You are on page 1of 149

A MODIFIED HOLE EROSION TEST (HET-P)

TO STUDY EROSION CHARACTERISTICS OF SOIL

by

MARCEL LUTHI

B.Eng., HSR University of Applied Sciences Rapperswil, 2005

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF


THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF APPLIED SCIENCE

in

The Faculty of Graduate Studies

(Civil Engineering)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

August 2011

© Marcel Luthi, 2011


Abstract

Abstract

Today’s increasing demand for energy and natural resources requires safe and reliable
infrastructure. This includes hydraulic earth structures like dikes, levees, or dams. Such structures are
susceptible to piping, a fundamental type of internal soil erosion. Piping is one of the principal causes
of failures and accidents affecting embankment dams. The Hole Erosion Test (HET) is based on soil
piping, and is used to determine the erodibility and critical shear stress of a soil. A soil specimen with a
preformed axial hole is subjected to a constant-head pressure flow, and the rate of enlargement of the
soil pipe is determined indirectly from flow rate and hydraulic gradient. This study presents a Modified
Hole Erosion Test (HET-P) that introduces a conventional Pitot-static tube to measure total energy
head and flow velocity of the exiting jet, which is correlated to a mean velocity within the axial hole. A
series of Modified Hole Erosion Tests (HET-P) was performed on non-erodible PVC specimens with
axial holes of constant, but different diameter, followed by HET-P tests on two types of soil, namely
glacial till material of a dam core and natural clay deposits from Ontario river banks. Results confirmed
that sidewall hydraulic head measurements to determine hydraulic gradients in the standard HET
overestimate the resulting axial wall shear stress by as much as an order of magnitude. Furthermore,
velocity measurements increase the confidence in test results as they allow for a more direct estimate
of the axial hole diameter at any time during a test. A Pitot-static tube used in the HET-P for velocity
and pressure measurement can easily be incorporated, and yields more transparent and reliable
results by eliminating or amending some of the limiting assumptions of the standard test. It is an easy,
fast, and economical approach that can be applied to soils in both constructed earth structures
including dams and embankments, and to natural river banks to determine their susceptibility to
internal and surface erosion.

ii
Table of contents

Table of contents

Abstract ................................................................................................................................................... ii

Table of contents................................................................................................................................... iii

List of tables .........................................................................................................................................vii

List of figures .......................................................................................................................................viii

List of symbols .....................................................................................................................................xii

List of abbreviations ...........................................................................................................................xiv

Acknowledgements..............................................................................................................................xv

Dedication ............................................................................................................................................xvi

1. Introduction ..................................................................................................................................... 1

1.1. Background ................................................................................................................................ 1

1.2. Objective and scope .................................................................................................................. 3

1.3. Outline of the thesis ................................................................................................................... 4

2. Literature review ............................................................................................................................. 5

2.1. Erosion tests category 1: Flow over surface .............................................................................. 6

2.2. Erosion tests category 2: Rotating cylinder ............................................................................... 7

2.3. Erosion tests category 3: Jet impact .......................................................................................... 8

2.4. Erosion tests category 4: Flow through defect .......................................................................... 9

2.5. Hole Erosion Test (HET) ..........................................................................................................10

2.5.1. UNSW Hole Erosion Test (HET) ....................................................................................10

2.5.2. HET theory .....................................................................................................................13

2.5.3. HET data analysis ..........................................................................................................15

2.5.4. Modifications to the HET ................................................................................................17

2.5.5. Main findings and conclusions .......................................................................................20

2.6. Summary and outline of the research program .......................................................................23

2.6.1. Outline of the research program ....................................................................................24

3. Modified Hole Erosion Test (HET-P) ...........................................................................................25

3.1. Hydraulic review of the HET ....................................................................................................25

iii
Table of contents

3.1.1. Friction head loss and hydraulic gradient .......................................................................25

3.1.2. Minor head losses and entrance length .........................................................................27

3.1.3. Flow conditions and axial hole velocity ..........................................................................28

3.2. HET-P apparatus .....................................................................................................................30

3.2.1. Test cell ..........................................................................................................................31

3.2.2. Pitot-static tube and differential pressure transducer system ........................................33

3.2.3. Water flow system ..........................................................................................................35

3.3. Experimental setup and test procedure ...................................................................................37

3.3.1. Calibration of measuring instruments .............................................................................37

3.3.2. Test preparation .............................................................................................................37

3.3.3. Test procedure ...............................................................................................................38

3.4. Modified analysis .....................................................................................................................39

3.4.1. HET-P: based on energy gradient ..................................................................................39

3.4.2. HET-P (V): based on energy gradient and flow velocity ................................................40

4. Non-erodible test specimens .......................................................................................................41

4.1. Research program ...................................................................................................................42

4.2. Results and analysis ................................................................................................................43

4.2.1. Head ratio and shear stress ...........................................................................................43

4.2.2. Flow coefficient and axial hole velocity ..........................................................................45

4.3. Discussion................................................................................................................................47

4.3.1. Head ratio and shear stress ...........................................................................................47

4.3.2. Axial hole velocity and hydraulic roughness ..................................................................48

4.3.3. Upstream total energy head ...........................................................................................49

4.3.4. Flow conditions and minor losses ..................................................................................50

4.3.5. Sources of errors ............................................................................................................51

4.3.6. Limitations ......................................................................................................................52

5. Erodible soil specimens ...............................................................................................................53

5.1. Research program ...................................................................................................................54

5.2. Results and analysis ................................................................................................................55

iv
Table of contents

5.2.1. Soil properties and description .......................................................................................55

5.2.2. Head ratio and shear stress ...........................................................................................57

5.2.3. Axial hole velocity and estimated diameter ....................................................................60

5.2.4. Erosion rate and critical shear stress .............................................................................63

5.2.5. Dam core material ..........................................................................................................64

5.2.6. Ontario clay samples ......................................................................................................65

5.3. Discussion................................................................................................................................68

5.3.1. Head ratio and shear stress ...........................................................................................68

5.3.2. Axial hole velocity and estimated diameter ....................................................................69

5.3.3. Erosion rate and critical shear stress .............................................................................69

5.3.4. Test soils ........................................................................................................................70

5.3.5. Sources of errors ............................................................................................................71

5.3.6. Limitations ......................................................................................................................72

6. Conclusions and recommendations ...........................................................................................73

6.1. Summary and conclusions .......................................................................................................73

6.2. Recommendations ...................................................................................................................75

References ............................................................................................................................................77

Appendices ...........................................................................................................................................83

Appendix A: Review of the standard Hole Erosion Test (HET) ..........................................................84

Appendix B: Engineering drawings .....................................................................................................85

Appendix C: Calibration of measuring instruments ............................................................................88

Differential pressure transducers ....................................................................................................88

Custom v-notch weir .......................................................................................................................89

Appendix D: HET-P test procedure ....................................................................................................91

Soil preparation for reconstituted specimens ..................................................................................91

Specimen preparation for reconstituted specimens........................................................................91

Specimen preparation for specimens from undisturbed (intact) soil samples ................................92

Test preparation ..............................................................................................................................92

Test procedure ................................................................................................................................93

v
Table of contents

Post-test procedure .........................................................................................................................94

Appendix E: Analysis methods ...........................................................................................................95

Appendix F: Experimental program ....................................................................................................96

Appendix G: Non-erodible test specimens .......................................................................................100

Appendix H: Soil properties ..............................................................................................................103

Appendix I: Erodible soil specimens .................................................................................................105

Dam core material .........................................................................................................................105

Ontario clay samples ....................................................................................................................109

vi
List of tables

List of tables

Table 1.1: Embankment-dam engineering and safety evaluation ........................................................ 1


Table 2.1: Qualitative relation of representative erosion rate index and progression of internal
erosion (Wan and Fell 2002, 2004a, 2004b) .....................................................................12
Table 3.1: Specifications SETRA Model 230 Bidirectional Wet-to-Wet Pressure Transducers .........34
Table 4.1: Summary of head ratio values, H / h , obtained from the three non-erodible test
specimens for turbulent flow with Re > 5000 .....................................................................43
Table 4.2: Summary of flow coefficient values, K , obtained from the three non-erodible test
specimens for turbulent flow with Re > 5000 .....................................................................45
Table 4.3: Summary of estimated axial hole diameters HET-P (V), t , obtained from the three non-
erodible test specimens for turbulent flow with Re > 5000.................................................46
Table 5.1: Summary of velocity ratio values obtained from the Ontario clay specimens for turbulent
flow with Re > 2000 for HET-P, respectively Re > 5000 for HET-P (V) ..............................61
Table 5.2: Summary of diameter ratio values obtained from the Ontario clay specimens for turbulent
flow with Re > 2000 for HET-P, respectively Re > 5000 for HET-P (V) ..............................62
Table 5.3: Summary of soil properties for erodible soil specimens ....................................................66
Table 5.4: Summary of test data and results for erodible soil specimens ..........................................67
Table A.1: Known challenges and issues of the standard HET with suggested improvements .........84
Table E.1: Step by step analysis of test data for different methods ....................................................95
Table F.1: Test number information ....................................................................................................96
Table F.2: Test program .....................................................................................................................97
Table G.1: Test data and results non-erodible test specimen D1-063 ..............................................100
Table G.2: Test data and results non-erodible test specimen D1-123 ..............................................101
Table G.3: Test data and results non-erodible test specimen D1-243 ..............................................102

vii
List of figures

List of figures

Figure 1.1: Internal erosion and piping process in earth dams .............................................................. 2
Figure 2.1: Schematic diagram flow over surface .................................................................................. 6
Figure 2.2: Schematic diagram rotating cylinder ................................................................................... 7
Figure 2.3: Schematic diagram jet impact.............................................................................................. 8
Figure 2.4: Schematic diagram flow through defect .............................................................................. 9
Figure 2.5: Schematic diagram Hole Erosion Test HET (adapted from Wan and Fell, 2004) .............11
Figure 2.6: Typical results Hole Erosion Test (schematic, no real test data) ......................................12
Figure 3.1: Flow pattern upstream and downstream of test specimen, including Energy Grade Line
(EGL), and Hydraulic Grade Line (HGL) at wall ................................................................27
Figure 3.2: Pitot-static tube with schematic turbulent velocity profile of jet exiting axial hole
immediately adjacent to test specimen ..............................................................................29
Figure 3.3: Schematic diagram of HET-P apparatus ...........................................................................30
Figure 3.4: Photograph of HET-P apparatus .......................................................................................31
Figure 3.5: a) Vertical HET-P test cell, b) Dismantled downstream part of HET-P test cell with bridge
element holding 6-mm wire mesh for soil specimen support, piezometer connection, and
Pitot-static tube ..................................................................................................................32
Figure 3.6: Differential pressure transducer system, setup for high range differential pressures .......34
Figure 3.7: Schematic hydraulic profile of Modified Hole Erosion Test (HET-P) with Energy Grade
Line (EGL) for initial hole diameter o = 6 mm at medium flow rate ..................................35
Figure 3.8: Downstream constant head tank immediately after failure of soil specimen illustrating
performance of flow conditioning plate and v-notch weir at 20-25 l/min ...........................36
Figure 3.9: Flowcharts describing test analysis of (a) HET, and (b) HET-P (V) (rectangle: measured
or deduced, parallelogram: assumed or affected by uncertainty) .....................................40
Figure 4.1: Non-erodible PVC specimens with axial holes of 6, 12, and 24 mm diameter..................42
Figure 4.2: Head ratio equals shear stress ratio versus flow rate for the three non-erodible test
specimens ..........................................................................................................................44
Figure 4.3: Wall shear stress HET-P using Pitot-static tube data versus wall shear stress HET from
sidewall hydraulic heads for the three non-erodible test specimens .................................44
Figure 4.4: Flow coefficient, K , versus flow rate for the three non-erodible test specimens ..............45
Figure 4.5: Mean flow velocity in axial hole using Pitot-static tube data versus mean flow velocity in
axial hole using continuity for the three non-erodible test specimens ...............................46
Figure 4.6: Estimated axial hole diameters HET-P (V), back calculated from Pitot-static tube data
versus flow rate for the three non-erodible test specimens ...............................................47

viii
List of figures

Figure 5.1: Head ratio equals shear stress ratio versus estimated mean hole diameter for the Ontario
clay specimens ..................................................................................................................58
Figure 5.2: Wall shear stress HET-P using Pitot-static tube data versus wall shear stress HET from
sidewall hydraulic heads for the Ontario clay specimens ..................................................59
Figure 5.3: Mean flow velocity in axial hole from HET-P and HET-P (V) using Pitot-static tube data
versus mean flow velocity in axial hole from HET using continuity for the Ontario clay
specimens ..........................................................................................................................61
Figure 5.4: Estimated axial hole diameter from HET-P and HET-P (V) back calculated from Pitot-
static tube data versus estimated axial hole diameter from HET for the Ontario clay
specimens ..........................................................................................................................62
Figure 5.5: Erosion rate versus time for successively increased test heads (S3-993.01) ...................63
Figure 5.6: Critical shear stress defined on flow rate versus shear stress diagram (S3-993.21) ........64
Figure C.1: H-U diagram differential pressure transducer #1 and #2 with fitted linear regression lines
to convert output voltage to differential pressure head .....................................................88
Figure C.2: H-Q diagram v-notch weir using Kindsvater-Shen relationship (curve fitting) ...................89
Figure C.3: H-Q diagram v-notch weir using Kindsvater-Shen relationship (head readings) ..............90
Figure H.1: Gradation curves dam core material (USCS) ..................................................................103
Figure H.2: Standard compaction test Dam MV4-Altered material (ASTM D698 Method A) .............104
Figure H.3: Standard compaction test Dam MV4-Core material (ASTM D698 Method B) ................104
Figure I.1: Dam MV4-Altered – S1-003.01 – Measured flow rate & results at point of failure ..........105
Figure I.2: Dam MV4-Altered – S1-003.01 – Dried specimen ..........................................................105
Figure I.3: Dam MV4-Altered – S1-003.02 – Measured flow rate & results at point of failure ..........106
Figure I.4: Dam MV4-Altered – S1-003.02 – Drained test cell US and detail with Pitot tube ...........106
Figure I.5: Dam MV4-Altered – S1-553.01 – Measured flow rate & results at point of failure ..........107
Figure I.6: Dam MV4-Altered – S1-553.01 – Drained test cell US and detail with Pitot tube ...........107
Figure I.7: Dam MV4-Core – S2-553.01 – Measured flow rate & results at point of failure .............108
Figure I.8: Dam MV4-Core – S2-553.01 – Drained test cell US and detail with Pitot tube ...............108
Figure I.9: Ontario Clay – Little Cataraqui – S3-993.01 – Prepared test specimen .........................109
Figure I.10: Ontario Clay – Little Cataraqui – S3-993.01 – Measured flow rate .................................109
Figure I.11: Ontario Clay – Little Cataraqui – S3-993.01 – Head difference and energy loss ............110
Figure I.12: Ontario Clay – Little Cataraqui – S3-993.01 – Head ratio ...............................................110
Figure I.13: Ontario Clay – Little Cataraqui – S3-993.01 – Axial hole diameter .................................111
Figure I.14: Ontario Clay – Little Cataraqui – S3-993.01 – Wall shear stress ....................................111
Figure I.15: Ontario Clay – Little Cataraqui – S3-993.01 – Erosion rate ............................................112
Figure I.16: Ontario Clay – Little Cataraqui – S3-993.01 – Flow rate versus shear stress ................112
Figure I.17: Ontario Clay – Little Cataraqui – S3-993.01 – Drained test cell US and plaster cast .....113
Figure I.18: Ontario Clay – Little Cataraqui – S3-993.01 – DS side of specimen pre- and post-test .113

ix
List of figures

Figure I.19: Ontario Clay – Little Cataraqui – S3-993.02 – Prepared test specimen .........................114
Figure I.20: Ontario Clay – Little Cataraqui – S3-993.02 – Measured flow rate .................................114
Figure I.21: Ontario Clay – Little Cataraqui – S3-993.02 – Head difference and energy loss ............115
Figure I.22: Ontario Clay – Little Cataraqui – S3-993.02 – Head ratio ...............................................115
Figure I.23: Ontario Clay – Little Cataraqui – S3-993.02 – Axial hole diameter .................................116
Figure I.24: Ontario Clay – Little Cataraqui – S3-993.02 – Wall shear stress ....................................116
Figure I.25: Ontario Clay – Little Cataraqui – S3-993.02 – Erosion rate ............................................117
Figure I.26: Ontario Clay – Little Cataraqui – S3-993.02 – Flow rate versus shear stress ................117
Figure I.27: Ontario Clay – Little Cataraqui – S3-993.02 – Drained test cell US ................................118
Figure I.28: Ontario Clay – Little Cataraqui – S3-993.02 – Test specimen after testing (DS) ............118
Figure I.29: Ontario Clay – Bear Brook – S3-993.12 – Prepared test specimen ................................119
Figure I.30: Ontario Clay – Bear Brook – S3-993.12 – Measured flow rate .......................................119
Figure I.31: Ontario Clay – Bear Brook– S3-993.12 – Head difference and energy loss ...................120
Figure I.32: Ontario Clay – Bear Brook – S3-993.12 – Head ratio .....................................................120
Figure I.33: Ontario Clay – Bear Brook – S3-993.12 – Axial hole diameter .......................................121
Figure I.34: Ontario Clay – Bear Brook – S3-993.12 – Wall shear stress ..........................................121
Figure I.35: Ontario Clay – Bear Brook – S3-993.12 – Erosion rate ..................................................122
Figure I.36: Ontario Clay – Bear Brook – S3-993.12 – Flow rate versus shear stress .......................122
Figure I.37: Ontario Clay – Bear Brook – S3-993.12 – Drained test cell US and plaster cast ...........123
Figure I.38: Ontario Clay – Bear Brook – S3-993.12 – Extracted test specimen after testing ...........123
Figure I.39: Ontario Clay – Wilton Creek – S3-993.21 – Prepared test specimen .............................124
Figure I.40: Ontario Clay – Wilton Creek – S3-993.21 – Measured flow rate.....................................124
Figure I.41: Ontario Clay – Wilton Creek – S3-993.21 – Head difference and energy loss ...............125
Figure I.42: Ontario Clay – Wilton Creek – S3-993.21 – Head ratio ...................................................125
Figure I.43: Ontario Clay – Wilton Creek – S3-993.21 – Axial hole diameter.....................................126
Figure I.44: Ontario Clay – Wilton Creek – S3-993.21 – Wall shear stress........................................126
Figure I.45: Ontario Clay – Wilton Creek – S3-993.21 – Erosion rate ................................................127
Figure I.46: Ontario Clay – Wilton Creek – S3-993.21 – Flow rate versus shear stress ....................127
Figure I.47: Ontario Clay – Wilton Creek – S3-993.21 – Drained test cell US and plaster cast .........128
Figure I.48: Ontario Clay – Wilton Creek – S3-993.21 – Test specimen after testing (DS) ...............128
Figure I.49: Ontario Clay – Wilton Creek – S3-993.22 – Prepared test specimen .............................129
Figure I.50: Ontario Clay – Wilton Creek – S3-993.22 – Measured flow rate.....................................129
Figure I.51: Ontario Clay – Wilton Creek – S3-993.22 – Head difference and energy loss ...............130
Figure I.52: Ontario Clay – Wilton Creek – S3-993.22 – Head ratio ...................................................130
Figure I.53: Ontario Clay – Wilton Creek – S3-993.22 – Axial hole diameter.....................................131
Figure I.54: Ontario Clay – Wilton Creek – S3-993.22 – Wall shear stress........................................131
Figure I.55: Ontario Clay – Wilton Creek – S3-993.22 – Erosion rate ................................................132

x
List of figures

Figure I.56: Ontario Clay – Wilton Creek – S3-993.22 – Flow rate versus shear stress ....................132
Figure I.57: Ontario Clay – Wilton Creek – S3-993.22 – Drained test cell US ...................................133
Figure I.58: Ontario Clay – Wilton Creek – S3-993.22 – Test specimen after testing (DS) ...............133

xi
List of symbols

List of symbols
2
= cross-sectional area of axial hole m
Ce = coefficient of soil erosion s/m
Cp = Pitot tube coefficient –
Cv = velocity coefficient –
D = diameter of flow chamber m
2
fL = friction factor for laminar flow conditions kg/m s
3
fT = friction factor for turbulent flow conditions kg/m
= forcing force N
= retaining force N
2
g = acceleration of gravity m/s
Gs = specific gravity of soil –
hd = measured downstream sidewall hydraulic head m
hf = friction head loss m
hu = measured upstream sidewall hydraulic head m
hv = centerline velocity head m
h = hydraulic head difference across test specimen m
Hd = measured downstream total energy head m
Hu = upstream total energy head m
H = energy head loss along test specimen m
i = H / L = energy gradient across the specimen in modified HET-P –
= HET erosion rate index –
K = CpCv = flow coefficient –
L = length of axial hole m
LT = transition length turbulent to laminar flow m
LL = liquid limit of soil %
2
p = fluid pressure N/m
2
ps = static pressure (Pitot-static tube) N/m
2
pt = total pressure (Pitot-static tube) N/m
2
pv = dynamic pressure (Pitot-static tube) N/m
p / g = pressure head m
P =  = wetted perimeter ≡ circumference of circular cross section m
PL = plastic limit of soil %
3
Q = measured flow rate m /s
Re = Reynolds number –

xii
List of symbols

s = h / L = hydraulic gradient across the specimen in standard HET –


Sr = degree of saturation %
t = elapsed time s
T = water temperature °C
umax = centerline jet velocity m/s
V = mean flow velocity in circular hole or pipe m/s
Vd = downstream flow velocity (flow chamber) m/s
Vt = estimated mean flow velocity in axial hole m/s
Vu = upstream flow velocity (flow chamber) m/s
wf = final water content (post-test) %
wo = initial water content (pre-test) %
wopt = optimum water content %

2
= erosion rate per unit surface area kg/s/m
2
= estimated erosion rate per unit surface area of the axial hole in HET kg/s/m
2
= estimated erosion rate per unit surface area of the axial hole in HET-P kg/s/m
 = Darcy friction factor –
 = kinematic viscosity (water at 20°C:  = 1.004E-06 m /s)
2 2
m /s
d = dry density of soil kg/m
3

d,max = standard maximum dry density kg/m


3

m = density of total (moist) soil specimen kg/m


3

w = density of water used as eroding fluid kg/m


3

g = specific weight of eroding fluid N/m


3

 = hydraulic shear stress along wetted area N/m


2

c = estimated critical shear stress for initiation of erosion N/m


2

2
= estimated wall shear stress along axial hole in HET N/m
2
= estimated wall shear stress along axial hole in HET-P N/m
o = initial shear stress N/m
2

 = hole diameter m
f = measured final hole diameter m
o = initial diameter of preformed hole m
t = estimated mean diameter of axial hole m

xiii
List of abbreviations

List of abbreviations

DHT = Drill Hole Test


EFA = Erosion Function Apparatus
EGL = Energy Grade Line
HET = Hole Erosion Test (Wan and Fell 2002, 2004a, 2004b)
HET-P = Modified Hole Erosion Test based on energy gradient
HET-P (V) = Modified Hole Erosion Test based on energy gradient and flow velocity
HGL = Hydraulic Grade Line
ICOLD = International Commission On Large Dams
JET = Jet Erosion Test
NEF = No Erosion Filter Test
RCT = Rotating Cylinder Test
UBC = University of British Columbia, Vancouver, Canada
UNSW = University of New South Wales, Sydney, Australia
USBR = United States Bureau of Reclamation
USCS = Unified Soil Classification System

xiv
Acknowledgements

Acknowledgements

I offer my sincere gratitude to my research supervisors Dr. Robert G. Millar and Dr. R. Jonathan
Fannin for their patient guidance throughout this research. I am very grateful for the opportunity, and
their continued encouragement and support of this challenging interdisciplinary research project.

I would like to extend my thanks to the faculty and staff of the Department of Civil Engineering and
the School of Music for the opportunity to learn and teach at UBC. I would like to express my special
thanks to Bill Leung, engineering technician at the Civil Engineering workshop, for his valuable time
and help in building and setting up of lab equipment. His creative ideas and cheerful soul were always
very much appreciated. I thank my fellow students for their friendship inside and outside the class
room and laboratory. I always enjoyed the interesting vivid discussions and exchange of ideas.

I would like to express particular thanks to Dr. Colin D. Rennie of the University of Ottawa who
provided the Ontario clay samples tested as part of this study, and to Dr. Kerry A. Mazurek and Daniel
Cossette of the University of Saskatchewan for carrying out soil property tests and providing the soil
property data for the Ontario clay samples.

I owe particular thanks to my colleagues at my former employer, Basler & Hofmann, Consulting
Engineers AG, Switzerland, for their wide support during the course of this program, both financially
and by providing me with valuable information to complete my studies. I am very grateful for the
generous financial support received from the Pestalozzi-Stiftung, Switzerland, with special thanks to
Käthi Schmidt and Barbara Schürmann. I am thankful for the financial assistance and support from
Swiss Engineering STV, the University of British Columbia Faculty of Graduate Studies, and Swiss
Association for Road and Transportation Experts (VSS).

My final and utmost thanks I owe to my family, especially my fiancée, who never stopped believing
in me, and whose invaluable support and love inspired me each day and made sad days joyful.

xv
Dedication

Dedication

To Yvonne, my fiancée
for her endless love and faith in me

xvi
Chapter 1. Introduction

1. Introduction

1.1. Background

Hydraulic earth structures, which include dikes, levees, or dams, are used worldwide to restrict
standing or flowing water within an assigned area. But the presence of water in such structures may
cause severe damage that could lead to a failure of the structure, eventually resulting in the loss of
lives and catastrophic damage. The three main mechanisms causing substantial damage are
structural causes and slope instability, overtopping, and internal erosion (Table 1.1). The latter
appears to be a main cause of dam instabilities. According to statistics by ICOLD (1995) and Foster et
al. (2000a, 2000b), about 30-50% of failures and accidents affecting embankment dams are caused
by piping as one of the fundamental types of internal soil erosion.

Table 1.1: Embankment-dam engineering and safety evaluation

Primary causes of failures of embankments Failure statistics for large dams


(ICOLD 1995) (Foster et al. 2000a, 2000b)
Mode of failure % Mode of failure %
Internal erosion (Piping) 31 Internal erosion (Piping) 48
- through embankment 17 - through embankment 31
- through foundation 14 - through foundation 15
- embankment to foundation --- - embankment to foundation 2
Overtopping 33 Overtopping 46
Structural causes 26 Slope instability 4
Other causes 11 Earthquake 2

The term internal erosion is used herein to describe conditions where seepage flow from the
reservoir erodes soil particles from within the structure and transports them downstream. Suffusion
and piping are the two fundamental types of internal erosion (e.g. Wan and Fell 2002) and describe
the manner in which eroded particles are transported downstream (Figure 1.1).

Suffusion, also referred to as internal instability, describes the selective erosion of fine particles
which are removed through the constrictions between the larger particles. This process leaves behind
an intact soil skeleton formed by the coarser particle fraction.

Piping involves the formation and development of a continuous tunnel or pipe within an earth
structure through erosion of the surrounding soil material driven by a hydraulic gradient. Piping failure

1
Chapter 1. Introduction

may occur in one of three modes (e.g. Foster et al. 2000a, 2000b), namely piping through the
embankment, piping from the embankment to the foundation, and piping through the foundation
(Figure 1.1). Fell et al. (2003) divided the process of internal erosion and piping into four phases; i)
initiation of erosion, ii) continuation of erosion, iii) progression to form a pipe, and iv) formation of a
breach. Piping may be initiated by means of backward erosion or concentrated leaks (e.g. Sherard et
al. 1984; Sherard and Dunnigan 1985; Sherard 1986). Backward erosion usually commences at the
downstream side where the hydraulic gradient is high enough to cause a displacement of soil
particles, and progressively continues upstream. Concentrated leaks originate within the structure, and
are caused by differential settlement (crack through core), hydraulic fracturing (crack jacked open by
penetrating water), poor compaction (high permeable zone, e.g. around concrete structures or rock
foundations), or digging animals and vegetation. Whether or not initiated erosion progresses and a
final breach is formed depends on various factors (e.g. Fell et al. 2003), like hydraulic gradient (implied
flow velocity and shear stress), the ability of the soil to sustain an open pipe, upstream conditions to
provide crack filling material, downstream conditions to receive or stop eroded soil (filter design
criteria, free opening), and the rate of enlargement of the hole (erosion resistance).

Suffusion Piping

Seepage

through
embankment embankment
to foundation

through foundation

Figure 1.1: Internal erosion and piping process in earth dams

2
Chapter 1. Introduction

Internal erosion and piping are potentially extremely dangerous. There may be little or no external
evidence that piping erosion has developed. Common signs are sand boils that may be hidden under
water, or sinkholes at the crest of a dam. Sophisticated surveillance and monitoring pore pressures
and seepage can help to warn about potential problems. In most cases however, early stages of
internal erosion and piping are very hard to detect (Fell et al. 2003).

