The magnitude and the vertical variation of the bed

density and shear strength have a significant influence on the rate of

erosion of cohesive sediment beds. Chapter 5 is devoted to describing

the bed density and shear strength variations of deposited beds.

3.5.2 Water Content

Lee (1979) measured the water content of deposited beds. For

consolidation periods varying from 1 to 10 days, the water content was

found to reduce from about 81 percent to 73 percent. Although this

appears to be a relatively small change, the erosion rate at 81 percent

water contents was one magnitude higher than the erosion rate with 73

percent water content. Postma (1967) reported that increased water

content in cohesive sediments resulted in a decreased shear stress

needed to initiate entrainment of sediment in the eroding fluid.

Southard et al. (1971) and Lonsdale and Southard (1974) found that the

critical shear stress for erosion of cohesive sediments decreased from

1.4 N/m2 to 0.08 N/2 when the water content was increased from 61

percent to 84 percent.

3.5.3 Critical Bed Shear Stress

The critical bed shear stress, Tcr, may be defined as the minimum

bed shear stress at which particles in the bed are dislodged and thereby

commence movement. The initial movement is often determined by visual

observation and is therefore subjective.

Different methods have been adopted by investigators to determine

T Dunn (1959) defined T as that shear stress at which the bed
cr cr
material is in general motion. While studying the erosion of

non-cohesive sediments, Shields (1936) defined T as that value of the
cr
shear stress for zero sediment discharge obtained by extrapolating a

line of observed sediment discharge versus bed shear stress. This

definition has also been used for cohesive sediments by Riley and

Arulanandan (1972), Arulanandan et al. (1972), Sargunam (1973) and

Alizadeh (1974). The definition of critical shear stress has been shown

schematically in Figure 3-1. The magnitude of the bed shear stress at

which erosion commences at the initial elevation of the bed-fluid

interface (at the beginning of an experiment) is denoted by Tcr. Design

of non-eroding irrigation channels based on the value of Tcr is quite

likely to be uneconomical since Tcr is generally very small. Hence a

value corresponding to Tc as shown in Figure-3-1 is often recommended

for use as critical shear stress (Partheniades and Paaswell, 1970).

Wo

cn /
LLJ

Tcr c Tch
SHEAR STRESS, Tb

Figure 3-1: Schematic Representation of Erosion Rate as a Function of
Bed Shear Stress.

The erosion of a sediment bed can be considered to begin when the

local time-mean bed shear stress exceeds the critical shear stress of

the bed material. This concept leads to the definition of the critical

bed shear stress as the time-mean value of the bed shear stress when

erosion is initiated (Christensen, 1973). The forces which cause a floc

to move are of a stochastic nature due to the turbulent flow near the

bed. Therefore, it may be possible to observe local bed failure due to

an instantaneous velocity fluctuation causing the destabilizing forces

i

to exceed the erosion-resisting forces although the time-mean bed shear

stress is less than the instantaneous bed shear stress. The probability

of erosion, P is determined by 1) the instantaneous hydrodynamic lift

force, L, and the instantaneous bed shear; and 2) the properties of

floc, namely, its buoyant weight W and the floc shear strength F' which

must be overcome in order to detach a floc from the bed. The probability

of erosion, Pe' can be expressed as follows:

P = probability of {l> + F (3-11)
e L

For further discussion on the probability of the erosion and deposition,

reference may be made to Mehta and Partheniades (1973). The cohesive

sediment bed is composed of a large number of flocs which have a variety

in terms of the number and the size of the primary particles, number and

type of inter-particle bonds (face to face or edge to face), and the

floc size and floc density. Hence all flocs do not have the same shear

strength. The magnitudes of the floc shear strength, F', and the

buoyant weight of the floc, W, have a statistical variation. F' and W

have a spatial variation along the bed surface and over the depth of

bed. F' also has a temporal variation while the bed is consolidating.

The measured value of flow-induced bed shear stress, b is a time-mean

value. The instantaneous value of Tb, denoted by Ti will, in general,

be different from the mean value (Section 3.2.1) and be accounted for in

relation to F'. A floc would be detached from the bed and entrained in

the fluid by turbulent diffusion if Equation (3-11) is satisfied.

Partheniades (1962) used placed beds as well as deposited beds made

from San Francisco Bay mud. The ratio of the shear strength of the

placed bed to that of the deposited bed was of the order of 100.

However, the minimum shear stress for commencement of erosion was about

the same for both the beds, i.e. 0.05 N/m2. The critical shear stress,

Tcr, does not increase with consolidation time. However, a change in the

physical and electro-chemical properties of the fluid can induce a

significant change in Tcr. An increase in water temperature results in a

decrease in Tcr (Section 3.4.3). Liou (1970) showed that Tcr for

compacted beds decreased from 2.73 N/m2 to 0.19 N/m2 when the pH was

increased from 5.6 to 8.2. The type of.sediment and fluid influence Tcr.

