Scour Potential of Cohesive Soils 
 
by 
 
Melvin Elmo Walker III 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master of Science 
 
Auburn, Alabama 
August 3, 2013 
 
 
 
 
Keywords: scour, erosion function, cohesive soils, erosion function apparatus, ultrasonic 
 
 
 
Copyright 2013 by Melvin Elmo Walker III 
 
 
Approved by 
 
J. Brian Anderson, Chair, Associate Professor of Civil Engineering 
Xing Fang, Professor of Civil Engineering 
Robert W. Barnes, James J. Mallett Associate Professor of Civil Engineering 
 
 
 
 
 
 
ii 
 
Abstract 
 
 
 Determination of erosion parameters in order to predict scour depth is imperative to 
designing safe, economic, and efficient bridge foundations.  Scour behavior of granular soils 
is generally understood, and design criteria have been established by the Federal Highway 
Administration.  The same is not true for cohesive soils, and because of their complexity, a 
universal scour prediction method has not been established by the industry.  The Erosion 
Function Apparatus (EFA) was created to determine the rate of scour of cohesive soils under 
known shear stresses, which can then be used to predict scour depths under similar 
conditions.  
 During this study, nine cohesive soil formations were sampled with the assistance of 
the Alabama Department of Transportation.  Six of these formations were scour tested in an 
updated EFA featuring an ultrasonic sensor for quantitative erosion measurements.  EFA 
tests were performed to determine erosion functions and whether any formations 
demonstrated scour resistance.  Geotechnical index tests were also performed on these 
formations to correlate scour to geotechnical properties. 
 Results of testing verified the performance of the ultrasonic sensor and updated EFA.  
Three of the tested formations were scour resistant, while three formations showed evidence 
of scour.  Velocity versus scour rate curves were generated for the scourable formations with 
scour rates upwards of 15 mm per hour.  The scour behavior observed was unique among 
formations limiting the ability to establish correlations between tests. 
iii 
 
Acknowledgments 
 
 
 I would like to thank Dr. Anderson for his patience and wisdom in guiding me 
through the thesis preparation process.  I am very grateful to Dr. Anderson for the 
opportunity to gain experience in research and engineering.  I would also like to thank my 
other committee members Dr. Fang and Dr. Barnes for their assistance throughout this 
process.  This project would not be possible without the support of the Alabama Department 
of Transportation.  The Auburn University Highway Research Center was also instrumental 
in making improvements to the Erosion Function Apparatus.   I would also like to thank 
Chris Jette of SeaTek for working with Dr. Anderson and me in designing the new ultrasonic 
sensor for the Erosion Function Apparatus.  Finally, I am very grateful for the support and 
patience of my wife, Brittany, and the rest of my family. 
iv 
 
Table of Contents 
 
 
Abstract ......................................................................................................................................... ii 
Acknowledgments........................................................................................................................ iii  
List of Tables ............................................................................................................................. viii  
List of Figures .............................................................................................................................. ix  
List of Abbreviations ................................................................................................................. xvi 
Chapter 1 Introduction   ................................................................................................................ 1 
 1.1 Background   ............................................................................................................... 1 
 1.2 Objectives   ................................................................................................................. 3 
 1.3 Scope of Study   .......................................................................................................... 3 
Chapter 2 Literature Review   ....................................................................................................... 4 
 2.1 Scour Background Information   ................................................................................. 4 
 2.2 Scour Rate in Cohesive Soils Method   ...................................................................... 8 
 2.3 Erosion Function Apparatus ..................................................................................... 11 
 2.4 Alternatives to the Erosion Function Apparatus   ..................................................... 14 
 2.5 Scour Relationships with Geotechnical Parameters ................................................. 19 
 2.6 Previous Scour Research at Auburn University........................................................ 20 
 2.7 Previous Scour Research by Others .......................................................................... 22 
Chapter 3 Testing System and Methods ..................................................................................... 24 
 3.1 EFA Testing Equipment   ......................................................................................... 24 
 v
 3.2 Ultrasonic Sensor   .................................................................................................... 27 
 3.3 Data Reduction.......................................................................................................... 36 
 3.4 Verification of Sensor Operation   ............................................................................ 39 
 3.5 Testing Procedure ..................................................................................................... 40 
  3.5.1 Sample Procurement .................................................................................. 40 
  3.5.2 Sample Preparation .................................................................................... 44 
  3.5.3 EFA Testing Procedure .............................................................................. 46 
 3.6 Testing Regimen ....................................................................................................... 52 
 3.7 Geotechnical Testing ................................................................................................ 54 
Chapter 4 Test Results ................................................................................................................ 55 
 4.1 General Sampling ..................................................................................................... 55 
 4.2 Bucatunna Clay ......................................................................................................... 56 
  4.2.1 Sampling .................................................................................................... 57 
  4.2.2 EFA Testing ............................................................................................... 58 
  4.2.3 Geotechnical Testing ................................................................................. 66 
 4.3 Yazoo Clay................................................................................................................ 67 
  4.3.1 Sampling .................................................................................................... 67 
  4.3.2 EFA Testing ............................................................................................... 68 
  4.3.3 Geotechnical Testing ................................................................................. 75 
 4.4 Demopolis Chalk ...................................................................................................... 76 
  4.4.1 Sampling .................................................................................................... 76 
  4.4.2 EFA Testing ............................................................................................... 78 
  4.4.3 Geotechnical Testing ................................................................................. 80 
 vi
 4.5 Mooreville Chalk ...................................................................................................... 82 
  4.5.1 Sampling .................................................................................................... 82 
  4.5.2 EFA Testing ............................................................................................... 83 
  4.5.3 Geotechnical Testing ................................................................................. 86 
 4.6 Prairie Bluff Chalk .................................................................................................... 87 
  4.6.1 Sampling .................................................................................................... 87 
  4.6.2 EFA Testing ............................................................................................... 88 
  4.6.3 Geotechnical Testing ................................................................................. 91 
 4.7 Porter?s Creek Clay ................................................................................................... 92 
  4.7.1 Sampling .................................................................................................... 92 
  4.7.2 EFA Testing ............................................................................................... 93 
  4.7.3 Geotechnical Testing ................................................................................. 99 
 4.8 Clayton Formation .................................................................................................. 100 
  4.8.1 Sampling .................................................................................................. 100 
 4.9 Nanafalia Clay Formation ....................................................................................... 101 
  4.9.1 Sampling .................................................................................................. 101 
 4.10 Naheola Clay Formation ....................................................................................... 102 
  4.10.1 Sampling ................................................................................................ 102 
Chapter 5 Results Discussion.................................................................................................... 104 
 5.1 Sampling Observations ........................................................................................... 104 
 5.2 EFA Testing Observations ...................................................................................... 106 
 5.3 Geotechnical Testing Observations ........................................................................ 108 
 5.4 Correlations in Testing ............................................................................................ 109 
 vii
Chapter 6 Conclusions and Recommendations ......................................................................... 114 
 6.1 Summary ................................................................................................................. 114 
 6.2 Conclusions ............................................................................................................. 115 
 6.3 Recommendations ................................................................................................... 116 
References   ............................................................................................................................... 119 
Appendix A Bucatunna Clay Results   ..................................................................................... 121 
Appendix B Yazoo Clay Results   ............................................................................................ 137 
Appendix C Demopolis Chalk Results   ................................................................................... 143 
Appendix D Mooreville Chalk Results   ................................................................................... 147 
Appendix E Prairie Bluff Chalk Results   ................................................................................. 153 
Appendix F Porter?s Creek Clay Results   ................................................................................ 158 
 
viii 
 
List of Tables 
 
 
Table 2-1: Summary of Crim?s Results ...................................................................................... 21 
Table 2-2: Summary of Navarro?s Results ................................................................................. 23 
Table 4-1: Bucatunna Clay Results at 0.6 m/s ............................................................................ 60 
Table 4-2: Bucatunna Clay Results at 1.0 m/s ............................................................................ 61 
Table 4-3: Bucatunna Clay Results at 1.5 m/s ............................................................................ 62 
Table 4-4: Bucatunna Clay Results at 2.0 m/s ............................................................................ 62 
Table 4-5: Bucatunna Clay Results at 3.0 m/s ............................................................................ 63 
Table 4-6: Yazoo Clay Results at 1.0 m/s ................................................................................... 72 
Table 4-7: Yazoo Clay Results at 1.5 m/s ................................................................................... 74 
Table 4-8: Yazoo Clay Results at 2.0 m/s ................................................................................... 74 
Table 4-9: Porter?s Creek Clay Results at 0.6 m/s ...................................................................... 94 
Table 4-10: Porter?s Creek Clay Results at 1.0 m/s .................................................................... 94 
Table 4-11: Porter?s Creek Clay Results at 1.5 m/s .................................................................... 95 
Table 4-12: Porter?s Creek Clay Results at 2.0 m/s .................................................................... 95 
Table 4-13: Porter?s Creek Clay Results at 3.0 m/s .................................................................... 96 
Table 6-1: Summary of Results and Observations ................................................................... 116 
 