Internal soil erosion and piping are complex processes that are very difficult to describe by
theoretical analysis, and are influenced by many hardly quantifiable factors. For this reason, various
researchers have developed a number of tests to investigate internal erosion characteristics of soil,
one of which is the Hole Erosion Test (HET). The HET is an accepted laboratory index test method,
developed to study piping erosion in concentrated leaks in earth dams (Wan and Fell 2002, 2004a,
2004b). In the HET, a soil specimen with preformed axial hole is subjected to head-controlled flow,
and measured hydraulic gradient and flow rate are used to indirectly determine the rate of
enlargement of the idealized soil pipe.

The main advantages of the HET are that it is simple and straightforward to use, and that tests
can be performed in an economical manner without the requirement for large amounts of soil. Thus, it
has been applied in several research projects, and there is a growing data base for erosion
characteristics of many different types of soil that will help to understand the relationship between
basic engineering properties and erosion characteristics. However, the development of the HET is still
a work in progress. Studies revealed various challenges and issues affiliated to the HET, including
systematic differences between the HET and other commonly used erosion test methods (Lim 2006;
Bonelli et al. 2006; Farrar et al. 2007; Bonelli and Brivois 2008; Wahl et al. 2008, 2009; Marot et al.
2011).

1.2. Objective and scope

To tie in with U.S. Bureau of Reclamation’s goal to “quantify the HET for standardization so
engineers now will have a readily available test for piping and internal erosion for risk analysis” (Farrar
et al. 2007), the main objective of this study is to improve the test configuration and procedure, and to
enhance the science-based framework for the interpretation of the Hole Erosion Test. Emphasis is
placed on expanding the focus to fluid mechanics, which influences erosion mechanisms both within
the soil specimen and at the upstream and downstream soil-water interface. This will increase the
potential of obtaining more accurate results, implying possible applications of the HET beyond index
testing.

3
Chapter 1. Introduction

Hydraulic effects and how they influence procedure and analysis of the HET will be investigated
by means of a hydraulic review of the standard HET to identify problems and possible improvements
that will guide the design and setup of a modified test apparatus. In question are in particular the
interpretation of the hydraulic gradient, so far believed to be responsible for erosion, and velocity
measurements with the potential to simplify the method of analysis. Laboratory tests not only on
erodible soil specimens but also using a novel method by means of non-erodible test specimens to
prove the applied changes will conclude this study.

1.3. Outline of the thesis

The structure of this report is based on the type of the different investigations as part of this study,
and is organized as follows:

Chapter 1: Introduction to the topic of internal soil erosion and piping, and description of the
objective and scope of this study.

Chapter 2: A review of research literature dealing with surface erosion as the principal
mechanism in piping with focus on the Hole Erosion Test (HET), as well as a
concluding summary and outline of the research program.

Chapter 3: Description of the Modified Hole Erosion Test (HET-P), including a hydraulic review
of the standard HET, followed by details about the modified design, setup, testing
procedure, and two new energy based methods of analysis of HET-P test data.

Chapter 4: Research program, results and analysis, and discussion of findings from HET-P
tests on three non-erodible test specimens as a novel approach to test the
applicability of the modified apparatus and suggested methods of analysis.

Chapter 5: Research program and main findings from HET-P tests on two types of erodible
soil specimens from four different origins, including a discussion of results and
description of the soil samples used in this study.

Chapter 6: Conclusions and recommendations for further research based on the two
completed series of laboratory tests presented in Chapter 4 and Chapter 5.

4
Chapter 2. Literature review

2. Literature review

This chapter presents a review of research literature related to soil erosion characteristics with
respect to laboratory tests to quantify critical shear stress and erosion rate for surface erosion. These
types of tests can be grouped into the following categories:
1. Flow over surface;
2. Rotating cylinder;
3. Jet impact;
4. Flow through defect.

These four categories, and the related testing procedures, are summarized in the following
sections 2.1, 2.2, 2.3, and 2.4, respectively. The main focus of the review, however, is on the Hole
Erosion Test (HET), an erosion test of category 4, and its modifications, which is described
comprehensively in Section 2.5. A final summary and outline of the research program is given in
Section 2.6.

Various other test methods had been developed and used by researchers to simulate other soil
erosion mechanisms, but were not meant to determine critical shear stress and erosion rate for
surface erosion. These include, but are not limited to, dispersivity tests like the Pinhole Test (Sherard
et al. 1976, ASTM D4647-06), the No Erosion Filter Test (NEF) in which dam core and filter materials
are tested at the same time (Sherard and Dunnigan 1985, 1989; Foster and Fell 2001; Soroush et al.
2008, 2009), or other various methods of internal erosion tests by means of flow through intact soil
samples using permeameter cells. They had been summarized extensively by others, and were not
addressed in detail in this study.

5
Chapter 2. Literature review

2.1. Erosion tests category 1: Flow over surface

- Gibbs (1962)
- Lyle and Smerdon (1965)
- Kandiah and Arulanandan (1974)
- Arulanandan and Perry (1983)
- Shaikh et al. (1988a, 1988b)
- Ghebreiyessus et al. (1994)
- Briaud et al. (1999, 2001)
- Ravens and Gschwend (1999)
- Zhang et al. (2003)
Figure 2.1: Schematic diagram flow over surface

This category comprises test methods for investigating surface erosion in river channels or unlined
canals, where water is flowing parallel to the soil surface at a certain speed and depth. This erosion
mechanism is modelled using hydraulic flume tests, where the specimens are placed on the flume bed
and subjected to open channel flow. Erosion rate is mostly only visually observed and described, while
hydraulic shear stress on the soil surface is deduced from flow velocity and water depth. While this
test is relatively simple, it requires large equipment and suffers from a lack of reproducibility because
of difficulties in controlling surface and soil properties.

Various criteria were developed to describe and evaluate erosion resistance of soil based on
Atterberg limits (Gibbs 1962; Lyle and Smerdon 1965), gradation (Gibbs 1962; Shaikh et al. 1988a),
void ratio (Lyle and Smerdon 1965), water content (Kandiah and Arulanandan 1974; Shaikh et al.
1988a), critical shear stress (Arulanandan and Perry 1983), chemistry (Kandiah and Arulanandan
1974; Shaikh et al. 1988b), or bulk density (Ghebreiyessus et al. 1994).

Flume tests have also been used to compare test results with other test methods like the Rotating
Cylinder Test (RCT). Kandiah and Arulanandan (1974) showed that the two tests produce similar
critical shear stresses, but different erosion rates. Others compared reconstituted and undisturbed soil
samples, showing that reconstituted specimens are less erosion resistant than undisturbed samples
(Zhang et al. 2003).

Special types of flow over surface type of tests are the Erosion Function Apparatus (EFA)
developed by Briaud et al. (1999, 2001) to test undisturbed thin wall tube specimens, or portable flume
tests for in-situ measurements (e.g. Ravens and Gschwend 1999).

6
Chapter 2. Literature review

2.2. Erosion tests category 2: Rotating cylinder

- Moore and Masch (1962)


- Masch et al. (1963)
- Arulanandan et al. (1973, 1975)
- Kandiah and Arulanandan (1974)
- Sargunan (1977)
- Arulanandan and Perry (1983)
- Chapuis (1986a, 1986b)
- Chapuis and Gatien (1986)
- Lim (2006)
Figure 2.2: Schematic diagram rotating cylinder

The several designs of erosion test apparatus for measuring erosion resistance of cohesive soils
using a rotating cylindrical fluid chamber are known as Rotating Cylinder Test (RCT). A soil specimen
is submerged and suspended inside the rotating cylindrical chamber. The chamber is rotated relative
to the soil specimen, which induces a flow around the specimen that applies shear stress to the soil
surface. Shear stress is deduced from the measured torque applied to the stationary specimen. The
most recent methods also allow for measurement of erosion rates by removing the eroded material
from the rotating chamber, and weighing the oven-dried mass. The RCT is difficult to perform, and the
apparatus is costly, but it provides accurate measures of erosion parameters (Lim 2006). However,
test results are influenced by the way the specimen is prepared as different sample preparation
methods may yield a different surface roughness of the specimen (Chapuis 1986a, 1986b).

The RCT was originally developed by Moore and Masch (1962) and Masch et al. (1963), and
similar designs have been used by others (see above). Chapuis (1986a, 1986b) and Chapuis and
Gatien (1986) developed a more advanced design that allowed testing both reconstituted as well as
undisturbed soil specimens, and more accurate measurement of erosion rates. Most recent, Lim
(2006) designed and used a more elaborate apparatus, which allowed for accurate shear stress
measurement, and easier control of the testing process. Lim (2006) also compared RCT results with
HET results, which is described in Section 2.5.5 below.

7
Chapter 2. Literature review

2.3. Erosion tests category 3: Jet impact

- Dunn (1959)
ASTM D5852-00
- Moore and Masch (1962)
Head
- Hanson (1991, 1992)
- Hanson and Robinson (1993)
- Hanson and Simon (2001)
- Hanson and Cook (2004)
- Hanson and Hunt (2006)
- Wahl et al. (2008, 2009)
- Marot et al. (2011)
Figure 2.3: Schematic diagram jet impact

Jet erosion tests are primarily designed to simulate erosion of spillway channels from a
submerged jet. The Jet Erosion Test (JET) – also called Submerged Jet Erosion Test – uses a nozzle
positioned above the center of the submerged specimen to produce a jet perpendicular to the soil
surface. The jet, driven by a constant water head, applies a shear stress to the soil surface, which in
turn experiences scouring. The measured geometry of the produced scour is used to determine both
shear stress and erosion rate, which are used for qualitative classification of the erodibility of soils. The
JET is relatively easy and straight forward to perform, and can be carried out on a wide range of
cohesive soils.

Dunn (1959) first proposed the use of the JET before Moore and Masch (1962) used the JET and
RCT in their research on cohesive soils. Hanson and his companion researchers further developed the
JET, which has become an ASTM standard (ASTM D5852-00). Hanson et al. (see above) developed
advanced and simplified designs to carry out in-situ and laboratory tests using reconstituted or
undisturbed tube samples.

Wahl et al. (2008, 2009) used the JET in their research at the United States Bureau of
Reclamation (USBR), which is developing tools for risk evaluation of piping and internal erosion
(Farrar et al. 2007). They have compared the JET with the Hole Erosion Test (HET), showing that the
two tests yield different results for shear stress and soil erodibility classification. Based on this work,
Marot et al. (2011) developed a new energy based method to obtain a unique soil erosion
classification for HET and JET.

8
Chapter 2. Literature review

2.4. Erosion tests category 4: Flow through defect

- Christensen and Das (1973)


- Hjeldnesa and Lavania (1980)
- Lefebvre et al. (1985, 1986)
Head
- Rohan et al. (1986)
- Wan and Fell (2002, 2004a, 2004b)

Figure 2.4: Schematic diagram flow through defect

Flow through defect erosion tests were categorized as internal erosion tests, and were set out to
simulate conditions along a crack or any other flow path in an earth structure. These types of tests
involve an undisturbed or reconstituted soil specimen in which a preformed defect is introduced prior
to testing. The prepared specimen is then subjected to head or flow rate controlled pressure flow,
while the hydraulic conditions are monitored in order to determine shear stress and erosion rate.

Various methods have been suggested to model the defect and monitor hydraulic conditions.
Christensen and Das (1973) used a 3 mm thick soil lining inside a brass tube to investigate the relation
between erosion rate and critical shear stress. Another method was developed by Hjeldnesa and
Lavania (1980), who formed a crack by applying tension to the soil specimen. But the use of this
method was very limited since the dimensions of the crack were not determined.

Lefebvre et al. (1985) and Rohan et al. (1986) presented a more sophisticated device, the Drill
Hole Test (DHT). They applied a flow rate controlled pressure flow to a cylindrical clay sample with a
predrilled axial hole. The friction head loss along the specimen was measured to determine shear
stress, and any eroded material was collected in a downstream sedimentation tank to calculate
erosion rate. Using the DHT, Lefebvre et al. (1986) showed that naturally structured clay was highly
erosion resistant and considerably less erodible than de-structured and reconstituted specimens.

Wan and Fell (2002, 2004a, 2004b) developed two head controlled devices, the Slot Erosion Test
(SET) and Hole Erosion Test (HET). The two tests are based on the same concepts, which are
comprehensively explained for the HET in Section 2.5. The SET used a larger 1 m long reconstituted
soil specimen with preformed slot along the side rather than in the center of the specimen.

9
Chapter 2. Literature review

2.5. Hole Erosion Test (HET)

The Hole Erosion Test (HET) was developed at the University of New South Wales (UNSW) as an
index test to model piping erosion in concentrated leaks in earth dams (Wan and Fell 2002, 2004a,
2004b). It is generally a faster and more economical alternative to the Slot Erosion Test (SET)
mentioned above with the goal to study relationships between basic engineering properties and
erosion characteristics of different types of soil. It allows testing of a much smaller soil specimen at
lower water heads. In a HET, a reconstituted soil specimen with predrilled axial hole is subjected to a
constant-head pressure flow. Erosion rate and shear stress are determined using measured flow rate,
hydraulic gradient, and final hole diameter. With most emphasis on the rate of erosion rather than
critical shear stress, test soils are characterized by a soil group number, which typically ranges from 1
(extremely rapid erosion) to 6 (extremely slow erosion) (Wan and Fell 2002, 2004a, 2004b).

Further work at UNSW (Lim 2006) and the United States Bureau of Reclamation USBR (Farrar et
al. 2007; Wahl et al. 2008, 2009) yielded technical improvements to the test apparatus and procedure,
and interpretation of test data. French researchers also developed a non-dimensional numerical model
for the interpretation of HET data, which simplifies post-test measurements and data analysis (Bonelli
et al. 2006; Bonelli and Brivois 2008). Most recently, Marot et al. (2011) developed an energy based
model for providing a unique erodibility ranking for different erosion tests.

2.5.1. UNSW Hole Erosion Test (HET)

In a standard HET, a soil specimen is reconstituted by compaction inside a standard compaction


mold at 95% maximum dry density and optimum water content. These properties are standardized
because test results are strongly influenced by the degree of compaction and water content (Wan and
Fell 2002, 2004a, 2004b). An axial hole with diameter o = 6 mm is drilled prior to testing. The
specimen is subjected to a constant-head pressure flow (Figure 2.5). The hydraulic head difference,
h , over the length of the soil specimen, L , describes the hydraulic gradient, s , which is increased
stepwise until progressive erosion (enlargement) of the hole is produced. Ideally, the upstream head is
initially set to a height where progressive erosion immediately starts. But choosing an appropriate
upstream head requires a few trial test runs. Once progressive erosion is produced, the upstream
head remains constant until the end of the test. A test is stopped before the eroded hole expands to
the side of the mold, or the maximum available flow rate is reached. During a standard HET, the
hydraulic gradient across the specimen, s , and flow rate, Q , are measured at select time intervals.
After the test, the size and shape of the eroded hole is carefully recorded, and a final hole diameter,
f , representative of the entire length of the eroded axial hole is defined.

10
Chapter 2. Literature review

Constant head tank,


adjustable in height

difference, h = hu - hd
Hydraulic head

(50-1200mm)
Two vertical
piezometer tubes,
connected to sidewall
Eroding of flow chambers,
fluid providing hydraulic
circulation head upstream, hu ,
system and downstream, hd L

Q
50 mm Upstream flow Compacted soil Downstream
diameter chamber filled with specimen with 6 mm flow chamber,
pipe 20-mm gravel axial hole empty

Figure 2.5: Schematic diagram Hole Erosion Test HET (adapted from Wan and Fell, 2004)

Measured hydraulic gradient across the soil specimen and flow rate are used to indirectly
determine the erosion rate per unit surface area, , and wall shear stress along the axial hole,
, at any time during the test. A plot of versus (Figure 2.6) is then used to graphically
determine critical shear stress, c , coefficient of soil erosion, Ce , and erosion rate index, ,
presuming:

(2.1)

Equation (2.1a) expresses soil erodibility in terms of erosion rate when the applied shear stress
exceeds the critical value. The rate of mass removal per unit surface area to represent the erosion rate
is considered most appropriate because porosity and density of the soil material is taken into account.
The coefficient of soil erosion is defined as the slope of the linear best-fit line where both and
increase. Critical shear stress describes the ease of erosion and has the physical meaning of the
value of shear stress at which erosion starts. It is defined as the x-intercept of the extrapolated linear
best-fit line as illustrated in Figure 2.6.

11
Chapter 2. Literature review

0.008
Estimated erosion rate per unit surface area of the
0.007 Early stage: Later stage:
disturbed and loose both ėHET and HET are increasing.
0.006 material is removed.

Coefficient of soil erosion, Ce defined


hole, ėHET [kg/s/m2]

0.005
as the slope of the linear best-fit line:
Ce = 3.96E-5 s/m
 d dt  d t
0.004 Erosion rate index, IHET = -log10(Ce) : eHET  
IHET = 4.4 2 dt 2 t
t
0.003  HET   w gs
4

0.002
Critical shear stress, c defined as the
0.001 x-intercept of the linear best-fit line:
c = 190 N/m2
0.000
100 150 200 250 300 350 400

Estimated wall shear stress, HET [N/m2]

Figure 2.6: Typical results Hole Erosion Test (schematic, no real test data)

The erosion rate index, IHET , defined in equation (2.1b) was introduced instead of Ce , which
-1 -6
usually is a small number ranging from 10 to 10 . Typical values of IHET range from 0 to 6, where a
low value indicates a soil that erodes more rapidly than a soil with a higher value. Test soils are
classified into six groups based on their representative erosion rate index, which describes the
standardized value of IHET at 95% maximum dry density and optimum water content (Table 2.1).

Table 2.1: Qualitative relation of representative erosion rate index and progression of internal
erosion (Wan and Fell 2002, 2004a, 2004b)

Group number Erosion rate index, IHET Progression of internal erosion


1 <2 Extremely rapid
2 2–3 Very rapid
3 3–4 Moderately rapid
4 4–5 Moderately slow
5 5–6 Very slow
6 >6 Extremely slow

12
Chapter 2. Literature review

2.5.2. HET theory

The underlying theory to derive shear stress is based on the force equilibrium on the body of
eroding fluid along a differential length of an axial circular hole:

(2.2)
where:
= retaining force, N
= forcing force, N
 = hydraulic shear stress along wetted area, N/m
2

P =  = wetted perimeter ≡ circumference of circular cross section, m


 = hole diameter, m
L = length of axial hole, m
2
= cross-sectional area of axial hole, m
2
p = fluid pressure, N/m

In a uniform circular cross section with fully developed flow, the differential pressure along
can be expressed by the energy head loss due to wall friction, or friction head loss, :

(2.3)
where:
g = specific weight of eroding fluid, N/m
3

Combining equations (2.2) and (2.3) gives

(2.4)

The term in equation (2.4) is called hydraulic radius, and is equal to for a
circular cross section. Equation (2.4) is integrated over the total length of the soil specimen by
introducing the following assumptions:
1. Flow through the soil matrix is negligible;
2. Only the soil surface along the preformed hole provides shear resistance;
3. Energy losses due to the eroding fluid entering and exiting the preformed hole are negligible;
4. Uniform circular cross section along the length of the soil specimen;
5. Fully developed flow profile in the preformed hole throughout the length of the soil specimen;
6. Hydraulic head difference across the soil specimen, h , equals total friction head loss, hf .

13
Chapter 2. Literature review

For a given diameter, the shear stress is then directly proportional to the friction head loss along
the length of the soil specimen, which is represented by the hydraulic gradient across the soil
specimen:

(2.5)

(2.6)

where:
HET = estimated wall shear stress along axial hole, N/m
2

w = density of water used as eroding fluid, kg/m


3

2
g = acceleration of gravity, m/s
t = estimated mean diameter of axial hole, m
L = length of axial hole, m
s = hydraulic gradient across the soil specimen in standard HET
2
hu = measured upstream sidewall hydraulic head, N/m
2
hd = measured downstream sidewall hydraulic head, N/m

Further, the erosion rate per unit surface area is given by:

(2.7)

where:
2
= erosion rate per unit surface area, kg/s/m
d = dry density of the soil, kg/m
3

2
= change in cross-sectional area with time, m /s
= change in diameter with time, m/s
= change in diameter t over a short time interval t , m/s

Equation (2.7) makes use of the following two additional approximations:


7. The change in radius , which yields and thus ;

8. The derivative can be approximated by ;

14
Chapter 2. Literature review

2.5.3. HET data analysis

Both shear stress, , and erosion rate, , depend on the diameter of the preformed and
eroded axial hole, t , which cannot be measured directly during a HET, and has to be indirectly
estimated from the measured flow rate, Q , and hydraulic gradient, s . It is assumed that the change in
diameter correlates to the change in assumed friction factors, fL and fT , which are based on continuity
and known common relations between shear stress,  , and velocity, V, for laminar and turbulent flow
conditions:

(2.8)

(2.9)

(2.10)

Equations(2.5), (2.8), (2.9), and (2.10) can be combined to:

(2.11)

Since diameter and flow conditions are known for both the beginning and the end of a test, the
initial and final friction factors can be determined, and interpolated over time to estimate diameter at
any particular time t during a HET:

(2.12)

The flow conditions to determine whether Equation (2.12a) or (2.12b) will be used, are governed
by the Reynolds number, Re :

(2.13)

where:
Vt = estimated mean flow velocity in axial hole, m/s
 = kinematic viscosity, m /s (water at 20°C:  = 1.004E-06 m /s)
2 2

15
Chapter 2. Literature review

Flow is laminar below a critical value of Re , defined below. For the use in HET, flow is considered
turbulent if the Reynolds number exceeds this critical value.

The above analytical solution to analyze HET data requires the following additional assumptions
and approximations:
2
9. Shear stress is proportional to V in turbulent flow conditions;
10. Pure laminar or turbulent flow throughout the entire test;
11. Friction factor fL or fT is linearly interpolated between its initial (t = 0) and final value (t = tf);
12. Flow conditions are turbulent if Reynolds number Re > 5000;
13. The measured final hole diameter, f , is representative of the entire length of the eroded hole.

The complete analysis of HET data involves the following steps:


i) Define initial and final flow conditions using Equations (2.8) and (2.13), and identify a
representative flow condition (laminar or turbulent) to be used for this test;
ii) Estimate the initial friction factor fL,o or fT,o based on the initial diameter of the preformed hole,
o = 6 mm, using Equation (2.11);
iii) Estimate the final friction factor fL,f or fT,f based on the measured final diameter of the eroded
hole, f , using Equation (2.11);
iv) Interpolate the friction factor fL or fT linearly between its initial (t = 0) and final value (t = tf);
v) Estimate the diameter of the axial hole, t , at any time during the test using Equation (2.12);
vi) Plot a curve of estimated diameter, t , against time, t ;
vii) Estimate the slope , if appropriate approximated by ;
viii) Estimate wall shear stress, HET , using Equation (2.5);
ix) Estimate erosion rate , , using Equation (2.7);
x) Plot against HET , and fit a linear straight line through the rising part of the curve;
xi) Determine coefficient of soil erosion, Ce , and erosion rate index, IHET , using Equation (2.1);
xii) Graphically obtain critical shear stress, c , as illustrated in Figure 2.6.

16
Chapter 2. Literature review

2.5.4. Modifications to the HET


Lim (2006), University of New South Wales UNSW
Lim (2006) noted some problems with the Hole Erosion Test during his PhD work at UNSW. The
preformed hole was introduced by drilling. It was observed that this method can lead to smearing and
remoulding of the surrounding soil, which lead to a denser surface layer and a higher critical shear
stress. Compaction of the soil into the test mold introduced inhomogeneity that could lead to vertical
layering with different densities. Both of these problems eventually lead to delayed erosion that
affected test results. Another problem was the effects of slaking, where soil particles detached from
the specimen due to the presence of water (hydrostatic conditions), rather than the applied shear
stress. This lead to a reduced hole length and difficulties defining a representative final hole diameter.
To avoid these problems, Lim (2006) proposed modifications to:
1. the estimation of friction factors fL and fT ;
2. the specimen preparation;
3. the interpretation if a test was affected by slaking.

It was found that the friction factors fL and fT do not vary linearly with time, and a better correlation
was found between the friction factor and diameter. It was assumed that the friction factors are linearly
proportional to the diameter of the eroded hole. However, this implied practical difficulties because the
diameter is not known until the analysis is complete. An iterative predictor approach was thus adopted,
introducing new assumptions and complexity to the analysis. It is further proposed herein that a better
correlation between friction factors and diameter can be assumed since diameter is used to calculate
friction factors (Equation (2.11)).

With respect to specimen preparation, it was recommended to increase the number of compaction
layers to produce a more homogeneous test specimen. To reduce the effect of remoulding during
drilling of the preformed hole, it was further recommended to use a slowly penetrating and sharp auger
drill. Emphasis was also placed on measuring the final hole diameter as precisely as possible, since
this measurement is crucial for the estimation of shear stress.

The effect of slaking and the resulting reduced hole length was investigated in terms of estimated
wall shear stress. It was found that small reductions of the hole length up to about 20 mm can be
ignored. Ignoring slaking up to about 40 mm resulted in minor errors of less than 10% for the
estimated wall shear stress, and negligible errors for the HET erosion rate index. Lim (2006) proposed
a method to correct for slaking if the amount was more than 40 mm. However, it is questionable how
representative a HET would be if almost half of the soil specimen is lost by failing mechanisms under
hydrostatic conditions.

17
Chapter 2. Literature review

Farrar et al. (2007), Wahl et al. (2008, 2009), the United States Bureau of Reclamation USBR
The Hole Erosion Test was further studied at the United States Bureau of Reclamation USBR
(Farrar et al. 2007; Wahl et al. 2008, 2009). During the research program at USBR, the apparatus and
data collection procedures proposed by Wan and Fell (2002, 2004a, 2004b) were considerably
improved:
- Flow rate was measured by a custom 10° v-notch weir at the downstream end of the setup;
- Automated head measurements upstream and downstream of the soil specimen, as well as at
the v-notch weir using pressure transducers and computerized data acquisition system with
recording intervals of 5 s throughout a test;
- High-head HET setup to produce test heads of up to 5400 mm;
- Successively doubled test head during a HET until progressive erosion was observed, starting
at a low test head of usually 50 mm.

The USBR further defined 5 major issues affecting the HET interpretation:
1. Identification of erosion regimes;
2. Curve fitting procedures;
3. Laminar versus turbulent flow;
4. Variation of friction factors;
5. Determination of final hole diameter and length.

Progressive erosion was indicated by an accelerating flow rate at a constant test head. The USBR
emphasized that only data collected during the period of progressive erosion should be considered in
the data analysis to determine erosion parameters. Data collected prior to progressive erosion
described the removal of disturbed and loose material, and were not useful for defining erosion
characteristics of the intact soil.

The development of the hole diameter, t , during progressive erosion was assumed to follow a
polynomial function, whose time derivative defines the erosion rate, , as shown in Equation (2.7).
The degree of the fitted polynomial curves varied between different types of soils. To find the
coefficient of soil erosion, Ce , from the plot, another linear curve fitting was necessary.
This need for data smoothening and adjustment adds complexity to the analysis procedure.

The USBR reported difficulties in justifying the use of the “virtual” friction factors fL and fT
introduced by Wan and Fell (2002, 2004a, 2004b) in regards of flow conditions. They also questioned
the high critical Reynolds number of Re,crit = 5000 used at the UNSW. They declared flow conditions as
turbulent if Reynolds number Re > 2000, as it is widely recognized in traditional fluid mechanics.

18
Chapter 2. Literature review

Friction factors were assumed to vary linearly with time. This could cause erroneous results,
especially if a test was started at a low head. In such cases, erosion was computed in the early stage
of a test, even though there was no observable sign of erosion. Also irregularities in computed hole
diameter suggested that the friction factors were incorrectly modeled. An independent investigation of
this problem revealed a better relationship between the friction factor and estimated hole diameter,
similar to the findings of Lim (2006). But the iterative solution method used by the USBR to determine
friction factors and hole diameter showed problems obtaining convergence. Thus, this method was
simplified by assuming that friction factors vary in proportion to for laminar flow and
for turbulent flow. These terms are substitutes for the hole diameter, t .

Also the USBR experienced difficulties in measuring the final hole diameter due to irregularities of
the eroded hole. These were observed especially at the entrance and exit of the soil specimen. These
points are prone to slaking and scouring due to eddies at the soil interface. To reduce these problems,
they introduced end plates with an orifice opening of 15 mm for stronger soils or 25 mm for weaker
soils. Unfortunately this method was less effective than expected. They also investigated several
methods to measure the final hole diameter, including measurements from plaster castings of the
eroded hole. To account for the changing length of the eroded hole in some tests, they assumed that
the length of the eroded hole varies linearly with time.

Bonelli et al. (2006), Bonelli and Brivois (2008), Cemagref, France


French researchers around Stéphane Bonelli developed a universal erosion model for permanent
flow over erodible soil. They adapted their model for piping erosion to be applicable to the Hole
Erosion Test by spatial integration of the system over a cylindrical volume representing the axial hole.
It provides an alternative method for analysing HET data, which allows the determination of the
estimated erosion characteristics without the need of measuring and interpolating the hole diameter.
This non-dimensional numerical model correlates a dimensionless hole radius with critical shear
stress, hydraulic gradient, and a dimensionless test time. It has only two unknown parameters that can
be solved with a simple non-linear numerical solver. The model is promising, but the list of
assumptions and limitations restricts its applicability to tests with (not conclusive):
- constant pressure drop (constant test head and hydraulic gradient);
- turbulent flow conditions (large Reynolds numbers);
- no variation of the friction factor;
- constant length and uniform diameter of the eroded axial hole;
- test heads exceeding critical shear stress, c ;
- data only collected during progressive erosion period.