In the present study it was found that deposited beds of kaolinite in 35
2
ppt salt water had Tcr=0.07 N/m2 whereas Lake Francis mud in

reconstituted lake water had Tcr =0.145 N/m2. An increase in salinity

did not affect Tcr for kaolinite. However, for the natural mud a

substantial increase in Tcr was noticed.Tcr for the lake mud increased

from 0.145 N/m2 to 0.25 N/m2 when the salinity was increased from 0 to

10 ppt. For deposited kaolinite beds in tap water, Tcr was 0.10 N/m2,

whereas in 35 ppt saline water, it was 0.04 N/m2. With the addition of

organic and inorganic nutrients mentioned in Appendix A, -cr of
2
kaolinite bed in tap water increased. to 0.22 N/m2. With the cultivation

of algae and microorganisms on this substrate, Tcr further increased to

0.35 N/ 2.

Thorn and Parsons (1979) reported that the critical shear stress

for commencement of erosion of Belawan mud was 0.13 N/m2. Arulanandan

et al. (1980) reported that undisturbed soil samples obtained from

field had a wide variation in Tcr, ranging from about 0.4 N/m2 to 4
2
N/2 They further mentioned that remolding the soil generally decreases

both the critical shear stress and the rate of change of erosion rate.

They have also shown that a decrease in salt concentration of the

eroding fluid decreased the critical shear stress.

Alizadeh (1974) has given expressions for computing the magnitude

of the critical shear stress based on certain parameters such as the

salt concentration, sodium adsorption ratio, dielectric dispersion and

percentage of water uptake. However, the relationships proposed by

Alizadeh (1974) are applicable only to the sediments and fluids used and

for the experimental conditions adopted in the tests because the

inter-particle bond strength is a function of additional parameters such

as the mineral composition, particle size distribution and CEC of the

sediment, pH of the eroding fluid, temperature, etc. It is believed

that the only reliable method available at present for determining the

critical shear stress for erosion of cohesive sediments is to simulate

the bed as close as possible to its natural environment, conduct

laboratory tests to measure erosion rates at different bed shear stress

magnitudes, and estimate the magnitude of Tcr from extrapolation of

results as shown in Figure 3-1.

3.5.4. Characteristic Bed Shear Stress

Attempts by researchers to measure or estimate the shear strength

of cohesive sediment beds have not always been successful. For example,

use of soil mechanical indices or measurement of shear strength carried

out with direct-shear-test apparatus do not provide the magnitude of

shear strength (see Section 3.3.1 and 5.4.1). Also, the bed density is

not necessarily correlated to the shear strength (see Section

5.4.1).Hence investigations were carried out at the University of

Florida in order to provide a procedure for measuring the shear strength

of deposited bed as a function of depth (Parchure,1980). The procedure

involved conducting an erosion experiment under a series of bed shear

stresses in Phase III (see Figure 4-2). Each successive value of Tb was

only slightly higher than the previous. The duration of the time step,

T, was typically 60 minutes. An important assumption was made in respect

of the suspended sediment concentration versus time relationships

obtained from the data. It was assumed that if the net rate of erosion

at the end of the time step was a small percentage (say less than 5

percent) of the rate of erosion at the beginning of the time step, then

the erosion was close to becoming arrested and the shear strength of bed

at that elevation of the sediment-water interface was equal to the

applied bed shear stress. Thus, each of the Tb values satisfying the

above condition yielded the depth of bed below the bed-fluid interface

at which the shear strength of the bed was equal to the corresponding Tb

value. The series of Tb values could be divided into two broad groups.

The first group of initial or lower Tb values resulted in low rates of

erosion whereas the higher Tb values resulted in relatively high rates

of erosion of the bed. Similar observations have been reported by other

investigators (Partheniades, 1962; Espey, 1963). The shear stress at

which Tb values appear to be divided into two such groups was denoted as

a characteristic shear stress, Tch, shown in Figure 3-1. Further

discussion on Tch is given in Section 5.2.4.

CHAPTER 4

PREVIOUS STUDIES ON EROSION

4.1 Introduction

Erosion of soil from land.surfaces has engaged the attention of

scientists and engineers in the field of agronomy for the past more than

half a century. Their main concern has been conservation of fertile top

soil from erosion caused by rainfall and runoff, and design of

irrigation canals with non-eroding banks. Extensive studies have been

carried out and a considerable amount of literature is available;

however, most of it is empirical in nature. Basic research on

properties and behavior of clay minerals under flow-induced shear stress

has been carried out only during the past two or three decades. From

time to time literature on cohesive sediments has been reviewed, as for

instance-by the Committee on Tidal Hydraulics (1960), Partheniades

(1964), Task Committee of ASCE (1968), and Paaswell (1973).