ix 
 
List of Figures 
 
 
Figure 1-1. Example of Bridge Scour  .......................................................................................... 1 
Figure 2-1. Erosion Function Apparatus  .................................................................................... 12 
Figure 2-2. Schematic of EFA Test  ........................................................................................... 13 
Figure 2-3. Typical Erosion Function  ........................................................................................ 13 
Figure 2-4. Rotating Erosion Test Apparatus  ............................................................................ 15 
Figure 2-5. Sediment Erosion Rate Flume (SERF)  ................................................................... 16 
Figure 2-6. SERF Sample Chamber and Stepping Motor .......................................................... 17 
Figure 2-7. Multiple Transducer Array Designed for SERF ...................................................... 18 
Figure 3-1. Schematic Showing Important Components of the EFA ......................................... 25 
Figure 3-2. Digital Readout from EFA ....................................................................................... 26 
Figure 3-3. Auburn Ultrasonic Sensor Schematic ...................................................................... 28 
Figure 3-4. CrossTalk Startup Dialog Box  ................................................................................ 30 
Figure 3-5. CrossTalk Connection Settings  ............................................................................... 31 
Figure 3-6. CrossTalk Script Dialog Box  .................................................................................. 31 
Figure 3-7. CrossTalk Script Dialog Box with Typical Settings ................................................ 32 
Figure 3-8. Temperature Calibration for EFA Thermistor  ........................................................ 36 
Figure 3-9. Sensor Tributary Areas  ........................................................................................... 38 
Figure 3-10. Sensor Verification Test  ........................................................................................ 40 
Figure 3-11. CME Continuous Sampling System ...................................................................... 43 
 x
Figure 3-12. Automatically Advancing Hacksaw Cutting Sample Tube  .................................. 44 
Figure 3-13. Extruding Sample  .................................................................................................. 45 
Figure 3-14. Cutting and Cleaning Surface of Sample  .............................................................. 45 
Figure 3-15. Soil Plunger Advancing Sample  ........................................................................... 50 
Figure 3-16. Soil Sample in EFA Level with Base of Flume  .................................................... 50 
Figure 3-17. Soil Sample in EFA Advanced 1mm  .................................................................... 51 
Figure 3-18. Text Import Wizard, Delimited Settings for CrossTalk File ................................. 51 
Figure 3-19. Import Properties for Data Reduction Spreadsheet  ............................................... 52 
Figure 4-1. Locations of Sampled and Tested Formations  ........................................................ 55 
Figure 4-2. Bucatunna Clay Sample Smaller Radially than Sample Tube ................................. 57 
Figure 4-3. ?Bucatunna 27.0_1? Results .................................................................................... 59 
Figure 4-4. ?Bucatunna 23.0_4? After Erosion Occurred  ......................................................... 64 
Figure 4-5. EFA Tests Results, Bucatunna Clay ........................................................................ 65 
Figure 4-6. Bucatunna Clay Velocity versus Scour Rate  .......................................................... 65 
Figure 4-7. Bucatunna Clay Grain Size Distribution  ................................................................. 66 
Figure 4-8. ?Yazoo Clay 18.5_1? Test Failure Wedge  .............................................................. 69 
Figure 4-9. ?Yazoo Clay 18.5_2? Test Swelling and Cracking  ................................................. 70 
Figure 4-10. ?Yazoo Clay 18.5_2? Test After Failure ............................................................... 70 
Figure 4-11. ?Yazoo Clay 18.5_2? Test Results......................................................................... 71 
Figure 4-12. ?Yazoo Clay 18.5_9? Test Scoured Area from Front ............................................ 73 
Figure 4-13. ?Yazoo Clay 18.5_9? Test Scoured Area from Behind ......................................... 73 
Figure 4-14. Yazoo Clay Formation Grain Size Distribution ..................................................... 76 
Figure 4-15. Demopolis Clay formation Sampling Location ..................................................... 77 
 xi
Figure 4-16. ?DemopolisChalk19.0_5? After Testing ................................................................ 80 
Figure 4-17. Demopolis Chalk Formation Grain Size Distribution ............................................ 81 
Figure 4-18. ?MoorevilleChalk22.0_7? After Testing ............................................................... 85 
Figure 4-19. Mooreville Chalk Formation Grain Size Distribution ........................................... 87 
Figure 4-20. Prairie Bluff Chalk Formation Cracking Example 1 .............................................. 90 
Figure 4-21. Prairie Bluff Chalk Formation Cracking Example 2 .............................................. 90 
Figure 4-22. Prairie Bluff Chalk Formation Grain Size Distribution ......................................... 92 
Figure 4-23. Porter?s Creek Clay Formation Scoured along Weathered Planes ......................... 97 
Figure 4-24. EFA Test Results, Porter?s Creek Clay Formation  ............................................... 98 
Figure 4-25. Porter?s Creek Clay Formation Velocity versus Scour Rate  ................................. 98 
Figure 4-26. Porter?s Creek Clay Formation Grain Size Distribution ...................................... 100 
Figure 5-1. Scour Observation using Moisture Content and SPT N Value  ............................. 111 
Figure 5-2. Scour Observation using Percent Passing #200 and SPT N Value  ....................... 112 
Figure 5-3. Scour Observation using Mean Grain Size and SPT N Value  .............................. 113 
Figure A-1. ?Bucatunna 27.0_3? Test Results.......................................................................... 121 
Figure A-2. ?Bucatunna 27.0_4? Test Results.......................................................................... 122 
Figure A-3. ?Bucatunna 27.0_5? Test Results.......................................................................... 122 
Figure A-4. ?Bucatunna 27.0_6? Test Results.......................................................................... 123 
Figure A-5. ?Bucatunna 27.0_7? Test Results.......................................................................... 123 
Figure A-6. ?Bucatunna 27.5_2? Test Results.......................................................................... 124 
Figure A-7. ?Bucatunna 27.5_3? Test Results.......................................................................... 124 
Figure A-8. ?Bucatunna 27.5_4? Test Results.......................................................................... 125 
Figure A-9. ?Bucatunna 27.5_5? Test Results.......................................................................... 125 
 xii
Figure A-10. ?Bucatunna 27.5_6? Test Results........................................................................ 126 
Figure A-11. ?Bucatunna 27.5_7? Test Results........................................................................ 126 
Figure A-12. ?Bucatunna 27.5_8? Test Results........................................................................ 127 
Figure A-13. ?Bucatunna 27.5_9? Test Results........................................................................ 127 
Figure A-14. ?Bucatunna 26.5_2? Test Results........................................................................ 128 
Figure A-15. ?Bucatunna 26.5_3? Test Results........................................................................ 128 
Figure A-16. ?Bucatunna 26.5_4? Test Results........................................................................ 129 
Figure A-17. ?Bucatunna 26.5_5? Test Results........................................................................ 129 
Figure A-18. ?Bucatunna 26.5_6? Test Results........................................................................ 130 
Figure A-19. ?Bucatunna 26.5_7? Test Results........................................................................ 130 
Figure A-20. ?Bucatunna 26.5_8? Test Results........................................................................ 131 
Figure A-21. ?Bucatunna 26.5_9? Test Results........................................................................ 131 
Figure A-22. ?Bucatunna 26.5_10? Test Results...................................................................... 132 
Figure A-23. ?Bucatunna 26.5_11? Test Results...................................................................... 132 
Figure A-24. ?Bucatunna 26.5_12? Test Results...................................................................... 133 
Figure A-25. ?Bucatunna 23.0_1? Test Results........................................................................ 133 
Figure A-26. ?Bucatunna 23.0_2? Test Results........................................................................ 134 
Figure A-27. ?Bucatunna 23.0_3? Test Results........................................................................ 134 
Figure A-28. ?Bucatunna 23.0_4? Test Results........................................................................ 135 
Figure A-29. Bucatunna Clay Liquid Limits Results ............................................................... 135 
Figure B-1. ?Yazoo Clay 18.5_2? Test Results ........................................................................ 137 
Figure B-2. ?Yazoo Clay 18.5_3? Test Results ........................................................................ 138 
Figure B-3. ?Yazoo Clay 18.5_4? Test Results ........................................................................ 138 
 xiii 
Figure B-4. ?Yazoo Clay 18.5_5? Test Results ........................................................................ 139 
Figure B-5. ?Yazoo Clay 18.5_6? Test Results ........................................................................ 133 
Figure B-6. ?Yazoo Clay 18.5_7? Test Results ........................................................................ 140 
Figure B-7. ?Yazoo Clay 18.5_8? Test Results ........................................................................ 140 
Figure B-8. ?Yazoo Clay 18.5_9? Test Results ........................................................................ 141 
Figure B-9. ?Yazoo Clay 18.5_2? Test Results ........................................................................ 141 
Figure B-10. Yazoo Clay Liquid Limits Results ...................................................................... 142 
Figure C-1. ?Demopolis Chalk 19.0_1? Test Results ............................................................... 143 
Figure C-2. ?Demopolis Chalk 19.0_2? Test Results ............................................................... 144 
Figure C-3. ?Demopolis Chalk 19.0_3? Test Results ............................................................... 144 
Figure C-4. ?Demopolis Chalk 19.0_4? Test Results ............................................................... 145 
Figure C-5. ?Demopolis Chalk 19.0_5? Test Results ............................................................... 145 
Figure C-6. Demopolis Chalk Liquid Limits Results ............................................................... 146 
Figure D-1. ?Mooreville Chalk 22.0_3? Test Results .............................................................. 147 
Figure D-2. ?Mooreville Chalk 22.0_4? Test Results .............................................................. 148 
Figure D-3. ?Mooreville Chalk 22.0_5? Test Results .............................................................. 148 
Figure D-4. ?Mooreville Chalk 22.0_6? Test Results .............................................................. 149 
Figure D-5. ?Mooreville Chalk 22.0_7? Test Results .............................................................. 149 
Figure D-6. ?Mooreville Chalk 22.0_8? Test Results .............................................................. 150 
Figure D-7. ?Mooreville Chalk 22.0_9? Test Results .............................................................. 150 
Figure D-8. ?Mooreville Chalk 22.0_10? Test Results ............................................................ 151 
Figure D-9. ?Mooreville Chalk 22.0_11? Test Results ............................................................ 151 
Figure D-10. Mooreville Chalk Liquid Limits Results ............................................................. 152 
 xiv
Figure E-1. ?Prairie Bluff Chalk 18.5_2? Test Results ............................................................ 153 
Figure E-2. ?Prairie Bluff Chalk 18.5_3? Test Results ............................................................ 154 
Figure E-3. ?Prairie Bluff Chalk 18.5_4? Test Results ............................................................ 154 
Figure E-4. ?Prairie Bluff Chalk 18.5_5? Test Results ............................................................ 155 
Figure E-5. ?Prairie Bluff Chalk 18.5_6? Test Results ............................................................ 155 
Figure E-6. ?Prairie Bluff Chalk 18.5_7? Test Results ............................................................ 156 
Figure E-7. ?Prairie Bluff Chalk 18.5_8? Test Results ............................................................ 156 
Figure E-8. Prairie Bluff Chalk Liquid Limits Results ............................................................. 157 
Figure F-1. ?Porter?s Creek 19.5_1? Test Results .................................................................... 158 
Figure F-2. ?Porter?s Creek 19.5_2? Test Results .................................................................... 159 
Figure F-3. ?Porter?s Creek 19.5_3? Test Results .................................................................... 159 
Figure F-4. ?Porter?s Creek 19.5_4? Test Results .................................................................... 160 
Figure F-5. ?Porter?s Creek 19.5_5? Test Results .................................................................... 160 
Figure F-6. ?Porter?s Creek 19.5_7? Test Results .................................................................... 161 
Figure F-7. ?Porter?s Creek 19.0_1? Test Results .................................................................... 161 
Figure F-8. ?Porter?s Creek 19.0_2? Test Results .................................................................... 162 
Figure F-9. ?Porter?s Creek 19.0_3? Test Results .................................................................... 162 
Figure F-10. ?Porter?s Creek 19.0_5? Test Results .................................................................. 163 
Figure F-11. ?Porter?s Creek 19.0_6? Test Results .................................................................. 163 
Figure F-12. ?Porter?s Creek 19.0_7? Test Results .................................................................. 164 
Figure F-13. ?Porter?s Creek 19.0_8? Test Results .................................................................. 164 
Figure F-14. ?Porter?s Creek 19.0_9? Test Results .................................................................. 165 
Figure F-15. ?Porter?s Creek 19.0_12? Test Results ................................................................ 165 
 xv
Figure F-16. ?Porter?s Creek 19.5_14? Test Results ................................................................ 166 
Figure F-17. Porter?s Creek Clay Liquid Limits Results .......................................................... 166 
 
xvi 
 
List of Abbreviations 
 
 
FHWA  Federal Highway Administration    
HEC  Hydraulic Engineering Circular 
SRICOS  Scour Rate in Cohesive Soils 
ALDOT   Alabama Department of Transportation 
E-SCRICOS Extended Scour Rate in Cohesive Soils 
S-SRICOS  Simple Scour Rate in Cohesive Soils 
USGS  United States Geological Survey 
RETA  Rotating Erosion Test Apparatus 
SERF  Sediment Erosion Rate Flume 
LVDT   Linear Variable Differential Transformer 
CME   Central Mining Equipment 
SPT  Standard Penetration Test 
USCS  Unified Soil Classification System 
LL  Liquid Limit 
PL  Plastic Limit 
PI  Plasticity Index   
1 
 
Chapter 1 Introduction 
1.1 Background 
In 2009, according to the Federal Highway Administration (FHWA), there were 
approximately 603,000 bridges in the National Bridge Inventory.  Of these 603,000 bridges, 
roughly 83 percent span water (Lagasse et al. 2007).  With such a large number of bridges 
crossing water, scour can be a major concern with accelerated flow conditions such as 
flooding.  Between 1961 and 1976, over 50 percent of the 86 major bridge failures were due 
to scour (Murillo 1987).  More recently, from 1989 to 2000 just over fifteen percent of all 
bridge failures were due to scour (Wardhana and Hadipriono 2003).  From these numbers it 
is evident that scour is a serious issue in bridge design and maintenance.  When scour does 
occur, remediation measures are extremely costly, due to potential instabilities in the bridge 
and river bed, as shown in Figure 1-1.  Estimates of scour are an important step in the bridge 
design process, as the estimated depth of scour is a driving force in the foundation system 
selection and penetration depth. 
 
Figure 1-1. Example of Bridge Scour 
2 
 
Current scour predictions of highway bridges are made using techniques reported in 
Hydraulic Engineering Circular 18 (Richardson and Davis 2001), abbreviated as HEC 18, 
and Hydraulic Engineering Circular 20, or HEC 20 (Lagasse et al. 2001).  These reports, 
published by the FHWA, estimate scour depth based on four major variables: channel 
configuration, stream velocity, soil grain size, and underlying bed material.  It is important to 
note that the methods defined in HEC 18 and HEC 20 were based on predicting scour in 
cohesionless bed material.  Alternatively, it is believed that the variables leading to scour in 
HEC 18, predominantly grain size, do not translate into accurately predicting scour in 
cohesive soils.  Much work has been completed pertaining to predicting the rate and 
magnitude of scour of cohesive soils, most notably by Briaud et al. (1999, 2001a, 2001b, 
2004).  The Erosion Function Apparatus, EFA, was created by Briaud?s research group, with 
the purpose of determining the rate of scour of cohesive soils.   
The EFA uses a pump and a flume to create a constant flow, and corresponding shear 
stress, which is exposed to a one millimeter protrusion of soil.  Determining the erosion rates 
of this one millimeter protrusion at different velocities, or bed stresses, creates an erosion 
function.  This erosion function is then used in accordance with Briaud?s Scour Rate in 
Cohesive Soils (SRICOS) method to predict the maximum depth of scour over flooding 
events (Briaud 1999).  The erosion rates created from the EFA are determined using a 
viewing window in the flume of the EFA.  An observer determines when the volume of the 
one millimeter protrusion has eroded and records a corresponding time stamp. 
Accurate scour predictions are a major contributing factor to the economic foundation 
design of bridges crossing bodies of water.  The depth of scour is a portion of the total depth 
of foundation needed to provide capacity to carry the bridge loads.  If the depth of scour is 
3 
 
overpredicted, the foundation length and construction costs of the bridge are unnecessarily 
increased.  This principle is the driving idea for better predictions of scour depth in cohesive 
soils.  This concept is directly tied to the SRICOS method as it is a relatively accepted 
method for predicting scour and is approved by the Federal Highway Administration. 
1.2 Objectives 
 The objectives of the study were 
? Adapt the EFA to measure scour rates automatically, including the design of an 
ultrasonic sensor, data acquisition system, and data reduction procedure, 
? Conduct EFA tests on cohesive soil samples that were collected from locations below 
the fall line in the coastal plain of Alabama, and   
? Determine if measured scour parameters correlate with common geotechnical 
parameters such as shear strength, Atterberg Limits, grain size, or Standard 
Penetration Test N values.   
1.3 Scope of Study 
 The scope of work included the following tasks: 
? Update the Auburn University Erosion Function Apparatus with an ultrasonic sensor 
that can volumetrically measure the mass of eroded material at any point during 
testing.  It is believed that this would add validity to current erosion testing practices, 
? Create a testing regimen that incorporates the new updated EFA, 
? Obtain samples from cohesive soil formations in the coastal plain of Alabama with 
the assistance of the Alabama Department of Transportation (ALDOT), 
? Perform EFA tests to determine erosion parameters, and 
? Perform geotechnical index tests to determine geotechnical parameters. 
4 
 
Chapter 2 Literature Review 
2.1 Scour Background Information 
Scour is defined as the result of the erosive action of flowing water, which can 
excavate and carry away material from the bed and banks of streams and from around the 
piers and abutments of bridges (Richardson and Davis 2001).  More specifically scour at 
bridges can be related to the following factors (Lagasse et al. 2001): 
? Channel slope and alignment 
? Channel shifting 
? Bed sediment size distribution 
? Antecedent floods and surging phenomena 
? Accumulation of debris, logs, or ice 
? Flow contraction, flow alignment, and flow depth 
? Pier and abutment geometry and location 
? Type of foundation 
? Natural or man-induced modification of the stream 
? Failure of a nearby structure 
Scour is divided into three different classifications: aggradation and degradation, 
general scour, and local scour.  Aggradation and degradation are based on long-term 
streambed elevation changes (Richardson and Davis 2001).  Aggradation is defined by the 
deposition of upstream material, resulting in the raising of a streambed.  Degradation is 
defined by the lowering of a streambed due to a deficit in the deposits from upstream.   
General scour is attributed to the lowering of a streambed across a stream at a bridge 
or large-scale impingement in flow.  General scour often refers to contraction scour, which 
5 
 
occurs when a streambed is narrowed, or contracted, by the addition of a bridge that 
increases flow velocities (Richardson and Davis 2001).  Another type of general scour 
involves scour around a bend in a river, as velocities tend to vary with respect to distance 
from the bend.  Typically contraction scour occurs across most of the streambed.  However 
general scour and contraction scour are not uniform throughout a given cross section.  
General scour differs from aggradation and degradation in that general scour is cyclical and 
often reflects flood activity (Richardson and Davis 2001). 
Local scour is defined as the removal of material around objects intercepting flow 
caused by the acceleration of flow around the objects.  With acceleration in flow around a 
bridge pier comes an increase in shear stress on the stream bed resulting in removal of bed 
material.  HEC 18 recommends calculating total scour at a bridge crossing by adding 
degradation, general scour, and local scour over the design life of the structure. 
Scour occurs when the shear stress exerted on the bed material by the flow of water 
exceeds the critical shear stress (?c) for the bed material.  As the shear stress increases 
beyond the critical shear stress of the bed material, a scour hole develops.  Scour can develop 
around an object, as in local scour, or across a channel in general scour.  As the scour 
increases and more particles are removed, the shear stress on the plane of the bed material 
decreases.  Maximum scour depth is reached once enough material has been removed to 
reduce the shear stress at the bottom of the scour hole to a level below the critical shear stress 
(Briaud et al. 1999).  This critical shear stress is proportional to the critical velocity (Vc) (in 
m/s for SI units and ft/s for English units) flowing through a channel.  HEC-18 defines the 
critical shear velocity by equation 2-1 below: 
3
1
16
1
DyKV uc ??=                                             (2-1) 
6 
 
  where: 
Vc  = Critical velocity above which bed material size D and smaller scours 
y = Average depth of flow upstream of bridge 
D  = Particle size correlated to Vc 
Ku = Curve fitting factor (6.19 for SI units and 11.17 for English units) 
Equation 2-1 states that the critical shear stress is a function of two different 
properties, depth of flow and grain size.  Grain size is also a variable in calculating the depth 
of contraction scour.  In live-bed contraction scour, grain size is a variable used to determine 
the mode that bed material is being transported (Richardson and Davis, 2001).  Similarly, in 
clear-water contraction scour the average depth of scour is proportional to the largest 
nontransportable particle in the bed material. 
For calculating the scour depth around a pier, HEC-18 suggests using equation 2-2.  
This equation was created based on flume experiments in sand using different pier 
configurations and shapes.   
43.0
1
35.0
1
43210.2 Fra
yKKKK
a
ys ?
??
?
??
??????=                   (2-2) 
where: 
ys = Scour depth 
y1 = Flow depth directly upstream of pier 
K1 = Correction factor for pier nose shape 
K2 = Correction factor for angle of attack of flow 
K3 = Correction factor for bed condition 
K4 = Correction factor for armoring by bed material size 
a = Pier width 
7 
 