19
Chapter 2. Literature review

Marot et al. (2011), Université de Nantes and the United States Bureau of Reclamation USBR
The Université de Nantes in collaboration with the USBR developed a new method to analyze
HET and Jet Erosion Test (JET) data based on fluid energy dissipation and measurement of eroded
mass to provide a unique erodibility classification of test soils. In HET, energy dissipated by erosion
(friction head loss) is correlated with measured hydraulic gradient and flow rate. An energy balance
equation for the fluid was applied between the upstream and downstream point where the hydraulic
heads are measured at the sidewall of the flow chamber (Figure 2.5). It takes into account minor head
losses due to fluid entering (flow contraction) and exiting (flow expansion) the axial hole, and assumes
same average velocities in both measuring sections. It was thus assumed that the hydraulic head
difference across the soil specimen, h , equals total energy head loss, which is the sum of friction
and minor energy head losses. Based on a series of HET on a non-erodible poly-acrylic model of the
specimen with predrilled 6-mm hole, an empirical constant was determined to isolate unknown friction
head losses responsible for erosion. It was found that, for this particular hole diameter, only about 25%
of the measured hydraulic head difference across the soil specimen, h , is due to wall friction.

2.5.5. Main findings and conclusions


Wan and Fell (2002, 2004a, 2004b), University of New South Wales UNSW
Wan and Fell (2002, 2004a, 2004b) did extensive tests on 13 different soil samples covering a
range of soil properties and origins. Most soils were from borrow areas used for dam construction in
Australia, but also included one each from the USA and New Zealand. The utility of the HET was
demonstrated by a total of 225 HET’s carried out in this research program. It was found to be fast,
simple, and easy to reproduce under identical test conditions.

They found that fine-grained and some plastic soils are more erosion resistant than coarse-
grained non-plastic soils and soils with low plasticity. Most of the soils showed higher erosion
resistance if they were compacted to the wet side of optimum water content and a higher dry density.
Results further suggested that the mineralogy, especially iron oxides content, plays an important role
regarding erosion resistance of a soil.

They suggested using an initial shear stress, o , rather than critical shear stress, c , in order to
describe initiation of erosion. Results of critical shear stress were mostly scattered, possibly due to
extrapolation of data, observed non-linearity in the coefficient of soil erosion, and other simplifications
and assumptions in the process of data analysis. The initial shear stress, o , refers to the shear stress
corresponding to the minimum water head at which erosion is first initiated. It is obtained
experimentally by multiple HET trial runs on identical test specimens.

20
Chapter 2. Literature review

Wan and Fell (2002, 2004a) further proposed predictive equations to estimate the HET erosion
rate index, IHET , based on multiple linear regression models obtained from statistical analysis of HET
data. They are not presented herein because it was recommended to use them with great caution as
they are based on a limited number of soils, and may not imply any strong relationship between the
erosion rate and the predictor variables.

In summary, the HET was considered a simple and rapid index test, but one that does not intend
to provide accurate quantitative measurements since all derived erosion parameters are indirectly
estimated from a few measured hydraulic parameters.

Lim (2006), University of New South Wales UNSW


Lim (2006) conducted a total of 139 HET’s at the UNSW on 9 natural clay soils and 5 engineered
artificial soil mixtures. The main findings were similar to those by Wan and Fell (2002, 2004a, 2004b).
A distinction was made between dispersive and non-dispersive erosion. Dispersive erosion, indicated
by dirty or cloudy outflow, was characterized by immediate erosion and fast enlargement of the hole,
implying considerable erosion at small shear stress. These soils were rarely affected by effects of
slaking, and erosion occurred evenly over the surface of the axial hole. Non-dispersive soils, on the
other hand, experienced severe slaking and reduction of the hole length, and the shape of the eroded
hole was usually irregular. Lim (2006) also claimed that the HET provides limited information about the
erosion behaviour of clay soils due to the reported deficiencies. A comparison with Rotating Cylinder
Test (RCT) results further revealed that the erosion rate indices from HET were significantly higher
than those from the RCT for non-dispersive soils, indicating slower erosion and higher shear stresses.

Farrar et al. (2007), Wahl et al. (2008, 2009), the United States Bureau of Reclamation USBR
The USBR tested a total of 10 soil samples. Test specimens were either reconstituted or
undisturbed. They claimed the HET to be difficult and less reproducible, especially for weak and very
strong soils, while it was good for intermediate soils. Collapsing and scouring were the main problems
in weak soils, while clogging of the axial hole and insufficient test head limited the success rate in
testing strong soils. These problems often lead to subjective interpretations and combinations of
analysis methods. They introduced a subjectivity index to quantify the level of uncertainty in test
results. It ranges from 0 to 3, with low values indicating high confidence in test results. They further
compared HET with Jet Erosion Test (JET) results. The JET indicated lower erosion resistance for all
test soils, including critical shear stresses that were as much as an order of magnitude smaller than
those from HET. With reference to Lim (2006), it was noted that differences between HET and JET are
of a similar order of magnitude as between HET and RCT.

21
Chapter 2. Literature review

Bonelli et al. (2006), Bonelli and Brivois (2008), Cemagref, France


The numerical model developed by Bonelli et al. (2006) was tested on 17 (18 in Bonelli and
Brivois, 2008) HET data sets from 9 different soil samples collected by Wan and Fell at the University
of New South Wales. Despite the many simplifying assumptions, they showed good agreement
between the analytical solution proposed by Wan and Fell (2002, 2004a, 2004b) and the new
numerical solution. This method was also successfully applied by the USBR (Farrar et al. 2007; Wahl
et al. 2008, 2009) and Marot et al. (2011), solely or in combination with the analytical solution.
Recognizing that the model is based on even more restrictive assumptions than the analytical solution,
and that it only considers processes within the cylindrical volume of the eroded hole, it deals with
similar issues as the analytical solution.

Marot et al. (2011), Université de Nantes and the United States Bureau of Reclamation USBR
The motivation for the new energy-based method developed by Marot et al. (2011) to obtain a
unique soil erosion classification for HET and JET was based on observed differences between the
two tests. Consistent with previously reported data by the USBR (Farrar et al. 2007; Wahl et al. 2008),
results from 17 paired HET and JET tests on seven different soil samples with a large range of
erodibility showed that erosion rate indices from HET were consistently larger than those from JET,
and that critical shear stresses from JET were on average about 50 times smaller than those from
HET. Thus, the two tests yielded different soil classifications. Based on the new method, Marot et al.
(2011) introduced an erosion resistance index, , used for soil erodibility classification similar to the
classification system proposed by Wan and Fell (Table 2.1). Test soils are classified into six categories
of soil erodibility, ranging from highly erodible for , to highly resistant for .They showed
that the new energy based method applied to the 17 paired tests yields the same soil classification for
both the HET and JET. However, it is important to note that this method only provides an erosion
resistance index for soil erodibility classification, and no absolute values of erosion rate and, more
importantly, critical shear stress.

22
Chapter 2. Literature review

2.6. Summary and outline of the research program

After the development of various field and laboratory tests to investigate different mechanisms of
soil erosion, the Hole Erosion Test (HET) was developed as a simple and economical laboratory test
to simulate soil erosion in a crack or pipe within an earth dam. Understanding the mechanisms
involved in the HET is a work-in-progress. After the development and first improvements of the HET at
the UNSW, the USBR automated and further improved the test, and also applied it in several cases.
Bonelli et al. added a mathematical background to the more pragmatic approach, while Marot et al.
introduced an energy-based method for providing a unique soil erosion ranking to be comparable with
JET. Extensive testing had been done using the HET with various attempts to improve specimen
preparation, testing and analysis procedures. Table A.1 in Appendix A provides a compiled list of
challenges and issues affecting the HET and offers suggestions how they might be addressed.

Most of this work focused on the behaviour of the axial hole, with the boundaries set at the soil-
water interface upstream and downstream of the test specimen. However, the HET involves more
considerations. The specimen is placed inside a hydraulic pipe system that interacts in more ways
than just applying a shear stress to the soil specimen. There is a need to combine soil and fluid
mechanics, and investigate the interacting mechanisms involved. Furthermore, very limited testing has
been done under controlled and steady flow conditions on non-erodible materials as a reference to
support findings from HET. The main question to be asked is:

Are measurements and what they represent in the standard HET influenced by any hydraulic
effects not captured in the current methods, and how could they be further improved?

The only measurements recorded during a HET are the hydraulic gradient, s , and flow rate, Q .
Since flow rate is usually measured at the downstream end of the system, it is most likely unaffected
by any local hydraulic effects. The hydraulic gradient, on the other hand, involves hydraulic head
measurements (Figure 2.5) that are affected by local turbulences, such as eddy formations due to a
change in flow area downstream of the soil specimen. The resulting wall shear stress, HET , will be
directly affected, if the hydraulic heads, and thus the hydraulic gradient, were in fact erroneous
(Equation (2.5)).

This question is supported by various researchers as described above (Lim 2006; Farrar et al.
2007; Wahl et al. 2008; Marot et al. 2011), who compared HET data with other surface erosion tests
like the Jet Erosion Test (JET) and Rotating Cylinder Test (RCT). All reported results agree that
standard HET generally produces higher erosion indices, and overestimates critical shear stress by up
to an order of magnitude or more. Given that all these tests were designed for similar purposes,

23
Chapter 2. Literature review

reasons for this behaviour were generally assumed to be in conceptual differences between the
compared tests, like the manner of hydraulic attack, geometric factors, and methods to measure or
estimate erosion parameters.

2.6.1. Outline of the research program

The main objectives of this research program were to investigate hydraulic effects and how they
influence data collection and analysis of the Hole Erosion Test, and to use these findings to improve
the testing procedure and analytical solution. This research consisted of the following four main parts:

1. Hydraulic review of the standard HET to identify problems and possible improvements;
2. Design and setup of a modified test apparatus incorporating suggested improvements;
3. Laboratory calibration and applicability tests using non-erodible PVC specimens;
4. Modified Hole Erosion Tests on reconstituted and undisturbed erodible soil samples.

A critical hydraulic review was conducted on the whole system with focus on the test cell including
test specimen and adjacent flow cambers (Figure 2.5). Hydraulic effects upstream and downstream of
the soil specimen were analysed to identify possible influences on measurements and interactions
with the soil specimen. The review was restricted to the ideal case of a test specimen with a straight
axial hole of uniform circular cross section.

Based on findings from the hydraulic review, improvements to the test apparatus were implied in a
modified design with additional measurements, accompanied by an improved and simpler method to
analyse HET data (analytical solution). The modified design was used in subsequent laboratory
experiments to test the proposed modifications.

The first set of laboratory experiments was conducted on non-erodible PVC specimens with axial
holes of different diameter. They were used to simulate various stages in a HET under controlled and
steady flow conditions. This assured a robust framework to test the modified test apparatus and
analysis procedure. Details of these tests are given in Chapter 4 below.

To demonstrate its applicability to soils testing, reconstituted and undisturbed soil samples from
different Canadian sites were tested in a final set of laboratory experiments using the modified test
apparatus. The improved method of analysis was used in conjunction with the originally proposed
analytical solution to identify advantages and weaknesses of the Modified Hole Erosion Test (HET-P).
Chapter 5 presents the details and findings from this set of tests.

24
Chapter 3. Modified Hole Erosion Test (HET-P)

3. Modified Hole Erosion Test (HET-P)

This chapter presents details about the Modified Hole Erosion Test (HET-P) that was developed
and used in this study. A hydraulic review of the standard HET is presented in Section 3.1, describing
shortcomings and suggested improvements of the standard test method. Section 3.2 describes the
design of the modified HET-P apparatus based on findings of the hydraulic review. The experimental
setup and test procedure of the HET-P is outlined in Section 3.3, including calibration of instruments
and test preparation. Two modified energy gradient-based analysis methods for HET-P test data are
introduced in Section 3.4.

3.1. Hydraulic review of the HET

3.1.1. Friction head loss and hydraulic gradient

Head loss due to wall friction is defined as the difference in total energy head across a uniform
test section (Bernoulli’s theorem, Equation (3.1)). The HET assumes that the friction head loss along
the axial hole, hf , is represented by the hydraulic gradient across the specimen, s , determined by the
difference between the hydraulic heads measured at the sidewall of the flow chamber just upstream
and downstream of the test specimen (Equation (2.5) and (2.6)). This approach is commonly used in
testing porous media, such as in the Gradient Ratio Test (ASTM D5101-01 2006), where velocity
2
heads, V / 2g , are small and can be ignored because the energy grade line (EGL) and hydraulic
grade line (HGL) remain parallel. As assumed by Marot et al. (2011) for the HET, this approach can
also be used in applications with similar velocities in both measuring sections, where the velocity
2 2
heads upstream and downstream are of equal value (Vu / 2g ≈ Vd / 2g), and the EGL and HGL have
similar slopes. In both cases, the difference between the sidewall hydraulic heads upstream and
downstream does provide a good measure of total energy head losses along the test section.

However, in the HET this approach is questionable for a number of reasons. Firstly, it is well
known from pipe-flow hydraulics (e.g. see Orifice Meter, Finnemore and Franzini 2002, p. 522) that an
abrupt expansion, such as on the downstream side of the specimen in a HET, leads to flow
recirculation and a reduction in the hydraulic head, hd , measured at the sidewall (Figure 3.1). Thus,
the sidewall hydraulic head may not be representative of the entire pipe section at that point.
Secondly, the velocity head downstream would not likely be equal to the upstream approach flow
because of the high-velocity jet exiting the axial hole. The velocity head downstream can be significant
and would cause a considerable additional reduction in the sidewall hydraulic head, hd . Thirdly,
expansion losses downstream of the test specimen are expected to be significant. These expansion

25
Chapter 3. Modified Hole Erosion Test (HET-P)

losses represent a total energy head loss, but not a friction head loss, and would not contribute to
erosion along the axial hole. The consequence for the standard HET is that the hydraulic head
difference across the soil specimen, h = (hu – hd) , is not necessarily representative of the friction
head loss along the axial hole of the test specimen, hf , and shear stress derived from this assumption
would likely be overestimated.

Applying Bernoulli’s theorem to the HET yields:

(3.1)

where:
= pressure head, m
z = elevation head, m
= velocity head, m
H = energy head loss along the test specimen, m
h = = hydraulic head, m
H = total energy head, m
u = subscript denoting upstream
d = subscript denoting downstream

To capture the friction head loss along the axial hole and exclude downstream expansion
anomalies, the upstream and downstream total energy head need to be measured close to the
specimen. The upstream total energy head, Hu , can be determined by adding the upstream velocity
2
head, Vu / 2g , to the measured upstream sidewall hydraulic head, hu (Equation (3.1)). The
downstream total energy head, Hd , can be measured using a Pitot tube positioned at the point with
the highest local energy head in the center of the specimen where the jet emerges from the axial hole.
The tip of the probe must be pointing directly into the flow (Figure 3.1).

26
Chapter 3. Modified Hole Erosion Test (HET-P)

EGL

HET-P:
HET:
HGL
(at wall)

Standard
HET:

Specimen
Flow profile

Modified
HET-P:

Upstream Downstream

Figure 3.1: Flow pattern upstream and downstream of test specimen, including Energy Grade
Line (EGL), and Hydraulic Grade Line (HGL) at wall

3.1.2. Minor head losses and entrance length

In turbulent flow, minor head losses develop in addition to friction head losses at the entrance and
exit of the axial hole due to eddy turbulences induced by a sudden change in cross section. Minor
entrance losses due to flow contraction can be assumed as negligible because head losses in
accelerating flow like this are usually small, and continue over a considerable length downstream
exceeding the length of the axial hole (L ≈ 117 mm). They are primarily the result of turbulences
created as the stream enlarges between the vena contracta (minimum flow area) and the end of the
entrance length upon the flow is fully developed. There is no single equation to estimate the entrance
length because it depends on various factors like entrance conditions or wall roughness, but it can be
assumed that flow is fully developed within 20 to 40 pipe diameters (e.g. see Viscous Sublayer in
Turbulent Flow, Finnemore and Franzini 2002, p. 271). Minor exit losses due to flow expansion are
avoided if the total energy head is measured at the point where the jet emerges from the axial hole.

27
Chapter 3. Modified Hole Erosion Test (HET-P)

3.1.3. Flow conditions and axial hole velocity

Conditions of either laminar or turbulent flow can be expected in the piping and flow chambers
upstream and downstream of the test specimen. While flow conditions in the piping are turbulent, it
may still be laminar in the flow chambers. The change in cross section from piping to the upstream
flow chamber will cause a transition from turbulent to laminar flow within the flow chamber at low to
medium flow rates (approx. 4-11 l/min). To assure fully-developed and uniform flow at the upstream
specimen interface, the flow chamber needs to be adequately long. A rule-of-thumb suggests a
transition length of LT ≈ (6-10)D , where D = diameter of flow chamber. Wan and Fell (2002) suggested
using flow chambers of the same size as the standard compaction mold the specimen is prepared in
(4x4 in), which is too short. In order to avoid positive feedback of downstream expansion anomalies
described above, the downstream flow chamber must be of adequate length as well, preferably the
same length as upstream.

With respect to measuring the downstream total head, Hd , of the exiting jet with a Pitot tube
(Figure 3.1), it would be practical to use a combined Pitot-static tube with the advantage of getting a
measure of the centerline jet velocity, umax , which can be correlated to the mean flow velocity in the
axial hole, Vt (Figure 3.2). This would simplify the analysis of HET data significantly, as described in
Section 3.4.2 below. The theoretical ratio of the mean pipe flow velocity to the centerline velocity in a
uniform circular pipe depends on the velocity profile, and can be described by the pipe factor for the
corresponding flow regime (e.g. see Laminar Flow in Circular Pipes, Finnemore and Franzini 2002, p.
264; Velocity Profile in Turbulent Flow, Finnemore and Franzini 2002, p. 276):

(3.2)

where:
 = Darcy friction factor

In laminar flow, the velocity profile is a parabola, and independent of pipe roughness. However,
due to the small hole diameter compared to relatively high flow rates, it is more likely that turbulent
flow conditions are observed in HET (Wan and Fell 2002, 2004b; Wahl et al. 2008). The shape of the
velocity profile in turbulent flow depends on Reynolds number, diameter, and pipe roughness, as
indicated by the Darcy friction factor,  , in Equation (3.2b). It is steeper near the center for a rough
pipe than for a smooth pipe. The numerical constant 1.326 in Equation (3.2b) is a theoretically derived
value that may be replaced by an empirical value of 1.44. Equation (3.2b) yields values for the pipe
factor of about 0.74 to 0.88 for rough pipes with typical values of  = 0.01-0.07.

28
Chapter 3. Modified Hole Erosion Test (HET-P)

Pipe Factor:

laminar:

turbulent: Turbulent 0.2


velocity umax Vt 0.4
profile
0.6
Total pressure, 0.8
Dynamic pressure,
1.0
Static pressure, Total pressure
probe
Holes for Sensing tip
Pitot-static tube;
static pressure
KIMO TPL-03-200

Figure 3.2: Pitot-static tube with schematic turbulent velocity profile of jet exiting axial hole
immediately adjacent to test specimen

29
Chapter 3. Modified Hole Erosion Test (HET-P)

3.2. HET-P apparatus

The HET-P apparatus incorporates a conventional Pitot-static tube in order to obtain additional
measurements of the total energy head, Hd , and velocity head, hv , close to the outlet of the axial hole,
so as to exclude downstream expansion anomalies, and to obtain a measure of the centerline jet
velocity, umax , for a more direct estimate of the mean flow velocity and diameter of the axial hole.

The general setup and design of the apparatus was based on the standard Hole Erosion Test
assembly by Wan and Fell (2002, 2004a, 2004b), and the more recent modifications by the U.S.
Bureau of Reclamation (Farrar et al. 2007; Wahl et al. 2008, 2009). A schematic diagram and
photograph of the HET-P apparatus is shown in Figure 3.3 and Figure 3.4, respectively. The
apparatus consisted of three major components;
1. Test cell, including flow chambers, test specimen with axial hole, and hydraulic head
measurements upstream and downstream of the specimen;
2. Pitot-static tube and differential pressure transducer system;
3. Water flow system, including upstream (US) constant head tank to control test head, and
downstream (DS) constant head tank with flow rate and temperature measurements.

Inflow

Constant head tank,


adjustable in height

Constant head and


Total Centerline Sidewall sedimentation tank,
energy velocity hydraulic head 10° v-notch weir
head head US, hu , and DS, hd for flow rate, Q

US
Hd hv hu Datum:
T Q
+/- 0.0

hd
Water
DS temperature
Pitot-static
tube
Gate valve DS
Test cell: flow
chambers and
Outflow test specimen Outflow

Figure 3.3: Schematic diagram of HET-P apparatus

30
Chapter 3. Modified Hole Erosion Test (HET-P)

Inflow High back pressure


tank

Upstream constant
head tank Piezometer board

Outflow
Low back pressure
tank

Flexible 50-mm
tubing
Differential pressure
transducer system

Test cell

Downstream
Gate valve DS constant head and
(Figure 3.8) sedimentation tank

Outflow (Figure 3.8)

10° v-notch weir


Scale: 0 0.2 0.4 0.6 0.8 1 m (center)
Figure 3.4: Photograph of HET-P apparatus

3.2.1. Test cell


The test cell is the core of the new HET-P apparatus, and includes the test specimen with axial
hole, and extended flow chambers with hydraulic head measurements upstream and downstream of
the specimen (Figure 3.5a). A detailed engineering drawing of the test cell is presented in Appendix B.
All hydraulic heads were measured relative to the common datum, located at mid-height of the test
specimen (Figure 3.3). This assured that elevation heads were included in the measurements, and
that measurements were not affected by the orientation of the apparatus (Equation (3.1)).

The mold containing the soil specimen was made of HDPE pipe, and had an inner diameter and
length of 100 mm. Intermediate flanges were used to fix the mold and hold it in place during
assembling of the test cell. Replaceable bridge elements were used between the mold and
intermediate flanges to easy accommodate different outside diameters of the mold and non-erodible

31
Chapter 3. Modified Hole Erosion Test (HET-P)

PVC specimens. The downstream bridge element had a slightly smaller inner diameter to hold a 6-mm
wire mesh with 30-mm center opening to support the soil specimen (Figure 3.5b). The center opening
assured that no gravel size material accumulated on top of the mesh.

The upstream and downstream flow chambers consisted of a smooth 100-mm acrylic pipe with a
length of 500 mm and attached flanges to securely tighten the cell. Gradual expansion and contraction
fittings were used for a smooth transition to the smaller diameter piping, connecting the test cell with
the constant head tanks. This setting assured minimal head loss, fully developed and uniform flow at
the upstream specimen interface, and reduced downstream turbulences as suggested in the hydraulic
review in Section 3.1. The clear acrylic pipe also allowed visual access inside the flow chambers.

Similar to previous studies, hydraulic heads hu and hd were measured 50 mm upstream and
downstream of the specimen using pressure tapping on the wall of the flow chambers (Figure 3.1),
connected to traditional vertical piezometer tubes for manual readout. The piezometer board was
equipped with a rough 100-mm overall measuring scale for fast readings, and a measuring tape at
each tube for accurate readings of hydraulic heads.

(a) (b) Piezometer connection

Figure 3.5: a) Vertical HET-P test cell, b) Dismantled downstream part of HET-P test cell with
bridge element holding 6-mm wire mesh for soil specimen support, piezometer
connection, and Pitot-static tube

32
Chapter 3. Modified Hole Erosion Test (HET-P)

3.2.2. Pitot-static tube and differential pressure transducer system

A conventional Pitot-static tube was installed in the HET-P test cell, positioned in the center at the
downstream side of the test specimen with the tip of the probe pointing directly into the flow. This
provided the total and static pressure of the exiting jet immediately (tip of probe 5 mm) downstream of
the specimen. The velocity head, hv , which is the difference of total and static pressure heads, was
measured directly using a differential pressure transducer connected to the Pitot-static tube. Using the
same transducer, the total energy head, Hd , was measured relative to the common datum.

The installed probe was a KIMO Pitot-static tube type L (KIMO TPL-03-200) as shown in Figure
1
3.2. It is a curved NPL model made of stainless steel 4/4 with an outside diameter of 3 mm and
ellipsoidal head. The Pitot tube coefficient, Cp , (see Section 3.4.2 below) provided by the
manufacturer is 1.0015, but it was recommended to carry out a calibration of the Pitot-static tube to
determine its exact coefficient. The accuracy of the probe is given as 1% for a ± 10° alignment to the
fluid flow, as Pitot-static tubes are most sensitive to flow angularity (yaw and pitch angle error). The
Pitot-static tube was permanently installed, and only periodically removed for inspection and cleaning.

The differential pressure transducer system to obtain pressure head readings from the Pitot-static
tube included two SETRA Systems Model 230 bidirectional wet-to-wet pressure transducers, valve
trees, short vinyl tubing, and two elevated back pressure tanks. The SETRA Systems Model 230 are
high output, low differential pressure transducers with a fast-response capacitance sensor that
provides a highly accurate, linear analog output proportional to pressure. Two transducers with
different pressure ranges were installed for the use in low head respectively high head tests (Table
3.1). The transducer in use was connected to a conventional voltmeter for manual voltage readout
corresponding to the pressure head difference between the high and low side of the transducer. The
total pressure port of the Pitot-static tube was permanently connected to the high side of the
transducer. The velocity head, hv , was measured by connecting the static pressure port of the Pitot-
static tube to the low side of the transducer. The total energy head, Hd , was measured relative to the
common datum by switching the low side of the transducer to a piezometer tube with constant water
level set at ± 0.0 (Figure 3.6). Two back pressure tanks, one just above the test cell, the other
somewhat higher than the maximum level of the upstream constant head tank (Figure 3.4), were used
for system saturation and applying a back pressure to avoid particles entering the Pitot-static tube
during setup. The high back pressure, connected to the tip of the probe, further allowed back flushing
during a test in case the Pitot-static tube got clogged by eroding particles. De-aerated water was used
to saturate and back flush the transducer system to avoid accumulation of dissolved air in the water
inside the system with time, which would negatively affect head readings.

1
Developed between 1952 and 1954 by the U.K. National Physical Laboratory (NPL)
33
Chapter 3. Modified Hole Erosion Test (HET-P)

Table 3.1: Specifications SETRA Model 230 Bidirectional Wet-to-Wet Pressure Transducers

Specification Transducer #1: low range Transducer #2: high range


Differential pressure range ± 1.0 psi ± 703 mm H2O ± 2.5 psi ± 1758 mm H2O
Proof pressure high side 40 psi 28.1 m H2O 100 psi 70.3 m H2O
Proof pressure low side 2.5 psi 1758 mm H2O 6.25 psi 4394 mm H2O
1
Accuracy RSS* (at const. temp.) ± 0.0025 psi ± 1.8 mm H2O ± 0.00625 psi ± 4.4 mm H2O
2
Resolution * 0.0002 psi 0.1 mm H2O 0.0005 psi 0.4 mm H2O
Response time 30-50 ms 30-50 ms
Excitation 24 VDC 13-30 VDC
Output 4-20 mA 0-10 VDC

1
* Residual sum of squares (RSS) of non-linearity, non-repeatability and hysteresis: ± 0.25% full scale (FS).
2
* Infinite, limited only by output noise level (0.02% FS).

Transducer #1, low range,


NOT connected

High back pressure for


system saturation and
back flushing

Transducer #2, high range

Total pressure (high) from


Pitot-static tube

Piezometer tube with


constant water level set at
system’s datum (±0.0)

Static pressure (low) from


Pitot-static tube Low back pressure for
system saturation
Figure 3.6: Differential pressure transducer system, setup for high range differential pressures

34
Chapter 3. Modified Hole Erosion Test (HET-P)

3.2.3. Water flow system

The water flow system was hydraulically optimized to confine the major energy head losses within
the test cell. This mainly involved reducing pipe length, pipe roughness and number of pipe fittings. A
schematic hydraulic profile with the major system components and corresponding relative energy head
losses is presented in Figure 3.7. The total energy head loss across the whole system represents the
difference between the upstream and downstream control water level, and is therefore referred to as
test head, which is not necessarily equal to the hydraulic head difference across the specimen, h .

EGL
Total Energy Head Loss
Constant Head Tank

Constant Head Tank


Flow Chamber US

Flow Chamber DS

Flow Contraction
Flow Expansion
Test Head =

Gate Valve
90° Elbow
Specimen
Energy Head

Upstream Test Cell Downstream

-2 50-mm flexible tubing 30


Chainage [not to scale]

Figure 3.7: Schematic hydraulic profile of Modified Hole Erosion Test (HET-P) with Energy
Grade Line (EGL) for initial hole diameter o = 6 mm at medium flow rate

In a manner similar to previous studies, flow was controlled by adjusting the height of the
upstream constant head tank. It was adjustable to a maximum height of 2400 mm above the center of
the specimen. Tap water was fed directly into the upstream tank, capable of providing a flow rate of up
to 40 l/min. The gate valve at the downstream side of the test cell was used to prevent water flow
during setup, and remained fully open throughout a test.