Since erosion of soil from land surfaces is not the subject under

consideration for the present study, only a sunmmary of relevant research

which provides information on the properties and erosional behavior of

clays is included in Section 4.2. The results of recent investigations

on the erosion of deposited beds have been mentioned at various places

in Chapters 5 and 6 in their relevant context. Hence they are not

reviewed here in order to avoid repetition. Only the types of beds used

in these investigations and the general procedure followed in conducting

the studies are described in this chepter in Sections 4.3 and 4.4,

respectively. A summary of earlier studies conducted at the University

of Florida is given in Section 4.5 because the present study is, in a

sense, a continuation and extension of these studies.

4.2 Soil Erosion

For engineering purposes, soil is defined as a natural aggregate of

mineral grains, moderately cohesive, inorganic or organic in nature,

that have the property of being separated by simple mechanical

processes, for instance by agitation by water (Grim, 1968). To the

agriculturist, loose mantle at the surface of earth which is capable of

supporting plant life consists of soil. The term soil may include a wide

range of materials such as sand, silt, clay, organic clay, humus etc.

Soil scientists have oriented their research in classification of soils

depending on their properties. In addition to the physical and chemical

parameters related to the soil, some have considered parameters such as

vegetation, intensity and duration of rainfall etc. while studying soil

erosion.

Bennett (1926) compared the chemical composition and the physical

properties of soils with respect to their erodibility. Middleton (1930)

gave an account of the properties of soils which influence their

erosion. The physical properties included moisture content, mechanical

analysis, liquid limit, specific gravity etc. Soil properties having the

greatest influence on erosion were characterized by various ratios

namely, dispersion ratio, erosion ratio etc. and limiting values of

these ratios were suggested for distinguishing erosive from non-erosive

soils. Middleton (1932) compiled data from eight experimental stations.

Physical characteristics affecting erosional behavior were determined

and a study was made of their relation to each other, to the chemical

composition of the colloid and to field erosional behavior. However, no

quantitative expression to describe the erosional behavior was

presented. Lutz (1934) prepared a report for the Agricultural Experiment

Station of the University of Missouri on the physico-chemical properties

of soils affecting their erosion. Studies along similar lines were

continued in the subsequent years, for example,.by Peele (1937), Rost

and Rowels (1940), Musgrave, (1947), among others. However, none of

them give any quantitative expression to correlate soil erosion to other

parameters.

Laflen and Beasley (1960) were among the first researchers to point

out the effects of soil compaction on the critical tractive force in

cohesive soils. One of the widely quoted studies on tractive resistance

of cohesive channels is that of Dunn (1959). He used soil samples from

channel beds in Nebraska, Wyoming and Colorado, and developed an

expression for critical tractive force as a function of mean grain size,

percent of particles finer than 0.06 mm and vane shear strength.

Reference must also be made to a subsequent study reported by Lyle and

Smerdon (1965). These investigators pointed out that the concept of

maximum permissible velocity suggested by Lane (1955) to avoid channel

bank erosion did not adequately take into account soil properties which

control erodibility.- They indicated the possibility of occurrence of

soils with the same texture displaying widely different erosive

characteristics. Lyle and Smerdon (1965) presented eight general

expressions, each giving the critical shear stress for erosion as a

function of one or more of the eight measured parameters, namely,

plasticity index, dispersion ratio, percent organic matter, vane shear

strength, cation exchange capacity, mean particle size, calcium-sodium

ratio and percent clay. They also stated that the best correlation was

obtained in the order in which they are listed above.

In spite of decades of research, studies on soil erosion did not

lead to a comprehensive understanding of erosion mechanism because of

the empirical. approach followed in many such. studies. Due to the

fundamental differences in the behavior of cohesive and non-cohesive

sediments in the soil, a progress in the understanding of erosive

properties was made only in the subsequent studies which were conducted

on cohesive sediments alone. Results of these studies have been cited in

Chapters 5 and 6.

4.3 Types of Beds in Erosion Studies

Three different types of cohesive sediment beds have been used for

conducting erosion studies:

a. Uniform bed (also called placed bed or remolded bed): the principal

feature of this bed is that the shear strength is uniform over

depth. One way to prepare such a bed is to pour a thick slurry of

sediment and fluid and allow it to settle under gravitu. Another

way is to remold the clay in a laboratory apparatus. Remolding is

essentially the result of repeated shear deformations throughout

the clay mass, which tends to destroy the flocculated fabric and

causes a parallel particle orientation known as dispersed fabric.