L = Length of pier 
 Fr1 = Froude Number directly upstream = 
1yg
Velocity
?
 
 g = Acceleration due to gravity 
In equation 2-2 grain size is a variable in the constant K4, a correction factor for 
armoring the pier by bed material.  This correction factor reduces the amount of pier scour if 
the pier is considered protected by heavier coarse-grained materials such as gravels.  
Throughout HEC-18, scour prediction equations are heavily influenced by grain size.  This 
concept is based on the assumption that soils modeled in these scour predictions erode 
particle by particle.  Erosion of coarse-grained material such as sand and gravel occurs 
typically by rolling, sliding, or plucking of the particle.  Close observations and slow motion 
cameras have determined that scour of these particles is typically a combination of these 
mechanisms (Briaud et al. 1999).  Erosion of sands and gravels is mainly resisted by gravity 
forces from the weight of the particles.  Using this logic, the larger the particle, the more 
shear stress it can resist.  For example, it is understood that the shear stress needed to erode a 
grain of sand is much less than the shear stress needed to erode a boulder.  However, HEC-18 
uses these principles to predict the depth of scour of fine-grained cohesive soils.  HEC-18 
does acknowledge that the rate of scour may be much lower for cohesive soils, but states that 
the maximum scour depth should still be calculated using the equations above.  In this 
regard, cohesive soils are considered to have a lower critical shear stress than sands using 
equation 2-1.  Also fine-grained cohesive soils do not provide any protection against pier 
scour using the correction factor K4 in equation 2-2.  In essence, the lack of understanding 
regarding scour of cohesive soils has resulted in a penalty using current design standards. 
 
8 
 
2.2 Scour Rate in Cohesive Soils Method 
Numerous studies have been performed to create a better understanding of scour of 
cohesive soils.  Most of these studies involved channel or flume tests that simulated a flood 
event, in which scour was closely monitored.  HEC-18 does acknowledge one study in 
particular as a method for predicting pier scour in cohesive soils.  This research, performed at 
Texas A&M University, resulted in the Scour Rate in Cohesive Soils (SRICOS) method, and 
started by determining the scour mechanisms in sands, clays, and rock (Briaud et al. 1999). 
Water velocity ranges from 0.1 to 3.0 m/s in most rivers and streams.  This velocity 
results in average bed shear stresses ranging from 1 to 50 N/m2.  Since the bed shear stresses 
resulting in scour are much less than the shear strength parameters typically found in clay, it 
is presumed that scour is a result of a cyclical failure, which occurs after long-term exposure 
to relatively low shear stresses (Briaud et al. 1999).  Observations of scour in sands showed a 
rapid, immediate failure compared to the cyclical failure observed in cohesive soils.  The 
different scouring methods of sand and clay suggest that the forces resisting scour are not 
similar.  As previously stated the main force that resists scour in sand is gravity.  
Gravitational force is relatively small, depending on particle size. The resultant scour rates 
found in sand and resisted by gravitational forces are represented in meters per hour.  
Research has shown van der Waals forces, which hold clay particles together, can better 
resist the constant cyclical loading found in streams, resulting in scour rates in the order of 
millimeters per hour (Briaud et al. 1999).  Similarly, when scour occurs in clay, it does not 
occur particle by particle but in larger groups of particles or chunks. 
The SRICOS method was created to predict the depth of pier scour with respect to 
time for a known flow velocity in a uniform cohesive soil.  The SRICOS method involves 
9 
 
testing a site specific soil sample in an EFA, or Erosion Function Apparatus, and recording 
scour rates for a range of velocities.  The velocities tested in the EFA should encompass the 
expected shear stresses around the pier under flooding conditions.  From EFA tests, an initial 
scour rate is established as the scour rate corresponding to the maximum shear stress 
expected during the flood event.  A maximum scour depth is calculated based on site specific 
geometry and flow rates.  Using the maximum scour depth, initial scour rate, and flood 
information, a flood specific scour depth is established. 
The SRICOS method allows for an informed scour prediction to be made in cohesive 
soil.  However, the SRICOS method makes certain assumptions that are not realistic in all 
streams.  For instance the SRICOS method produces a depth of scour for one flood event, in 
a uniform soil layer (Briaud et al. 1999).  This is not realistic as bridge piers will resist 
several floods with varying velocities over a design service period, and commonly are 
embedded into varying soil layers.  To improve the accuracy of the SRICOS method, Briaud 
et al. (2001a) expanded the SRICOS to encompass the full hydrograph throughout the design 
period.  Also the Briaud et al. (2001a) expanded the SRICOS method to predict the 
maximum depth of scour in different soil layers.  Since the combinations of varying soil 
layers and flood events can become very complex, a computer program was developed to 
calculate pier scour using the new Extended SRICOS (E-SRICOS) method.  A Simple 
SRICOS (S-SRICOS) method was also created to accommodate scour predictions similar to 
the E-SRICOS method without the use of the SRICOS computer program (Briaud et al. 
2001a) 
The procedure for the E-SRICOS method is similar to the original SRICOS method.  
The maximum scour depth is still calculated using the mean flow velocity, diameter of the 
10 
 
pier, and viscosity of the water.  Samples are collected and tested in the same manner as the 
SRICOS method, and a velocity versus scour rate curve is created.  The flow hydrograph is 
then created for the bridge, typically using a discharge hydrograph from the United States 
Geological Survey (USGS) near the bridge location (Briaud et al. 2001a).  The flow 
hydrograph is then transformed into a velocity hydrograph using profile of the stream at the 
bridge crossing.  Using the velocity hydrograph, the velocity versus scour rate curve created 
in the EFA, soil profiles, and general bridge properties, the SRICOS program calculates the 
depth of pier scour throughout the entire design life of the bridge (Briaud et al. 2001a).  
Through several case studies Briaud et al. (2001a) found the maximum scour depth 
calculated, using HEC 18 standards, was not reached throughout the design life hydrograph  
As previously stated the S-SRICOS method was created to produce similar results to 
the E-SRICOS method without the use of the SRICOS computer program.  The maximum 
scour depth is calculated, usually related to the 100-year flood conditions.  Site specific 
samples are collected and tested in the EFA to obtain erosion functions.  It is important that 
samples are obtained and tested from all soil layers within the maximum scour depth (Briaud 
et al. 2001a).  A single equivalent erosion function is then created by averaging the functions 
from all soils within the maximum scour depth.  Using the flow hydrograph and bridge 
information, the maximum shear stress is calculated.  The calculated maximum shear stress 
and combined erosion function are used to determine the scour rate corresponding to the 
maximum shear stress.  An equivalent time factor is needed to reduce the number of 
iterations performed by the SRICOS computer program.  The equivalent time is the time 
required for the maximum velocity in the hydrograph to create the same scour depth as the 
complete hydrograph (Briaud et al. 2001a).  The equivalent time factor is a function of the 
11 
 
design life of the bridge, the maximum velocity of the river, and the initial scour rate 
corresponding to the maximum shear stress.  The total pier scour depth is then calculated 
using the equivalent time, initial scour rate, and maximum scour depth. 
2.3 Erosion Function Apparatus 
The quality of the SRICOS prediction is based on the results from the EFA test that is 
performed on site specific samples.  The Erosion Function Apparatus is a closed-channel, 
flume-like machine equipped with a pump and a stepping motor (Briaud et al. 2001b).  The 
bottom of the flume has a circular opening for testing a Shelby tube sample with a diameter 
of 76.2 mm.  A watertight seal is created between the Shelby tube and the flume by an O-
ring.  The cross section of the rectangular flume is 101.6 mm x 50.8 mm.  The total length of 
the flume measures 1.22 m.  The pump is regulated by a valve on the front of the EFA, and 
generates velocities ranging from 0.1 to 6 meters per second (m/s).  The Shelby tube is held 
flush at the bottom of the flume during testing.  A piston attached to the stepping motor 
protrudes the sample from the Shelby tube into the flume in 1 mm increments (Briaud et al. 
2001b).  The EFA can be viewed in Figure 2-1 below.  The EFA is instrumented with a 
thermistor and a paddle flow meter to electronically monitor the temperature and flow rate 
using a computer. 
12 
 
 
Figure 2-1. Erosion Function Apparatus 
The first step in performing an EFA test proposed by Briaud is to place the sample in 
the EFA, fill the flume with water and wait an hour to mimic stream conditions.  Next the 
flume velocity is set to 0.3 m/s and the sample was advanced 1 mm into the flume.  The EFA 
is equipped with a viewing glass so that a technician can observe the erosion of the sample.  
After the sample is advanced 1 mm, as shown in Figure 2-2, the technician measures how 
much time it takes for the 1 mm sample protrusion to erode.  After 1 mm has eroded, or after 
one hour of testing, the sample is advanced to the 1 mm protrusion location and the velocity 
increased is to 0.6 m/s.  The erosion of the sample is again monitored, and the time for the 
current1 mm protrusion to erode is recorded.  The sample is sequentially advanced and the 
erosion timed for velocities of 1 m/s, 1.5 m/s, 2.0 m/s, 3/0 m/s, 4.5 m/s, and 6 m/s (Briaud et 
al. 2001b).  A schematic of a prepared EFA test is shown in Figure 2-2 (Mobley 2009). 
13 
 
 
Figure 2-2. Schematic of EFA Test 
The results from the EFA are used to create an erosion function, which is used in the 
scour calculations of the SRICOS method.  An example of an erosion function can be viewed 
in Figure 2-3 (Mobley 2009).  The erosion function is a derived from a velocity versus 
erosion rate relationship that is created from the EFA data.  An erosion function shows the 
relationship between erosion rate (z) and shear stress (t) for a given soil.  Velocity is related 
to shear stress by using the geometry of the flume, density of water, and friction factor 
obtained from a Moody Chart.  It is imperative the erosion function encompass all shear 
stresses expected in the stream bed during design flood conditions.   
 
Figure 2-3. Typical Erosion Function 
14 
 
Other than the relationship between shear stress and erosion rate, an erosion function 
provides two critical values that are used in scour predictions.  The critical shear stress (?c) is 
the bed shear stress at which scour first occurs (Mobley 2009).  The critical shear stress is 
found by observation while performing an EFA test.  During erosion testing flow is gradually 
increased until erosion begins.  This velocity is converted to a shear stress, recorded, and 
plotted on the erosion function.  The initial erodibility (Si) measures the rate of scour just 
after the critical shear stress is reached.  The initial erodibility is calculated by drawing a line 
tangent to the erosion function through the critical shear stress (Crim 2003).  Generally, 
initial erodibility is higher in sands compared to cohesive soils.   
Since the results of EFA tests are important to producing quality scour estimates 
using the SRICOS method, it is integral that scour rates are calculated with accuracy.  The 
determination of scour rates can be a subjective process as erosion is typically not uniform in 
cohesive soils.  Previous research suggested the technician performing an EFA test should 
estimate when one millimeter of scour has occurred by taking an average across the surface 
(Crim 2003).  However, averaging is subjective and not consistent with different technicians.  
Several attempts have been made to reduce the subjectivity in erosion testing.  The 
University of Florida created two different apparatuses to objectively measure scour of 
cohesive soil and soft rock. 
2.4 Alternatives to the Erosion Function Apparatus 
The Rotating Erosion Test Apparatus (RETA) was designed to determine the volume 
of erosion of stiff clay, sandstone, and limestone.  A schematic of the RETA is shown in 
Figure 2-4.  The RETA performs an erosion test on a 4 in. long cylindrical soil sample with 
either a 2.4 in. or 4 in. diameter (Sheppard et al. 2005).  A ? in. diameter hole is drilled 
15 
 
vertically through the sample and a support shaft that attaches to the RETA is inserted 
through the sample.  The opposite end of the shaft is attached to a torque cell and clutch that 
is fixed to a surface (Sheppard et al. 2005).  A larger diameter cylinder is placed over the 
sample, and water is added to fill the annulus between the sample and the outer cylinder.  
The outer cylinder is rotated using a motor and pulley system to create a shear stress on the 
surface of the sample.  After a known duration of time the test is stopped and the test sample 
is measured radially and any eroded material is oven dried to determine the mass of the 
scoured soil.  The shear stress is calculated by knowing the amount of torque applied to the 
sample along with the initial radius and length of the test specimen.  The erosion rate is 
calculated using the change in radius, duration of the test, mass density of sample, and 
original sample geometry (Sheppard et al. 2005). 
 
Figure 2-4. Rotating Erosion Test Apparatus (RETA) 
 
The RETA is an acceptable testing apparatus for determining an erosion rate and 
shear stress relationship.  The RETA is not ideal, however, because it does not test the 
16 
 
appropriate failure plane.  A soil specimen in a stream bed will be exposed to shear stress and 
eroded from the top down, where the RETA measures scour on a radial plane.  Also, the 
sample preparation is difficult for the RETA specimen, as it can be difficult to create a 
consistent curvature around the entire outer diameter of the sample.  If the radius of the 
sample is not consistent, the shear stress will not be equally distributed throughout the 
sample, causing irregular erosion. 
The Sediment Erosion Rate Flume (SERF) is the second apparatus created at the 
University of Florida for the purpose of measuring relationships between erosion rate and 
shear stress.  The SERF functions similar to the EFA.  Like the EFA, the SERF uses a flume 
with a rectangular cross section, a water pump, a reservoir, and a stepping motor.  The SERF 
does have a larger water capacity than the EFA, as it utilizes two pumps and has an 1100 gal 
reservoir.  The SERF is much more automated than the EFA, using multiple instruments and 
a Labview program.  Images of the SERF are shown in Figure 2-5 and Figure 2-6. 
 