35
Chapter 3. Modified Hole Erosion Test (HET-P)

The downstream constant head tank was fixed in height, and the water level was controlled by a
custom thin-plate 10° v-notch weir for measuring flow rate. The minimum water level was set 100 mm
above the center of the specimen to ensure the specimen was submerged at all times. The tank was
divided into three compartments; 1) inflow, 2) weir, and 3) outflow (Figure 3.8). To account for changes
in viscosity, water temperature was measured in the inflow compartment intended for sedimentation of
soil particles and dissipation of vertical kinetic energy. A perforated vertical flow conditioning plate
between the inflow and weir compartment assured horizontal flow and reduced turbulences in the weir
compartment, yielding a still water level behind the v-notch weir for more accurate water level readings
to determine flow rate. A piezometer connected to the tank and a fixed ruler next to the v-notch weir
provided two redundant water level readings in the weir compartment. Concrete blocks were placed
behind the weir to reduce the change of volume necessary to adjust the water level for varying flow
rates, while maintaining a large flow area between the inflow and weir compartment. The outflow
compartment mainly collected the effluent, discharging it to the sewer system. An overflow pipe was
inserted to the outlet as needed to create a water pocket for further particle settlement.

Scale: 0 0.2 0.4 0.6 0.8 1 m (center)


Vertical flow Still water
conditioning plate level behind
v-notch weir

2
Energy dissipation 1
in inflow 3
compartment

Horizontal
flow in weir
compartment

Gate
valve DS Outflow

Figure 3.8: Downstream constant head tank immediately after failure of soil specimen
illustrating performance of flow conditioning plate and v-notch weir at 20-25 l/min

36
Chapter 3. Modified Hole Erosion Test (HET-P)

3.3. Experimental setup and test procedure

3.3.1. Calibration of measuring instruments

The differential pressure transducers connected to the Pitot-static tube were calibrated before
installation to obtain the relation between differential pressure head and output voltage from the
transducers. Different water pressure heads were applied to the high side of the transducers, while a
constant reference water pressure head was maintained at the low side of the transducers. Applied
differential pressure heads ranged from ± 0 mm to 1000 mm for the low range transducer #1 and
± 0 mm to 2400 mm for the high range transducer #2 with a resolution of 10-20 mm at low heads
respectively 50-100 mm at high heads. The transducers were subjected to both pressure paths low-
high and high-low to detect possible hysteresis. The plot of differential pressure head versus output
voltage showed a linear output proportional to pressure (Figure C.1 in Appendix C) for both pressure
paths with only very little hysteresis. Linear regression trend lines were fitted to the data, yielding
equations to convert output voltage to differential pressure head.

Calibration was also necessary in order to obtain the relation between flow rate through the
custom v-notch weir and water level reading in the downstream constant head tank. The flow rate was
calculated for various constant water levels by measuring the time to collect a control volume of water
overflowing the weir. The water levels in the downstream constant head tank were set to water heads
relative to the crest of the weir ranging from ± 0.0 mm to 120 mm with an average resolution of about
2
5 mm. The Kindsvater-Shen relationship (ASTM D5242-92 (2007)) was adapted to fit a curve through
measured data of water head relative to the crest versus discharge (Figure C.2 in Appendix C). Zero
readings at the piezometer connected to the tank and at the fixed ruler next to the v-notch weir were
used to convert water level readings to water heads relative to the crest.

3.3.2. Test preparation

Test preparation included preparing the test specimen, transferring the specimen to the HET-P
apparatus, installing measuring instruments, assembling the test cell, and saturating the water flow
system. Details of this procedure are given in Appendix D, and are summarized below.

Test specimens used in this study were either non-erodible PVC specimens or erodible soil
specimens. Each PVC specimen had a preformed axial hole and did not need any further preparation.
Reconstituted soil specimens were sieved to remove any particles larger than about one-tenth of the

2
V-notch weir equation suggested by various organizations for standardization like ASTM or ISO.
37
Chapter 3. Modified Hole Erosion Test (HET-P)

mold diameter, and compacted in three layers inside the test mold at 95% maximum dry density and
optimum water content (or any other desired degree of compaction) using the custom compaction
device shown in Appendix B. The soil was compacted around a vertical 6-mm rod in the center of the
mold, which was removed prior to testing to open up the preformed hole. This method was preferred
over drilling to avoid problems like smearing and remoulding that could lead to a dense surface layer
of the axial hole. Undisturbed soil samples were carefully trimmed and centered in the mold. The
remaining gap between soil and mold was sealed with modelling clay and topped with epoxy glue in a
manner that no sealant would penetrate into the soil matrix where it affects the strength of the soil. The
preformed hole was introduced by drilling using a drill press and wood auger bit.

Preparing the apparatus started by choosing the appropriate differential pressure transducer
according to expected pressure heads, connecting it to the Pitot-static tube, and saturating the
transducer system and piezometer tubes with de-aerated water. Starting at the downstream constant
head tank, the whole system was then carefully saturated by constantly raising the water level. The
control valve downstream was closed after the water level reached the crest of the weir, and saturation
continued from the upstream side. The test specimen was carefully placed on top of the downstream
flow chamber in a manner to prevent any air pockets to be trapped underneath the specimen. Finally,
the test cell was securely tightened and connected to the upstream constant head tank after placing
the upstream flow chamber and connecting the piezometer tubes.

3.3.3. Test procedure

The HET-P test procedure was consistent with methods described by Wan and Fell (2002, 2004a,
2004b), and Wahl et al. (2008, 2009), and described in detail in Appendix D. Starting with an upstream
hydraulic head (tank elevation) of 150 mm above datum, the specimen was subjected to a constant-
head pressure flow. Elapsed time, hydraulic head upstream and downstream of the specimen, total
and velocity head from the Pitot-static tube, height of water level above v-notch weir (flow rate) and
water temperature were taken at select time intervals. The upstream tank was raised to increase the
test head if the flow rate became steady. Upstream hydraulic head increments were generally 20 mm
or 100 mm for PVC specimens and about 30-50% of the previous head for soil specimens. The
upstream hydraulic head was maintained for several minutes once progressive erosion initiated in soil
specimens, indicated by an increasing flow rate. Ideally, the test was stopped before the axial hole
expanded to the sidewall of the mold. After the test, the system was drained, and the mold containing
the soil specimen carefully removed for visual inspection of the eroded hole. A representative portion
of the remaining soil specimen was oven-dried to determine water content after a plaster cast of the
eroded hole was made to determine the final hole diameter and effective length.

38
Chapter 3. Modified Hole Erosion Test (HET-P)

3.4. Modified analysis

3.4.1. HET-P: based on energy gradient

According to the findings in Section 3.1.1 as illustrated in Figure 3.1, an energy gradient based
method was introduced, herein after referred to as HET-P. It uses the difference of upstream total
energy head, Hu , and downstream total energy head, Hd , measured with the Pitot-static tube.
Equation (2.5) and (2.6) were modified by redefining the friction head loss along the axial hole, hf , and
replacing the hydraulic gradient, s , by the energy gradient, i . This assumed that minor head losses
due to the fluid entering the axial hole in turbulent flow conditions are negligible:

(3.3)

(3.4)

The upstream total energy head, Hu , was determined according to Equation (3.1), using the
velocity in the upstream flow chamber, Vu , based on continuity (Equation (2.8)). Equation (2.11) and
(2.12) were modified accordingly by replacing s by i :

(3.5)

(3.6)

The HET-P analysis procedure is based on the original analytical method introduced by Wan and
Fell (2002, 2004a, 2004b). Modifications regarding interpolation of friction factors (Lim 2006; Wahl et
al. 2008, 2009) were not applied in this study. Using measured flow rate, Q , energy gradient, i , and
final hole diameter, f , the analysis of test data remains the same as outlined in Section 2.5.3, except
for the following adjustments:
- Flow conditions are considered turbulent if Reynolds number Re > 2000 (Wahl et al. 2008);
- Length of axial hole is assumed to vary linearly with time where applicable (Wahl et al. 2008);
- Equation (2.5), (2.11) and (2.12) are replaced by Equation (3.3), (3.5) and (3.6), respectively.

Table E.1 in Appendix E presents a step by step analysis of test data for the different methods.

39
Chapter 3. Modified Hole Erosion Test (HET-P)

3.4.2. HET-P (V): based on energy gradient and flow velocity

The combined Pitot-static tube also provided an independent measure of the mean axial hole
velocity, Vt , which in turn was used to estimate the diameter of the axial hole, t . This method
combined with the energy gradient based method introduced above is herein after referred to as
HET-P (V). The velocity head of the jet exiting the axial hole, hv , was used to estimate Vt :

(3.7)

where:
Vt = estimated mean flow velocity in axial hole, m/s
Cp = Pitot tube coefficient (1.0015 for the installed Pitot-static tube)
Cv = velocity coefficient
hv = velocity head, m

The Pitot tube coefficient, Cp , is a coefficient of instrument, and accounts for both local directional
velocity fluctuations due to turbulences, and reading errors due to the shape of the sensing tip. The
velocity coefficient, Cv , was introduced to convert measured center line velocity to the mean flow
velocity in the axial hole. It is comparable with the theoretical pipe factor described above.

Using measured Q together with Vt from the Pitot-static tube (Equation (3.7)), the diameter of the
axial hole, t , can be simply estimated from continuity (Equation (2.8)) assuming a cylindrical axial
hole, while there is no need to determine the final hole diameter. This simplified analysis of test data
significantly by eliminating steps i) to iv) outlined in Section 2.5.3 (Figure 3.9). The remaining steps are
the same as for the HET-P method, except that Equation (3.6) is replaced by Equation (3.7) and (2.8).

Q dt / dt Ce
fT,o ėHET
END
HET

s f fT  t t(fT) IHET
fT,f HET
o c
(a)
HET-P (V)

Q dt / dt ėHET-P Ce
t
END

i IHET
HET-P
Vt c
(b)

Figure 3.9: Flowcharts describing test analysis of (a) HET, and (b) HET-P (V) (rectangle:
measured or deduced, parallelogram: assumed or affected by uncertainty)

40
Chapter 4. Non-erodible test specimens

4. Non-erodible test specimens

This chapter presents experimental studies conducted on non-erodible test specimens. This series
of experiments was performed in order to test the applicability of the modified apparatus and
suggested methods of analysis. The approach of using non-erodible test specimens eliminates the
uncertainty in shape and size of the eroded axial hole, and assures a robust framework to test
suggested procedures under controlled conditions.

Section 4.1 presents an overview of the research program and specified test procedure. Details
about the experimental program can be found in Appendix F, including test number information and a
detailed table of tests performed in this and the subsequent series of experiments. Results and
analysis of test data are described in Section 4.2, showing the differences between standard HET and
modified HET-P tests, and presenting advantages of the modified methods over the standard HET.
Detailed tables with test data and results for all non-erodible specimens are presented in Appendix G.
Findings, sources of errors and limitations of the modified methods are discussed in Section 4.3.

41
Chapter 4. Non-erodible test specimens

4.1. Research program

Non-erodible PVC specimens with an axial hole of constant diameter were used in this series of
tests. Three different specimens were tested, with axial holes of 6, 12 and 24 mm, drilled into 100 mm
long solid PVC cylinders (Figure 4.1). These diameters covered the range expected for erodible soil
specimens in a typical HET. Each PVC specimen was placed directly into the test cell as described
above, replacing the mold otherwise containing the soil specimen.

Scale: 0 2 4 6 8 10 cm 0 1 2 3 4 5 in
Figure 4.1: Non-erodible PVC specimens with axial holes of 6, 12, and 24 mm diameter

A series of three independent tests was performed on each PVC specimen using the modified
apparatus described in Section 3.2 with the high range differential pressure transducer #2 (Table 3.1).
Applied test heads as defined in Figure 3.7 ranged from 5 mm to 2240 mm with a minimum upstream
sidewall hydraulic head of hu ≥ 150 mm. With the 24-mm specimen, test heads could not exceed
160 mm at maximum flow rate, providing limited data of low accuracy using the high range differential
pressure transducer #2. Thus, an additional test was performed on the 24-mm specimen using the low
range differential pressure transducer #1 to increase accuracy at these low test heads. Measurements
were taken as described above for each test head when the water level in the downstream tank
reached steady state. This could take several minutes at low flow rates.

42
Chapter 4. Non-erodible test specimens

4.2. Results and analysis

Test data were analysed according to the following three methods as introduced above:
1. HET Original method proposed by Wan and Fell (2002, 2004a, 2004b);
2. HET-P New method based on energy gradient;
3. HET-P (V) New method based on energy gradient and flow velocity.

Observed flow rates ranged from about 1 l/min to 8 l/min for the 6-mm specimen and about 2 l/min
to 36 l/min for the 12 and 24-mm specimen. Only turbulent flow (Re > 2000) was observed for the
applied test heads. Results were generally less scattered for Reynolds numbers Re > 5000. Thus, a
further distinction was made regarding flow regime in the results presented below. More specifically, a
distinction was made between a critical zone with 2000 < Re < 5000, where flow could be either
laminar or turbulent, and complete turbulent flow with Reynolds numbers Re > 5000.

4.2.1. Head ratio and shear stress

The differences between the HET and HET-P tests can be described by the head ratio, defined
here as H / h . The value of head ratio was found to be approximately constant for each of the test
specimens, but varies with hole diameter (Figure 4.2). It tended to decrease with increasing diameter
for the range of tested diameters. The values obtained for the three non-erodible test specimens are
summarized in Table 4.1.

Table 4.1: Summary of head ratio values, H / h , obtained from the three non-erodible test
specimens for turbulent flow with Re > 5000

Hole diameter, o Head ratio, H / h


[mm] Mean Standard deviation Number of observations
6 0.27 0.01 22
12 0.10 0.01 23
24 0.03 0.02 23

The head ratio also reflects the relative values of shear stress obtained from the HET-P and HET
tests, and equals the shear stress ratio, HET-P / HET , for a given diameter. Figure 4.2 and Figure 4.3
show that the shear stress values from HET-P tests, HET-P , were up to one, almost two orders of
magnitude smaller than those from HET tests, HET , for all three specimens.

43
Chapter 4. Non-erodible test specimens

0.4 0.4

Shear Stress Ratio, HET-P /HET [ - ]


Head Ratio, H/h [ - ]

0.3 0.3

0.27
0.2 0.2

0.1 0.1
0.10
0 0.03 0
0 5 10 15 20 25 30 35 40
Flow Rate, Q [l/min]

6 mm (Re>5000) 12 mm (Re>5000) 24 mm (Re>5000)


6 mm (Re=2000-5000) 12 mm (Re=2000-5000) 24 mm (Re=2000-5000)
6 mm Mean (Re>5000) 12 mm Mean (Re>5000) 24 mm Mean (Re>5000)

Figure 4.2: Head ratio equals shear stress ratio versus flow rate for the three non-erodible test
specimens

1000
6 mm (Re>5000)
Wall Shear Stress HET-P, HET-P [N/m2]

6 mm (Re=2000-5000)
100

12 mm (Re>5000)

10 12 mm (Re=2000-5000)

24 mm (Re>5000)

1
24 mm (Re=2000-5000)

1:1 Line
0.1
0.1 1 10 100 1000

Wall Shear Stress HET, HET [N/m2]

Figure 4.3: Wall shear stress HET-P using Pitot-static tube data versus wall shear stress HET
from sidewall hydraulic heads for the three non-erodible test specimens

44
Chapter 4. Non-erodible test specimens

4.2.2. Flow coefficient and axial hole velocity


For the Pitot-static tube to be of use in HET-P (V) tests, it is essential that the product CpCv in
Equation (3.7) is constant throughout a test. For simplicity, this product was called flow coefficient, K ,
and was determined using Equation (3.7) and flow velocities obtained from continuity (Equation (2.8))
using measured flow rate and known axial hole diameter of the non-erodible test specimens.
Measurements from the three non-erodible specimens across a range of discharges indicated an
essentially constant value for the flow coefficient, K , of about 0.825 as summarized in Table 4.2.
However, a slight but noticeable increase in K was observed with increasing flow rate (Figure 4.4).

Table 4.2: Summary of flow coefficient values, K , obtained from the three non-erodible test
specimens for turbulent flow with Re > 5000

Hole diameter, o Flow coefficient, K = CpCv


[mm] Mean Standard deviation Number of observations
6 0.816 0.008 22
12 0.830 0.013 23
24 0.827 0.015 23
All 0.825 0.014 68

1.0
Flow Coefficient, K [ - ]

0.8

0.6

0.4
Vcontinuity
0.2 Vt  C p Cv 2 ghv  K 2 ghv  K  K = 0.825
2 ghv
0.0
0 5 10 15 20 25 30 35 40
Flow Rate, Q [l/min]

6 mm (Re>5000) 12 mm (Re>5000) 24 mm (Re>5000)


6 mm (Re=2000-5000) 12 mm (Re=2000-5000) 24 mm (Re=2000-5000)
Linear best fit

Figure 4.4: Flow coefficient, K , versus flow rate for the three non-erodible test specimens

45
Chapter 4. Non-erodible test specimens

Using a constant value K = 0.825, good agreement was achieved for mean flow velocity in the
axial hole from Equation (3.7), and calculated from continuity (Equation (2.8)) based on the known
axial hole diameters of the non-erodible test specimens (Figure 4.5).

6.0
6 mm (Re>5000)
Velocity from Pitot-static tube [m/s]

5.0
6 mm (Re=2000-5000)

4.0
12 mm (Re>5000)

3.0 12 mm (Re=2000-5000)

2.0 24 mm (Re>5000)

24 mm (Re=2000-5000)
1.0
1:1 Line
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Velocity from Continuity [m/s]

Figure 4.5: Mean flow velocity in axial hole using Pitot-static tube data versus mean flow
velocity in axial hole using continuity for the three non-erodible test specimens

Accordingly, estimated axial hole diameters HET-P (V), back calculated from continuity and mean
flow velocities in the axial hole using Pitot-static tube data, corresponded well with the known axial
hole diameter of the non-erodible test specimens, which is summarized in Table 4.3 and illustrated in
Figure 4.6.

Table 4.3: Summary of estimated axial hole diameters HET-P (V), t , obtained from the three
non-erodible test specimens for turbulent flow with Re > 5000

Hole diameter, o Estimated axial hole diameter HET-P (V), t


[mm] Mean Standard deviation Number of observations
6 5.97 0.03 22
12 12.04 0.10 23
24 24.02 0.17 23

46
Chapter 4. Non-erodible test specimens

30
Hole Diameter HET-P (V), t [mm]

25

20 24

15

10 12
5
6
0
0 5 10 15 20 25 30 35 40
Flow Rate, Q [l/min]

6 mm (Re>5000) 12 mm (Re>5000) 24 mm (Re>5000)


6 mm (Re=2000-5000) 12 mm (Re=2000-5000) 24 mm (Re=2000-5000)
6 mm Diameter 12 mm Diameter 24 mm Diameter

Figure 4.6: Estimated axial hole diameters HET-P (V), back calculated from Pitot-static tube
data versus flow rate for the three non-erodible test specimens

4.3. Discussion

4.3.1. Head ratio and shear stress

The head ratio H / h was used to describe differences between the HET and HET-P tests.
Results showed that for a given diameter, the head ratio equals the shear stress ratio HET-P / HET .
This is because the hydraulic shear stress on the boundary of the preformed hole is directly
proportional to the friction head loss, hf (Equation (2.4)  (2.5) and (3.3)). A ratio of 1.0 would indicate
complete agreement between the two methods. But the head ratio was found to be consistently less
than unity, indicating that shear stress values obtained from the standard HET were overestimated.
This finding is a consequence of the energy gradient, i , rather than the hydraulic gradient, s , being
used to calculate shear stress.

The average head ratio for the 6-mm specimen (0.27, Table 4.1) was similar to the result reported
by Marot et al. (2011), who found that, based on calculated energy losses, roughly 25% of the
measured hydraulic gradient, s , is transformed into friction and erosion. This is equivalent to a head
ratio of H / h = 0.25. Unlike Marot et al. (2011), results of the current study showed that this value is

47
Chapter 4. Non-erodible test specimens

not constant, but varies with hole diameter (Figure 4.2). While for the axial hole diameters tested
herein the head ratio decreased with increasing diameter, the value is expected to ultimately increase
with hole diameter and converge to unity in the extreme case where the diameter of the axial hole
approaches the diameter of the flow chamber (100 mm).

The findings that HET-P was less than HET may explain the observation of Lim (2006) that the
erosion rate index from HET was significantly higher than that from the RCT for non-dispersive soils,
implying slower erosion and higher shear stresses, and the companion observation by the USBR
(Farrar et al. 2007; Wahl et al. 2008) and Marot et al. (2011) that the JET also yields critical shear
stress values that are significantly less than those obtained from the standard HET.

4.3.2. Axial hole velocity and hydraulic roughness

In Hole Erosion Tests, flow and geometric conditions of the axial hole change with time. The
approximately constant value for the flow coefficient K = 0.825, found independent of flow rate and
hole diameter, and a good agreement between velocities from Pitot-static tube data and continuity,
confirmed the applicability of a Pitot-static tube for velocity measurements downstream of the test
specimen.

The velocity coefficient, Cv , introduced in Equation (3.7), represents the ratio of the mean pipe
flow velocity to the centerline velocity (which is measured using the Pitot-static tube), given by the pipe
factor defined in Equation (3.2). Using K = CpCv = 0.825 and the value of the Pitot tube coefficient
provided by the manufacturer, Cp = 1.0015 ≈ 1.0, a constant velocity coefficient could be determined
as Cv = 0.825, which is within the range expected for turbulent pipe flow (Figure 3.2).

The pipe factor in turbulent flow, and thus Cv depends on Reynolds number, diameter, and pipe
roughness (Equation (3.2)). For a given diameter and roughness, Cv is expected to increase with
Reynolds number and thus flow rate until complete turbulent flow is developed, where the Darcy
friction factor is constant and independent of Re . Complete turbulence was not achieved using the
PVC specimens due to the low pipe roughness. This explains the slight increase in K and thus Cv
observed for the three non-erodible test specimens (Figure 4.4).

Ideally, the theoretical value of Cv in HET-P (V) tests should be either constant, or at least
experience only a minimum rate of change. Referring to the Moody diagram, this would be the case for
higher Reynolds numbers in rougher pipes of smaller diameter. In typical Hole Erosion Tests on
erodible soil specimens, Reynolds number, axial hole diameter and pipe roughness all increase with

48
Chapter 4. Non-erodible test specimens

time. Thus, and given the very small observed change in Cv for the PVC specimens, the change in Cv
is expected to be insignificantly small, and assuming a constant value is a reasonable approximation
in testing erodible soil specimens.

For certain applications, Pitot-static tubes can also be aligned in the direction of the flow, rather
than having the tip of the probe pointing directly into the flow. However, these applications are rather
limited, and tests on the non-erodible PVC specimens using a Pitot-static tube at the upstream side of
the specimen did not reveal any consistent results to be of use in HET-P tests.

As introduced with the HET-P (V) method, the velocity obtained using the Pitot-static tube
(Equation (3.7)) can be used to determine the mean diameter of the axial hole from continuity during
testing of erodible soil specimens. This is a crucial aspect of the analysis, because the erosion rates
are determined from the increase in the hole diameter (Equation (2.7)). Because it is not possible to
directly measure the diameter during testing, it has been necessary to back-calculate the diameter
using an assumed hydraulic roughness (Wan and Fell 2002, 2004a, 2004b), which has been reported
as problematic (Lim 2006; Wahl et al. 2008, 2009). The Pitot-static tube can therefore provide a more
direct estimate of axial hole velocity and diameter, and potentially simplify the analysis of HET data.

4.3.3. Upstream total energy head


2
The HET-P method includes the velocity head in the upstream flow chamber, Vu / 2g , to calculate
the upstream total energy head, Hu (Equation (3.4)) needed to determine shear stress based on the
energy gradient, i . Using continuity (Equation (2.8)), a maximum upstream velocity head of
2
Vu / 2g ≈ 0.4 mm was deduced from measured flow rate in the later stage of testing. This value
appears insignificantly small compared to the resolution of measured hydraulic heads, and one could
argue that the upstream sidewall hydraulic head, hu , could be approximated as total energy head.
However, neglecting the upstream velocity head introduces a significant error at low heads and high
2
flow rates as observed with the 24-mm specimen. Absolute values of Vu / 2g could take up as much
as 15-20% of the energy head loss along the test specimen in HET-P tests if the upstream velocity
2
head was neglected. Thus, it is recommended to include Vu / 2g in all calculations of Hu .

Wan and Fell (2002, 2004a, 2004b) suggested filling the upstream flow chamber with 20-mm
gravel to diffuse the fluid and assure an even pressure/flow distribution upstream of the test specimen.
This would cause an unknown reduction in effective cross sectional area and an increase in velocity.
Assuming equally distributed flow paths, this would result in a decrease in the sidewall hydraulic head,
hu , and thus introduce an error in the upstream total energy head, Hu . It is therefore necessary not to

49
Chapter 4. Non-erodible test specimens

fill the upstream flow chamber with any material. This goes along with findings of Wahl et al. (2008),
who found that there was no need to place gravel into the upstream flow chamber for testing erosion-
resistant soils.

4.3.4. Flow conditions and minor losses


In HET-P tests, flow conditions were considered turbulent if Reynolds number Re > 2000. This is a
widely recognized lower critical value (e.g. see Critical Reynolds Number, Finnemore and Franzini
2002, p. 256), and was also suggested by Wahl et al. (2008) to be used in HET. However, there is no
guarantee that flow is turbulent right above this value. Going from low to high Reynolds numbers, the
laminar-flow zone is followed by a critical zone, in which laminar flow could be maintained up to an
upper critical Reynolds number of about 4000 or higher. The analysis of test data revealed difficulties
in getting meaningful results for data sets within this critical zone, even up to Reynolds numbers of
about 5000. This finding is consistent with the original suggestion of Wan and Fell (2002, 2004a,
2004b), who used Re = 5000 as critical value above which flow was considered turbulent. Even though
Reynolds numbers most likely exceed 5000 in testing erodible soil specimens, it is suggested to
distinguish between complete turbulent flow and a critical zone with 2000 < Re < 5000 in the analysis
and interpretation of HET data.

The minor losses due to the eroding fluid entering the axial hole in turbulent flow conditions were
estimated by calculating the major friction head losses along the axial hole using the Darcy-Weisbach
equation and an explicit approximation of the Colebrook-White equation for pipe roughness (Haaland
1983), and subtracting this from the measured total head loss across the specimen. This indicated that
the minor entrance losses were insignificant, and can be ignored. One reason for this is that the
entrance length over which the minor entrance losses develop exceeded the length of the specimen
by as much as a factor of 10. This indicated that the test specimen was too short for the flow to
become fully developed, which is a problem already stated by Bonelli and Brivois (2008), because the
interpretation of test data assumes uniform flow conditions.

Furthermore, minor entrance losses very much depend on the conditions at the entrance of the
axial hole. The upstream end of the axial hole in erodible soil specimens usually develops a rounded
or funnel shaped form during a test. This implies a gradual contraction with almost no vena contracta,
and therefore no minor entrance losses. The minor exit losses are avoided by measuring the
downstream total energy head using the Pitot-static tube close to the point where the jet exits the axial
hole.

50
Chapter 4. Non-erodible test specimens

4.3.5. Sources of errors

The main sources of errors in HET-P and HET-P (V) tests were associated with:
- Location of measurements;
- Flow irregularities;
- Measuring instruments.

There were uncertainties and implied errors regarding the location of measurements and their
assigned meaning in the analysis and interpretation of test data. Including the upstream velocity head,
2
Vu / 2g , in the calculation of the upstream total energy head, Hu , is only correct if the difference
2
between the EGL and HGL equals Vu / 2g . But the relative magnitude of the HGL at the location of
measurement hu was unsure, because the sidewall hydraulic head increases, and its value converges
to that of the total energy head at the corner of the wall and specimen, due to centrifugal action
caused by curved stream lines (Figure 3.1). It was in fact unknown at which point upstream of the
specimen that the HGL started to rise. Depending on flow conditions, it has to be assumed that
measurements of hu are slightly affected by this local hydraulic phenomenon, resulting in
overestimated values of Hu .

The Pitot-static tube was located in a zone with a relatively high local energy gradient and
increasing HGL, depending on axial hole diameter and flow rate (Figure 3.1). This implied two sources
of error. Firstly, the tip of the Pitot tube was placed 5 mm downstream of the specimen, but the
measurement was assigned to the downstream end of the specimen. This means that measurements
of Hd would have been slightly underestimated, and resulting head ratios overestimated. Secondly,
the holes for static pressure were located even further downstream (30 mm from specimen) at a point
with an increased static pressure due to the expanding jet and decreased flow velocity. Combing the
two sources would imply a slightly underestimated measured velocity head, hv , and estimated mean
flow velocity in the axial hole, Vt .

Flow irregularities in the system caused slight fluctuations in all readings, except in the
downstream tank and water temperature, which both had a relatively slow response time. Common
practice was to record mean values observed over a certain period of time appropriate to current flow
conditions. This method also introduced possible human errors.