Erosion of remolded beds has been studied by Partheniades (1965),

and Kelly and Gularte (1981) among others.

b. Compacted bed: sediment with a low water content is compacted in

the test apparatus by using external mechanical force. This type of

bed has been used for erosion studies by Espey (1963), Christensen

and Das (1973), Alizadeh (1974), and Raudkivi and Hutchison (1974)

among others. Properties of compacted bed have been discussed by

Lambe (1958) and Seed and Chan (1959). Although compacted beds may

have a uniform.density and shear strength.over depth, they are

different from uniform beds in regards to the water content, void

ratio etc. and hence must be classified separately.

c. Deposited bed (also called stratified bed or flocculated bed):

sediment in suspension in the test apparatus is allowed to deposit

and form a bed either under quiescent condition (no flow) or under

a very low flow-induced bed shear stress which permits all or most

of the sediment to deposit. The former is called a statically

deposited bed whereas the latter is called a flow-deposited bed.

With reference to estuarine conditions, as a result of the temporal

and spatial variations in the magnitude and the direction of tidal

current, at least a part of the fine sediment is suspended,

transported and deposited in a cyclic manner. Hence bottom

surficial estuarine sediments tend to have the characteristics of a

deposited bed. This type of bed has been used by Krone (1962),

Partheniades (1962), Fukuda (1978), Lee (1979), Yeh (1979) Parchure

(1980) and Dixit (1982). The properties of deposited bed are

described in Chapter 5.

For purposes of the present study, the terms uniform bed and

deposited bed refer to uniform saturated (with water and without any

trapped gases such as air) bed and deposited saturated bed,

respectively.

4.4 Procedures in Erosion Studies

In general, the following procedure is adopted by investigators for

conducting laboratory erosion studies: 1) selection and characterization

of selected sediment and fluid, 2) calibration of apparatus 3) formation

of bed in the apparatus, and 4) erosion of bed. The usual objective of

these studies has been to obtain an erosion rate expression for a known

flow-induced shear stress on a sediment bed which has been formed using

a certain sediment and fluid mixture. Very few (for instance Sargunam

1973; Arulanandan, 1975) have used a pore fluid of different

electro-chemical properties from the eroding fluid. Some investigators

however did not adequately perform all appropriate characterization

tests. In fact, missing data pose a serious problem in comparing the

results obtained by various investigators. The diversity of selections

made in different studies has greatly enhanced the complexity of the

subject matter, thus making the task of comparing the results of

different studies not only difficult but also of questionable value.

Hunt (1981) has given a comparative review of the parameters selected by

different investigators for conducting studies on the erosion of

cohesive sediments. Mehta (1981) has described the difficulties

encountered in comparing the results.

1

Grissinger (1966) conducted experiments on molded beds by mixing

measured quantities of kaolinite, illite and montmorillonite to Granada

silt loam which was used as stock material. Erosion tests were conducted

on compacted beds in a small horizontal flume. Samples were eroded for

a duration varying from less than 3 minutes to 10 minutes. It was

concluded that the stability of cohesive material against erosive forces

of flowing water varied with the type and amount of clay minerals,

sample bulk density, antecedent water in the sample and temperature of

the eroding fluid.

The above is a typical example of an erosion study for cohesive

sediment beds. Similar investigations have been carried out by others

making a choice of sediment, fluid, type of bed, apparatus, test

duration, test procedure, and so on. Each investigator has pointed out

the influence of certain parameters which characterize the sediment,

fluid, or bed structure on the shear strength of bed or on the rate of

erosion. Results of investigations that are relevant to the present

study are included in Chapters 5 and 6 in their appropriate context.

4.5 Previous Studies at the University of Florida

4.5.1 Objective

Investigations have been carried out at the University of Florida

over the past decade to study the erosional and depositional behavior of

fine sediments. Both kaolinite and marine muds have been used as

sediment, and the studies have been conducted in a recirculating

straight flume as well as in a rotating annular flume. Distilled water,

salt water and tap water have been used as the eroding fluid to examine

the effect of water quality on the erosion rate. The erosion studies

were carried out with particular reference to the structure of the bed

and the variation of bed erodibility with depth. These studies are

briefly described here.

4.5.2 Apparatus

Experiments have been conducted in two types of flumes:

1. Rotating annular flume: This consists of an annular channel 20 cm in

width, 46 cm in depth and 76 cm in mean radius (Figure 4-1). The channel

is made of. 0.95. an thick fiberglass.. Four. plexiglass windows, each 7.6

cm x 5 an, are provided symmetrically along the circumference of the

channel to facilitate visual observations at the bed-fluid interface.

Inside the channel, a 20 cm wide plexiglass ring is suspended by means

of four vertical supports which are attached to a central vertical shaft

by means of four horizontal supports. The ring can be rotated by

actuating the central shaft by means of an electrical motor. A second

electric motor is used to rotate the annular channel. Each motor is

provided with a control unit with an indicator panel to enable its

running at the desired speed. The equipment was calibrated for obtaining

the required flow-induced bed shear stress up to 0.9 N/m2. The ring and

the channel are rotated in opposite directions to minimize the effect of

secondary currents and to provide a uniform flow in the channel. Water

depth in the channel can be varied from 0 to 33 cm and the ring

elevation adjusted to match the water depth. The bottom surface of the

ring is adjusted precisely in order to provide complete contact between

the water surface and the bottom of the ring. When the ring is rotated,

shear is transmitted to the sediment bed through the water column.