Figure 2-5. Sediment Erosion Rate Flume (SERF) 
17 
 
 
Figure 2-6. SERF Sample Chamber and Stepping Motor 
The SERF utilizes a Multiple Transducer Array (MTA) created by SeaTek to scan 
across a Shelby tube sample (Sheppard et al. 2005).  This array consists of twelve ultrasonic 
transducers that are used to measure the distance from the top of the flume to the surface of 
the sample.  A schematic of the ultrasonic sensor used in the SERF is shown in Figure 2-
7(Sheppard et al 2005).  The ultrasonic transducers are spaced so that both a 7.3 cm and 9.5 
cm soil/rock sample can be scanned (Sheppard et al. 2005).  The data collected by the 
ultrasonic transducers are used in a Labview program to control the positioning of the sample 
being tested.  Unlike the EFA, the soil sample does not protrude from the bottom of the 
flume, but is kept level with the base of the flume.  Once the sample averages a total scour of 
0.5 mm, as detected by the ultrasonic sensor, the sample is automatically advanced by the 
stepping motor.  After the test is completed, the erosion rate is calculated by dividing the 
length of sample advanced by the stepping motor by the amount of time the test was 
18 
 
performed.  A video camera is used to verify the results from the SERF and to determine the 
method of erosion.   
 
Figure 2-7.  Multiple Transducer Array Designed for SERF 
Researchers at the Georgia Institute of Technology created and used another 
alternative to the Erosion Function Apparatus.  A rectangular, tilting, recirculating flume was 
altered and utilized with the same general function of the EFA and SERF machines (Navarro 
2004).  The flume used was considerably larger than the ones used in the EFA and SERF 
with a length of 6.1 m, width of 0.38 m, and depth of 0.38 m.  Two variable-speed pumps 
generated flow to the flume.  Adhered to the flume bed were gravel particles with a mean 
diameter of 3.3 mm (Navarro 2004).  These gravel particles were used to assure fully 
developed and fully rough turbulent flow (Navarro 2004).  The basin, which fed both pumps, 
had a holding capacity of approximately two cubic meters. 
19 
 
Shelby tube samples were collected using formations consisting of both cohesive and 
cohesionless materials.  The Shelby tube was placed below a circular opening in the bottom 
of the flume, and the measurement of erosion was determined using a linear variable 
differential transformer (LVDT) attached to a piston which advanced the sample into the 
flume (Navarro 2004).  An operator controlled the apparatus that advanced a piston during 
erosion testing, to ensure the sediment surface was level with the top of the gravel bed in the 
flume.  This procedure is similar to the procedure for operating the SERF.  These tests were 
run continuously at varying velocities throughout the entire length of a Shelby tube sample to 
mirror typical field stratification (Navarro 2004).   
Erosion rates are calculated by converting the vertical displacement measured by the 
LVDT into eroded mass, using the dry density of the tested sample.  The product of the 
vertical displacement and dry density of the sample is divided by the time interval over 
which the erosion occurs to determine the erosion rate at a given velocity or shear stress.       
2.5 Scour Relationships with Geotechnical Parameters 
Generally speaking the critical shear stress of fine-grained soils increases when the 
soil unit weight increases, plasticity index increases, unconfined compressive strength 
increases, void ratio decreases, swell decreases, percent fines increases, temperature 
decreases, and water temperature increases (Briaud et al. 1999).  Erosion rates are influenced 
by the hydraulic shear stress applied, the clay content, the soil and water temperature, the soil 
and water chemical composition, the soil water content, the plasticity index, the soil unit 
weight, the soil undrained shear strength, and the mean grain size (Briaud et al. 1999).  
Briaud et al. (1999) found general qualitative relationships relating critical shear stress and 
initial erodibility to plasticity index, undrained shear strength, and percent fines (percent 
20 
 
passing #200 sieve).  Briaud et al. (1999) also plotted a correlation between critical shear 
stress and initial erodibility showing that as critical shear strength increases, initial erodibility 
decreases.  Research hints that a correlation between scour parameters and geotechnical 
parameters is complex and involves a combination of many parameters.  Research at the 
University of Florida developed a correlation between cohesive strength and erosion rates of 
limestone cores (Sheppard et al. 2006).  This function relates erosion rate to applied 
hydraulic shear stress, unconfined compressive strength, and splitting tensile strength.  
Further research at the Georgia Institute of Technology suggests a correlation of critical shear 
stress with the percent fines and median particle diameter (Sturm et al. 2004).  This 
relationship suggests that the smaller the particle size of a given formation, the greater the 
influence of interparticle forces on the erodibility of the formation.  Although these 
relationships are relatively unproven, it does show that a correlation is possible between 
cohesive materials and erosion rates.   
2.6 Previous Scour Research at Auburn University 
Auburn University has owned an Erosion Function Apparatus since 2001.  Crim 
(2003) detailed work to develop a better understanding of the inner workings of the EFA, 
published a procedure, and created erosion functions as defined in the SRICOS method.  
With Shelby tube samples provided by the Alabama Department of Transportation 
(ALDOT), Crim constructed erosion functions of five different Alabama cohesive soils.  
Crim tested samples from Goose Creek in Wilcox County, culverts on US 84 in Covington 
county, Linden Bypass in Marengo County, County Road 5 over Cheaha Creek in Talladega 
County, and Alabama State Road 123 over Choctawatchee River in Dale County.  A 
summary of the results obtained by Crim (2003) are shown in Table 2-1.  
21 
 
Table 2-1:  Summary of Crim?s Results
  
Overall, Crim witnessed scour rates ranging from 0 mm/hour upwards of 100 
mm/hour.  Crim plotted the results from the erosion functions and geotechnical tests 
comparing critical shear stress and plasticity index between samples with varying Standard 
Penetration Test N Values.  Crim also plotted results compatible with Briaud?s relationship 
between initial erodibility and critical shear stress. 
Mobley (2009) continued research with the EFA at Auburn.  This work also showed 
erosion testing is highly variable and that few similarities exist in testing different 
formations.  Mobley tested soils from Talladega County, Sumter County, and Dallas County.  
Mobley witnessed a high variability in erosion rates from the Talladega County soils, 
encompassing three orders of magnitude.  Generally, the critical shear stress of the Talladega 
County soils ranged from 0.61 N/m2 to 4.5 N/m2.  Mobley tested rock core samples from 
Sumter County.  Mobley was unable to generate an erosion function for the Sumter County 
samples due to difficulty in sample preparation.  One test from Sumter County showed scour 
resistance with a flume velocity of 0.75 m/s.  The Dallas County samples consisted of soil 
from the Mooreville Chalk formation.  Mobley performed three EFA tests on the Dallas 
County samples at the highest velocity generated by the EFA, approximately 6 m/s.  None of 
these tests exhibited scour over a duration of two hours.  Mobley stated the critical shear 
stress of the Mooreville Chalk was in excess of 45 N/m2.  Mobley also declared scour 
Stream Crossing County Critical Shear Stresses (N/m2) Initial Erodibility (mm/hr/N/m2)
Goose Creek Wilcox 0.4 to 4.5 0.2 to 5.6
US 84 Covington 1.1 to 3.1 1.7 to 6.1
Linden Bypass Marengo 0.9 to 4.7 0.4 to 11.4
County Road 5 Talladega 0.6 to 2.8 0.4 to 410
State Road 123 Dale 1.2 to 2.5 1.0 to 1.2
22 
 
performance was highly variable between formations and was somewhat depended on 
sample preparation. 
Due to limited soil samples, Mobley tested model soils with varying densities and 
clay content.  Mobley determined that model soils with higher density and higher clay 
content resulted in higher critical stresses and lower erosion rates.  
2.7 Previous Scour Research by Others 
 Sheppard et al. (2006) determined the critical shear stresses, using the SERF, for 
uniform sands ranging from 0.08 N/m2 to 0.75 N/m2.  The critical shear stresses measured 
were increasing with mean grain size.  Erosion rates for these tests exceeded 300 mm/s.   
 Sheppard et al. (2006) also created erosion rates for natural limestone core samples.  
Eight of these cores were tested in the RETA with once core tested in the SERF.  The SERF, 
limited by shorter test duration, observed no scour during testing.  Erosion functions for the 
limestone cores were created using the RETA.  These tests yielded a minimum critical shear 
stress of 20 N/m2 ranging upwards of 35 N/m2.  The erosion rates calculated on the limestone 
were much lower than those observed in the uniform sands with a maximum erosion rate of 
0.6 centimeters per year at a shear stress of 70 N/m2.   
 In addition to the tests described above Sheppard et al. (2006) performed erosion tests 
on cemented sands in an effort to mimic the scour behavior of sands with cohesive materials.  
The results of these tests varied due to the difficulty of sample preparation.  The critical shear 
stress of these samples ranged from 5 N/m2 to 10 N/m2. An estimate of erosion rate was 
created for these samples using the erosion function, splitting tensile strength, and 
unconfined compression strength. 
23 
 
 Navarro (2004) performed erosion tests in a large scale flume, as described above, to 
produce critical shear stresses of soil samples obtained in Georgia.  The results of these tests 
are summarized in Table 2-2 below.  The critical shear stresses observed by Navarro (2004) 
are compatible with critical shear values ranging from 0.5 N/m2 to 5.0 N/m2 reported by 
Briaud et al. (1999). 
Table 2-2:  Summary of Navarro?s Results 
 
 
County Critical Shear Stresses (N/m2) Classification
Murray 3 to 21 Silty Sand
Towns 6.8 to 21 Sandy Silt
Habersham 2.5 to 21 Sandy Silt
Haralson 3 to 12 Silty Sand
Bibb 3.3 to 9.7 Sand to Lean Clay
Wilkinson 0.4 to 21 Sand to Fat Clay
Effingham 3.24 to 21 Sand to Clayey Sand
Decatur 2.5 to 7.9 Clayey Sand
Berrien > 21 Clayey Sand and Fat Clay
McIntosh 17.17 Clayey Sand
24 
 
Chapter 3 Testing System and Methods 
3.1 EFA Testing Equipment 
As stated earlier, the Erosion Function Apparatus is the device Briaud (1999) created 
at Texas A&M University in order to collect the scour rates used to construct erosion 
functions.  These erosion functions were used to estimate the depth of scour over the design 
life of a bridge.  Figure 2-1 shows an image of an EFA.  Figure 3-1 shows a schematic of an 
EFA featuring all of the integral components used for erosion testing (Crim 2003).  The EFA 
contains a reservoir, a pump, and a flume.  Water is pumped from the reservoir, through the 
flume, and across the sample during testing before being recirculated back into the reservoir.  
The valve on the front of the EFA is used to regulate the flow velocity during testing.  A flow 
straightener is used to develop a full flow condition before the water flows over the soil 
sample.  
 
25 
 
 
Figure 3-1. Schematic Showing Important Components of the EFA 
The EFA is instrumented with a flow meter and temperature sensor in order to 
monitor flow velocity and water temperature during testing.  The sample tube is positioned in 
the flume using the crank wheel attached to the piston platform.  The piston platform is then 
locked into place so that the sample tube positioning does not change during testing.  A 
motor is attached to the piston and the sample can be brought to the proper elevation in the 
flume by using a control switch.  During testing, scour is observed through the observation 
window.  Other controls shown in Figure 3-1 include a power switch to the EFA, a power 
switch to the pump, and a power switch to a sump pump, which is located in the reservoir. 
26 
 
EFA tests are instrumented to monitor the temperature and flow rate of the water used 
during a test.  It is important for the flow velocity to be monitored and kept consistent during 
an EFA test as it directly reflects the shear stress imposed on a sample.  Temperature is 
monitored to ensure the conditions in the EFA during testing are consistent with those in a 
stream or river.  The EFA software consists of a Labview readout in which the temperature 
and flow velocity can be monitored and recorded.  Figure 3-2 shows an image of the digital 
readout provided by the EFA software. 
 
Figure 3-2.  Digital Readout from EFA 
 
The top window in Figure 3-2 shows the readout of the current flow velocity and 
current temperature.  The readout reports the flow velocity by dividing the flow rate 
calculated from the flow meter by the cross sectional area of the flume.  The sample can also 
be advanced in 0.5 mm increments by clicking the ?Push Up? button.  A test can be started 
by setting the logging rate and clicking ?start the test?.  During a test the attached data 
acquisition system records the temperature, flow velocity, and soil push location at the 
specified logging interval.  This data logging method was used to calculate erosion rates in 
previous research at Auburn University by Crim (2003) and Mobley (2009).  The screen in 
27 
 
the background of Figure 3-2 shows a graph that logs the velocity and temperature with 
respect to time during an EFA test.   
3.2 Ultrasonic Sensor 
Similar to the work performed by Sheppard et al. (2005) at the University of Florida, 
an ultrasonic sensor was designed and installed on Briaud?s Erosion Function Apparatus at 
Auburn University.    The SERF created by Sheppard and Bloomquist was designed to not 
only assist in calculating erosion rates, but to obtain a pattern of erosion.  The ultrasonic 
sensor on the SERF consists of twelve transducers.  Eight of the transducers are used in 
measuring scour of a 7.3 cmsample, while all twelve transducers are used to measure scour 
of a 9.5 cm sample.  The design criteria for the EFA ultrasonic sensor was to create a similar 
array of transducers that will measure scour in cohesive soils, instead of rock, and consist of 
a tight grouping of transducers creating a clearly mapped soil surface.  This ultrasonic sensor 
was designed and created with the help of Seatek and consists of sixteen transducers.  The 
sensor is connected to a stainless steel housing, and mounted to the top of the flume of the 
EFA above the location of the test sample.  A schematic of the ultrasonic sensor designed by 
Seatek is shown below in image 3-3. 
 