Errors introduced by measuring instruments included piezometer level readings, accuracy of the
differential pressure transducers, and form and alignment of the Pitot-static tube. Piezometer levels
were recorded with a resolution of ±0.5 mm H2O over the whole range of pressure heads. The

51
Chapter 4. Non-erodible test specimens

introduced error depends on the relative difference between compared piezometer levels and varies
with test head and flow rate. The downstream total energy head, Hd , was occasionally affected by a
rising water level in the reference piezometer tube caused by a volume change due to the flexibility of
the measuring system. It was not always deemed necessary or possible to reset the water level to the
system’s datum before each reading. The two differential pressure transducers used in this study had
different accuracies depending on the differential pressure range. Transducers were not switched
during a test. Thus, readings in tests covering a wide range of differential pressure heads were less
accurate, introducing reading errors especially at low differential pressure heads.

Possible errors introduced by the Pitot-static tube could have different origins. Firstly, and most
importantly, the probe could have been slightly out of angle causing the flow not being parallel to the
sensing tip. Secondly, the installed probe was not affected by Mach number errors, since velocities in
HET were below critical values, and there were no compressive effects on the fluid. Thirdly, the Pitot-
static tube was not affected by viscosity effects because Reynolds numbers exceeded a minimum
value of about 30-70. Fourthly, even though the probe was placed outside the hole adjacent to the
specimen, and a sufficient distance away from solid boundaries, static pressure could have been
affected by locally accelerated flow due to the presence of the test specimen or confined jet, forming a
Venturi-type passage around the sensing tip. Lastly, errors introduced by an increased response time
due to the small diameter were reduced to a minimum by using short tubing and observing voltage
outputs over an appropriate period of time before recording the data.

4.3.6. Limitations

The present study was limited to turbulent flow with Reynolds numbers larger than 2000.
Generally, interpreted results are further limited to flow conditions of complete turbulence with
Reynolds numbers exceeding 5000.

The experimental studies were limited to non-erodible PVC specimens with uniform axial holes
and a limited number of three different axial hole diameters of up to 24 mm. Results may vary for
larger diameters and non-uniform axial holes.

The repeatability of the test methods was demonstrated by consistent results from a series of
three, respectively four tests on each PVC specimen. However, all test results were linked to the same
limitations regarding accuracy of measuring instruments and other sources of errors implied by the test
apparatus used in this study.

52
Chapter 5. Erodible soil specimens

5. Erodible soil specimens

After the successful testing of non-erodible PVC specimens, the Modified Hole Erosion Test
(HET-P) was used to test erodible soil specimens of different type and erosion resistance. The
objective of this test series was to confirm findings from previous tests on PVC specimens, and to
demonstrate the applicability of the implemented modifications to real soil testing.

Presented in this chapter is a summary of the research program (Section 5.1), which is
supplemented by Appendix F that includes test number information and a detailed list of tests
performed in this and the preceding series of experiments. The main findings from tests on erodible
soil specimens are presented in Section 5.2. It ranges from the description of soil samples used in this
study to observed differences between the HET and HET-P methods regarding head ratio, shear
stress, estimated axial hole diameter, and critical shear stress. These differences are further
discussed and compared to previous studies in Section 5.3 that also includes errors and limitations
associated with this study.

53
Chapter 5. Erodible soil specimens

5.1. Research program

The laboratory experiments conducted in this test series included basic material property tests and
Modified Hole Erosion Tests (HET-P) on five different soil samples. Two types of soil were tested in
this study, reconstituted glacial till material from a dam core and undisturbed clay samples of natural
river bank deposits from various Ontario rivers.

The soil samples were classified using the Unified Soil Classification System (USCS) according to
ASTM Standard D2487-11. Grain size distribution tests were conducted using mechanical sieve and
hydrometer analysis following ASTM Standards D421-85 (2007), D422-63 (2007) and D6913-04
(2009). The Atterberg Limits of test soils, namely liquid limit, LL , and plastic limit, PL , were
determined as described in ASTM Standard D4318-10. Standard compaction tests were performed
according to ASTM Standard D698-07 to determine optimum water content, wopt , and standard
maximum dry density, d,max , for reconstituted samples. If possible, standard compaction test Method
B was used with material passing the 3/8 in (9.5 mm) sieve to be consistent with the maximum particle
size used in HET-P tests. Water content of the soil samples respectively prepared specimens was
taken before (wo) and, if possible, after testing (wf) following ASTM Standard D2216-05. Prior to
testing, the moist density, m , and dry density, d , of the prepared soil specimen were determined by
the direct measurement method (Method B) as described in ASTM Standard D7263-09.

Using the modified apparatus, HET-P tests were conducted on soil specimens prepared from each
soil sample according to the procedure described in Section 3.3. A total of four soil specimens from
the reconstituted dam core material, and five soil specimens from undisturbed Ontario clay samples
were tested. Expected progressions of internal erosion, according to Wan and Fell (Table 2.1), ranged
from rapid for the glacial till material to slow for the clay samples.

54
Chapter 5. Erodible soil specimens

5.2. Results and analysis

As with the non-erodible test specimens, HET-P test data were analysed according to the
following three methods as introduced in Chapter 2 and 3 above:
1. HET Original method proposed by Wan and Fell (2002, 2004a, 2004b);
2. HET-P New method based on energy gradient;
3. HET-P (V) New method based on energy gradient and flow velocity.

Observed flow rates ranged from about 1 l/min to 40 l/min. At failure, flow rate could occasionally
exceed the system’s maximum of 40 l/min, but would not maintain a constant upstream water head.
Laminar flow (Re < 2000) was rare, and only observed in the very beginning of a test. A critical
Reynolds number of Re = 2000 was applied to all tests, with precaution to interpretation regarding data
points falling within the critical zone where 2000 < Re < 5000.

Presented below is first a description of the soils used in this study with obtained soil properties,
followed by the results from HET-P testing. Tests on the glacial till material from the dam core were
very difficult to perform, and did not yield enough data for a complete analysis. Thus, the general
outcome is presented using solely results from tests on Ontario clay samples, while tests on the dam
core material are described separately.

A summary of soil properties for all tested samples described below, and a summary of test data
and results for all the erodible soil specimens tested in this study is provided at the end of this section
in Table 5.3 and Table 5.4, respectively. Appendix H and Appendix I provide additional soil property
information and detailed test data and results for each specimen in form of figures and photos.

5.2.1. Soil properties and description


Dam core material
The first type of soil tested in this study was material obtained from the core of a dam. The dam is
located on the Columbia River in British Columbia, Canada. The dam was designed with a wide core,
constructed of clayey glacial till material from local borrow areas within today’s reservoir. The material
was placed in layers of 25 cm thickness or less, and compacted in eight passes by a 57-ton
pneumatic-tired roller. Water content was within 2% below and 1% above optimum.

The soil sample used in this study, labelled as MV4-Core, was excavated from the existing dam
core at a depth of approximately 7 m (23 ft) from the road elevation on August 25, 2003. Protected by

55
Chapter 5. Erodible soil specimens

a plastic cover to prevent moisture loss, the soil was delivered to the UBC geotechnical laboratory in a
steel storage drum, in which it was stored ever since. The soil had a medium brown color, and was
classified as SC-SM; silty, clayey sand with gravel. The grain size distribution and Atterberg Limits of
the MV4-Core material were obtained by Abelardo A. Julio at UBC in 2009. The reproduced gradation
curve is presented in Appendix H, Figure H.1. Compaction data were obtained by the standard
compaction test Method B (maximum particle size 9.5 mm) and dry preparation of the sample (Figure
H.3). The water content of the whole sample was taken from various locations within the steel drum
prior to any other testing, and ranged from 5.2% to 8.2%. Densities and water content listed in Table
5.3 are for the one test specimen prepared for HET-P testing.

Due to a limited amount of soil available for HET-P testing, three specimens were prepared and
tested beforehand using a dummy soil consisting of dam core material previously tested as part of
ongoing research at UBC. This altered sub-sample, denoted by MV4-Altered, had a maximum particle
size passing the 3 in (75 mm) sieve and a significantly reduced fines content of about 7%. Atterberg
Limits were not determined for this sample. The standard compaction test Method A (maximum
particle size 4.75 mm) was used to obtain compaction data (Figure H.2). Table 5.3 also gives
estimates of maximum dry density and optimum water content corrected for a maximum particle size
of 9.5 mm. The presence of oversize fraction in the test specimen would results in a higher maximum
dry density and lower optimum water content.

A rapid progression of internal erosion was expected for specimens prepared from the dam core
material (MV4-Core and MV4-Altered). According to Wan and Fell (2002, 2004a, 2004b), most soils
show higher erosion resistance if they were compacted to the wet side of optimum water content.
Thus, the test specimens were prepared with tendency to the wet side with water contents not
exceeding 1% above optimum. One specimen was prepared at 3% above optimum to test the erosion
behaviour for varying water contents.

Ontario clay samples


The rest of the tested soil samples were taken from natural river bank deposits of various Ontario
rivers. The samples were provided by Dr. Colin D. Rennie of the University of Ottawa. Samples were
collected by driving a 100-mm diameter stainless steel tube into the soil. Two samples each from five
different sites were collected, sealed with cling wrap, and shipped to the UBC geotechnical laboratory.
Later inspection showed that some cling wrap seals were broken during the transport, and the soil
samples probably lost some moisture content. Upon inspection, all samples were re-sealed with cling
wrap and stored in a moist room until they were tested. The delivered samples provided just enough
material to carry out the HET-P tests. Thus, soil property tests were performed on different samples at

56
Chapter 5. Erodible soil specimens

the University of Saskatchewan Department of Civil & Geological Engineering. They are described in
Cossette and Mazurek (2011), and are summarized in Table 5.3, which shows mean values from tests
on two samples of each origin.

Five of the provided samples, two each from Jock River and Raisin River, and one from Bear
Brook, were not suitable for HET-P testing. Sample Bear Brook 1/2 fell apart upon extrusion form the
steel tube, leaving intact portions too small for testing. Samples from Jock River and Raisin River were
all penetrated by heavy vegetation and organic matter, and in some instances contained insufficient
material for testing.

The tested samples from Little Cataraqui were classified as CL; lean clay with sand. They
contained some organic matter like little pieces of wood, and light to no vegetation. The soil had a
grey-brown colour with red-brown speckled areas, and showed some layering in form of horizontal
cracks, probably caused by the sampling method and possible loss of moisture content.

Sample Bear Brook 2/2, classified as CL-ML; sandy silty clay, had a red speckled dark brown
color with a sandy texture, and a medium vegetation and organic matter content. Exposing the plaster
cast after testing revealed a single 3-mm root crossing the specimen diagonally through the center at
about a third from the bottom. The root was separated by drilling the preformed hole, but hold in place
during the test by the remaining surrounding soil.

The two soil samples from Wilton Creek had a slightly green tinted grey-brown color with red
tinted/speckled patches. This soil was classified as CL; sandy lean clay. The samples had medium to
heavy vegetation and organic matter present. Soil erosion during the HET-P test on sample Wilton
Creek 2/2 revealed a large 5-mm root with complex root hairs at the bottom of the specimen.

5.2.2. Head ratio and shear stress

The head ratio, H / h , which was used to describe the differences between HET and HET-P
tests, varied during each test because of the changing axial hole diameter (Figure 5.1). The value of
head ratio first decreased with increasing axial hole diameter as seen with the non-erodible test
specimens. After a certain diameter, the reverse behaviour was observed, and values of head ratio
increased with increasing diameter. The rate of initial decrease and the point of change were different
between soil samples and test specimens. Initial values of head ratio ranged from 0.3 to 0.9 and
decreased to a minimum of about 0.1 to 0.7. Test S3-993.21, Wilton Creek, showed a different
behaviour with no initial decrease in head ratio.

57
Chapter 5. Erodible soil specimens

1.2 1.2

Shear Stress Ratio, HET-P /HET [ - ]


1.0 ? 1.0
Head Ratio, H/h [ - ]

0.8 ? 0.8

0.6 0.6
S3-01/02: Little
Cataraqui
0.4 0.4
S3-11/12: Bear
Brook
0.2 0.2
S3-21/22: Wilton
Creek
0.0 0.0
6 10 14 18 22 26 30

Estimated Mean Diameter, t [mm]

Figure 5.1: Head ratio equals shear stress ratio versus estimated mean hole diameter for the
Ontario clay specimens

The shear stress ratio, HET-P / HET , which is interchangeable with the head ratio, was smaller than
unity in all cases. This is also reflected in Figure 5.2, where shear stresses from all three methods,
HET , HET-P , and HET-P (V) , were directly compared. Shear stress values from HET-P and HET-P (V)
tests were up to one order of magnitude smaller than those from HET tests, especially at lower values,
where results were more scattered due to the accuracy of the measuring devices. At higher values,
data points gathered around the linear relationship HET-P = 0.5 HET , with a trend towards the 1:1 line
with increasing shear stress resulting from an increase in axial hole diameter as the test progresses.
Generally, HET-P and HET-P (V) tests yield similar results with only very little deviation compared to
HET tests.

58
Chapter 5. Erodible soil specimens

1000 1000

Wall Shear Stress HET-P (V), HET-P (V) [N/m2]


Wall Shear Stress HET-P, HET-P [N/m2]

100 100

10 10

1 1

0.1 0.1
0.1 1 10 100 1000

Wall Shear Stress HET, HET [N/m2]

S3-01/02: Little Cataraqui - HET-P S3-01/02: Little Cataraqui - HET-P (V)


S3-11/12: Bear Brook - HET-P S3-11/12: Bear Brook - HET-P (V)
S3-21/22: Wilton Creek - HET-P S3-21/22: Wilton Creek - HET-P (V)
1:1 Line Linear: y=0.5x

Figure 5.2: Wall shear stress HET-P using Pitot-static tube data versus wall shear stress HET
from sidewall hydraulic heads for the Ontario clay specimens

59
Chapter 5. Erodible soil specimens

5.2.3. Axial hole velocity and estimated diameter

Mean flow velocities in the axial hole were obtained from Pitot-static tube data (Equation (3.7)),
and used in the HET-P (V) method to estimate axial hole diameters. Velocity data from the Pitot-static
tube were analysed using a constant value for the flow coefficient of K = 0.825, as found by testing
non-erodible specimens (Section 4.2.2).

Using continuity (Equation (2.8)), values of mean velocity and diameter of the axial hole were
determined for all three methods, and compared among each other. Differences between methods
were described by the velocity ratio respectively diameter ratio, defined here as the ratio of values
from the corresponding modified method to values obtained according to the standard HET. Values of
both ratios were close to 1.0 for most of the successful tests, indicating relatively good agreement
between methods. However, with a deviation of up to 51%, the HET-P (V) method does not agree well
with the other two methods for flow conditions with Reynolds numbers smaller than about 5000. A
statistical summary of the two ratios is presented in Table 5.1 (velocity) and Table 5.2 (diameter).

A direct graphical comparison between mean flow velocities (Figure 5.3) respectively diameters
(Figure 5.4) obtained using the HET-P methods and standard HET supported these findings.
Excluding data with Re < 5000 for the HET-P (V) method, most data points did fall within a deviation of
about 20% (velocity) and 10% (diameter) from the 1:1 line of agreement, respectively.

60
Chapter 5. Erodible soil specimens

Table 5.1: Summary of velocity ratio values obtained from the Ontario clay specimens for
turbulent flow with Re > 2000 for HET-P, respectively Re > 5000 for HET-P (V)

Velocity Ratio HET-P and HET-P (V), Vt,HET-P / Vt,HET and Vt,HET-P (V) / Vt,HET
Soil Sample
Mean Standard deviation Number of observations
1 1 1
Test # HET-P HET-P (V) * HET-P HET-P (V) * HET-P HET-P (V) *
2 23
S3-993.01 * 0.76 n/a 0.09 n/a 29
(6)
0.93 0.09 16
S3-993.02 0.90 0.17 20
(0.62) (0.04) (4)
0.95 0.07 17
S3-993.12 1.06 0.06 21
(0.76) (0.09) (4)
3 1.43 0.11 61
S3-993.21 * 0.79 0.10 73
(1.18) (0.10) (12)
0.84 0.12 13
S3-993.22 1.04 0.04 18
(0.49) (0.13) (5)
1
* HET-P (V): values in parentheses represent data points where 2000 < Re < 5000
2
* S3-993.01: Pitot-static tube measured flow through intact axial hole at the bottom of the specimen (Figure I.18).
3
* S3-993.21: 2 flow paths, Pitot-static tube measured mainly flow through primary axial flow path (Figure I.47).

4 4 S3-01/02: Little
Velocity from Continuity HET-P, Q /AHET-P [m/s]

Velocity Pitot-static tube HET-P (V), Vt [m/s] Cataraqui - HET-P

2 2 S3-11/12: Bear Brook


- HET-P
S3-21/22: Wilton
1 1 Creek - HET-P
S3-01/02: Little
Cataraqui - HET-P (V)
0.5 0.5
S3-11/12: Bear Brook
- HET-P (V)
0.25 0.25 S3-21/22: Wilton
HET-P (V): Re > 5000 Creek - HET-P (V)

0.125 0.125 1:1 Line

0.125 0.25 0.5 1 2 4


1:1 ±20%
Velocity from Continuity HET, Q /AHET [m/s]

Figure 5.3: Mean flow velocity in axial hole from HET-P and HET-P (V) using Pitot-static tube
data versus mean flow velocity in axial hole from HET using continuity for the
Ontario clay specimens

61
Chapter 5. Erodible soil specimens

Table 5.2: Summary of diameter ratio values obtained from the Ontario clay specimens for
turbulent flow with Re > 2000 for HET-P, respectively Re > 5000 for HET-P (V)

Diameter Ratio HET-P and HET-P (V), t,HET-P / t,HET and t,HET-P (V) / t,HET
Soil Sample
Mean Standard deviation Number of observations
1 1 1
Test # HET-P HET-P (V) * HET-P HET-P (V) * HET-P HET-P (V) *
2 23
S3-993.01 * 1.15 n/a 0.07 n/a 29
(6)
1.04 0.05 16
S3-993.02 1.06 0.09 20
(1.27) (0.04) (4)
1.03 0.04 17
S3-993.12 0.97 0.03 21
(1.15) (0.07) (4)
3 0.84 0.04 61
S3-993.21 * 1.14 0.08 73
(0.92) (0.04) (12)
1.10 0.09 13
S3-993.22 0.98 0.02 18
(1.46) (0.18) (5)
1
* HET-P (V): values in parentheses represent data points where 2000 < Re < 5000
2
* S3-993.01: Pitot-static tube measured flow through intact axial hole at the bottom of the specimen (Figure I.18).
3
* S3-993.21: 2 flow paths, Pitot-static tube measured mainly flow through primary axial flow path (Figure I.47).

24 24 S3-01/02: Little
Estimated Diameter HET-P (V), t,HET-P (V) [mm]
Estimated Diameter HET-P, t,HET-P [mm]

Cataraqui - HET-P
S3-11/12: Bear Brook
- HET-P
S3-21/22: Wilton
Creek - HET-P
12 12 S3-01/02: Little
Cataraqui - HET-P (V)
S3-11/12: Bear Brook
- HET-P (V)
S3-21/22: Wilton
HET-P (V): Re > 5000 Creek - HET-P (V)

6 6 1:1 Line
6 12 24
1:1 ±10%
Estimated Diameter HET, t,HET [mm]

Figure 5.4: Estimated axial hole diameter from HET-P and HET-P (V) back calculated from
Pitot-static tube data versus estimated axial hole diameter from HET for the Ontario
clay specimens

62
Chapter 5. Erodible soil specimens

5.2.4. Erosion rate and critical shear stress

Starting at a low upstream hydraulic head (tank elevation), and using successively increased head
increments during HET-P tests, revealed a distinct behaviour of observed erosion rate (Figure 5.5).
Before reaching critical conditions, each increase of the test head resulted in a peak in erosion rate
with following decrease at constant test head. Absolute peak values generally decreased with
increasing test heads, resulting in a flattening curve of erosion rate until a critical point at which failure
occurred in the form of progressive erosion.

0.018
Estimated erosion rate per unit surface area of

0.016 HET

0.014 Peak and decreasing HET-P


the axial hole, ε̇ [kg/s/m2]

erosion rate with


0.012 each test head.
0.01

0.008

0.006 Critical point: c


start of failure
0.004

0.002

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure 5.5: Erosion rate versus time for successively increased test heads (S3-993.01)

Plots of erosion rate, , versus shear stress,  , rarely presented themselves in the form described
by Wan and Fell (Figure 2.6), and it was not possible at all to fit a reasonable linear best-fit line in
order to determine the coefficient of soil erosion, Ce , erosion rate index, IHET , and critical shear stress,
c . Thus, critical shear stress was not defined as the x-intercept of the extrapolated linear best-fit line
on a plot, but rather at the point where the erosion rate progressively increased. It was found
that this point was distinct on a plot of flow rate, Q , versus shear stress,  , as illustrated in Figure 5.6.
The initially steep curve did flatten as the test progressed, and an increase of the slope was observed
at the critical point where failure started. Thus, critical shear stress was defined at the point with the
minimum slope on a Q –  plot.

63
Chapter 5. Erodible soil specimens

20
Critical point: c,HET-P
18
start of failure
Measured Flow Rate, Q [l/min]

16
c
14
12
10
8
Critical point: c,HET
6
start of failure
4
2 HET HET-P HET-P (V)
0
0 100 200 300 400 500 600 700 800 900 1000
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure 5.6: Critical shear stress defined on flow rate versus shear stress diagram (S3-993.21)

5.2.5. Dam core material

All four specimens prepared from the dam core material experienced immediate erosion, together
with very turbid outflow. The soil started to collapse right upon wetting during placement of the
specimen. Mostly fine material and occasional small aggregates of particles separated from the
specimen and dispersed into the downstream flow chamber during setup. This was visually observed
by a major non-transparent discolouration inside the downstream flow chamber and later of the
effluent during testing.

In the case of test S1-553.01 (MV4-Altered), this pre-test failure caused the preformed axial hole
to collapse and become completely blocked after placing the specimen. No flow was observed up to a
test head of about 2300 mm, at which the specimen suddenly collapsed and completely eroded within
less than a minute. Shear stress and erosion rate analysis was not possible for this test.

The other three tests all showed similar behaviour. Very rapid erosion started right after the first
test head increment from about 50 mm to 160 mm (doubling of upstream hydraulic head from 150 mm
to 300 mm). The specimens completely eroded within 1 to 3 minutes, except in test S1-003.01, where

64
Chapter 5. Erodible soil specimens

the supporting wire mesh clogged and retained a significant portion of the specimen. It was not
possible to obtain plots of erosion rate, , versus shear stress,  , to determine erosion parameters
(Figure 2.6). Thus, critical shear stress was defined at the point where sudden progressive erosion
started, and the progression of internal erosion was estimated based on observations rather than from
deduced coefficient of soil erosion, Ce , and erosion rate index, IHET . HET analyses revealed very low
2
critical shear stresses of less than about 30-40 N/m (Table 5.4). HET-P and HET-P (V) analyses were
not possible in most cases because the Pitot-static tube was either blocked or internally clogged. The
progression of internal erosion for this soil was estimated as very rapid.

5.2.6. Ontario clay samples

All specimens prepared from Ontario clay samples showed non-dispersive erosion behaviour
(relatively clear outflow) with an estimated progression of internal erosion of “very slow” to “moderately
slow” (according to Wan and Fell, cf. Table 2.1). Test S3-993.12, Bear Brook, was the only completely
successful test. Problems with all other tests on the Ontario clay samples partially affected test results.
During test S3-993.01, Little Cataraqui, a portion of the bottom part of the specimen remained intact,
including the preformed axial hole. This resulted in invalid velocity data from the Pitot-static tube.

Results of both test S3-993.02, Little Cataraqui, and S3-993.22, Wilton Creek, showed only a
small increase in diameter over time before the specimen suddenly collapsed. Further, diameters are
in relatively good agreement between all three methods of analysis (Table 5.2). Along with continuous
erosion observed in the lab and post-test examination of the eroded specimens, this suggested that a
cavity developed along the wall of the mold by backward erosion, which collapsed upon further
increase of the test head.

Test S3-993.21, Wilton Creek, showed a similar behaviour, but with the difference that axial hole
diameters determined from Pitot-static tube data were constantly low, also reflected in a low diameter
ratio for the HET-P (V) method (Table 5.2). This was consistent with test S3-993.01, and suggested
that backward erosion created a second flow path through the specimen, slowly enlarging from the
bottom until failure occurred (sudden enlargement on top).

65
Chapter 5. Erodible soil specimens

Table 5.3: Summary of soil properties for erodible soil specimens

Grain size distribution (USCS) Atterberg limits Standard compaction test Prepared soil specimen
United Soil Gravel Sand Silt Clay Standard Optimum Initial
Soil sample Classification Liquid Plastic Moist Dry
> > > < maximum water water
System Limit, Limit, density, density,
4.75 0.075 0.002 0.002 dry density, content, content,
LL PL m d
mm mm mm mm d,max wopt wo
3 3 3
Test # USCS [%] [%] [%] [%] [%] [%] [kg/m ] [%] [kg/m ] [kg/m ] [%]
Dam MV4-Altered

S1-003.01 2331 2106 10.7


1 1
2121 * 7.8 *
S1-003.02 SP-SC 41 52 7 – – 2373 2187 8.5
(2170-2200) (7.4-7.1)
S1-553.01 2268 2094 8.3

Dam MV4-Core

S2-553.01 SC-SM 35.3 36.5 20.5 7.7 22 15 2141 8.2 2186 2023 8.1

Ontario Clay – Little Cataraqui

S3-993.01 1895 1445 31.1


CL 0 17 67 16 46 27 – –
S3-993.02 1817 1401 29.7

Ontario Clay – Bear Brook

S3-993.12 CL-ML 0 33 57 10 25 20 – – 1880 1651 13.9

Ontario Clay – Wilton Creek

S3-993.21 1899 1558 21.9


CL 0 33 58 9 26 18 – –
S3-993.22 1766 1507 17.2

1
* Compaction data based on Method A; dmax = 4.75 mm, yielding lower dry density and higher water content than Method B.
Values in parentheses are corrected for dmax = 9.5 mm.

66
Chapter 5. Erodible soil specimens

Table 5.4: Summary of test data and results for erodible soil specimens

Specimen Axial hole Post-test measurements Critical shear stress


Soil sample
fixed f Lf wf c,HET c,HET-P Failure mechanism
reconstituted undisturbed drilled center 2 2
Test # rod [mm] [mm] [%] [N/m ] [N/m ]
Dam MV4-Altered
very rapid
S1-003.01 X X 45 10 – ≤ 40 ≤ 32
progression of internal erosion
very rapid
S1-003.02 X X – – – ≤ 27 –
progression of internal erosion
very rapid
S1-553.01 X X – – – – –
progression of internal erosion
Dam MV4-Core
very rapid
S2-553.01 X X – – – ≤ 28 –
progression of internal erosion
Ontario Clay – Little Cataraqui
moderately slow
S3-993.01 X X 28.3 93 30.0 742 297
progression of internal erosion
structural failure
S3-993.02 X X 9 75 38.9 92 26
(backward erosion)
Ontario Clay – Bear Brook
very slow
S3-993.12 X X 16.3 73 20.4 141 80
progression of internal erosion
Ontario Clay – Wilton Creek
structural failure
S3-993.21 X X 11.6 58 28.9 785 556 nd
(backward erosion + 2 flow path)
1 1 structural failure
S3-993.22 X X 9.8 79 32.8 89-149 * 58-86 *
(backward erosion)

1
* Lower value = last measurement at lower test head of 300 mm, higher value = estimated values at higher test head of about 500 mm at which failure occurred.

67
Chapter 5. Erodible soil specimens

5.3. Discussion

5.3.1. Head ratio and shear stress

The results from testing erodible soil specimens agree with the findings from tests on the non-
erodible PVC specimens, as it was shown that the head ratio, and hence shear stress ratio, was also
consistently less than 1.0 (Figure 5.1). This confirmed that shear stress values obtained from the
standard HET are generally overestimated due to the use of the hydraulic gradient, s , which is
affected by various downstream flow expansion anomalies (Chapter 3).

As observed with the non-erodible test specimens and unlike assumed by Marot et al. (2011),
values of head ratio were not constant throughout a test, and generally decreased with increasing
diameter of the axial hole. Based on theoretical considerations regarding uniform axial holes, it was
surmised earlier that the values of head ratio would converge to unity in the extreme case where the
diameter of the axial hole approaches the diameter of the flow chamber. Results confirmed that head
ratio increases with diameter after a certain point of change (Figure 5.1), but suggested that it would
converge to unity even earlier in a test. The reason for this behaviour is most likely in the fact that soil
specimens developed a funnel shaped entrance and exit to the axial hole (also Wan and Fell 2002,
2004a, 2004b; Lim 2006; Wahl et al. 2008), which reduces hydraulic turbulences downstream and
eventually yields a head ratio of 1.0.

Values of head ratio were generally higher compared to those from non-erodible PVC specimens.
This was due to a significantly higher hydraulic pipe roughness in the soil specimens. This increased
the portion of the hydraulic gradient, s , responsible for friction losses, and yielded better agreement
between HET and HET-P tests.