Samples of suspended sediment are collected through four taps provided

Refilling Well

Drain
Suspension
Collector

and Brush Block Assembly
Strain Gages
-Ring Suspending Blode
Aft r-Refilling Funnel
fiberglass Channel.

-Fiberglass Stiffener
Sample Top

Driving Motors

Figure 4-1:

Schematic View of Annular Flume Facility
(Modifications made for conducting Sediment
Flushing Experiment not shown)

on the outer vertical face of the channel. The concentration of the

sediment is determined by using Millipore filter apparatus and an

electronic balance with digital readout which can measure weights to an

accuracy of 0.1 milligram. For additional details reference may be made

to Mehta and Partheniades (1973).

Modifications to the flume were made in the present study to

provide regulated inflow and outflow of water in order to conduct an

experiment involving flushing of the suspended sediment out of the

channel while rotating. The speed.controller units for the electric

motors driving the ring and the channel were replaced by new

computer-compatible units (Model MW 25/75 E of Browning Company,

Maysville, Kentucky) to avoid their manual operation. A desk top

mini-computer, Hewlett-Packard Model HP-85, along with two programmable

digital-to-analog converter units (Model HP 59501 A of Hewelett-Packard

Company, Orlando, Florida) were provided to exercise precise control on

the rate of application as well as on the magnitude and duration of bed

shear stress during the course of the experiments. Photographs of the

rotating channel and the control equipment are given in Figure 4-2 (a)

and 4-2 (b).

2. Recirculating straight flume: This is a steel flume, 18 m long, 0.6

m wide and 0.9 m deep. One side of the flume is provided with glass

panels to facilitate visual observations. The flume can be tilted

longitudinally to a maximum slope of 0.02. An underflow type control

gate is provided at the downstream end of the flume. A centrifugal pump
3
with a maximum capacity of 0.164 m /s is used to generate flow in the

flume. A downstream sump and a 200 nn diameter return pipe are provided

to enable recirculation of water in the flume. An orifice meter is used

64

aM

Channel

Ring

Taps for Sampling Suspended Sediment

Inflow Pipe

Outflow Pipes

Circular Tray

Figure 4-2(a): Photograph of the Rotating Channel Facility

G : bbtor Controller Units

H : HP-85 Computer

I : Digital-to-analog Converters

Figure 4-2(b): Photograph of Control Equipment

1

to measure discharge. Details of this facility are given by Dixit

(1982).

4.5.3 Methodology

Figure 4-3 is a schematic representation of the procedure used for

conducting the erosion experiments. Phase I and Phase II together

represent the pre-erosion stress history of the deposited bed, whereas

Phase III represents the erosion phase.

PhoseI -PhRse 11 Phosel -
Tm

--Tm --- Td, -i--Td. -c-----d -T-,-T2---T3-+T4-T5- Time

Pr---- e-Erosion Stress Hislory Resuspension

Figure 4-3: Schanatic Representation of Procedure used for Bed
Preparation and Erosion Tests in Experimental Series 3
and 4.

Phase I is a mixing phase. The sediment and the fluid in the

apparatus are mixed under a shear. stress Tm for a duration Tm.

Phase II is the deposition phase. The shear stress is reduced from

Tm to Tdland maintained over a duration Tdl so that a part of the
material deposits. It is then further reduced to Td and maintained over
2
a duration Td2 which results in deposition of some more sediment. The

flow is then completely stopped so that the remaining sediment in

suspension is deposited. A zero flow condition is next maintained over a

duration Tdc which is the gravitational consolidation time for the

deposited bed. When a bed is formed under Td over a duration of Td or
1di Td1

under additional steps such as Td2 over a duration Td2 etc., it is

called a flow-deposited bed. If the shear stress is brought to zero at

the beginning of Phase II, the entire deposition occurs under quiescent

condition and the bed is called a statically deposited bed.

Phase III is the resuspension phase. A series of bed shear stresses

(Tbl, Tb2, etc.), each slightly greater than the previous in its
magnitude, is applied to the deposited bed. Tb1 occurs over a duration

TI, Tb2 over a duration T2 and so on. Typically the duration T1, T2,

etc. (referred to as time-steps) is the same, e.g., 30 minutes, 60

minutes etc. in one experiment. The manner in which Tb is varied is

characterized either by the ratio of consecutive bed shear stresses

given by

Tr = T+1 (4-1)
bn
or by the normalized incremental bed shear stress given by

(ATbn) = Tbn+l Tbn (4-2)

where n = 1, 2, 3,...etc. stands for the serial number of a time-step.It

is noted that

(Abn) = Tr I (4-3)

When the bed shear stress is compared in comparison with the shear

strength of the bed, the term normalized excess shear stress (Tb)ex may

be used.