 
28
 
 
Figure 3-3.  Auburn Ultrasonic Sensor Schematic
 
29 
 
The transducers function at 5 MHz, with a physical diameter of 0.5 cm and an 
acoustic footprint of 0.8 cm at a distance of 5 cm (Jette 2010).  The ultrasonic sensor was set 
into an aluminum cover, which replaced the original acrylic cover of the EFA.  The ultra-
sonic sensor was sealed to the aluminum cover by using two O-rings.  Careful considerations 
were made to ensure that the ultrasonic sensor was flush with the bottom on the EFA 
aluminum cover.  The ultrasonic sensor could not protrude from the EFA cap, as this would 
result in irregular flow conditions.  The ultrasonic sensor also could not be raised above the 
EFA cap, as the transducers must be submerged to function properly.  The ultrasonic sensor 
is able to collect data from 16 different points across the area of a Shelby tube soil sample 
(71.12 mm).  Also 12 data points can be obtained from standard rock coring soil samples, 
which is convenient if a soil formation is too stiff to be sampled by pushing Shelby Tubes.   
The data acquisition system used with the ultrasonic sensor is capable of collecting 
data from all 16 transducers, along with the ability to sample up to 4 external analog 
channels.  If any external analog channels are used, the output voltage must range between 0 
V and 4 V.  The electronics package is able to communicate via an RS232 (serial) 
connection.  The software used to communicate with the data acquisition system was 
CrossTalk.  CrossTalk is used to set the datalogging parameters for the system and produces 
an output ASCII text file from the data received that can easily be imported into a 
spreadsheet.  The CrossTalk software and data acquisition system are set up on an updated 
secondary computer that is separate from the computer that acquires the temperature and 
flow velocity from the EFA. 
When opening the CrossTalk program, a new session should be started, as seen in 
Figure 3-4.  Next, the Connection Settings window should be open to ensure that the proper 
 
30 
 
parameters are input so CrossTalk can record the signal from the data acquisition system.  
The connection should be set to ?Local COM Port (Direct Connect)?.  The correct COM port 
should be chosen along with a Baud rate of 19200.  Also, the Flow Control should be set to 
?XONXOFF?.  The Connection Settings window should match the settings displayed in 
Figure 3-5.  When setting up the CrossTalk software for an EFA test, the correct script 
should be chosen.  This script is entitled ?SEATEK.AU SCRIPT.XWC? and can be chosen 
as shown in Figure 3-6.  The CrossTalk Script Dialog box will appear after the Seatek script 
has been selected, and ?Begin data collection? should be selected.  The Script Dialog box 
completed with typical test setup settings can be viewed in Figure 3-7. 
 
Figure 3-4.  CrossTalk Startup Dialog Box 
 
31 
 
 
Figure 3-5.  CrossTalk Connection Settings 
 
 
Figure 3-6.  CrossTalk Script Dialog Box 
 
 
32 
 
 
Figure 3-7.  CrossTalk Script Dialog Box with Typical Settings 
The script dialog box sets the parameters on each individual scan along with setting 
the duration of the entire test.  For the ease of data reduction, transducer 1 was set to the 
starting transducer and transducer 16 was the ending transducer.  The transducers were 
recorded in an increment of 1.  The scan rate was typically set to 1500 corresponding to a 
scan every 15 seconds.  It is important that acceptable distance parameters were established 
in order to qualify the scan readings from the ultrasonic sensor.  The minimum or blanking 
distance for the ultrasonic sensor was 3 cm.  The maximum distance was typically 
determined by a value that is not expected to be exceeded during an individual run.  For most 
tests, the minimum distance was 3 cm and the maximum distance was 10 cm.  As stated 
above these limits qualify the readings reported by the data acquisition system.  If an errant 
reading was recorded that was outside of the acceptable limits, the reading was recorded as 0, 
disregarded, and not calculated in the data reduction.  
 
33 
 
The data acquisition system was able to sample up to 20 individual range readings, or 
pings, for every recorded reading.  This capability greatly reduced the variability between 
recorded values by eliminating errant readings.  If 20 pings were used and returned 
acceptable ranges, the three shortest and the three longest ranges were excluded from the data 
reading.  The remaining 14 readings were then averaged, and the average was reported for 
each transducer.  The ultrasonic sensor records a range measurement when the processed 
acoustic return exceeds a threshold voltage (Jette 2010).  This threshold voltage was set 
through trial and error by adjusting the threshold voltage until distance fluctuations were a 
minimum.  Since the ultrasonic sensor was used to measure scour in cohesive soils and 
cohesive soils are often smooth in nature in comparison to granular soils, the threshold 
voltage was typically set to a relatively low value of 500 mV. 
The ultrasonic sensor designed by Seatek for Auburn University was created with the 
idea of improving scour measurements taken using Briaud?s EFA.  However, the 16 
transducer ultrasonic sensor did have a few constraints that were addressed before a 
cohesive-soil testing regimen was adopted.  As previously stated the bottom of the ultrasonic 
sensor, where the transducers contact the flow of water, could neither be elevated above the 
water or protruding into the flow of water.  The ultrasonic sensor was originally designed to 
be mounted on top of the current acrylic cover of the EFA.  This preliminary design left 
approximately a 13 mm hole above the flow of water allowing large air pockets to build up 
around the transducers.  It was determined that the acoustic pulsing of the ultrasonic sensor 
was disturbed by these air pockets resulting in many blank readings.  As a result a 
counterbore was machined into the acrylic cover of the EFA to recess the ultrasonic sensor 
flush with the bottom of the EFA cover.  This design worked in theory, but the acrylic cover 
 
34 
 
did not allow for a waterproof seal to be formed around the ultrasonic sensor.  In an attempt 
to tighten the seal between the EFA cover and the ultrasonic sensor the acrylic in the 
ultrasonic sensor was cracked allowing more water to leak through the seal.  A third EFA 
cover was designed by counterboring the ultrasonic sensor into an aluminum stock plate.  
The aluminum plate was identical to the original acrylic cover, and strong enough to form a 
waterproof seal with the ultrasonic sensor.   
Along with assuring the ultrasonic sensor was mounted flush in the EFA cover 
creating a waterproof seal, additional care was taken to ensure the transducers were not 
damaged.  It was important the cables were not crimped, pulled, or forced into a small radius 
bend at any point during testing.  When the cables are coiled, the signal-to-noise ratio may 
decrease resulting in blank or disturbed readings (Jette 2010).  The transducers should also be 
supported at all times, and the cable alone should never support the transducers.  To ensure 
that the cables were always supported, a wooden frame was built onto some of the flume 
supports of the EFA and the cables were draped over the frame. 
The installed sensor measures distance readings by using ultra-sonic transducers; 
therefore, temperature effects must be considered.  The water temperature is an input in the 
CrossTalk ?Script Dialog? box as seen in Figure 3-7.  However, the data acquisition system 
does not allow for any changes in water temperature throughout the duration of a test.  This 
presents an issue as the data acquisition system assumes a certain water temperature in 
calculating the wave speed dictates the scanned distance.  A six-hour continuous test showed 
that the water temperature in the EFA rose from 26 ?C to 54 ?C.  This temperature change is 
related to the EFA?s pump energy and flume friction.  The EFA contains a water tank that is 
approximately 1.36 cubic meters.  To reduce the magnitude of temperature change, water 
 
35 
 
was continuously circulated through the EFA tank.  Water was circulated into the EFA tank 
using a water hose from a spigot, while the water was pumped out by using the EFA?s drain 
and sump pump.  During a three hour test with water continuously circulated through the 
EFA the water temperature ranged from 19.2 ?C to 21.3 ?C.  It is important to keep the water 
temperature in a reasonable range to mimic stream conditions as closely as possible.  Due to 
the effects of temperature on wave speed, a temperature correction was necessary to 
overcome the variance in temperature.  During a typical test the temperature was set at 20 ?C, 
and a temperature correction was performed during data reduction.   
As previously stated, the EFA contained a separate data acquisition system and 
software package intended to detect and record flow rate and water temperature.  It was 
determined the thermistor installed on the EFA fit the voltage parameters of the analog 
channels on the ultrasonic sensor data acquisition package.  The EFA thermistor was then 
installed and calibrated as an analog input for the ultra-sonic sensor.  During the setup for a 
typical test, it was necessary to include one analog channel on each scan in the CrossTalk 
?Script Dialog? box shown in Figure 3-7.   
The thermistor was calibrated using a voltmeter and the digital readout from the EFA.  
This calibration was then plotted and equation relating temperature to voltage was derived.  
The temperature calibration can be viewed in equation 3-1, and an image of the temperature 
calibration can be viewed in Figure 3-8.  The Auburn EFA flow meter was calibrated by 
Mobley (2008).  This calibration was performed manually by measuring the amount of water 
coming out of the spout of the EFA and comparing it to the velocity shown in the EFA 
software.  Mobley concluded there is up to 10 percent error in velocity readings less than 1.0 
 
36 
 
meter per second.  At higher velocities there is not any appreciable error between the true 
flume velocity and the measured velocity by the EFA?s flow meter.   
83.1331232.0 ??=? mVF                  (3-1) 
 
Figure 3-8.  Temperature Calibration for EFA Thermistor 
3.3 Data Reduction 
For each test and scan performed, data reduction was necessary to determine a change 
in mass.  Each scan from the Seatek sensor contained a time stamp, a distance measurement 
from all 16 transducers, and a voltage reading from the thermistor.  Each time stamp was 
converted into an elapsed time, so that erosion rates may be calculated.  If a distance 
measurement returns a blank value (reading of 0.00), an average distance of neighboring 
transducers was taken and recorded over the blank value.  The voltage recorded from the 
thermistor was also converted to a temperature, using equation 3-1, so that temperature 
changes could be monitored throughout testing. 
Since the Seatek sensor utilized ultrasonic wave propagation, any changes in water 
temperature affected distance measurements.  Therefore, a temperature correction was 
y = 0.1232x -133.83
R? = 0.9991
0
20
40
60
80
100
120
140
0 500 1000 1500 2000 2500
Te
mp
era
tur
e (
????F)
Voltage (mV)
 
37 
 
required with every scan.  The speed of sound was calculated by using Equation 3-2 below 
(Seatek 2010), where speed of sound is represented in meters per second, and temperature 
(T) is given in degrees Celsius. 
g1845g1841g1845 =0.0029g1846g2871 ?0.055g1846g2870 +4.95g1846+1402.3                                       (3-2) 
An initial speed of sound (SOS1) was calculated, in meters, based on the input 
temperature given in the data acquisition setup as shown in Figure 3-7.  The final speed of 
sound (SOS2) was calculated based on the temperature at the time of each scan.  The final 
speed of sound represents the true water temperature in the EFA at the time a reading is 
taken.  Using the initial speed of sound, the elapsed time, in seconds, of each scan was 
calculated by using Equation 3-3. 
  g1831g1864g1853g1868g1871g1857g1856uni0009g1846g1861g1865g1857 =[ g3020g3016g3020g2869?g2869g2868g2868g3010g3041g3036g3047g3036g3028g3039uni0009g3014g3032g3028g3046g3048g3045g3032g3040g3032g3041g3047]g2879g2869                       (3-3) 
The elapsed time represents the time for the ultrasonic wave to travel from the sensor 
to the surface of the sample and back to the sensor.  This time is used by the data acquisition 
system to calculate a measured distance according to the input parameters in Figure 3-7.  The 
elapsed scan time was used to calculate the corrected distance measurement, as shown in 
Equation 3-4. 
  g1829g1867g1870g1870g1857g1855g1872g1857g1856uni0009g1830g1861g1871g1872g1853g1866g1855g1857 =g1845g1841g18452?g1831g1864g1853g1868g1871g1857g1856uni0009g1846g1861g1865g1857?100                  (3-4) 
The transducers reported an average distance of 20 pings per scan, and scans were 
performed every 15 seconds to reduce the level of variability between readings.  Four 
consecutive corrected distance measurements were averaged to represent the measured 
distance per elapsed minute. 
The final step of data reduction process was to convert the measured distances at any 
given time, to a volume of erosion.  Each transducer was assigned a tributary area based in 
 
38 
 
scan area such that a weighted average could be taken.  Since the ultrasonic sensor 
commonly measured two different sample sizes, two tributary areas were assigned to each 
transducer.  Figure 3-9 maps the tributary areas for both the 5.72 cm and 6.35 cm samples.  
Tributary areas were only calculated for transducers 1 through 12, as transducers 13 through 
16 were outside the sample area. 
 
Figure 3-9. Sensor Tributary Areas 
During any given scan, an initial distance measurement, or datum, was established 
over the first minute of scans, for each transducer.  The corrected average distance at any 
given time was subtracted from the initial datum, and multiplied by the tributary area.  The 
sums of these products were summed over transducers 1 through 12.  The volume of erosion 
at any time interval, in cubic centimeters, can be calculated by using Equation 3-5. 
( )?
=
??
12
1n
tfi ADD                            (3-5) 
 
39 
 
The volume of erosion was converted to cubic millimeters, and an average height of 
erosion was calculated, by dividing the volume of erosion by the area of the sample.  The 
average height of erosion represents a height change assuming all calculated scour was 
uniform.  For EFA testing, once the overall height change reflected a decrease of one 
millimeter, the sample was advanced.  This average height of erosion is precisely calculated 
adding an advantage of traditional visual observations presented by Briaud.  The elapsed time 
for one millimeter of erosion can then be converted to an erosion rate. 
3.4 Verification of Sensor Operation 
Verification of the ultra-sonic sensor was performed through several trial scans, using 
a non-erodible sample.  The non-erodible sample was created by applying a sand surface to 
an aluminum rod with a diameter of 6.35 cm.  The goal of each trial scan was to successfully 
show sample movement by advancing the sample at a known time and distance.  The trial 
scans mirrored the results and procedure of a typical EFA test.  Since the EFA advances the 
sample from the bottom, the non-erodible sample trial scans displayed movement as the 
target became closer to the sensor.   In a typical EFA test, the sample will be advanced and 
then progressively move farther away, as scour occurs, from the sensor located at the top of 
the flume.  For the sake of verifying the effectiveness of the ultrasonic sensor it was 
determined that as long as movement can be accurately measured, the relative direction of 
movement did not matter.   
Figure 3-10 below shows the results of a trial scan.  The non-erodible aluminum 
sample was placed in the EFA and scanned for approximately 150 minutes.  During this scan 
two separate one millimeter pushes were induced, at 42 and 128 minutes respectively.  Figure 
3-10 proves that the ultrasonic sensor, data acquisition package, and data reduction methods 
 
40 
 
produced the expected results, as one millimeter jumps are observed at 42 and 128 minutes.  
It is important to notice that variation in these readings were small, with a maximum of 0.08 
mm from the target distance.  These small variations could be attributed to variations in 
errant scans or temperature corrections. 
 