Some specimens experienced a sudden structural failure involving extensive “uncontrolled”


erosion of the whole specimen, rather than “controlled” progressive erosion around the axial hole. In
these cases, the point of change in Figure 5.1 is misleadingly shifted to the left because of a lack of
intermediate measuring points, and the estimated maximum diameter does not necessarily reflect the
final diameter of the eroded hole.

Even though less significant than with PVC specimens, it was found that HET-P was less than
HET , which further confirms that observed differences between the HET and RCT (Lim 2006), and the
HET and JET (Farrar et al. 2007; Wahl et al. 2008; Marot et al. 2011) may be due to an incorrect
interpretation of the hydraulic gradient assigned to friction head losses in the standard HET.

68
Chapter 5. Erodible soil specimens

5.3.2. Axial hole velocity and estimated diameter

The relatively good agreement found for deduced velocities and diameters between all three
methods of analysis, HET, HET-P and HET-P (V) (Table 5.1 and Table 5.2), confirmed the applicability
of velocity measurements using a Pitot-static tube at the downstream end of the specimen. However, it
also showed that the standard HET method yields reasonably accurate estimates of axial hole
diameters, which suggests that velocity measurements are not as crucial for the interpretation of HET
data as the downstream total energy head. However, the Pitot-static tube does provide a more direct
estimate of axial hole velocity and diameter, and potentially simplifies the analysis of HET data.
Considering both methods in parallel would increase confidence in the interpretation of HET results.

5.3.3. Erosion rate and critical shear stress

Wan and Fell (2002, 2004a, 2004b) described the early stage in a HET, in which decreasing
erosion rate was observed with increasing shear stress, as the period where disturbed and loose
material is being removed (Figure 2.6). However, erosion rate behaved differently using successively
increased test heads during HET-P tests (Figure 5.5). The initial period of increasing and decreasing
erosion rate at lower test heads may reflect a process of successive armouring, where smaller and
looser particles are removed from the area directly adjacent to the axial hole, while the course soil
matrix remains intact. This creates a rougher and more erosion-resistant surface with time. Once
shear stress exceeds an upper critical value, which is high enough to destroy and remove the
developed armouring, the whole specimen starts to erode because the less erosion-resistant intact soil
underneath is now exposed to an excessive force.

This raises the question whether or not a model scaling factor should be applied to HET-P results.
Scalping the gradation curve of test soils to a maximum grain size of 9.5 mm or less, as it is required
for test specimens of that size, does not necessarily reflect field conditions. Thus, obtained critical
shear stresses may not be directly applicable to actual building materials, and need to be adjusted to
the corresponding coarser fraction. However, no adjustment is necessary if the HET-P was used as a
simple index test to characterize test soils in terms of relative progression of internal erosion as initially
suggested by Wan and Fell (2002, 2004a, 2004b).

Determining the critical shear stress based on a plot of erosion rate versus shear stress according
to the standard HET method is associated with various problems and uncertainties, including scattered
results for repeated tests, non-linearity in the coefficient of soil erosion, data smoothening,
extrapolation of data, and complex curve fitting procedures (Section 2.5). Wan and Fell (2002, 2004a,

69
Chapter 5. Erodible soil specimens

2004b) already suggested to use an initial shear stress, o , corresponding to the minimum upstream
hydraulic head at which erosion is first initiated, rather than critical shear stress, c , in order to
describe initiation of erosion. The newly introduced definition of critical shear stress being at the point
with the minimum slope on a Q –  plot follows up on this suggestion, with the only difference that the
corresponding minimum hydraulic head was not found by multiple HET trial runs on identical test
specimens. It also depicts the reasoning on the process of surface armouring. The critical shear stress
would represent the erosion resistance of an exposed coarser fraction of the soil matrix. This method
further involved considerably less uncertainties because a highly-biased variable was replaced by the
measured flow rate, and shear stress could be obtained more directly using the HET-P (V) method.

5.3.4. Test soils

Some of the Ontario clay specimens experienced unexpected erosion behaviour like backward
erosion and the development of a second flow path, while the preformed axial hole remained more or
less intact. Several reasons could be responsible for that. Firstly, drilling may have introduced a
disturbed and denser surface layer around the axial hole that was more erosion-resistant than the
surrounding soil. Secondly, possible lost of moisture content and disturbance of the sample caused by
the sampling method may have introduced weak points in form of cracks or fractures. Thirdly, the gap
between soil and mold due to a slightly smaller diameter of the soil samples may not have been
properly sealed, creating a point of weakness to initiate erosion.

It is known that soil specimens developed a funnel shaped entrance and exit to the axial hole in
the course of a HET. This was also observed for the Ontario clay specimens. The amount of
remoulding on the upstream side was usually smaller that at the downstream side. Other researchers
assigned this phenomenon to a process called slaking, where soil particles detached from the
specimen due to the presence of water (hydrostatic conditions), rather than the applied shear stress.
Recalling the hydraulic review of the HET (Section 3.1), it is most likely that these geometries are the
results of corresponding flow patterns (Figure 3.1), forcing the soil into hydraulically optimal shapes.
Higher turbulences and flow recirculation at the downstream side may explain the larger amount of
remoulding at the bottom of the specimen, and increases the risk of backward erosion.

The erosion behaviour of the two types of soil tested in the current study is comparable with
observations for clay soils by Lim (2006), who distinguished between dispersive and non-dispersive
soils. Soil specimens prepared from the dam core material showed immediate and very rapid erosion
with turbid outflow, which is consistent with the erosion behaviour of dispersive clay soils observed by
Lim (2006). The Ontario clay specimens, on the other hand, had a relatively clear outflow, and showed

70
Chapter 5. Erodible soil specimens

a slower rate of erosion, together with a considerable reduction of the hole length and irregular shape
of the eroded axial hole. This erosion behaviour is characteristic for non-dispersive unsaturated clay
soils (Lim 2006).

According to Soroush et al (2008), a formed crack may collapse and self-heal upon wetting, and
would rarely sustain in soils with 15% or less fines passing the #200 sieve based on an adjusted
gradation curve with a maximum particle size of 4.75 mm (#4 sieve). This may explain the problems
experienced with the dam core material. Dam MV4-Altered had an adjusted fines content of about
12%, which is less than the reported threshold of 15%, and a collapsing of the hole as it occurred in
test S1-553.01 could be expected. The response of pressure gauges right after the start of test S2-
553.01 on Dam MV4-Core material indicated that the preformed axial hole did stay open upon wetting,
which can be expected with an adjusted fines content of about 44%. However, this does not guarantee
that a soil is erosion resistant. In fact, Wan and Fell (2002, 2004a, 2004b) characterized the
progression of internal erosion of soils with fines content up to 42% as extremely to moderately rapid.

5.3.5. Sources of errors

The main sources of errors have already been described in Chapter 4 on the non-erodible test
specimens. Additional sources related to tests on erodible soil specimens include non-uniformity of the
eroded axial hole, remoulding of the downstream side of the specimen, and a delay in flow rate
measurement.

The preformed axial hole could develop a wide range of non-cylindrical shapes, which made it
challenging to assign a representative final diameter to the eroded hole. This problem was also
reported and discussed extensively by others (Section 2.5), and is not further addressed herein.

A funnel shaped downstream end of the eroded hole as it developed during HET-P tests causes
the distance from the end of the effective hole length (begin of downstream funnel) to the Pitot-static
tube to increase. This may cause an error in downstream head and velocity readings. This would
cause a lower total energy head and thus an increase in the energy gradient across the test specimen,
i , and thus estimated shear stress. The velocity, determined from the Pitot-static tube reading, would
decrease and a higher estimated diameter would be the result.

At very low flow rates, it took several minutes for the downstream tank to adjust the water level to
the corresponding height above the v-notch weir. This caused a delay in flow rate measurements,
which was reflected in a temporary decrease of estimated axial hole diameter. This delay decreased

71
Chapter 5. Erodible soil specimens

rapidly with increasing flow rates, and was minimized by reducing the tank volume using concrete
blocks placed behind the v-notch weir. This error cannot be completely eliminated and needs to be
accounted for in the interpretation of HET-P results. However, it could be significantly reduced by
using an in-line flow meter an appropriate distance downstream of the test specimen, where the
measurements would not be influenced by flow disturbances caused by the test specimen.

5.3.6. Limitations

All test results were limited to the accuracy of measuring instruments and other sources of errors
implied by the test apparatus used in this study and described above.

As for the non-erodible test specimens, this test series on erodible soil specimens was limited to
turbulent flow with Reynolds numbers larger than 2000. Using a constant value for the flow coefficient,
K , did limit the applicability of velocity measurements to flow conditions with Reynolds numbers
exceeding 5000. It was possible to use Pitot-static tube data below this threshold, but it required
caution and additional judgement in the interpretation of results as additional errors were introduced.

Absolute values of critical shear stress were limited to conditions created and observed in the lab.
This especially involves scalping of the gradation curve by limiting the maximum particle size for
testing. Test results should be corrected for oversize fraction if absolute values should be used in field
applications.

Tests on highly erodible and dispersive soils (dam core) were difficult to perform, because the
erosion process started right upon wetting during placing of the specimen, and progressed very rapid
after the test was started. This made it nearly impossible to collect enough high quality data for a
representative and accurate analysis, which reduced the confidence in interpreted results.

Furthermore, clogging of the Pitot-static tube by eroding particles was a common problem in
testing highly erodible and dispersive soils (dam core). The incorporated back flush mechanism before
each reading could clear any particles from the Pitot-static tube without disturbing the downstream
side of the soil specimen. However, back-flushing was not always completely successful, and the
partially or totally clogged Pitot-static tube produced erroneous HET-P data.

72
Chapter 6. Conclusions and recommendations

6. Conclusions and recommendations

6.1. Summary and conclusions

The standard Hole Erosion Test (HET) was modified by incorporating a conventional Pitot-static
tube at the downstream side of the test specimen that allows for an improved energy based analysis of
test data. The modified test is referred to as the HET-P. The HET-P provides an additional measure of
the total energy head, Hd , and centerline velocity head, hv , at the point where the fluid jet exits the
axial hole. Two modified methods of analysis (2, 3) were introduced and compared to the method
originally proposed by Wan and Fell (1):

1. HET Original method proposed by Wan and Fell (2002, 2004a, 2004b);
2. HET-P New method based on energy gradient using Hd ;
3. HET-P (V) New method based on energy gradient and flow velocity using Hd and hv .

Two series of tests, one on non-erodible PVC specimens (10 tests), and one on erodible soil
specimens (9 tests), yield the following conclusions:

- The differences between the modified HET-P and the standard HET are described by the
head ratio, H / h , which is usually smaller than unity. Further, unlike suggested by Marot et
al. (2011), values of H / h are not constant throughout a test, but vary with axial hole
diameter and hydraulic pipe roughness. Values are closer to unity for higher hydraulic pipe
roughness. With increasing diameter, values of H / h initially decrease to a minimum, after
which they start converging unity as the axial hole enlarges and approaches the diameter of
the flow chamber. The development of a funnel shaped exit to the axial hole in soil specimens
reduces downstream hydraulic turbulences, causing an earlier and faster rate of increase.

- The head ratio, H / h , also reflects the relative values of energy gradient and hydraulic
gradient across the test specimen, i / s , as well as shear stress obtained from the HET-P and
HET tests, HET-P / HET . The deduced shear stress from HET-P is significantly smaller than
that obtained from the HET. This is mainly a result of the total energy head measured with the
Pitot-static tube at the exit of the axial hole exceeding the hydraulic head measured at the
sidewall of the device just downstream of the test specimen, in a region affected by high
turbulence, flow separation and eddies. The downstream sidewall hydraulic head yields a
hydraulic gradient that is not representative of the friction head loss along the axial hole of the
test specimen. This may explain differences observed between the HET and RCT (Lim 2006),

73
Chapter 6. Conclusions and recommendations

and between HET and JET (Farrar et al. 2007; Wahl et al. 2008; Marot et al. 2011). It further
confirms the findings by Marot et al. (2011) that only a portion of the measured hydraulic
gradient, s , is transformed into friction and erosion in the standard HET.

- In addition, the Pitot-static tube also provides a direct measure of the centerline jet velocity
escaping the axial hole, which can be used to determine the mean velocity in the axial hole,
Vt . This yields a simplified and more direct estimate of the axial hole diameter, t , and
corresponding erosion parameters without any assumptions regarding the hydraulic
roughness of the axial hole, which are required by the standard HET (Wan and Fell 2002,
2004a, 2004b; Wahl et al. 2008, 2009). However, velocity measurements are affected by flow
conditions, and it has to be distinguished between laminar and turbulent flow.

- This study assumed turbulent flow conditions if Reynolds numbers exceed 2000 (Wahl et al.
2008). However, experience showed that results were less variable and could be interpreted
with higher confidence in flow conditions with Reynolds numbers exceeding 5000. The range
where 2000 < Re < 5000 is called the critical zone, in which flow could be either laminar or
turbulent. In this highly complex hydraulic system, it should be distinguished between
complete turbulent flow (Re > 5000) and this critical zone in the analysis and interpretation of
HET-P data.

- Hole Erosion tests, standard and modified, are difficult to perform on highly erodible and
dispersive soils like the dam core material used in this study. This is consistent with results
reported by others (Wan and Fell 2002, 2004a, 2004b; Lim 2006; Farrar et al. 2007; Wahl et
al. 2008, 2009). Further difficulties arise if the desired uniform erosion of the axial hole is
superseded or accompanied by other unintended erosion mechanisms like backward erosion
or the development of a second flow path. These may interfere with analysis assumptions or
create additional need for subjective judgement in the process of data analysis and
interpretation.

- With more confidence in the interpretation of the critical shear stress, c , using velocity
measurements and the Q- diagram, the HET-P could be used beyond index testing. With c
from HET-P tests representing a critical shear stress corresponding to the maximum particle
size used in testing, laboratory results could be scaled to values to be used in the field for
design purposes or safety evaluations of existing structures.

74
Chapter 6. Conclusions and recommendations

- The tests performed in this study showed a significant potential for the introduction of a
conventional Pitot-static tube in HET for pressure and velocity measurements downstream of
the test specimen. Incorporating a Pitot-static tube into a HET apparatus requires modest
technical effort, and is relatively easy to implement. Operation of the instrument is simple, and
it can easily be integrated in an automated data acquisition system. Overall, the Pitot-static
tube is a simple low cost method that yields more transparent and reliable results, making the
Modified Hole Erosion Test (HET-P) a technically and economically feasible method to study
erosion characteristics of soil with applications in both constructed earth structures including
dams and embankments, and natural river banks.

6.2. Recommendations

The following recommendations for further research are made from experiences in this study on
the Modified Hole Erosion Test (HET-P):

- To further investigate the hydraulic component of the Hole Erosion Test and its implications on
erosion behaviour and data analysis, it is recommended to carry out tests on non-erodible test
specimens with non-uniform axial holes. That is testing axial hole geometries as they may
develop in soil specimens during the course of a test, especially the often observed funnel
shaped deformation of the entrance and exit to the axial hole (e.g. Figure I.17 or Figure I.37)
that implies changes to hydraulic conditions and flow patterns at the points of measurement.

- Besides hydraulic conditions and flow patterns, the hydraulic pipe roughness inside the axial
hole plays a major role regarding erosion resistance. In stable conditions where
, roughness describes the resistance to flow, and governs the relationship
between the energy head loss along the test specimen, H , and flow rate, Q . PVC
specimens and specimens from different soil samples experience different hydraulic pipe
roughness, which implies different H-Q relationships. Further, roughness tends to increase
during a Hole Erosion Test as fines are eroded or washed out more easily than the coarser
particles. Testing non-erodible specimens with different hydraulic pipe roughness would help
to obtain a better understanding of the implications of changing hydraulic pipe roughness
inside the axial hole.

- The HET-P apparatus has been tested on limited soil samples. In order to obtain a
comprehensive data base for this test, and to improve the confidence in the testing procedure,
more tests should be performed on a number of different soil samples.

75
Chapter 6. Conclusions and recommendations

- This study showed that the differences between the standard HET and the modified HET-P
could explain the differences observed between the standard HET and other soil erosion test
methods (Lim 2006; Farrar et al. 2007; Wahl et al. 2008; Marot et al. 2011). To prove this
statement, paired testing of HET-P and RCT, HET-P and JET, and other possible test
methods would be necessary.

76
References

References

ASTM Standard D421-85, 1985 (2007), Standard Practice for Dry Preparation of Soil Samples for
Particle-Size Analysis and Determination of Soil Constants, ASTM International, West
Conshohocken, PA, 2007, www.astm.org.
ASTM Standard D422-63, 1963 (2007), Standard Test Method for Particle-Size Analysis of Soils,
ASTM International, West Conshohocken, PA, 2007, www.astm.org.
ASTM Standard D698-07e1, 2007, Standard Test Methods for Laboratory Compaction Characteristics
of Soil Using Standard Effort (12 400 ft-lbf/ft3 (600 kN-m/m3)), ASTM International, West
Conshohocken, PA, 2007, www.astm.org.
ASTM Standard D2216-05, 2005, Standard Test Methods for Laboratory Determination of Water
(Moisture) Content of Soil and Rock by Mass, ASTM International, West Conshohocken, PA,
2005, www.astm.org.
ASTM Standard D2487-11, 2011, Standard Practice for Classification of Soils for Engineering
Purposes (Unified Soil Classification System), ASTM International, West Conshohocken, PA,
2011, www.astm.org.
ASTM Standard D4318-10, 2010, Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity
Index of Soils, ASTM International, West Conshohocken, PA, 2010, www.astm.org.
ASTM Standard D4647-06, 2006, Standard Test Method for Identification and Classification of
Dispersive Clay Soils by the Pinhole Test, ASTM International, West Conshohocken, PA, 2006,
www.astm.org.
ASTM Standard D5101-01, 2001 (2006), Standard Test Method for Measuring the Soil-Geotextile
System Clogging Potential by the Gradient Ratio, ASTM International, West Conshohocken, PA,
2006, www.astm.org.
ASTM Standard D5242-92, 1992 (2007), Standard Test Method for Open-Channel Flow Measurement
of Water with Thin-Plate Weirs, ASTM International, West Conshohocken, PA, 2007,
www.astm.org.
ASTM Standard D5852-00, 2000 (2007), Standard Test Method for Erodibility Determination of Soil in
the Field or in the Laboratory by the Jet Index Method, ASTM International, West Conshohocken,
PA, 2007, www.astm.org.
ASTM Standard D6913-04, 2004 (2009), Standard Test Methods for Particle-Size Distribution
(Gradation) of Soils Using Sieve Analysis, ASTM International, West Conshohocken, PA, 2009,
www.astm.org.
ASTM Standard D7263-09, 2009, Standard Test Methods for Laboratory Determination of Density
(Unit Weight) of Soil Specimens, ASTM International, West Conshohocken, PA, 2009,
www.astm.org.

77
References

Arulanandan, K., Sargunam, A., Loganathan, P., & Krone, R.B. (1973). Application of chemical and
electrical parameters to prediction of erodibility. In Soil Erosion: Causes and Mechanisms;
Prevention and Control (Special Report 135). Highway Research Board, 42-51.
Arulanandan, K., Loganathan, P., & Krone, R.B. (1975). Pore and eroding fluid influences on surface
erosion of soils. Journal of the Geotechnical Engineering Division, ASCE. 101(GT1), 55-66.
Arulanandan, K., & Perry, E.B. (1983). Erosion in relation to filter design criteria in earth dams. Journal
of the Geotechnical Engineering Division, ASCE. 109(GT5), 682-698.
Bonelli, S., & Brivois, O. (2008). The scaling law in the hole erosion test with a constant pressure drop.
International Journal for Numerical and Analytical Methods in Geomechanics, 32(13), 1573-1595.
Bonelli, S., Brivois, O., Borghi, R., & Benahmed, N. (2006). On the modelling of piping erosion.
Comptes Rendus Mécanique, 334(8-9), 555-559.
Briaud, J.L., Ting, F.C.K., Chen, H.C., Gudavalli, R., Perugu, S., & Wei, G. (1999). Sricos: prediction of
scour rate in cohesive soils at bridge piers. Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, 125(4), 237-246.
Briaud, J.L., Ting, F.C.K., Chen, H.C., Cao, Y., & Han, S.W. (2001). Erosion function Apparatus for
score rate predictions. Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
127(2), 105-113.
Chapius, R.P. (1986a). Quantitative measurement of the scour resistance of natural solid clays.
Canadian Geotechnical Journal, 23, 132-141.
Chapius, R.P. (1986b). Use of rotational erosion device on cohesive soils. Transportation Research
Record No. 1089, 23-28.
Chapius, R.P., & Gatien, T. (1986). An improved rotating cylinder technique for quantitative
measurements of the scour resistance of clays. Canadian Geotechnical Journal, 23, 83-87.
Christensen, R.W., & Das, B.M. (1973). Hydraulic erosion of remolded cohesive soils. In Soil Erosion:
Causes and Mechanisms; Prevention and Control (Special Report 135). Highway Research
Board, 8-19.
Colebrook, C. F. (1939). Turbulent flow in pipes, with particular reference to the transition region
between the smooth and rough pipe laws. Journal of the Institution of Civil Engineers, London,
11(4), 133-156.
Cossette, D., & Mazurek, K.A. (2011). Scour by a Submerged Vertical Circular Turbulent Impinging Jet
in Natural Cohesive Soils, 20th Canadian Hydrotechnical Conference, June 14-17, 2011, Ottawa,
Ontario.
Dunn, I. S. (1959). Tractive resistance of cohesive channels. ASCE Journal of the Soil Mechanics and
Foundations Division, 85(SM3), 1-24.
Farrar, J. A., Torres, R. L., & Erdogan, Z. (2007). Bureau of reclamation erosion testing for evaluation
of piping and internal erosion of dams. Geo-Denver 2007: New Peaks in Geotechnics, (167) 3.

78
References

Fell, R., Wan, C. F., Cyganiewicz, J., & Foster, M. (2004). Discussion of time for development of
internal erosion and piping in embankment dams by robin fell, chi fai wan, john cyganiewicz, and
mark foster. Journal of Geotechnical and Geoenvironmental Engineering, 130(9), 980-981.
Fell, R., Wan, C. F., Cyganiewicz, J., & Foster, M. (2003). Time for development of internal erosion
and piping in embankment dams. Journal of Geotechnical and Geoenvironmental Engineering,
129(4), 307-314.
Finnemore, E. J., & Franzini, J. B. (2002). Fluid Mechanics with Engineering Applications, 10th Ed.,
McGraw Hill, Boston, USA, 790p.
Foster, M., & Fell, R. (2001). Assessing embankment dam filters that do not satisfy design criteria.
Journal of Geotechnical and Geoenvironmental Engineering, 127(5), 398-407.
Foster, M., Fell, R., & Spannagle, M. (2000a). Statistics of embankment dam failures and accidents.
Canadian Geotechnical Journal, 37(5), 1000-1024.
Foster, M., Fell, R., & Spannagle, M. (2000b). Method for assessing the relative likelihood of failure of
embankment dams by piping. Canadian Geotechnical Journal, 37(5), 1025-1061.
Ghebreiyessus, Y. T., Gantzer, C. J., Alberts, E. E., & Lentz, R. W. (1994). Soil erosion by
concentrated flow: shear stress and bulk density. Transactions of the ASAE, 37(6), 1791-1797.
Gibbs, H.J. (1962). A Study of Erosion and Tractive Force Characteristics in Relation to Soil
Mechanics Properties (Division of Engineering Laboratories, Soils Engineering Report No. EM-
643). Denver, Colorado: U.S. Department of the Interior, Bureau of Reclamation.
Haaland, S. E. (1983). SIMPLE AND EXPLICIT FORMULAS FOR THE FRICTION FACTOR IN
TURBULENT PIPE FLOW. Journal of Fluids Engineering, Transactions of the ASME, 105(1), 89-
90.
Hanson, G.J. (1991). Development of a Jet Index to characterize erosion resistance of soils in earthen
spillways. Transactions of the ASAE, 34(5), 2015-2020.
Hanson, G.J. (1992). Erosion resistance of compacted soils. Transportation Research Record No.
1369, 26-30.
Hanson, G.J., & Robinson, K.M. (1993). The influence of soil moisture and compaction on spillway
erosion. Transactions of the ASAE, 36(5), 1349-1352.
Hanson, G.J., & Simon. A. (2001). Erodibility of cohesive streambeds in the loess area of the
Midwestern USA. Hydrological Processes, 15, 23-28.
Hanson, G.J., & Cook, K.R. (2004). Apparatus, test procedures, and analytical methods to measure
soil erodibility in situ. Applied Engineering in Agriculture, 20(4), 455-462.
Hanson, G.J., & Hunt, S.L. (2006). Lessons learned using laboratory jet method to measure soil
erodibility of compacted soils. Applied Engineering in Agriculture, 23(3), 305-312.
Hjeldnes, E. I., & Lavania, B. V. K. (1980). Cracking, leakage, and erosion of earth dam materials.
Journal of the Geotechnical Engineering Division, ASCE, 106(GT2), 117-135.
ICOLD (1995). Dam Failures – Statistical Analysis, Bulletin No. 99, ICOLD, Paris, France.

79
References

Indraratna, B., Muttuvel, T., Khabbaz, H., & Armstrong, R. (2008). Predicting the erosion rate of
chemically treated soil using a process simulation apparatus for internal crack erosion. Journal of
Geotechnical and Geoenvironmental Engineering, 134(6), 837-844.
Jain, R. K., & Kothyari, U. C. (2009). Cohesion influences on erosion and bed load transport. Water
Resources Research, 45, W06410.
Kandiah, A., & Arulanandan, K. (1974). Hydraulic Erosion of Cohesive Soils. Transportation Research
Record No. 497, 60-68.
Kothyari, U. C., & Jain, R. K. (2008). Influence of cohesion on the incipient motion condition of
sediment mixtures. Water Resources Research, 44(4)
Lefebvre, G., Rohan, K., & Douville, S. (1985). EROSIVITY OF NATURAL INTACT STRUCTURED
CLAY: EVALUATION. Canadian Geotechnical Journal, 22(4), 508-517.
Lefebvre, G., Rohan, K., & Milette, J. (1986). EROSIVITY OF INTACT CLAY: INFLUENCE OF THE
NATURAL STRUCTURE. Canadian Geotechnical Journal, 23(4), 427-434.
Lim, S. S. (2006). Experimental investigation of erosion in variably saturated clay soils. (A thesis
submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy, School
of Civil and Environmental Engineering, The University of New South Wales, Australia).
Lyle, W.M., & Smerdon, E.T. (1965). Relation of compaction and other soil properties to erosion
resistance of soils. Transactions of the ASAE, 8(3), 419-422.
Marot, D., Regazzoni, P., & Wahl, T. L. (2011). An Energy Based Method for Providing Soil Surface
Erodibility Rankings. Paper accepted for publication in Journal of Geotechnical and
Geoenvironmental Engineering, American Society of Civil Engineers.
Masch, F.D., Jr., Espey, W.H., Jr., & Moore, W.L. (1963). Measurements of the Shear Resistance of
Cohesive Sediments. Proceedings of the Federal Inter-Agency Sedimentation Conference,
Agricultural Research Service, Publication No. 970, Washington, D.C., 151-155.
Moore, W.L., & Masch, F.D., Jr. (1962). Experiments on the Scour Resistance of Cohesive Sediments.
Journal of Geophysical Research, 67(4), 1437-1449.
Ravens, T. M., & Gschwend, P. M. (1999). Flume measurement of sediment erodibility in Boston
Harbor. Journal of Hydraulic Engineering, ASCE, 125(10), 998-1005.
Rohan, K., Lefebvre, G., Douville, S., & Milette, J. (1986). NEW TECHNIQUE TO EVALUATE
EROSIVITY OF COHESIVE MATERIAL. Geotechnical Testing Journal, 9(2), 87-92.
Sargunan, A. (1977). Concept of Critical Shear Stress in Relation to Characterization of Dispersive
Clays, In Dispersive Clays, Related Piping, and Erosion in Geotechnical Projects. ASTM Special
Technical Publication 623, Sherard, J.L. and Decker R.S. Eds., 390-397.
Schmertmann, J. H. (2004). Time for development of internal erosion and piping in embankment
dams" by robin fell, chi fai wan, john cyganiewicz, and mark foster. Journal of Geotechnical and
Geoenvironmental Engineering, 130(9), 980-980.