(ATb)ex b Ts (4-4)
Ts

The sequence of bed shear stresses may be pre-selected with Tr
n
increasing or decreasing or remaining constant for consecutive Tb

values. Several erosion experiments were also carried out with only one

value of the bed shear stress over the entire duration of the test

ranging form 24 to 200 hours.

The methodology described above has been used by Mehta and

Partheniades (1979), Yeh (1979), Parchure (1980) and Dixit (1982) for

conducting erosion experiments. These are described in Sections 4.5.4

through 4.5.7 as Series 1 through Series 4, respectively. Only the

important results of the various experiments are summarized here.

4.5.4 Results of Studies under Series 1

These experiments were carried out in the annular flume using

kaolinite in distilled water (Mehta and Partheniades, 1979). Two types

of bed were used. The first was a flow-deposited bed prepared in

accordance with Phase II described in Section 4.5.3. The second was a

placed bed formed by making a slurry of water and sediment outside the

flume at a density close to the average density of the flow-deposited

bed and placing this slurry in the annular flume. The deposited bed had

a variation in bed density and shear strength as a function of depth

below the sediment-water interface, whereas the placed bed had a more or

less uniform structure over its depth. The deposited bed was subjected

to erosion under a bed shear stress, (T), of 0.207 N/rrt, whereas the

placed bed was eroded under Tb = 0.413 N/in. The concentration-time

relationship observed for each is shown in Figures 4-4 and 4-5,

respectively. It is seen that the time-averaged rate of change of

concentration for the deposited bed continuously decreased with time,

whereas the same rate was constant in the case of the uniform bed. The

significance of this observation is discussed later in Section 5.2.1.

8 006
j 0.04 b= 0.207 N/m2
La 0.02 7
(ro
< 0 2 4 6 8 10 1214 16 18 20 22 24
TIME (Hrs)

Figure 4-4: Relative Suspended Sediment Concentration against
Time, t, for Erosion of a Deposited Bed using
Kaolinite in Distilled Water at Tb = 0.207 N/m2
(Based on the Data of Mehta and Partheniades,
1979).

o OO . ... . . . :
0- 600
En 500 a
0 oE40

I 00 . .. . .. I

TIME (Hrs)

Figure 4-5: Suspended Sediment Concentration, C, against
Time, t, for Erosion of a Uniform Bed using
Kaolinite in Distilled Water at Tb= 0.413
N/m2. (Based on the Data of Mehta and
Partheniades, 1979).

4.5.5 Results of Studies under Series 2

These experiments were carried out on statically deposited beds

formed in the annular flume using kaolinite in distilled water as well

as in salt water (Yeh 1979). Stratified beds were formed in the flume

according to the methodology described earlier, and were subjected to

erosion under bed shear stress values of 0.43, 0.34, 0.24 and 0.15 N/n?.

Each experiment was conducted over a duration of approximately 200

hours. Representative results of these experiments for kaolinite in

distilled water are shown in Figure 4-6. The rate of change of

suspension concentration continuously decreased over the duration of the

experiment, while the total concentration continually increased with

time. The time-concentration relationship followed the exponential law

given by

C = CS (1 e-ft) (4-5)

where Cs = steady state concentration, 8 = an empirical coefficient. C =

concentration of suspension at time t. Tests using kaolinite in salt

water, as well as those using a natural mud in salt water confirmed the

same general trend of results. The erosion rate expression for these

studies has been given in Table 6-7 in Chapter 6.

4.5.6 Results of Studies under Series 3

All the experiments under this series were conducted to study

erosion of deposited beds formed by using kaolinite in salt water with

35 ppt concentration. The erosion of a deposited bed was studied under a

series of bed shear stresses Tbl' Tb2, Tb3. (see Phase III shown in

Figure 4-3). The variation in consecutive magnitudes of Tb was

characterized by the normalized excess bed shear stress (Tbn)* defined

by Equation (3-2).

Three different patterns of variation of (Tbn) were studied,

namely, (Tn), increasing with time, decreasing with time and remaining

constant with time. Also, a variation in the magnitudes of the bed shear

stresses was achieved by selecting the value 0.2, 0.5 and 1.0 for (ATbn)*

in different experiments. The duration, T, was either 30 minutes or 60

minutes and was maintained constant in each individual experiment. The

structure of the deposited bed was varied by changing the parameters Td,

Td and Tdc in Phase II (see Figure 4-3). It was noted that beds with

longer consolidation times had lower rates of erosion. Thus, the

importance of pre-erosion stress history in the erosion rate of

deposited beds was experimentally demonstrated. The concentration-time

relationship obtained for a typical experiment in which Tdc was 40 hours

is shown in Figure 4-7.