Figure 3-10.  Sensor Verification Test 
3.5 Testing Procedure 
3.5.1 Sample Procurement 
Drilling operations were coordinated with the Alabama Department of Transportation 
(ALDOT) to acquire cohesive soil samples for testing.  ALDOT stated that these hard 
cohesive soils and chalks would not be conducive to sampling via Shelby Tube.  Since the 
materials to be sampled were so stiff, a typical all terrain vehicle or truck drill rig would not 
be able to advance a Shelby tube sample without damaging the tube.  Two sampling 
0.0
0.5
1.0
1.5
2.0
2.5
0 20 40 60 80 100 120 140 160
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
 
41 
 
alternatives were derived to ensure that samples could both be acquired with relative ease and 
tested using the ultrasonic sensor at Auburn.  Option one involved using either 44.5 mm or 
47.6 mm rock core samples, while option two involved using 57.2 mm continuous samples, a 
new sampling method acquired by ALDOT.  Preliminary tests were performed with the EFA 
using rock core samples, and it was determined that the rock core samples were not ideal and 
should only be used as an alternative.  This was decided as it was necessary for rock core 
samples to be completely vertical and plumb or else the samples would not fit into the 
previously created EFA testing tube.  Also, it was determined that short segments could not 
be used with rock core samples as the flume created a ?suction? like force pulling the loose 
sample towards the ultrasonic sensor.  This sample movement upward could not be tolerated 
as any movement other than scour and planned sample advancements recorded would 
conflict with the results presented by the ultrasonic sensor. 
Therefore, it was decided to sample with the new continuous sampling technique that 
ALDOT recently acquired.  This sampler easily fits onto an ATV- or truck-mounted drill rig 
and can be used to acquire undisturbed soil samples in difficult to sample soil formations.  
The sampling technique involves first drilling down to the sample level with a hollow-stem 
auger.  The 1.52 m sampler is attached to either AW or NW threaded drill rod and lowered to 
the bottom of the hollow stem auger.  Once connected the sampler is pneumatically advanced 
as the hollow stem auger cuts around the obtained soil sample.  A sealed bearing assembly 
conveys the thrust from the drill rig to the tube and sampler shoe while isolating the sample 
tube from rotation of the auger (CME 2012).  An image of the Central Mining Equipment 
(CME) continuous sampling setup used is shown in Figure 3-11.  A sampling shoe is used to 
contain the sample in the tubes for a maximum recovery.  The sampler itself is a split spoon 
 
42 
 
sampler that includes two acrylic 57.2 mm diameter tubes that are each 762 mm long.  Once 
samples are gathered lids are placed on top and bottom of each sample allowing tubes to be 
tested and stored separately. 
During a typical sampling trip ALDOT geologists located possible drilling sites 
containing the target formations.  ALDOT geologists also determined the depth where the 
target formation was reached.  Once the target formation was confirmed from drill cuttings 
and split spoon samples, a Standard Penetration Test (SPT) was performed to obtain the SPT 
N value for each formation.  Typically two to three standard penetration tests were performed 
in each formation depending on the approximate depth of the formation provided by the 
ALDOT geologist.  Once the SPT was completed, the AW rod was removed from the boring 
and a hollow stem auger was used to the depth of the previously drilled hole.  Typically a 
Shelby Tube was pushed inside the hollow stem auger to ensure that any loose cuttings 
would not be sampled.  Sampling was then performed as described above with the CME 
continuous sampling system.  Typically 2-3 continuous samples were gathered in each 
formation resulting in 4-6 sample tubes.  It was believed that obtaining 1.5-2.3 m of each 
formation would be sufficient for both EFA testing and geotechnical testing.  After 
collection, sample tubes were capped and taped to maintain field moisture conditions.  Since 
EFA and geotechnical testing could take place months after sampling, samples were stored in 
a curing room to preserve field moisture conditions.  Sample tubes were also marked 
according to depth and formation prior to storage. 
 
43 
 
 
Figure 3-11. CME Continuous Sampling System 
 
 
 
 
 
44 
 
3.5.2 Sample Preparation 
A testable EFA sample was not cracked radially or longitudinally with a length of at 
least 101 mm.  It was important that testable sections were not cracked as cracks would have 
induced shear planes that would alter the results of testing.  A typical sample tube had an 
outside diameter of 63.5 mm, an inside or sample diameter of 57.2 mm, and a length of 762 
mm.  The pushing arm of the EFA could accommodate a sample tube length up to 457 mm.  
An automatically advancing hacksaw was used to cut the sample tubes to accommodate the 
dimensions of the EFA.  A testable section for the EFA was located and marked on the 
sample tube.  The section was then loaded into the saw and clamped to ensure that a square 
cut was made.  It was important the clamp was not so tight to distort the acrylic tube.  Figure 
3-12 below shows a sample section being cut with the automatically advancing hacksaw.  
 
Figure 3-12.  Automatically Advancing Hacksaw Cutting Sample Tube 
 
45 
 
Once the sample section was cut, the sample was extruded approximately 5 mm in 
order to prepare the surface for testing.  Using a spatula the surface of the sample was 
cleaned of any loose debris.  The sample was also cut level with the sample tube, and the 
surface was smoothed with the spatula. At times a wire saw was used along with a spatula to 
prepare the surface for testing.  Figure 3-13 shows the sample being extruded, and Figure 3-
14 shows the sample surface being prepared for testing.   
 
Figure 3-13.  Extruding Sample 
 
 
Figure 3-14.  Cutting and Cleaning Surface of Sample 
 
46 
 
3.5.3 EFA Testing Procedure 
Once the EFA test sample was cut, extruded, and cleaned the below procedure was 
followed for performing a typical EFA test. 
1. The EFA tank was rinsed of any suspended material or debris to clean out the tank. 
2. The EFA tank was filled with clean water, approximately 2/3 of the tank height. 
3. The drainage valve at the bottom of the EFA was opened to allow water to drain from 
the EFA tank. 
4. The sump pump on the EFA was turned on to assist the drainage valve in draining 
water from the tank. 
5. Once the tank was full, drainage valve open, and sump pump on, clean water was 
added to the tank.  This completed the continuous circulation of clean water in the 
EFA to control temperature levels. 
6. The water level was monitored throughout the testing process to ensure that the water 
level did not drop below the pump intake or rise within 300 mm of the top of the tank.   
7. The blank sample was loaded into EFA, and the pump was turned on.  The flow 
velocity was adjusted to the target testing velocity.  This step, suggested by Crim 
(2003), allowed for velocity to be adjusted between tests without exposing the soil 
sample to varying shear stresses. 
8. Once the target velocity was reached the pump was turned off and the blank sample 
was removed from the EFA. 
9. The prepared soil sample was then loaded into the EFA, assuring that the sample tube 
was flush with the lip of the base plate of the flume. 
 
47 
 
10. Using the stepping motor, the sample was advanced through the base plate until the 
sample is level with the base of the flume.  Figure 3-15 shows the soil plunger being 
pushed against the base of the soil sample, and Figure 3-16 shows the sample pushed 
level with the base of the flume.   
11. The CrossTalk program was started, connection settings entered, and the script dialog 
box was filled out according the settings shown in Figure 3-7.  It is important to note 
that this step occurred before the flume was started.  Also it is important the 
CrossTalk script was simply set-up and not started, as blank scans at the beginning of 
a test would interfere with data reduction. 
12. The EFA pump was started and the water velocity was allowed to accelerate until the 
target velocity was reached.  The velocity was monitored using the software shown in 
Figure 3-2. 
13. The CrossTalk script was quickly started by pressing ?OK? on the Script Dialog Box 
shown in Figure 3-7 and naming the data file according to the sample being tested.  It 
is important that this step be started immediately after the flume fills with water, as 
any soil mass lost before the CrossTalk script was started would not be recorded. 
14. Depending on the scour characteristics of the soil being tested, the Crosstalk script 
was usually allowed to run approximately three minutes with the soil sample level 
with the base of the flume.  This step was essential is establishing a clear baseline 
before the sample was advanced and allowed to scour.  However, when testing at high 
velocities that resulted in higher scour rates this amount of time was reduced to 1.5 
minutes to ensure the soil mass did not scour prior to being advanced. 
 
48 
 
15. The sample was advanced 1mm into the flume using the EFA software.  Figure 3-17 
shows a sample advanced into the flume. 
16. Throughout the entire test, scour behavior was visually monitored and recorded to 
verify the results provided by the ultrasonic sensor.  Scour was also visually 
monitored to determine any trends in the scour mechanism of each formation tested.   
17. Once the data acquisition system and EFA test were running smoothly, the real time 
data reduction Excel spreadsheet template was started.  It is important to note that this 
step is not necessary as data reduction can be performed with the created data file 
after the test is completed.  However, one of the benefits of CrossTalk software and 
data acquisition system is that data can be reduced in real time and scour results can 
be presented complimentary to visual confirmation. 
18. With the template Microsoft Excel data reduction file open, the ?Data? tab was 
selected.  The ?From Text? icon was selected and the CrossTalk text file was 
imported.  It was determined that the text should be delimited by tab and space so that 
each data column would be imported separately.  The ?Properties? tab was opened 
when prompted to select the cell in which the data will be placed.  In the ?External 
Data Range Properties? window the ?Save query definition?, ?Adjust Column 
Width?, and ?Preserve Cell Formatting? boxes were selected.  Under the ?Refresh 
Control? heading, the box was selected to refresh the data prompt every 2 minutes.  
The Excel sheet was set to refresh the data from the text file every 2 minutes in order 
to monitor scour every in real time.  Screen shots from this step are shown in Figure 
3-18 and Figure 3-19. 
 
49 
 
19. Scour was monitored in the ?Erosion Height Change? tab of the Excel template sheet. 
Typically the 1 mm push and subsequent scour was clearly visible.  Once the 
?Erosion Height Change? tab showed erosion totaling 1 mm the test was ended.  
Again it was important to visually confirm that 1 mm of scour occurred before 
stopping the CrossTalk software or the EFA pump. 
20. Once erosion occurred and the test was completed, the CrossTalk software was 
stopped by pressing ?Control-C?.  The EFA pump and the water supply were stopped 
once the test ended.   
21. The water was allowed to drain from the flume before the sample was removed from 
the base plate of the EFA.  The sample surface was cleaned similarly to the method 
described in the ?Sample Preparation? section.  This action was necessary to ensure 
that water did not percolate through the sample over time compromising future tests 
from being performed at field conditions. 
22. Once the sample surface was cleaned and level with the sample tube the sample was 
capped, sealed with tape, and placed back into the curing room until the next EFA test 
was performed. 
23. The EFA was drained and rinsed of any debris that resulted from erosion testing.  It 
was important that the amount of suspended soil particles in the EFA were minimized 
as erosion could result by the collision of particles in future tests. 
24. The erosion rate for each test was determined by dividing the height of erosion (1 
mm) by the amount of time it took to achieve the height of erosion.  This could be 
clearly mapped using the Excel data reduction spreadsheet. 
 
50 
 
25. The CrossTalk text file and the Excel data reduction spreadsheet were saved with the 
title of each test so that any additional post processing could be performed if 
necessary. 
 
Figure 3-15.  Soil Plunger Advancing Sample 
 
Figure 3-16.  Soil Sample in EFA Level with Base of Flume 
 
51 
 
 
Figure 3-17.  Soil Sample in EFA Advanced 1mm 
 
 
 
Figure 3-18. Text Import Wizard, Delimited Settings for CrossTalk File 
 
52 
 
 
Figure 3-19.  Import Properties for Data Reduction Spreadsheet 
 
3.6 Testing Regimen 
Based on the original EFA reports by Briaud et al. (1999), previous work at Auburn 
University (Crim 2003; Mobley 2009), and the maximum stream velocity expected in 
Alabama rivers, a testing system was created to include six different testing velocities.  These 
EFA testing velocities include 0.3 m/s, 0.6 m/s, 1.0 m/s, 1.5 m/s, 2.0 m/s, and 3.0 m/s.  
Typically a formation was first tested at 0.3 m/s, and the velocity was gradually increased 
until scour occurred.  A formation was considered to resist scour at a certain velocity if scour 
did not occur with one hour of testing.  After it was determined that a given formation was 
scour resistant at a certain velocity, the velocity was increased to the next highest velocity.  
Once scour occurred at a given velocity, numerous tests were performed at each of the 
 
53 
 
remaining testing velocities so that averages could be established.  Typically a minimum of 
three EFA tests were performed at each velocity step greater than the threshold velocity 
related to the critical shear stress.  The amount of tests performed at each velocity was 
dependent upon the amount of testable soil recovered during sampling. 
After testing was completed across all of the testing velocities, a threshold velocity 
test was performed to determine the velocity that correlates to the critical shear stress.  This 
test was started by exposing the soil sample to the highest velocity that did not show any sign 
of erosion.  The flume velocity was then steadily and carefully increased until scour started.  
Once scour started the velocity was recorded and used as the threshold velocity in the 
creation of erosion functions. 
If a formation was determined to be resistant to scour at all of the listed EFA testing 
velocities, a multiple events test was performed.  This test attempted to model the 
performance of a formation against changing shear stress cycles.  For the purposes of this 
research, the multiple events test included running an EFA test on a sample for one hour at 
3.0 m/s, reducing the velocity to 1.0 m/s for 30 minutes, and increasing the flume velocity to 
3.0 m/s for another hour.  The multiple events test was not performed on any formation with 
a threshold velocity less than 3.0 m/s.  It was determined that adequate scour data could not 
be collected on a sample that had already scoured 1 mm and was simply advanced another 
millimeter.  In previous work at Auburn University (Mobley 2009) it was noted that scour in 
cohesive soils does not occur in a uniform fashion.  Therefore, advancing a sample 1 mm 
after it has already eroded would not expose an even shear stress across the plane of the 
sample as some of the test specimen would be higher than the necessary 1 mm and parts of 
the specimen would be lower than the necessary 1 mm. 
 
54 
 
3.7 Geotechnical Testing 
As previously stated, a major goal of this research was to correlate scour parameters 
and behavior to common geotechnical parameters.  In an attempt to determine any trends 
between scour behavior and geotechnical parameters, several geotechnical laboratory and 
field tests were performed.  During sampling Standard Penetration Tests were performed on 
each formation.  Natural moisture contents were taken from sampling tubes prior to EFA 
tests.  Excess soil from sampling was used to perform a sieve analysis (ASTM D422), 
hydrometer, liquid limit and plastic limit test (ASTM D318) for each formation.  It was 
intended that unconfined compression tests be performed on each formation in an attempt to 
correlate scour parameters to shear strength.  However, after studying the sample tubes, there 
was not a sufficient uncracked sample length of 11.43 cm to perform an unconfined 
compression test among any of the sampled formations. 
 