80
References

Shaikh, A., Ruff, J.F., & Abt, SR. (1988a). Erosion rate of compacted NA-montmorillonite soils. Journal
of Geotechnical Engineering, ASCE, 114(3), 296-305.
Shaikh, A., Ruff, J.F., Charlie, W.A., & Abt, S.R. (1988b). Erosion rate of dispersive and nondispersive
clays. Journal of Geotechnical Engineering, ASCE, 114(5), 589-600.
Sherard, J. L., & Dunnigan, L. P. (1989). Critical filters for impervious soils. Journal of Geotechnical
Engineering, 115(7), 927-947.
Sherard, J. L. (1986). HYDRAULIC FRACTURING IN EMBANKMENT DAMS. Journal of Geotechnical
Engineering, 112(10), 905-927.
Sherard, J. L., & Dunnigan, L. P. (1985). FILTERS AND LEAKAGE CONTROL IN EMBANKMENT
DAMS. Seepage and Leakage from Dams and Impoundments. Proceedings of a Symposium in
Conjunction with the 1985 ASCE National Convention. 1-30.
Sherard, J. L., Dunnigan, L. P., & Decker, R. S. (1976). IDENTIFICATION AND NATURE OF
DISPERSIVE SOILS. American Society of Civil Engineers, Journal of the Geotechnical
Engineering Division, 102(4), 287-301.
Sherard, J. L., Dunnigan, L. P., Decker, R. S., & Steele, E. F. (1976). PINHOLE TEST FOR
IDENTIFYING DISPERSIVE SOILS. American Society of Civil Engineers, Journal of the
Geotechnical Engineering Division, 102(1), 69-85.
Sherard, J. L., Dunnigan, L. P., & Talbot, J. R. (1984a). BASIC PROPERTIES OF SAND AND
GRAVEL FILTERS. Journal of Geotechnical Engineering, 110(6), 684-700.
Sherard, J. L., Dunnigan, L. P., & Talbot, J. R. (1984b). FILTERS FOR SILTS AND CLAYS. Journal of
Geotechnical Engineering, 110(6), 701-718.
Soroush, A., Aminzadeh, A., & Shourijeh, P. T. (2008). Identifying low-fines soils not suited to NEF
testing. Proceedings of the Institution of Civil Engineers: Geotechnical Engineering, 161(4), 181-
188.
Soroush, A., & Shourijeh, P. T. (2009). A review of the no erosion filter test. Geotechnical Testing
Journal, 32(3), 209-218.
Wahl, T. L., Regazzoni, P., & Erdogan, Z. (2008). Determining erosion indices of cohesive soils with
the hole erosion test and jet erosion test (Hydraulic Laboratory (HL) No. DSO-08-05). Denver,
Colorado: U.S. Department of the Interior, Bureau of Reclamation, Technical Service Center.
Wahl, T. L., Regazzoni, P., & Erdogan, Z. (2009). Practical improvements for the hole erosion test.
33rd IAHR Congress: Water Engineering for a Sustainable Environment, Vancouver. 6383-6390.
Wan, C. F., & Fell, R. (2002). Investigation of internal erosion and piping of soils in embankment dams
by the slot erosion test and the hole erosion test (UNICIV Report No. R-412). Sidney, Australia:
The University of New South Wales.
Wan, C. F., & Fell, R. (2004a). Investigation of rate of erosion of soils in embankment dams. Journal of
Geotechnical and Geoenvironmental Engineering, 130(4), 373-380.

81
References

Wan, C. F., & Fell, R. (2004b). Laboratory tests on the rate of piping erosion of soils in embankment
dams. Geotechnical Testing Journal, 27(3), 295-303.
Zhang, G. H., Liu, B. Y., Liu, G. B., He, X. W., & Nearing, M. A. (2003). Detachment of undisturbed
soils by shallow flow. Soil Science Society of America Journal, 67(3), 713-719.

82
Appendices

Appendices

83
Appendices

Appendix A: Review of the standard Hole Erosion Test (HET)

Table A.1: Known challenges and issues of the standard HET with suggested improvements

Stage Reported challenge or issue Suggested improvement


Specimen Introduced inhomogeneity by specimen Increase number of compaction layers from
preparation reconstitution 3 to 6 or more
Dense surface layer due to smearing and Use sharp auger drill at low speed, slowly
remoulding during drilling advanced, and repeatedly cleaned
Different initial roughness of the preformed Clean out drilled hole in a uniform manner,
hole introduced by drilling e.g. with wire brush

Testing Temporary or partial blockage of the Restrict particle size, interrupt test to free
preformed hole (soil dependent) preformed hole if possible
Effects of slaking (detachment of particles End plates with orifice openings of a size
due to the presence of water) appropriate to the erosion resistance of
Scouring at entrance and exit due to the test soil, e.g. 15 mm for stronger and
eddies at the soil interface 25 mm for weaker soils
Change of geometry of preformed hole due Control curing time of prepared specimen,
to swelling (soil dependent) time of introducing hole, and test time
Non-uniform cross section along the length
of the soil specimen
The length of the preformed hole can Interpolate hole length over time in data
change analysis
Specimen too short for establishing fully
developed flow
Flow conditions are mostly turbulent, but
can change throughout a test
Difficult to assign a representative final Advanced measuring technique, e.g. from
hole diameter to the eroded hole plaster casting of the eroded hole
Numerical model by Bonelli et al.
Different critical hydraulic gradients Good quality soil samples, standardized
between soil samples and/or specimens procedures and experienced
of the same sample technicians
Absent or significantly low normal stress

Analysis & 13+ assumptions and approximation in Best engineering practice unless otherwise
Interpretation analytical solution noted
The length of the eroded hole is assumed
to vary linearly with time
Curve fitting procedures and extrapolation Numerical model by Bonelli et al.
of data
Only data collected during the period of
progressive erosion are considered
useful to determine erosion parameters
Virtual, not physically existing friction Numerical model by Bonelli et al.
factors fL and fT
Non-linearity in friction factors fL and fT
with time Friction factors assumed to be linear
Increase of estimated diameter, t , even at proportional to eroded hole diameter
times with no erosion
Non-linearity in the coefficient of soil
erosion, Ce (time dependent)
Graphical extrapolation to find critical shear Use of initial shear stress, o , to describe
stress, c (imprecise results) initiation of erosion

84
Appendices

Appendix B: Engineering drawings

85
Appendices

86
Appendices

87
Appendices

Appendix C: Calibration of measuring instruments

Differential pressure transducers

2500
3.5

y = 671.650x + 31.539
R² = 1.000
3.0
2000
Differential Pressure Head, h [mm H2O]

y = 669.061x + 32.834 2.5


R² = 1.000

Differential Pressure, p [psi]


1500
2.0
y = 353.221x - 1'829.847
R² = 1.000

1.5
1000

y = 352.678x - 1'824.900
R² = 1.000 1.0

500

0.5

0 0.0
0 2 4 6 8 10 12 14
Voltage, U [V]

Differential Pressure Transducer #1: low-high Differential Pressure Transducer #1: high-low
Differential Pressure Transducer #2: low-high Differential Pressure Transducer #2: high-low

Figure C.1: H-U diagram differential pressure transducer #1 and #2 with fitted linear regression
lines to convert output voltage to differential pressure head

88
Appendices

Custom v-notch weir

0 5 10 15 20 25 30 35 40
130 130

120 120

110 110

100 100

90
Flow Rate: 90
Kindsvater-Shen Relationship
0-40 l/min Fitted Curve:
Water head relative to crest, h [mm]

Water head relative to crest, h [mm]


q 
80 Q  2.36  C  tan    h  k  2 80
5

2
Q = discharge over weir [m3/s]
70 C = effective discharge coefficient [ - ] 70
C = 0.6700
h = head on the weir [m]
60 k = head correction factor [m] 60
k = 0.0000 m
q = angle of v-notch [degrees]
q = 9.5° (measured)
50 50

40 40

Best Fit
30 30
C3-001.01: Blue
C3-002.01: Blue
20 20
C3-001.01: Red
Flow Rate: C3-002.01: Red
10 0-4 l/min C3-001.01: MM 7
10

C3-002.01: MM 7
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Flow Rate, Q [l/min]

Figure C.2: H-Q diagram v-notch weir using Kindsvater-Shen relationship (curve fitting)

89
Appendices

0 5 10 15 20 25 30 35 40
148 235

138 225

128 215

118 205

108
Flow Rate: 195
Kindsvater-Shen Relationship
0-40 l/min Fitted Curve:
Head Reading Tank - Blue Scale [mm]

Head Reading Piezometer 7 [mm]


q 
98 Q  2.36  C  tan    h  k  2 185
5

 
2
Q = discharge over weir [m3/s]
88 C = effective discharge coefficient [ - ] 175
C = 0.6700
h = head on the weir [m]
78 k = head correction factor [m] 165
k = 0.0000 m
q = angle of v-notch [degrees]
q = 9.5° (measured)
68 155

58 145

48 Flow Rate: 135

0-4 l/min
38 125

28 115

18 105
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Flow Rate, Q [l/min]

Figure C.3: H-Q diagram v-notch weir using Kindsvater-Shen relationship (head readings)

90
Appendices

Appendix D: HET-P test procedure

Soil preparation for reconstituted specimens

1. Take a sub-sample of the soil selected for testing, large enough to obtain about 3 kg of
material passing the 9.5 mm (3/8 in) sieve.
2. Depending on the type of soil, oven-dry the sub-sample or proceed with step 3.
3. Sieve the sub-sample through a 9.5 mm (3/8 in) sieve.
4. If the sub-sample was NOT oven-dried, take the water content of the material passing the
9.5 mm sieve according to ASTM Standard D2216-05.
5. If the water content is too high, air-dry the sub-sample and repeat step 4 and 5 as necessary.
6. Calculate the desired compaction parameters:
a. Mass of soil (air- or oven-dry, or water content as received);
b. Mass of water to be added;
c. Mass of wet soil per compaction layer;
d. Height of each compaction layer.
7. Place the soil in a tray over a balance and add the calculated amount of water.
8. Thoroughly mix the soil and water, and place the prepared soil in an air-tight container or
plastic bag to cure (preferably in a moist room). Curing time depends on the type of soil.
Overnight (16 h) is appropriate in most cases (e.g. ASTM Standard D698-07e1).

Specimen preparation for reconstituted specimens

1. Clean the test mold, base, collar and 6-mm center brass rod.
2. Assemble the base, test mold and center brass rod.
3. Take the mass of the assembled mold (base + test mold + center rod).
4. Attach the collar on top of the test mold.
5. Evenly distribute and compact the first layer of soil to the calculated height using the custom
compaction device (Appendix B).
6. Repeat step 5 for layer 2 and 3.
7. Remove the collar, and carefully trim the compacted specimen using a straight edge to form a
plane surface even with the top of the test mold.
8. Take the mass of the compacted specimen including base, test mold and center brass rod.
9. Take the water content of the trimmed material according to ASTM Standard D2216-05.
10. Seal the prepared specimen in plastic bags and/or cling wrap to prevent moisture loss, and
cure overnight before testing.

91
Appendices

Specimen preparation for specimens from undisturbed (intact) soil samples

1. Clean the test mold, base and collar.


2. Assemble the base and test mold.
3. Take the mass of the assembled mold (base + test mold).
4. Trim the soil sample to a length 10-20 mm longer than the test mold and if necessary to a
diameter as close to the diameter of the test mold as possible.
5. Take the mass of the trimmed sample including base, test mold and collar.
6. Place the collar onto the base and underneath the test mold in order to elevate the test mold.
7. Place the trimmed sample centrically inside the test mold, and fix it such that it equally
overtops on either side of the test mold. To do so, place wooden or plastic elements with a
diameter equal to that of the test mold inside the collar to fill the space between base and soil.
8. Record the size of the gap between the soil and the wall of the test mold.
9. Close the gap between soil and test mold with modeling clay 5-10 mm below the top of the
test mold, and fill the remaining gap with epoxy sealant.
10. Seal the assembled mold and trimmed sample in plastic bags and/or cling wrap to prevent
moisture loss, and let the epoxy sealant harden.
11. Carefully turn the trimmed sample and test mold upside down and repeat step 9 and 10.
12. Take the mass of the trimmed sample including filler, base, test mold and collar.
13. Calculate the mass of the filler (modelling clay and epoxy sealant).
14. Carefully trim the soil using a straight edge to form a plane surface even with the top and the
bottom of the test mold.
15. Take the mass of the prepared specimen including base and test mold.
16. Take the water content of the trimmed material according to ASTM Standard D2216-05.
17. Drill an axial 6 mm (1/4 in) hole centrically through the specimen using a drill press and wood
auger bit.

Test preparation

1. Prepare the bridge elements and rubber o-rings, and assemble the downstream flow chamber,
intermediate flange and bridge element with supporting wire mesh.
2. Decide on an appropriate differential pressure transducer according to test heads expected for
the prepared soil specimen, and connect it to the Pitot-static tube.
3. Saturate the transducer system and piezometer tubes with de-aerated water. Make sure that
there is no air left in the system before proceeding to the next step.

92
Appendices

4. Check the alignment of the Pitot-static tube, and slightly open the valves connecting the back
pressure tanks to the Pitot-static tube to allow a low back flow in order to prevent any particles
from entering the Pitot-static tube during setup.
5. Fill the downstream constant head tank with water, slightly overtopping the v-notch weir.
6. Carefully open the downstream gate valve until the water level reaches the top of the
downstream bridge element.
7. Connect the downstream piezometer tube as soon as the connection is fully submerged.
8. Carefully remove the center rod from the soil specimen if necessary, and place the mold
containing the prepared soil specimen on top of the bridge element and wire mesh, making
sure no air is trapped underneath the soil specimen.
9. Place the upstream intermediate flange, and fix the mold to hold it in place centrically above
the Pitot-static tube.
10. Assemble the upstream flow chamber, and securely tighten the test cell.
11. Proceed saturating the system by slowly filling the upstream flow camber with water.
12. Connect the upstream piezometer tube as soon as the connection is fully submerged.
13. Raise the upstream constant head tank to its maximum height, and connect it to the upstream
flow chamber.
14. Proceed saturating the system by slowly filling the upstream constant head tank with water,
and making sure that there are no air pockets trapped inside the flexible tubing.
15. Position the upstream constant head tank at the desired starting head level, and adjust the
inflow to an appropriate rate. Tests usually start at a tank level of 150 mm.
16. Connect the voltmeter, turn it on, and close all back pressure valves.

Test procedure

1. Open the gate valve DS, and start the timer.


2. Take the following readings in intervals of 1-2 minutes:
a. Elapsed time, t ;
b. Sidewall hydraulic head upstream, hu , and downstream, hd ;
c. Total energy head downstream, Hd (voltage output);
d. Centerline velocity head, hv (voltage output);
e. Tank level downstream for flow rate, Q ;
f. Water temperature, T .
3. The intervals may be reduced if erosion is rapid, or increased if there is no noticeable erosion.
4. Raise the upstream constant head tank to increase the test head soon after the flow rate
stabilizes at a given head, and repeat steps 2 through 4. The upstream tank level is usually

93
Appendices

doubled, unless there are reasons to believe that critical conditions may occur at an
intermediate level.
5. Maintain the upstream tank level until the end of the test if the flow rate increases.
6. Stop the test by closing the gate vale DS if one of the following applies:
a. Maximum test head is maintained for at least one hour with no noticeable erosion;
b. At least several minutes of accelerating flow is observed;
c. Flow rate exceeds inflow;
d. The eroded axial hole expanded to the wall of the test mold.

Post-test procedure

1. Disconnect all electronic devices.


2. Drain the apparatus in reverse order according to steps 5 through 14 of the test preparation.
3. Take pictures of the upstream and downstream side of the eroded soil specimen inside the
test mold with appropriate labels.
4. Roughly sketch the shape of the eroded hole and estimate the final hole diameter.
5. Assemble the base and the test mold with the eroded soil specimen, and fill the eroded hole
with Plaster of Paris if reasonable. Otherwise proceed with step 7, 9 and 10.
6. Seal the assembled mold containing soil and Plaster of Paris in plastic bags and/or cling wrap
to prevent moisture loss, and let the plaster cast harden.
7. Carefully extract the plaster cast, and take the water content of the remaining soil specimen
according to ASTM Standard D2216-05.
8. Determine the exact shape, volume, final axial hole diameter, f , and effective hole length, L ,
from the plaster cast. Take at least five measurements in each direction along the length
respectively circumference of the plaster cast.
9. Remove all the remaining soil from the test apparatus and test mold. Make sure to use
sedimentation tanks in combination with sediment traps to prevent soil particles entering the
sewer system.
10. Clean and reassemble everything as necessary to prepare the equipment for the next test.

94
Appendices

Appendix E: Analysis methods

Table E.1: Step by step analysis of test data for different methods

HET HET-P HET-P (V)


Step
(Wan and Fell 2002, 2004) based on energy gradient, i based on i and velocity, Vt
i) Define initial and final flow Define initial and final flow
conditions using Equation conditions using Equation
(2.8) and (2.13), and identify (2.8) and (2.13), and identify
a representative flow a representative flow
condition to be used for this condition to be used for this
test: test:
- laminar; Re < 5000 - laminar; Re < 2000
- turbulent; Re > 5000 - turbulent; Re > 2000
ii) Estimate the initial friction factor Estimate the initial friction factor
fL,o or fT,o based on the initial fL,o or fT,o based on the initial
diameter of the preformed diameter of the preformed
hole, o = 6 mm, using hole, o = 6 mm, using
Equation (2.11) Equation (3.5)
iii) Estimate the final friction factor Estimate the final friction factor
fL,f or fT,f based on the fL,f or fT,f based on the
measured final diameter of measured final diameter of
the eroded hole, f , using the eroded hole, f , using
Equation (2.11) Equation (3.5)
iv) Interpolate the friction factor fL Interpolate the friction factor fL
or fT linearly between its or fT linearly between its
initial (t = 0) and final value initial (t = 0) and final value
(t = tf) (t = tf)
v) Estimate the diameter of the Estimate the diameter of the Estimate the diameter of the
axial hole, t , at any time axial hole, t , at any time axial hole, t , at any time
during the test using during the test using during the test using
Equation (2.12) Equation (3.6) Equation (3.7) and (2.8)
vi) Plot a curve of estimated Plot a curve of estimated Plot a curve of estimated
diameter, t , against time, t diameter, t , against time, t diameter, t , against time, t
vii) Estimate the slope , if Estimate the slope , if Estimate the slope , if
appropriate approximated by appropriate approximated by appropriate approximated by

viii) Estimate wall shear stress, Estimate wall shear stress, Estimate wall shear stress,
, using Equation (2.5) , using Equation (3.3), , using Equation (3.3),
assuming the length of the assuming the length of the
axial hole varies linearly with axial hole varies linearly with
time where applicable time where applicable
ix) Estimate erosion rate, , Estimate erosion rate, , Estimate erosion rate, ,
using Equation (2.7) using Equation (2.7) using Equation (2.7)
x) Plot, , against , and fit Plot, , against , Plot, , against ,
a linear straight line through and fit a linear straight line and fit a linear straight line
the rising part of the curve through the rising part of the through the rising part of the
curve curve
xi) Determine coefficient of soil Determine coefficient of soil Determine coefficient of soil
erosion, Ce , and erosion rate erosion, Ce , and erosion rate erosion, Ce , and erosion rate
index, IHET , using Equation index, IHET , using Equation index, IHET , using Equation
(2.1) (2.1) (2.1)
xii) Graphically obtain critical shear Graphically obtain critical shear Graphically obtain critical shear
stress, c , as illustrated in stress, c , as illustrated in stress, c , as illustrated in
1
*
Figure 2.6 Figure 2.6 Figure 2.6

* Alternatively, critical shear stress, c , may be obtained graphically using the Q- plot as illustrated in Figure 5.6.
1

95
Appendices

Appendix F: Experimental program

Table F.1: Test number information

a) Test number

A Type Name abc


x Type No. 0-9
Ax - y1y2y3 . z1z2 y Series No. 0-9
z Test No. 0-9

b) Universal nomenclature

y3 Setup No. 0 n/a (P) No data recorded (T, C, D, S)


1 DS Pitot Tube, "large" manometer tubes, 10° V-Notch.
2 DS Pitot Tube, p transducers at Pitot Tube,
concrete blocks in inflow tank DS to reduce volume.
3 as #2, concrete blocks in inflow tank DS modified,
total head DS using MM 4 as datum (±0.00)
R Reproduced using data by others
z1
Continuous two-digit test number
z2

c) Specific nomenclature

Type Type No. Series Numbers


T Test runs 1 Preliminary y1 n/a
2 Stage 1
3 Stage 2 y2 n/a

C Calibration 1 Tank US y1 n/a


2 Manometers
3 Tank DS y2 n/a
4 Equipment
D PVC specimens 1 Uniform defect y1 0 2-digit # Hole diameter 1: 6 mm
1 2-digit # Hole diameter 2: 12 mm
2 2-digit # Hole diameter 3: 24 mm
y2 6 2-digit # Hole diameter 1: 6 mm
2 2-digit # Hole diameter 2: 12 mm
4 2-digit # Hole diameter 3: 24 mm
P Soil properties 1 Soil 1 Dummy 11 Gradation Sieving
y1y2

2 Soil 2 Dam 12 Hydrometer


3 Soil 3 O-Clay 13 Combined Methods
21 Compaction Standard
y1y2 y1y2

22 Modified
31
32
S Soil specimens 1 Soil 1 Dummy y1 0 Special conditions (water content)
2 Soil 2 Dam 5 Optimum water content
3 Soil 3 O-Clay 9 Water content as received
y2 0 Special conditions (dry density)
5 95% standard max. dry density
9 Dry density as received

96
Appendices

Table F.2: Test program

Test number Date Audio Data sheet Analysis Notes


Test runs
T1-001.01 02.12.2009 X   V-notch weir preliminary design (10°)
T1-000.01 03.12.2009 X X  V-notch weir improved preliminary design (10°)

T2-000.01 12.01.2010 X X  Setup 1, empty soil mold Leakages


T2-000.02 15.01.2010 X X  Setup 1, empty soil mold Leakages
T2-000.03 22.01.2010 X X  Setup 1, empty soil mold Test okay
T2-001.01 26.01.2010 X   Setup 1, D1-06 Pitot slow

T3-002.01 22.02.2010 X   Setup 2, D1-06, Differential p. transducers Test okay

Calibration
C1-001.01 22.01.2010 X   Maximum overflow capacity

C2-001.01 26.01.2010 X   Zero readings piezometer board setup 1


C2-002.01 22.02.2010 X   Zero readings piezometer board setup 2
C2-002.02 26.02.2010 X  V1.1 Calibration 1 Differential Pressure Transducers
C2-002.03 02.03.2010 X  V1.1 Calibration 2 Differential Pressure Transducers
C2-003.01 27.04.2010 X  V1.1 Calibration 3 Differential Pressure Transducers
C2-Summary X X X V3.1 Summary Differential Pressure Transducers
C2-003.02 03.05.2010 X X  Height instruments and connections setup 3

C3-001.01 22.01.2010 X  V2.1 Calibration 1 v-notch weir


C3-002.01 19.02.2010 X  V2.1 Calibration 2 v-notch weir
C3-Summary X X X V1.2 Summary v-notch weir
C3-003.01 10.05.2010 X  V1.3 Head-Volume Relationship Inflow Tank DS setup 3

C4-003.01 22.07.2010 X   Dimensions Compaction Mold (white) and Center Brass Rod

97
Appendices

Test number Date Audio Data sheet Analysis Notes


Non-erodible test specimens
D1-061.01 27/28.01.10 X  V1.3 2-day full range (US head) test setup 1 Pitot slow, air
D1-062.01 12.03.2010 X  V2.4 Full range (US head) test setup 2: two DT DT1 error

D1-063.01 09.04.2010   Test okay


Uniform defect: diameter = 6 mm
D1-063.02 27.04.2010   V6.3 No flow restrictions Test okay
D1-063.03 27.04.2010   Max flow rate approx. 7.8 l/min Test okay

D1-123.01 03.05.2010   Test okay


Uniform defect: diameter = 12 mm
D1-123.02 03.05.2010   V6.3 No flow restrictions Test okay
D1-123.03 03.05.2010   Max flow rate approx. 36 l/min Test okay

D1-243.01 04.05.2010   Low accuracy


Uniform defect: diameter = 24 mm
D1-243.02 04.05.2010   V6.3 No flow restrictions Low accuracy
D1-243.03 04.05.2010   Max flow rate approx. 39 l/min Low accuracy
D1-243.04 17.06.2011   V6.3 Low range p transducer for better accuracy Test okay

D1-Summary X X X V6.3 Comparison of all three dummy specimens (6,12,24 mm)

Soil properties
P1-110.01 07.07.2010 X  V3.3 Gradation - Sieving - ASTM D 6913-04 (2009)
P1-210.01 08.07.2010 X  V3.3 Standard Compaction Test - ASTM D 698-07e1 - Method A
P1-210.02 21.07.2010 X  V3.3 Compaction Test - adapted from ASTM D 698-07e1 - Method B

P2-13R.01 19.07.2010 X  V3.1 Gradation - Sieving and Hydrometer Analysis Reproduced


P2-13R.02 29.07.2010 X  V1.2 Gradation - Sieving and Hydrometer Analysis Reproduced
P2-210.01 16.07.2010 X  V4.2 Standard Compaction Test - ASTM D 698-07e1 - Method B

Soil property data received from D. Cossette,


P3 09.06.2011 X X X
Civil & Geological Engineering, University of Saskatchewan

98
Appendices

Test number Date Audio Data sheet Analysis Notes


Erodible soil specimens
S1-003.01 11.05.2010   V6.3 Supporting wire mesh clogged Failed
S1-003.02 12.05.2010   V6.3 Specimen collapsed, Pitot DS clogged Failed (blowout)

S1-553.01 22.07.2010   X Specimen collapsed, Pitot DS clogged Failed (blowout)

S2-553.01 23.07.2010   X Specimen collapsed, Pitot DS clogged Failed (blowout)

S3-Little Cataraqui
S3-993.01a 27.08.2010   X Part 1: hole blocked Test aborted
S3-993.01b 27.08.2010   V6.3 Part 2: bottom part intact => V unusable, Plaster cast made Partial test
S3-993.02 08.11.2010   V6.3 Spec. suddenly collapsed at H = 0.5 m, t = 1:00 h Structural failure

S3-Bear Brook
S3-993.11 07.10.2010 X X X Sample disturbed and cannot be tested Failed
S3-993.12 12.10.2010   V6.3 Duration too short, Plaster cast made Test okay

S3-Wilton Creek
S3-993.21 12.11.2010   V6.3 Spec. suddenly collapsed at H = 2.0 m, Plaster cast made Structural failure
S3-993.22 15.10.2010   V6.3 Spec. suddenly collapsed at H = 0.6 m Structural failure

S3-Raisin River X X X Not tested: lots of organic matter

S3-Jock River X X X Cannot be tested: specimen too small, lots of organic matter

S3-Summary X X X V6.3 Comparison of qualitative results for all soils type S3

99
Appendices

Appendix G: Non-erodible test specimens

Table G.1: Test data and results non-erodible test specimen D1-063

o Head Q hu hd hv Hd Re CpCv Vt t s i H/h HET HET-P HET-P (V)


2 2 2
[mm] [mm] [l/min] [mm] [mm] [mm] [mm] [-] [-] [m/s] [mm] [-] [-] [-] [N/m ] [N/m ] [N/m ]
6 20 0.8 151 131 14 145 2.62E+03 0.878 0.43 6.2 0.20 0.06 0.29 3.0 0.9 0.9
6 23 0.8 154 131 14 145 2.36E+03 0.922 0.43 6.3 0.23 0.09 0.41 3.4 1.4 1.5
6 67 1.2 202 135 42 180 3.39E+03 0.804 0.75 5.9 0.68 0.22 0.33 9.9 3.3 3.2
6 156 2.1 299 142 109 251 5.62E+03 0.829 1.21 6.0 1.58 0.48 0.31 23.2 7.1 7.1
6 255 2.6 400 146 184 325 8.32E+03 0.808 1.57 5.9 2.57 0.76 0.30 37.7 11.2 11.0
6 345 3.1 494 148 251 397 8.97E+03 0.827 1.83 6.0 3.49 0.98 0.28 51.3 14.4 14.4
6 451 3.6 602 151 344 491 9.76E+03 0.812 2.14 6.0 4.56 1.12 0.25 66.9 16.5 16.4
6 550 3.9 703 152 410 553 1.15E+04 0.806 2.34 5.9 5.57 1.52 0.27 81.7 22.3 22.0
6 640 4.2 795 154 470 618 1.19E+04 0.824 2.50 6.0 6.47 1.79 0.28 95.1 26.3 26.2
6 739 4.6 895 156 565 713 1.26E+04 0.820 2.75 6.0 7.47 1.83 0.25 109.7 26.9 26.9
6 842 4.8 999 156 622 770 1.35E+04 0.811 2.88 5.9 8.52 2.31 0.27 125.1 34.0 33.7
6 949 5.2 1107 157 692 851 1.41E+04 0.825 3.04 6.0 9.60 2.58 0.27 140.9 38.0 38.0
6 1048 5.4 1207 158 784 943 1.47E+04 0.808 3.24 5.9 10.60 2.67 0.25 155.7 39.2 38.8
6 1137 5.7 1298 159 848 996 1.54E+04 0.818 3.36 6.0 11.51 3.05 0.27 169.0 44.8 44.6
6 1234 5.7 1395 159 897 1060 1.56E+04 0.804 3.46 5.9 12.48 3.39 0.27 183.3 49.8 49.1
6 1343 6.0 1505 160 978 1144 1.63E+04 0.805 3.61 5.9 13.59 3.64 0.27 199.6 53.5 52.8
6 1436 6.3 1599 161 1070 1215 1.72E+04 0.813 3.78 6.0 14.53 3.88 0.27 213.3 57.0 56.6
6 1533 6.5 1697 161 1134 1275 1.78E+04 0.816 3.89 6.0 15.52 4.26 0.27 227.8 62.6 62.2
6 1635 6.7 1800 161 1190 1356 1.83E+04 0.818 3.99 6.0 16.56 4.48 0.27 243.1 65.8 65.5
6 1738 7.0 1904 162 1293 1445 1.86E+04 0.822 4.15 6.0 17.59 4.64 0.26 258.3 68.1 68.0
6 1843 7.1 2009 162 1353 1505 1.95E+04 0.816 4.25 6.0 18.66 5.10 0.27 273.9 74.8 74.4
6 1925 7.3 2092 162 1420 1568 2.00E+04 0.818 4.35 6.0 19.49 5.29 0.27 286.3 77.7 77.3
6 2035 7.5 2202 163 1515 1664 2.00E+04 0.816 4.50 6.0 20.60 5.44 0.26 302.4 79.9 79.4
6 2132 7.8 2300 163 1568 1731 2.12E+04 0.827 4.58 6.0 21.59 5.75 0.27 317.0 84.5 84.5
6 2242 7.8 2410 163 1649 1798 2.13E+04 0.810 4.69 5.9 22.70 6.18 0.27 333.3 90.8 90.0