The rate of change of the depth-averaged sediment concentration,

dC/dt, which could be readily measured, is related to the rate of

erosion, e, by:

E = dC (4-6)
dt

where h is the depth of flow.

The measured rate of change of concentration is also related to the

bed density, p by:

dC p dz (4
d --t (4-7)
dt h dt

The bulk density of bed as a function of depth z below the bed-fluid

interface must be measured in order to obtain dz/dt from the measured

dC/dt and h.

Two apparatuses (Figure 4-8) were developed to determine bed

density. Apparatus I simulated the condition of static deposition of

sediment from an initial uniform suspension outside the rotating

channel. Apparatus II was developed for in-situ measurements in the

rotating channel where a flow-deposited bed could be formed under a low

shear stress. Details on the use of these apparatuses to measure the bed

dinsity are described by Parchure (1980). The variation of bed density

with depth measured with the use of Apparatus I is given in Figure 4-9

for the test corresponding to Figure 4-7.

During the course of various experiments, it was noticed that for

Tb< Tch, the bed continued to erode over a certain period of time
during which the concentration of sediment in suspension continuously

increased and then reached constant magnitude. Under such a condition,

it was assumed that erosion had stopped because the bed shear stress was

equal to the bed shear strength. Since the bed elevation for this

condition could be determined from the data on bed density and total

sediment in suspension, a plot of bed shear strength versus depth could

be obtained when an experiment was conducted under a series of bed shear

stresses with small increments (see Section 5.4.1 for further

informationn. Figure 4-10 shows such a relationship corresponding to

the results of the erosion experiment presented in Figure 4-7. The

suspension concentration at the end of each time step of 60 minutes

(denoted as C60) is plotted as a function of Tb in Figure 4-11. It is

noted that Tch increased with increasing Tdc. The significance of Tch,

density variation and shear strength variation are discussed in Chapter

5.

4.5.7 Results of Studies under Series 4

Dixit (1982) conducted erosion experiments in the straight flume

using kaolinite in tap water to form deposited beds. The consolidation

time was varied from 2 hours to 240 hours. It was shown that Tch

increased with consolidation time increasing from 2 hours to 120 hours

and then appeared to be almost invariant for consolidation times greater

than 120 hours (Figure 4-12). The relationship between erosion rate,e ,

and the normalized excess shear stress was expressed as

Tb Ts(z)
C = EO expel Ts (z) (4-8)

TIME (Hours)

Figure 4-6:

z
0
o 7
crc
z. E

Z

LwE

Li

) C

U)

Figure 4-7:

Suspended Sediment Concentration versus Time
Observed by Yeh (1979) for Erosion of Deposited
Kaolinite Beds using Distilled Water

TIME (Hours)
7 8

9 10

2 3
TIME (Hours)

Variation of Suspended Sediment Concentration
With Time during a Test using Kaolinite in Salt
Water following the Experimental Methodology
shown in Figure 4-3 (Parchure, 1980)

1

74

APPARATUS I

---Top Cylinder 15cm dia.

Plastic Tubes d various heights,
0.95 cm dia. glued to the
bottom plate --

Bottom
Cylinder
15cm di.-

APPARATUS II

-2 cm dia plastic tube

-- 15 cm dia. plexiglass cylinder
-2.5 cm dia metal tube

Annular space for mixture of alcohol
and dry ice.
jiment
Filled with ice abes

Piston with Screw Rod

Figure 4-8:

Apparatus for Measurement of Density
Function of Depth for Deposited Beds
1980)

as a
(Parchure,

T
225 cm

L

T
15cm
JL

Porcelein
Dish -

i

DENSITYp (gm/e)

Figure- 4-9:

Variation of-Bed Density, p, with.Depth, ,z
in Test corresponding to Figure 4-7

SHEAR STRENGTH Ts (N/m2)

Figure 4-10:

Variation of Bed Shear Strength, Ts, with
Depth z, in Test corresponding to Figure 4-7

I I I I r

(Arbn) = 0.2
rchT =0.21 N/ m2
rch2=0.29N/m2
rch3=0.34N/m2
rd N/m2 T(hrs)
S0.050 24
S0.015 40
0 135

'(60) Vlues

6/

-C(60) Volues

BED SHEAR STRESS rb (N/m2)

Figure 4-11:

Suspension Concentration versus Bed Shear Stress
for Different Flow Deposited Beds (Parchure, 1980)

E

Z'
0

z
o
a
r-

in
z
0

z
0
z

1,
vn

/
I!x
11/

where a and E, are empirical coefficients, Tb is the bed shear stress

and Ts (z) is the bed shear strength as a function of depth z below the
T T
bed-fluid interface. The relationship between n (E/:,) and a ( b- s)
Ts
is shown in Figure 4-13 for Tdc = 48 hours as an illustration. (Mehta

and Partheniades, 1982). Similar results were obtained for other values

of Tdc. It was noted that Es and a appear to vary somewhat with the

time-step, and therefore with depth below the bed surface. However, the

two coefficients were found to be independent of the consolidation time

Tdc. Data obtained in experiments conducted with kaolinite and salt.

water were analyzed in a similar manner. While the trends were similar,

average values of a and cowere different; a (tap water) was 1.6 times

a (salt water) and -g (tap water) was about 2 times Eo (salt water)

where a and soindicate average values over the duration Tdc. The

significance of a and Eo is described in Section 6.3. By adopting the

same procedure which is described under Series 3, the structure of bed

in terms of Ts as a function of depth below surface was determined for

the deposited bed using kaolinite in tap water. The results are given by

Dixit (1982).