55 
 
Chapter 4 Test Results 
4.1 General Sampling 
 As previously stated, the sampling program for this research used the CME continuous 
sampling system.  Sampling was coordinated with the assistance of ALDOT geologists and drill 
crews.  Nine total cohesive soil and chalk formations were sampled with the intent to determine 
scour and geotechnical parameters.  These drilling locations along with the seven previous drill 
sites tested at Auburn are shown below in Figure 4-1.   
 
Figure 4-1.  Locations of Sampled and Tested Formations 
Previously Tested at AU 
 
Soils Acquired for Testing 
 
56 
 
 The current study?s drill sites were located in the southern and western portions of 
Alabama, specifically in the coastal plains and prairies.  These locations were drilled between 
April and June of 2012. 
 Typical tests were annotated using a combination of the formation, depth, and test 
number of each sample.  For example, a test name was titled ?Bucatunna27.0_1?.  The first word 
in the title represents the formation of the sample, the Bucatunna Clay formation.  The first set of 
numbers in the title represents the approximate depth (in feet) the sample was procured, or 27 
feet in the example title above.  Finally, the number after the underscore represents the test 
number performed on any given sample. 
4.2 Bucatunna Clay 
The first formation tested using the updated EFA featuring the ultrasonic sensor was the 
Bucatunna Clay formation located in southern Alabama.  The Bucatunna Clay formation was 
sampled using the CME continuous sample system.  EFA tests along with geotechnical tests 
were also performed on the Bucatunna Clay formation.  
4.2.1 Sampling 
The Bucatunna Clay formation was sampled on April 5, 2012 in Monroe County, 
Alabama.  An ALDOT geologist classified the Bucatunna Clay formation as a dark gray to 
brown clay.  The formation was located with the assistance of an ALDOT geologist at a depth of 
3.35 m below the ground surface.  A Standard Penetration Test was performed between 3.5 and 
3.96 m below the ground surface, yielding a SPT N value of 6.  Three different continuous 
samples were taken from depths 4.11 to 5.64 m, 5.64 to 7.16 m, and 7.16 to 8.69 m.  Since this 
was the first sample procured using the CME continuous sample system and because the 
Bucatunna Clay was relatively soft, a Shelby Tube was advanced from 8.69 to 9.30 m.  The 
 
57 
 
Shelby Tube sample was acquired as a back-up in case EFA testing could not be performed with 
the continuous samples.  Lastly, another Standard Penetration Test was performed between 9.30 
and 9.75 m with a SPT N value of 9. 
The continuous sample from 4.11 m to 5.64 m was determined to be unusable for EFA 
testing.  The top of the sample was not usable as it had a plug missing from the Standard 
Penetration test taken above.  The sample gathered from 4.11 to 5.64 m was also smaller radially 
than the sample tube at the base of the sample as shown in Figure 4-2.   
 
Figure 4-2.  Bucatunna Clay Sample Smaller Radially than Sample Tube 
The second sample of the Bucatunna Clay formation ranged from 5.64 to 7.16 m and was 
mostly unusable.  The entire top half of the sample was severely fractured in all directions and 
could not be used for EFA testing.  However, the bottom third of the bottom half of the sample 
was mostly uncracked and sections could be used for EFA testing.  The third and final sample 
from 7.16 to 8.69 m was wrinkled at the top with the bottom 20 to 30 mm un-cracked.  It 
appeared in the second and third samples that the soil was failing in shear during sampling along 
 
58 
 
the tube walls.  This would be possible if the rotating head of the continuous sampler did not 
fully separate the sample from the rotation of the hollow stem auger and the sample was shoved 
and twisted into the tubes.  This phenomenon was observed and monitored throughout the 
sampling of all formations and these disturbed and cracked regions were not tested in the EFA.  
These regions were however used in geotechnical testing where soil was processed prior to 
testing. 
4.2.2 EFA Testing 
  The Bucatunna Clay was the first formation tested at Auburn University using the 
ultrasonic sensor and data acquisition system.  In total 32 EFA tests were performed on the 
Bucatunna Clay formation with 27 tests providing results for determining a scour rate versus 
velocity relationship.  The first sample used for EFA testing was located at approximately 8.22 m 
below the ground surface.  The first test titled ?Bucatunna27.0_1? was performed to determine a 
starting velocity for testing.  The data collected from ?Bucatunna27.0_1? is shown below in 
Figure 4-3.   
 
59 
 
 
Figure 4-3.  ?Bucatunna 27.0_1? Results 
  The ?Bucatunna 27.0_1? test specimen began at a starting velocity of 0.3 m/s, and was 
advanced 1mm into the flume after approximately 5 minutes.  This can be viewed in the bump at 
the five minute mark in Figure 4-3, and further proves the validity of the ultrasonic sensor.  The 
?Bucatunna 27.0_1? sample actually showed a net gain in height over the next hour while the 
velocity was held at 0.3 m/s.  This seemed odd that the sample was actually rising into the flume 
as the shear stress was applied to the sample, but this movement was visually confirmed.  After 
one hour, no scour was observed and the sample had raised an average of 2.5 mm.  It was 
believed that this swelling was due to the Bucatunna Clay being a swelling clay formation, and 
this issue would be confirmed during geotechnical testing.  Once the flume velocity was 
increased to 0.6 m/s the ?Bucatunna27.0_1? sample scoured drastically, as shown in Figure 4-3.  
-2
-1
0
1
2
3
4
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
0.3 m/s
0.6 m/s
Push 1mm
 
60 
 
From the results of ?Bucatunna 27.0_1? it was determined that the Bucatunna Clay formation did 
not scour at 0.3 m/s, the threshold velocity was between 0.3 and 0.6m/s, and that further scour 
testing should start at a flume velocity of 0.6 m/s.   
  A total of five EFA tests were performed on the Bucatunna Clay formation at a flume 
velocity of 0.6 m/s.  Table 4-1 below shows the test results obtained by visual observation along 
with the ultrasonic sensor.  
Table 4-1:  Bucatunna Clay Results at 0.6 m/s 
 
  As seen in Table 4-1, the elapsed time for erosion as determined by the EFA were similar 
to the elapsed time observed.  The largest difference in the observed and measured elapsed time 
was two minutes as seen in test ?Bucatunna 27.0_4?.  The scour rates for the five tests were 
similar with ?Bucatunna 27.0_5? having a slightly higher scour rate of 5.0 mm/hour and 
?Bucatunna 27.0_6? having a slightly lower scour rate.  The average scour rate at 0.6 m/s as 
calculated by the ultrasonic sensor was 3.91 mm/hr.   
  A total of 7 EFA tests were performed at a flume velocity of 1.0 meters per second.  
Below, Table 4-2 shows the results obtained for the tests performed at 1.0 m/s.  The elapsed time 
and subsequent scour rates observed at 1.0 m/s were more variable than those observed at 0.6 
meters per second.  Seven tests were performed as two tests ?Bucatunna 27.5_3? and ?Bucatunna 
27.5_4? were uncharacteristic of the rest of the data set with elapsed times measured at 7 and 38 
minutes respectively.  As seen in Table 4-2, the ?Bucatunna 27.5_3? test was unable to be 
Buccatunna 27.0_3 14 13 4.29 4.62
Buccatunna 27.0_4 17 15 3.53 4.00
Buccatunna 27.0_5 12 13 5.00 4.62
Buccatunna 27.0_6 22 22 2.73 2.73
Buccatunna 27.0_7 15 15 4.00 4.00
Scour Rate 
Visual 
(mm/hr)
Scour Rate 
Ultrasonic 
(mm/hr)
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
 
61 
 
visually observed, but visual observations were once again relatively close to ultrasonic sensor 
calculations.  It was also observed that the ?Bucatunna 27.5_3? test appeared to be softer than 
insitu conditions prior to testing, explaining the high scour rate.  The average erosion rate for the 
7 EFA tests performed on the Bucatunna Clay formation at a flume velocity of 1.0 m/s was 4.07 
mm/hr.   
Table 4-2:  Bucatunna Clay Results at 1.0 m/s 
   Five tests were performed at a flume velocity of 1.5 m/s, and the results can be viewed in 
Table 4-3.  As expected the recorded scour rates increased compared to those recorded at lower 
velocities, with an average scour rate of 5.59 mm/hr.  The data set acquired from this velocity 
had a low variability with the exception on the ?Bucatunna 26.5_3? test that had a higher scour 
rate.  It was encouraging that this data set was relatively close with respect to scour rates as it 
was the first data set tested using two different sample depths.  It was noted during testing that a 
sand seam was located approximately 50 mm above the ?Bucatunna 26.5_3? test sample.  
However, subsequent tests using this sample showed the same scour characteristics as previous 
test samples.   
 
 
 
 
Buccatunna 27.5_2 13 14 4.62 4.29
Buccatunna27.5_3 7 N/A 8.57 N/A
Buccatunna27.5_4 38 24 1.58 2.50
Buccatunna27.5_5 15 16 4.00 3.75
Buccatunna27.5_6 17 17 3.53 3.53
Buccatunna27.5_7 18 18 3.33 3.33
Buccatunna27.5_8 21 17 2.86 3.53
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
62 
 
Table 4-3:  Bucatunna Clay Results at 1.5 m/s 
 
  Five EFA tests were performed on the Bucatunna Clay formation at a flume velocity of 
2.0 m/s.  The results from these tests can be viewed below in Table 4-4.  The test titled 
?Bucatunna 26.5_7? was the lone outlier in this data set with an elapsed time of 22 minutes.  
However, there were not any oddities or special notes recorded with regards to this test, it just 
seemed as though this test was more scour resistant than other tests at this velocity.  Overall the 
average scour rate at a flume velocity of 2.0 m/s was 6.67 mm/hr.   
Table 4-4:  Bucatunna Clay Results at 2.0 m/s 
 
  The final velocity used in EFA testing of the Bucatunna Clay formation was at 3.0 m/s.  
Five EFA tests were performed at this velocity and two separate samples were used, similar to 
the tests performed at 1.5 m/s.  The results from the tests performed at 3.0 m/s can be viewed 
below in Table 4-5.  At this high velocity, erosion occurred rather quickly averaging less than 6 
minutes per test.  Since the time for erosion was so brief, instead of advancing the sample after 
five minutes, the sample was advanced into the flume after only being exposed to flow for two 
minutes.  This minimized the possibility of erosion prior to advancing the sample.  The average 
Buccatunna 27.5_9 14 11 4.29 5.45
Buccatunna 26.5_2 12 11 5.00 5.45
Buccatunna26.5_3 7 8 8.57 7.50
Buccatunna26.5_4 13 14 4.62 4.29
Buccatunna26.5_5 11 11 5.45 5.45
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
Buccatunna 26.5_6 10 10 6.00 6.00
Buccatunna26.5_7 22 20 2.73 3.00
Buccatunna26.5_8 10 10 6.00 6.00
Buccatunna26.5_9 7 8 8.57 7.50
Buccatunna26.5_10 6 8 10.00 7.50
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
63 
 
scour rate at 3.0 m/s was 11.01 mm/hr, a jump of 5.0 mm/hr above the average scour rate at a 
flume velocity of 2.0 m/s. 
Table 4-5:  Bucatunna Clay Results at 3.0 m/s 
 
  The final test performed on the Bucatunna Clay formation was aimed at determining the 
critical shear velocity of the formation.  Since previous tests on the Bucatunna Clay formation 
was scour resistant, and actually swelled, at 0.3 meters per second and scoured at 0.6 m/s, the 
critical shear velocity was located between the two velocities.  The test titled ?Bucatunna 
23.0_4? was started with a flume velocity of 0.3 m/s and gradually increased until scour was 
observed.  The flume velocity was held between 0.35 and 0.4 m/s for approximately 15 minutes, 
and again swelling was observed instead of scour.  Slowly the flume velocity was increased 
between 0.45 and 0.5 m/s.  At this velocity a large chunk eroded from the sample.  Figure 4-4 
below shows the sample immediately after erosion occurred. This image agrees with the 
ultrasonic sensor, while portions of the sample are located above the base of the flume, the 
majority of the sample has eroded in deep pockets below the flume.  From this test it was 
determined that the threshold velocity was 0.45 m/s. 
Buccatunna 26.5_11 8 10 7.50 6.00
Buccatunna26.5_12 7 8 8.57 7.50
Buccatunna 23.0_1 5 8 12.00 7.50
Buccatunna 23.0_2 5 5 12.00 12.00
Buccatunna 23.0_3 4 5 15.00 12.00
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
64 
 
 
Figure 4-4.  ?Bucatunna 23.0_4? After Erosion Occurred 
  As previously stated a total of 27 EFA tests were performed on the Bucatunna Clay 
formation at five different velocities.  Figure 4-5 below shows the all of the results for the EFA 
test results with respect to scour rate and flume velocity.  For the most part the data sets are 
tightly bunched with similar scour rates at each velocity interval.  Figure 4-6 below shows the 
drawn-in velocity versus scour rate curve for the Bucatunna Clay formation as derived from the 
mean scour rate at each velocity.   
 
65 
 
 
Figure 4-5.  EFA Tests Results, Bucatunna Clay 
 
 
Figure 4-6.  Bucatunna Clay Velocity versus Scour Rate 
0
2
4
6
8
10
12
14
16
0 0.5 1 1.5 2 2.5 3 3.5
Sc
ou
r R
ate
 (m
m/
hr
)
Velocity (m/s)
0.6 m/s
1.0 m/s
1.5 m/s
2 m/s
3.0 m/s
0
2
4
6
8
10
12
0 0.5 1 1.5 2 2.5 3 3.5
Sc
ou
r R
ate
 (m
m/
hr
)
Velocity (m/s)
AVG
 
66 
 
4.2.3 Geotechnical Testing 
As part of the scope of this research several common geotechnical index tests were 
performed on the Bucatunna Clay formation.  Two initial moisture contents were taken to obtain 
the insitu moisture content of the formation prior to EFA testing.  These tests yielded moisture 
contents of 49.5 and 46.2 percent, with an average of 47.9 percent.  Using ASTM D421, several 
kilograms of the soil obtained during sampling was processed and oven dried.  Using this 
processed soil, a full grain size distribution was performed on the formation following ASTM 
D422.  The results of the grain size analysis are shown in Figure 4-7.  Since this research 
revolved around fine-grained materials, a coarse sieve analysis was not performed as the 
maximum particle size diameter for this formation was approximately 0.43 mm.  From the grain 
size analysis the mean particle diameter (D50) was calculated to be 0.033 mm.  Approximately 65 
percent of the sample passed the number 200 sieve, classifying the formation in the silt and clay 
family.   
 