100
Appendices

Table G.2: Test data and results non-erodible test specimen D1-123

o Head Q hu hd hv Hd Re CpCv Vt t s i H/h HET HET-P HET-P (V)


2 2 2
[mm] [mm] [l/min] [mm] [mm] [mm] [mm] [-] [-] [m/s] [mm] [-] [-] [-] [N/m ] [N/m ] [N/m ]
12 8 1.9 149 142 7 148 2.55E+03 0.770 0.30 11.6 0.08 0.01 0.11 2.2 0.2 0.2
12 14 2.6 160 147 12 159 3.47E+03 0.790 0.40 11.7 0.13 0.01 0.09 3.9 0.4 0.4
12 46 4.9 204 158 40 198 6.71E+03 0.814 0.74 11.9 0.46 0.06 0.13 13.6 1.7 1.7
12 134 8.4 304 170 122 293 1.15E+04 0.802 1.27 11.8 1.35 0.11 0.08 39.7 3.3 3.2
12 219 11.0 397 177 198 371 1.50E+04 0.824 1.62 12.0 2.22 0.26 0.12 65.1 7.7 7.7
12 313 13.1 496 182 286 463 1.73E+04 0.814 1.95 11.9 3.17 0.34 0.11 93.1 9.9 9.9
12 408 14.8 595 185 381 551 1.97E+04 0.799 2.26 11.8 4.14 0.45 0.11 121.6 13.1 12.9
12 505 16.7 696 189 466 653 2.22E+04 0.815 2.49 11.9 5.13 0.43 0.08 150.5 12.7 12.6
12 606 18.5 800 191 558 749 2.45E+04 0.824 2.73 12.0 6.15 0.52 0.08 180.7 15.3 15.2
12 699 20.2 896 193 646 833 2.68E+04 0.838 2.94 12.1 7.10 0.63 0.09 208.5 18.6 18.7
12 806 21.5 1006 195 738 936 2.85E+04 0.833 3.14 12.1 8.19 0.71 0.09 240.6 20.8 20.9
12 905 22.8 1107 196 830 1014 3.02E+04 0.833 3.33 12.1 9.20 0.94 0.10 270.2 27.7 27.9
12 995 23.9 1199 198 901 1102 3.16E+04 0.836 3.47 12.1 10.11 0.98 0.10 296.9 28.8 29.0
12 1091 25.1 1297 198 996 1190 3.32E+04 0.836 3.65 12.1 11.10 1.08 0.10 326.0 31.7 31.9
12 1192 26.0 1399 200 1095 1278 3.45E+04 0.828 3.82 12.0 12.11 1.22 0.10 355.7 35.8 35.9
12 1299 27.3 1508 200 1194 1370 3.51E+04 0.830 3.99 12.0 13.21 1.39 0.11 388.0 40.9 41.0
12 1387 28.3 1598 203 1275 1448 3.74E+04 0.832 4.13 12.1 14.09 1.52 0.11 413.8 44.5 44.7
12 1482 29.5 1694 202 1346 1554 3.80E+04 0.847 4.24 12.2 15.07 1.42 0.09 442.6 41.6 42.1
12 1590 30.4 1804 202 1473 1642 3.91E+04 0.833 4.43 12.1 16.18 1.64 0.10 475.2 48.0 48.3
12 1690 31.3 1905 206 1540 1724 4.14E+04 0.839 4.53 12.1 17.16 1.84 0.11 504.0 53.9 54.3
12 1783 32.1 1999 205 1635 1780 4.13E+04 0.835 4.67 12.1 18.12 2.21 0.12 532.2 65.0 65.4
12 1876 33.2 2094 206 1709 1900 4.27E+04 0.845 4.78 12.1 19.07 1.96 0.10 560.0 57.6 58.3
12 1983 34.1 2202 208 1815 2003 4.52E+04 0.843 4.92 12.1 20.14 2.02 0.10 591.5 59.2 59.9
12 2086 35.0 2306 208 1900 2084 4.50E+04 0.844 5.04 12.1 21.19 2.25 0.11 622.3 66.0 66.7
12 2193 35.9 2414 208 2003 2197 4.62E+04 0.845 5.17 12.1 22.28 2.20 0.10 654.4 64.5 65.3

101
Appendices

Table G.3: Test data and results non-erodible test specimen D1-243

o Head Q hu hd hv Hd Re CpCv Vt t s i H/h HET HET-P HET-P (V)


2 2 2
[mm] [mm] [l/min] [mm] [mm] [mm] [mm] [-] [-] [m/s] [mm] [-] [-] [-] [N/m ] [N/m ] [N/m ]
24 0 3.7 153 153 2 154 2.62E+03 0.704 0.16 22.2 0.00 -0.01 n/a 0.0 -0.8 -0.8
24 0 3.7 153 153 2 154 2.62E+03 0.704 0.16 22.2 0.00 -0.01 n/a 0.0 -0.4 -0.4
24 5 7.1 172 167 5 171 5.01E+03 0.816 0.27 23.9 0.05 0.00 0.07 2.7 0.2 0.2
24 5 7.1 172 168 5 171 4.87E+03 0.816 0.27 23.9 0.05 0.01 0.18 2.7 0.5 0.5
24 12 10.6 190 179 12 190 7.26E+03 0.808 0.40 23.7 0.11 0.00 0.00 6.7 0.0 0.0
24 12 10.6 190 179 12 190 7.26E+03 0.808 0.40 23.7 0.11 0.00 0.00 6.7 0.0 0.0
24 23 14.6 210 189 22 209 9.92E+03 0.815 0.54 23.8 0.22 0.01 0.06 12.8 0.7 0.7
24 24 14.3 210 188 22 209 9.77E+03 0.808 0.54 23.8 0.22 0.01 0.03 12.9 0.3 0.3
24 24 14.6 211 189 22 209 9.92E+03 0.815 0.54 23.8 0.22 0.02 0.07 13.1 0.9 0.9
24 24 14.6 211 189 22 209 9.92E+03 0.821 0.54 23.9 0.22 0.02 0.07 13.1 0.9 0.9
24 36 17.9 230 197 33 229 1.19E+04 0.827 0.66 24.0 0.33 0.01 0.02 19.6 0.5 0.5
24 37 17.9 231 197 33 230 1.19E+04 0.827 0.66 24.0 0.34 0.01 0.04 20.2 0.9 0.9
24 52 21.2 252 205 45 250 1.41E+04 0.830 0.78 24.1 0.47 0.02 0.04 27.7 1.1 1.1
24 52 21.2 252 205 46 250 1.41E+04 0.827 0.78 24.0 0.47 0.02 0.04 27.7 1.1 1.1
24 65 23.7 270 211 58 269 1.57E+04 0.816 0.88 23.9 0.60 0.01 0.02 35.0 0.6 0.6
24 65 23.7 270 211 58 269 1.57E+04 0.816 0.88 23.9 0.60 0.01 0.02 35.0 0.6 0.6
24 80 26.7 289 216 72 286 1.77E+04 0.828 0.98 24.0 0.74 0.03 0.04 43.3 1.8 1.9
24 81 26.7 290 216 71 286 1.77E+04 0.832 0.97 24.1 0.75 0.04 0.06 43.9 2.4 2.5
24 96 29.3 309 222 86 307 1.94E+04 0.832 1.07 24.1 0.88 0.02 0.03 51.9 1.4 1.4
24 96 29.3 309 222 86 307 1.94E+04 0.832 1.07 24.1 0.88 0.03 0.03 51.9 1.6 1.6
24 111 31.5 327 226 99 326 2.08E+04 0.832 1.15 24.1 1.02 0.02 0.02 59.9 0.9 0.9
24 112 31.5 328 226 99 326 2.08E+04 0.833 1.15 24.1 1.03 0.02 0.02 60.5 1.1 1.1
24 129 34.2 349 231 116 347 2.27E+04 0.835 1.25 24.2 1.19 0.02 0.02 70.0 1.2 1.3
24 129 34.2 349 231 116 347 2.27E+04 0.834 1.25 24.1 1.19 0.02 0.02 70.0 1.2 1.3
24 133 35.1 354 232 120 352 2.32E+04 0.842 1.27 24.2 1.23 0.02 0.02 72.4 1.2 1.2
24 133 35.1 354 232 121 353 2.32E+04 0.839 1.27 24.2 1.23 0.02 0.01 72.4 1.0 1.0
24 134 35.8 356 233 121 353 2.37E+04 0.855 1.27 24.4 1.24 0.03 0.03 73.0 1.8 1.9

102
Appendices

Appendix H: Soil properties

SAND SIZE GRAVEL SIZE COBBLE


FINES
FINE MEDIUM COARSE FINE COARSE SIZE

Sieve No. 200 100 60 40 20 10 4 ⅜'' ¾'' 1'' 1½''2'' 3''


100

90

80

70
Percent Finer Than

60

50

40

30

20 Dam MV4-Core

10 Dam MV4-Altered

0
0.0001 0.001 0.01 0.1 1 10 100
Grain Size [mm]
Figure H.1: Gradation curves dam core material (USCS)

103
Appendices

21.5
Compaction Data
7.8%; 20.80 Compaction Curve
21.0
Optimum
95% Max. Std. Dry Unit Weight
20.5
Dry unit weight [kN/m3]

Water Content Limits


Sr = 1.0 (Zero Air Voids Curve)
20.0
4.0%; 19.76 11.3%; 19.76
19.5
Gs = 2.72 (assumed)
19.0 wreceived = 0.2 %
Std-wopt = 7.8 %
Std-gd,max = 20.80 kN/m3
18.5 Std-d,max = 2121 kg/m3
Std-Sr,opt = 76 %
18.0
0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22%
Water Content [%]
Figure H.2: Standard compaction test Dam MV4-Altered material (ASTM D698 Method A)

21.5
8.2%; 21.00 Compaction Data
Compaction Curve
21.0
Optimum
95% Max. Std. Dry Unit Weight
20.5
Dry unit weight [kN/m3]

Water Content Limits


Sr = 1.0 (Zero Air Voids Curve)
20.0 2.7%; 19.95 11.0%; 19.95

19.5
Gs = 2.72 (assumed)
19.0 wreceived = 5.2-8.2 %
Std-wopt = 8.2 %
Std-gd,max = 21.00 kN/m3
18.5 Std-d,max = 2141 kg/m3
Std-Sr,opt = 83.2 %
18.0
0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22%
Water Content [%]
Figure H.3: Standard compaction test Dam MV4-Core material (ASTM D698 Method B)

104
Appendices

Appendix I: Erodible soil specimens

Dam core material

40

35
Measured Flow Rate, Q [l/min]

30 Sudden very rapid progressive erosion:


h = 160 mm
25 H = 130 mm
t = 10 mm
20 L = 99 mm
c,HET ≤ 40 N/m2
15 c,HET-P ≤ 32 N/m2

10

0
00:00:00 00:15:00 00:30:00 00:45:00
Time, t [hh:mm:ss]

Figure I.1: Dam MV4-Altered – S1-003.01 – Measured flow rate & results at point of failure

Upstream (US) Downstream (DS)

Figure I.2: Dam MV4-Altered – S1-003.01 – Dried specimen

105
Appendices

40

35
Measured Flow Rate, Q [l/min]

30

25

20 Sudden very rapid progressive erosion:


h = 148 mm
15 H = null mm
t = 7-8 mm
10 L = 99 mm
c,HET ≤ 27 N/m2
5 c,HET-P ≤ null N/m2

0
00:00:00 00:05:00 00:10:00 00:15:00 00:20:00 00:25:00 00:30:00
Time, t [hh:mm:ss]

Figure I.3: Dam MV4-Altered – S1-003.02 – Measured flow rate & results at point of failure

Figure I.4: Dam MV4-Altered – S1-003.02 – Drained test cell US and detail with Pitot tube

106
Appendices

40

35
Measured Flow Rate, Q [l/min]

30 Sudden very rapid progressive erosion:


h = 2300 mm max test head
25 H = null mm
t = 0.0 mm hole collapsed
20 L = 99 mm
c,HET ≤ null N/m2 h not
15 c,HET-P ≤ null N/m2 representative

10

0
00:00:00 00:05:00 00:10:00 00:15:00 00:20:00 00:25:00 00:30:00
Time, t [hh:mm:ss]

Figure I.5: Dam MV4-Altered – S1-553.01 – Measured flow rate & results at point of failure

Figure I.6: Dam MV4-Altered – S1-553.01 – Drained test cell US and detail with Pitot tube

107
Appendices

40

35
Measured Flow Rate, Q [l/min]

30

25

20 Sudden very rapid progressive erosion:


h = 187 mm
15 H = null mm
t =6 mm
10 L = 99 mm
c,HET ≤ 28 N/m2
5 c,HET-P ≤ null N/m2

0
00:00:00 00:05:00 00:10:00 00:15:00 00:20:00 00:25:00 00:30:00
Time, t [hh:mm:ss]

Figure I.7: Dam MV4-Core – S2-553.01 – Measured flow rate & results at point of failure

Figure I.8: Dam MV4-Core – S2-553.01 – Drained test cell US and detail with Pitot tube

108
Appendices

Ontario clay samples

Figure I.9: Ontario Clay – Little Cataraqui – S3-993.01 – Prepared test specimen

50
45
Measured Flow Rate, Q [l/min]

40
35
30
25
Critical point:
20
start of failure
15
10
5
0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.10: Ontario Clay – Little Cataraqui – S3-993.01 – Measured flow rate

109
Appendices

2000 2000
Hydraulic head difference across test specimen

1800 HET 1800

Energy head loss along test specimen


HET raw Critical point:
1600 1600
HET-P start of failure
1400 HET-P raw 1400

(HET-P), H [mm]
(HET), h [mm]

1200 HET-P (V) 1200


1000 HET-P (V) raw 1000
800 800
600 600
400 400
200 200
0 0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.11: Ontario Clay – Little Cataraqui – S3-993.01 – Head difference and energy loss

1.2

1.0
Head Ratio, H/h [ - ]

Critical point:
0.8 start of failure

0.6

0.4

0.2 Head Ratio (Re>2000)

Head Ratio raw


0.0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.12: Ontario Clay – Little Cataraqui – S3-993.01 – Head ratio

110
Appendices

70
Estimated mean diameter of axial hole, t [mm]
HET HET-P (V): The Pitot-static tube
60 HET raw measured a velocity according
HET-P to an almost constant
50 diameter, because the bottom
HET-P raw of the specimen remained Critical point:
HET-P (V) partially intact, including the start of failure
40
HET-P (V) raw preformed axial hole.

30

20

10

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.13: Ontario Clay – Little Cataraqui – S3-993.01 – Axial hole diameter

1600
HET
Estimated wall shear stress,  [N/m2]

1400
HET raw
1200 HET-P Critical point:
start of failure
HET-P raw
1000
HET-P (V)
800 HET-P (V) raw

600

400

200

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.14: Ontario Clay – Little Cataraqui – S3-993.01 – Wall shear stress

111
Appendices

0.04
Estimated erosion rate per unit surface area of

HET
0.035
HET raw
0.03 Peak and decreasing HET-P
the axial hole, ε̇ [kg/s/m2]

erosion rate with


HET-P raw
0.025 each test head.
HET-P (V)
0.02 HET-P (V) raw
Critical point:
0.015
start of failure
0.01

0.005

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00
Time, t [hh:mm:ss]

Figure I.15: Ontario Clay – Little Cataraqui – S3-993.01 – Erosion rate

50
45 HET HET raw
HET-P HET-P raw
Measured Flow Rate, Q [l/min]

40
HET-P (V) HET-P (V) raw
35
30
Critical point:
25
start of failure
20
15
10
Critical point:
5
start of failure
0
0 200 400 600 800 1000 1200 1400 1600
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure I.16: Ontario Clay – Little Cataraqui – S3-993.01 – Flow rate versus shear stress

112
Appendices

Figure I.17: Ontario Clay – Little Cataraqui – S3-993.01 – Drained test cell US and plaster cast

Post-test: a portion of the bottom part of


the specimen remained intact.

Figure I.18: Ontario Clay – Little Cataraqui – S3-993.01 – DS side of specimen pre- and post-test

113
Appendices

Figure I.19: Ontario Clay – Little Cataraqui – S3-993.02 – Prepared test specimen

50
45
Specimen
Measured Flow Rate, Q [l/min]

40
suddenly
35 collapsed!
30
25
Critical point:
20
start of failure
15
10
5
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.20: Ontario Clay – Little Cataraqui – S3-993.02 – Measured flow rate

114
Appendices

350 350
Hydraulic head difference across test specimen

HET Critical point:

Energy head loss along test specimen


300 HET raw start of failure 300
HET-P
250 250
HET-P raw

(HET-P), H [mm]
Specimen
(HET), h [mm]

200 HET-P (V) suddenly


200
HET-P (V) raw collapsed!
150 150

100 Critical point: 100


start of failure
50 50

0 0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.21: Ontario Clay – Little Cataraqui – S3-993.02 – Head difference and energy loss

1.2
Head Ratio (Re>2000) Specimen
1 suddenly
Head Ratio raw
collapsed!
Head Ratio, H/h [ - ]

0.8

0.6 Critical point:


start of failure

0.4

0.2

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.22: Ontario Clay – Little Cataraqui – S3-993.02 – Head ratio

115
Appendices

40
Estimated mean diameter of axial hole, t [mm]

HET Specimen
35
HET raw suddenly
30 HET-P collapsed!
HET-P raw
25
HET-P (V)
Critical point:
20 HET-P (V) raw
start of failure
15

10

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.23: Ontario Clay – Little Cataraqui – S3-993.02 – Axial hole diameter

100
90 HET
Estimated wall shear stress,  [N/m2]

HET raw Critical point:


80
HET-P start of failure
70 HET-P raw
60 HET-P (V)
50 HET-P (V) raw

40
30
Critical point:
20 start of failure Specimen
suddenly
10 collapsed!
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.24: Ontario Clay – Little Cataraqui – S3-993.02 – Wall shear stress

116
Appendices

0.09
Estimated erosion rate per unit surface area of

HET
0.08
HET raw
Critical point:
0.07 HET-P start of failure
the axial hole, ε̇ [kg/s/m2]

0.06 HET-P raw Specimen


HET-P (V) suddenly
0.05 collapsed!
HET-P (V) raw
0.04
Peak and decreasing
0.03
erosion rate with
0.02 each test head.

0.01

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00
Time, t [hh:mm:ss]

Figure I.25: Ontario Clay – Little Cataraqui – S3-993.02 – Erosion rate

50
HET
45
HET raw
Measured Flow Rate, Q [l/min]

40 HET-P
35 HET-P raw
HET-P (V) Specimen
30 suddenly
HET-P (V) raw
collapsed!
25
Critical point:
20
start of failure
15
Critical point:
10 start of failure
5
0
0 10 20 30 40 50 60 70 80 90 100
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure I.26: Ontario Clay – Little Cataraqui – S3-993.02 – Flow rate versus shear stress

117
Appendices

Figure I.27: Ontario Clay – Little Cataraqui – S3-993.02 – Drained test cell US

Figure I.28: Ontario Clay – Little Cataraqui – S3-993.02 – Test specimen after testing (DS)

118
Appendices

Figure I.29: Ontario Clay – Bear Brook – S3-993.12 – Prepared test specimen

20
18
Critical point:
Measured Flow Rate, Q [l/min]

16
start of failure
14
12
10
8
6
4
2
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.30: Ontario Clay – Bear Brook – S3-993.12 – Measured flow rate

119
Appendices

450 450
Hydraulic head difference across test specimen

HET
400 400

Energy head loss along test specimen


HET raw
350 350
HET-P
Critical point:

(HET-P), H [mm]
300 HET-P raw start of failure 300
(HET), h [mm]

250 HET-P (V) 250

200 HET-P (V) raw 200

150 150
Critical point:
100 100
start of failure
50 50

0 0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.31: Ontario Clay – Bear Brook– S3-993.12 – Head difference and energy loss

1.2

1
Critical point:
Head Ratio, H/h [ - ]

start of failure
0.8

0.6

0.4
Head Ratio (Re>2000)
0.2
Head Ratio raw
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.32: Ontario Clay – Bear Brook – S3-993.12 – Head ratio

120
Appendices

20
Estimated mean diameter of axial hole, t [mm]

18
Critical point:
16
start of failure
14
12
10 HET
8 HET raw
HET-P
6
HET-P raw
4
HET-P (V)
2 HET-P (V) raw
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.33: Ontario Clay – Bear Brook – S3-993.12 – Axial hole diameter

250
HET
Critical point:
Estimated wall shear stress,  [N/m2]

HET raw start of failure


200
HET-P
HET-P raw
150
HET-P (V)
HET-P (V) raw
100

50
Critical point:
start of failure
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.34: Ontario Clay – Bear Brook – S3-993.12 – Wall shear stress

121
Appendices

0.018
Estimated erosion rate per unit surface area of

HET
0.016
HET raw
0.014 Peak and decreasing
erosion rate with HET-P
the axial hole, ε̇ [kg/s/m2]

0.012 each test head. HET-P raw


HET-P (V)
0.01
HET-P (V) raw
0.008
Critical point:
0.006
start of failure
0.004

0.002

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 01:30:00
Time, t [hh:mm:ss]

Figure I.35: Ontario Clay – Bear Brook – S3-993.12 – Erosion rate

20
18
Measured Flow Rate, Q [l/min]

16
Critical point:
14
start of failure
12
10 HET
8 HET raw
6 HET-P
Critical point: HET-P raw
4
start of failure HET-P (V)
2
HET-P (V) raw
0
0 50 100 150 200 250
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure I.36: Ontario Clay – Bear Brook – S3-993.12 – Flow rate versus shear stress

122
Appendices

Figure I.37: Ontario Clay – Bear Brook – S3-993.12 – Drained test cell US and plaster cast

Figure I.38: Ontario Clay – Bear Brook – S3-993.12 – Extracted test specimen after testing

123
Appendices

Figure I.39: Ontario Clay – Wilton Creek – S3-993.21 – Prepared test specimen

18
Critical point:
16
start of failure
Measured Flow Rate, Q [l/min]

14

12

10

8
Specimen
6
suddenly
4 collapsed!

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.40: Ontario Clay – Wilton Creek – S3-993.21 – Measured flow rate

124
Appendices

2000 2000
Hydraulic head difference across test specimen

HET Critical point:


1800 1800

Energy head loss along test specimen


HET raw start of failure
1600 1600
HET-P
1400 HET-P raw 1400

(HET-P), H [mm]
(HET), h [mm]

1200 HET-P (V) 1200


1000 HET-P (V) raw 1000
800 800
600 600
400 400
Specimen
200 200
suddenly
0 collapsed! 0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.41: Ontario Clay – Wilton Creek – S3-993.21 – Head difference and energy loss

1.2
Head Ratio (Re>2000)
Critical point:
1 Head Ratio raw start of failure
Head Ratio, H/h [ - ]

0.8

0.6

0.4

Specimen
0.2 suddenly
collapsed!
0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.42: Ontario Clay – Wilton Creek – S3-993.21 – Head ratio

125
Appendices

20
Estimated mean diameter of axial hole, t [mm]
HET HET raw
18
HET-P HET-P raw Critical point:
16
HET-P (V) HET-P (V) raw start of failure
14
12
10
8
6
4 Specimen
suddenly
2 collapsed!
0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.43: Ontario Clay – Wilton Creek – S3-993.21 – Axial hole diameter

1000
HET Critical point:
900
Estimated wall shear stress,  [N/m2]

HET raw start of failure


800
HET-P
700 HET-P raw
600 HET-P (V)
500 HET-P (V) raw

400
300
200 Specimen
100 suddenly
collapsed!
0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.44: Ontario Clay – Wilton Creek – S3-993.21 – Wall shear stress

126
Appendices

0.02
Estimated erosion rate per unit surface area of

HET Specimen
HET raw suddenly
0.015 HET-P collapsed!
the axial hole, ε̇ [kg/s/m2]

HET-P raw
HET-P (V)
HET-P (V) raw Critical point:
0.01
start of failure
Peak and decreasing
erosion rate with
0.005 each test head.

0
00:00:00 00:30:00 01:00:00 01:30:00 02:00:00 02:30:00 03:00:00
Time, t [hh:mm:ss]

Figure I.45: Ontario Clay – Wilton Creek – S3-993.21 – Erosion rate

20
Critical point:
18
start of failure
Measured Flow Rate, Q [l/min]

16
14
12
10 Specimen
suddenly
8 Critical point: collapsed!
start of failure
6
HET HET raw
4
HET-P HET-P raw
2
HET-P (V) HET-P (V) raw
0
0 100 200 300 400 500 600 700 800 900 1000
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure I.46: Ontario Clay – Wilton Creek – S3-993.21 – Flow rate versus shear stress

127
Appendices

Figure I.47: Ontario Clay – Wilton Creek – S3-993.21 – Drained test cell US and plaster cast

Figure I.48: Ontario Clay – Wilton Creek – S3-993.21 – Test specimen after testing (DS)

128
Appendices

Figure I.49: Ontario Clay – Wilton Creek – S3-993.22 – Prepared test specimen

30

25
Measured Flow Rate, Q [l/min]

20
Increase
15 test head:
Specimen
Critical point: suddenly
10 start of failure collapsed!

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.50: Ontario Clay – Wilton Creek – S3-993.22 – Measured flow rate

129
Appendices

350 350
Hydraulic head difference across test specimen

HET

Energy head loss along test specimen


300 HET raw 300
HET-P
250 250
HET-P raw

(HET-P), H [mm]
(HET), h [mm]

200 HET-P (V) 200


HET-P (V) raw Increase
150 test head: 150
Specimen
100 suddenly 100
collapsed!
50 Critical point: 50
start of failure
0 0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.51: Ontario Clay – Wilton Creek – S3-993.22 – Head difference and energy loss

1.2
Re < 5000
1
Head Ratio, H/h [ - ]

0.8

0.6
Critical point:
0.4 start of failure Increase
test head:
Specimen
0.2 Head Ratio (Re>2000)
suddenly
Head Ratio raw collapsed!
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.52: Ontario Clay – Wilton Creek – S3-993.22 – Head ratio

130
Appendices

35
Estimated mean diameter of axial hole, t [mm]

HET
30 HET raw
HET-P Critical point:
25
HET-P raw start of failure

20 HET-P (V)
HET-P (V) raw Increase
15 Re < 5000 test head:
Specimen
10 suddenly
collapsed!
5

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.53: Ontario Clay – Wilton Creek – S3-993.22 – Axial hole diameter

100
HET
90
Estimated wall shear stress,  [N/m2]

HET raw
80
HET-P
70 HET-P raw
Increase
60 HET-P (V)
test head:
50 HET-P (V) raw Specimen
suddenly
40
collapsed!
30
Critical point:
20
start of failure
10
0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.54: Ontario Clay – Wilton Creek – S3-993.22 – Wall shear stress

131
Appendices

0.06
Estimated erosion rate per unit surface area of

HET
0.05 HET raw
HET-P
the axial hole, ε̇ [kg/s/m2]

Critical point:
0.04 HET-P raw
start of failure
HET-P (V) Increase
HET-P (V) raw test head:
0.03
Specimen
Peak and decreasing suddenly
0.02 erosion rate with collapsed!
each test head.
0.01

0
00:00:00 00:15:00 00:30:00 00:45:00 01:00:00
Time, t [hh:mm:ss]

Figure I.55: Ontario Clay – Wilton Creek – S3-993.22 – Erosion rate

30
HET Increase
25 HET raw test head:
Measured Flow Rate, Q [l/min]

HET-P Specimen
suddenly
20 HET-P raw collapsed!
HET-P (V)
15 HET-P (V) raw Critical point: Critical point:
start of failure start of failure
10

0
0 10 20 30 40 50 60 70 80 90 100
Critical shear stress defined at Estimated wall shear stress,  [N/m2]
point with minimum slope.

Figure I.56: Ontario Clay – Wilton Creek – S3-993.22 – Flow rate versus shear stress

132
Appendices

Figure I.57: Ontario Clay – Wilton Creek – S3-993.22 – Drained test cell US

Figure I.58: Ontario Clay – Wilton Creek – S3-993.22 – Test specimen after testing (DS)

133

You might also like