4.6 Conclusions of Chapter 4

Studies on soil erosion conducted during the years 1930 to 1960

were mainly based on empirical approaches and did not lead to a

comprehensive understanding of the erosion mechanism. Conclusions of

subsequent studies focussing attention to the behavior of clays in

particular are contained in Chapters 5 and 6 of this study. The studies

carried out at the University of Florida were primarily concerned with

(M
E

z0.2
l-

0

Figure 4-12:

CONSOLIDATION TIME (hrs)

Variation of Tch with Consolidation Time (Dixit, 1982)

2.5 -- i --

-=Te T
2.0- eo

; 1-5 /
o

y.
Series 4

1.01-

Tdc -48 hr

Time c xtO
Slep Tb
i (Nm'z) (g -m' mn' )
I 00.030 &6 0.47
o 2 0.050 8.5 0.73
3 0.087 5.5 0.43
o 4 0183 6.7 0.48
1 1 1

0.5

F]/

C(Tb -TS)

Figure 4-13:

Variation of Zn( /o/E) with (Tb-Ts)/Ts for a Test using
Kaolinite in Tap Water with Tdc = 48 Hours (Based on
the data of Mehta and Partheniades, 1982)

r

I I

I
\r

9:

investigating the erosion rate of cohesive sediment beds as a function

of the bed shear stress and the bed shear strength. Conclusions of

these studies are given in this section.

1. Erosion of a uniform bed under a constant bed shear stress resulted

in a linear variation of suspension concentration as a function of

time. From the corresponding constant rate of erosion, it was

concluded that the eroded material did not deposit, because

simultaneous erosion and deposition would ultimately result in a

state of equilibrium in which the rates of deposition and erosion

would be equal and the suspension concentration will achieve a

constant value.

2. The rate of erosion was found to increase with the bed shear

stress. The rate was constant for a uniform bed, but it decreased

continuously in the case of a deposited bed. From this observation

it was concluded that the shear strength of a deposited bed

increases with depth.

3. The constancy of final suspension concentration observed in the

erosion of a deposited bed was attributed to arrested erosion. It.

was therefore concluded that under such conditions, the bed shear

stress was equal to the bed shear strength at the bed-fluid

interface. This concept was further used for developing a

methodology of conducting experiments in which a deposited bed was

eroded layer by layer with small increments in bed shear stress and

the depth of erosion was estimated from the bed density and

suspension concentration data.

4. The Pre-erosion stress history was found to be an important factor

in determining the structure of deposited bed, and hence the rate

of erosion.

5. Erosion tests conducted by using the methodology mentioned above

showed that the characteristic bed shear stress increases with

consolidation time up to about 5 days. For longer consolidation

times, further increase is not very significant.

6. Measurements of bed density showed that the bed density increases

with depth.

7. The erosion rate, of a deposited bed is a function of bed shear

stress, Tb and bed shear strength, Ts, which can be expressed as

ex [ b --s (4-9)
ep [ TS
"o 's

where a and so are empirical coefficients.

1

CHAPTER 5

EROSION MECHANISM AND BED STRUCTURE

5.1 Introduction

As noted later in Chapter 6, the rate of erosion is a function of

the difference between the forces which cause erosion and the forces

which resist erosion. The erosive force is represented by the bed shear

stress, Tb whereas the resistive force is represented by the

inter-particle attractive force which in turn is characterized by the

cohesive shear strength, Ts with respect to erosion. The dependence of

the erosion rate on Tb Ts has been confirmed by Thorn and Parsons

(1980) and by studies conducted at the University of Florida. The bed

shear stress is an independent parameter which has been discussed in

Section 3.21.. The bed shear strength, on the other hand, is a function

of different parameters. Various investigators have measured erosion

rates of cohesive sediments as a function of several parameters (Hayter

and Mehta, 1982). Examples of some of the parameters are given in

Chapter 3.

The fundamental mechanisms responsible for the bonding of cohesive

sediment particles are 1) an overlap of diffuse double layer (Section

2.4) 2) particle arrangement (Section 6.6), and 3) physical binding of

particles by cementing agents. Through these mechanisms, the influence

of any parameter brings about a change in the inter-particle force,

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
45	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file