Figure 4-7.  Bucatunna Clay Grain Size Distribution 
0
10
20
30
40
50
60
70
80
90
100
0.0010.010.1110
%
 Pa
ssi
ng
Particle Diameter (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.033 mm
 
67 
 
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Bucatunna Clay formation.  The procedure presented by ASTM D318 was 
followed when performing these Atterberg limits.  The liquid limit (LL) was determined to be 
approximately 68, while the plastic limit (PL) was found to be 39.  Using the Atterberg limits, 
the plasticity index was 29.  Using these values along with the grain size distribution in Figure 4-
7, the Bucatunna Clay formation was classified as a sandy elastic silt (USCS Classification MH).  
The high plasticity index of 29 is greater than a plasticity index of 10, which is commonly used 
as the threshold to classify a clay with swelling potential.  This explains the swelling 
phenomenon observed in a few tests at velocities below the critical shear velocity.  
As previously stated, the recovered sample was not conducive to being used in an 
unconfined compression test as the sample was cracked at intervals tighter than the minimum test 
length of 115 mm.  Individual EFA test and geotechnical test results can be viewed in Appendix 
A. 
4.3 Yazoo Clay 
The next formation tested using the updated EFA featuring the ultrasonic sensor was the 
Yazoo Clay formation located in southern Alabama.  The Yazoo Clay formation was sampled 
using the CME continuous sample system.  EFA tests along with geotechnical index tests were 
also performed on the Yazoo Clay formation.  
4.3.1 Sampling 
The Yazoo Clay formation was sampled on April 6, 2012 in Conecuh County, Alabama.  
An ALDOT geologist classified the Yazoo Clay formation as a light colored stiff gray clay.  The 
Yazoo Clay formation was sampled above a box culvert crossing a stream with a visible outcrop 
of the formation viewed in the streambed.  An ALDOT geologist identified the formation in drill 
 
68 
 
cuttings and split spoon samples at approximately 4.11 m below the ground surface.  A Standard 
Penetration Test was performed at the top of the formation yielding an N value of 15 blows.  
Three total CME continuous sample runs were performed on the Yazoo Clay formation with 
varying results.   
The first run was taken from 4.27 to 5.79 m below ground surface.  The top half was 
disturbed and unusable for EFA testing due to the Standard Penetration Test which was 
terminated at 4.6 m.  The bottom half of the first run provided several sections that were testable 
in the EFA.  The second continuous sample run was performed between 5.79 and 7.32 m below 
the ground surface.  The top tube of the run had many cracked sections that were not able to be 
used in the EFA, but could be used for classification and index testing.  The bottom tube of the 
sample run from 6.55 to 7.32 m had several EFA testable sections.  The third and final sample 
run was performed between 7.32 and 8.84 m below ground surface.  After this continuous 
sample was obtained it was a sandy material was noticed at a depth of 8.2 m.  The Yazoo clay 
material obtained in this third run was completely unusable as a transition into the sandy material 
below was apparent.  Due to the change in material, another Standard Penetration Test was not 
performed after sampling. 
4.3.2 EFA Testing 
A total of twelve EFA tests were performed on the Yazoo Clay formation.  After the 
samples were brought back to Auburn University and stored for testing, it was determined that 
much less of the formation was able to be tested in the EFA than originally thought.  The best 
section for EFA testing was approximately one third of a meter long between 5.48 and 5.80 m 
below the ground surface.  It was noted during sample preparation that the Yazoo Clay formation 
 
69 
 
was much tougher to cut and create a smooth sample surface.  It was apparent that the sand 
content in the Yazoo Clay formation was much higher than the Bucatunna Clay formation.   
The first test performed on the Yazoo Clay formation was used to help determine the 
starting test velocity and was titled ?Yazoo Clay 18.5_1?.  The ?Yazoo Clay 18.5_1? test started 
with a flume velocity of 0.3 m/s, and scour was not observed.  After one hour the flume velocity 
was increased to 0.45 m/s and immediately a wedge of the sample scoured away.  This wedge 
developed shortly after the test was started and grew in size throughout the test, and image of this 
wedge can be viewed in Figure 4-8.  However, since this method of failure had not yet been 
observed a second test titled ?Yazoo Clay 18.5_2? was performed starting at a flume velocity of 
0.45 m/s.   
 
Figure 4-8.  ?Yazoo Clay 18.5_1? Test Failure Wedge 
The ?Yazoo Clay 18.5_2? test did not perform similarly to the previous test but exhibited 
a swelling characteristic similar to that observed in the Bucatunna Clay formation.  The test was 
started at a flume velocity of 0.45 m/s and did not scour, after one hour.  The flume velocity was 
then increased to 0.6 m/s for an hour in which scour was not observed.  The flume velocity was 
then increased temporarily to 0.8 m/s for fifteen minutes before being increased once again to 1.0 
 
70 
 
m/s.  During this time the test sample rose and average of 1.75 mm, and developed a large crack 
in the center of the sample as shown in Figure 4-9.  After the test sample was exposed for 25 
minutes at a flume velocity of 1.0 m/s, the sample scoured well below the bottom of flume as 
shown in Figure 4-10.  The test results showing the gradual swell in the ?Yazoo Clay 18.5_2? 
test can be viewed in Figure 4-11.  The test was analyzed and it was determined that the testing 
regiment should be started at a flume velocity of 1.0 m/s.     
 
Figure 4-9.  ?Yazoo Clay 18.5_2? Test Swelling and Cracking 
 
Figure 4-10.  ?Yazoo Clay 18.5_2? Test After Failure 
 
71 
 
  
Figure 4-11.  ?Yazoo Clay 18.5_2? Test Results 
Five EFA tests were performed on the Yazoo Clay formation at a flume velocity of 1.0 
m/s.  The average scour rate among these five tests was 7.33 mm/hr, but included very high 
variability.  Four of the five tests at this velocity resulted in scour times ranging from five to 
eight minutes.  The fifth test, titled ?Yazoo Clay 18.5_7?, at this velocity did not scour under a 
flume velocity of 1.0 m/s after one hour.  After an hour the flume velocity was increased to 1.2 
m/s.  At this velocity thin flakes began to scour immediately after the velocity was increased, but 
stopped, and the flume velocity was increased to 1.5 m/s.  After fifteen minutes at a flume 
velocity of 1.5 m/s, no major scour was observed and the flume velocity was increased once 
again to 2.0 m/s.  Five minutes later at a flume velocity of 2.0 m/s a large chunk of material 
scoured from the surface of the test.  Once the ?Yazoo Clay 18.5_7? test was complete and 
results analyzed, it was determined that the testing regiment should be continued at a flume 
-3
-2
-1
0
1
2
3
4
0 20 40 60 80 100 120 140 160 180
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
0.3 m/s
0.6 m/s
0.8m/s
1.0 m/s
 
72 
 
velocity of 1.5 m/s.  A summary table of the tests performed at a flume velocity of 1.0 m/s is 
shown below in Table 4-6. 
Table 4-6:  Yazoo Clay Results at 1.0 m/s 
 
Two EFA tests were performed at a flume velocity of 1.5 m/s, and the results mirrored 
the ?Yazoo Clay 18.5_7? test results.  The first test performed at a flume velocity of 1.5 m/s 
showed minimum scour for the first 40 minutes of the test; however a huge chunk immediately 
scoured away after 43 minutes resulting in a scour rate of 1.40 mm/hr.  The second EFA test 
performed at a flume velocity of 1.5 meters per second had varying results mostly due to 
connection issues with the ultrasonic sensor.  Once the second test, ?Yazoo Clay 18.5_9?, was 
started approximately one third of the sample scoured on the back half of the sample.  This scour 
was not noticed much ultrasonically, due to high peaks in the scour surface being located directly 
under the transducers.  Transducers 9 and 10 which were located directly above the location the 
scour occurred continued to record readings on par with the rest of the transducers.  Transducers 
11 and 12 showed very erratic readings during a fifteen minute stretch of the test.  These erratic 
readings were most likely due to connection issues with the ultrasonic sensor and were rarely 
noted during testing.  After one hour the ?Yazoo Clay 18.5_7? test showed very little signs of 
scour other than the portion that eroded immediately.  Images of this scoured section can be 
viewed below in Figure 4-12 and Figure 4-13.  It is very plausible that the high readings 
Yazoo Clay 18.5_3 5 5 12.00 12.00
Yazoo Clay 18.5_4 8 8 7.50 7.50
Yazoo Clay 18.5_5 7 7 8.57 8.57
Yazoo Clay 18.5_6 7 7 8.57 8.57
Yazoo Clay 18.5_7* N/A N/A 0.00 0.00
Scour Rate 
Visual 
(mm/hr)
Scour Rate 
Ultrasonic 
(mm/hr)
* Did Not Scour
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
 
73 
 
observed in transducers 9 and 10 are directly associated with high peaks and steep valleys in the 
scoured area.   
Since the two samples that were tested at a flume velocity of 1.5 m/s showed smaller 
amounts of scour than the tests at 1.0 m/s, the testing schedule was continued at a flume velocity 
of 2.0 m/s.  Overall the average scour rate of the two tests performed at a flume velocity of 1.5 
m/s was 0.70 mm/hr.  A summary of these two tests is shown in Table 4-7. 
 
Figure 4-12.  ?Yazoo Clay 18.5_9? Test Scoured Area from Front 
 
Figure 4-13.  ?Yazoo Clay 18.5_9? Test Scoured Area from Behind 
 
74 
 
Table 4-7:  Yazoo Clay Results at 1.5 m/s 
 
A total of three tests were performed on the Yazoo Clay formation at a flume velocity of 
2.0 m/s.  The first test, titled ?Yazoo Clay 18.5_10? scoured had a resulting scour rate of 4.29 
mm/hr.  However, the test did not scour at a constant rate or even in small chunks as observed in 
previous Yazoo Clay EFA tests.  The test showed minimum surface scouring for the first thirteen 
minutes of the test until a very large chunk scoured away.  Using the ultrasonic sensor this chunk 
encompassed the entire sample and was approximately ten millimeters deep into the flume.  Two 
additional tests were performed at a flume velocity of 2.0 m/s and both showed similar 
magnitudes of scour as the ?Yazoo clay 18.5_10? test immediately after the pump was started.  
Table 4-8 below shows the three test results at a flume velocity of 2.0 m/s. 
Table 4-8:  Yazoo Clay Results at 2.0 m/s 
 
The sporadic scour behavior of the Yazoo Clay formation lead to the pausing of the EFA 
testing schedule as most of the tests yielded contradictory scour results.  The only consistent 
results obtained from this formation were at a flume velocity of 1.0 m/s, which yielded an 
average scour rate of 7.33 mm/hr.  The test performed at both 1.5 and 2.0 m/s were highly 
Yazoo Clay 18.5_8 43 43 1.40 1.40
Yazoo Clay 18.5_9* N/A N/A 0.00 0.00
* Did Not Scour
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
Yazoo Clay 18.5_10 14 11 4.29 5.45
Yazoo Clay 18.0_1** 0 0 N/A N/A
Yazoo Clay 18.0_2** 0 0 N/A N/A
**Complete Failure Prior to Push
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
75 
 
variable and inconsistent yielding no consistent results.  Individual test results of the Yazoo Clay 
formation can be viewed in Appendix B.   
4.3.3 Geotechnical Testing 
Two initial moisture contents were taken to obtain the insitu moisture content of the 
formation prior to EFA testing.  These tests yielded moisture contents of 62.2 and 57.6 percent, 
with an average of 59.9 percent.  Using soil processed by following ASTM D421, a full grain 
size distribution was performed on the formation following ASTM D422.  The results of the 
grain size analysis are shown below in Figure 4-14.  Similar to the Bucatunna Clay formation 
one hundred percent of the formation passed through the number 10 sieve, thus negating the need 
for a coarse sieve analysis.  From the grain size analysis the mean particle diameter was 
calculated as 0.088 mm.  Approximately 44 percent of the sample tested passed the number 200 
sieve classifying the sample as a sand.   
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Yazoo Clay formation.  The liquid limit was determined to be approximately 57, 
and the formation was determined to be non-plastic.  Using these values along with the 
information gathered from the grain size distribution, the sample was classified as a non-plastic 
silty sand (USCS Classification SM).  
 
76 
 
 
Figure 4-14.  Yazoo Clay Formation Grain Size Distribution 
Due to the results of the geotechnical index tests along with the sporadic EFA test data, it 
was apparent that the sample acquired to represent the Yazoo Clay formation, was not purely 
Yazoo Clay, but had a high sand content.  This sand content was not evenly distributed 
throughout the sample as several EFA test results show potential for scour resistance.  Likewise, 
a high percentage of the sample was retained on the number 200 sieve and appeared to be a very 
fine silty sand.  Since this layer of clay was sampled between two sandy layers it appears the 
sample tested was not a pure sample of the Yazoo Clay formation. 
4.4 Demopolis Chalk 
4.4.1 Sampling 
The Demopolis Chalk formation was sampled on May 5, 2012 in Sumter County, 
Alabama.  An ALDOT geologist classified the Demopolis Chalk formation as a very stiff light 
gray chalk.  The Demopolis Chalk formation was sampled beside a bridge crossing the 
0
10
20
30
40
50
60
70
80
90
100
0.0010.010.1110
%
 Pa
ssi
ng
D (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.088 mm
 
77 
 
Tombigbee River.  An outcrop of the formation could be viewed in the bluffs comprising the 
river bed.  These bluffs were approximately 15 m above the river, entirely comprised of the 
Demopolis Chalk formation.  These bluffs can be viewed below in Figure 4-15.   
 
Figure 4-15.  Demopolis Clay Formation Sampling Location 
 
An ALDOT geologist identified the formation in drill cuttings and split spoon samples at 
approximately 4.27 m below the ground surface.  A Standard Penetration Test was performed at 
the top of the formation yielding a high N value of 92 blows.  The ALDOT drill crew performed 
two CME continuous sample runs on the Demopolis Chalk formation.  
The first run was taken from 4.57 to 6.10 m below ground surface.  The top half was 
disturbed and unusable for EFA testing due to the Standard Penetration Test which was 
terminated at approximately 4.65 m.  An EFA testable section was between 5.8 and 6.1 m below 
ground surface.  The middle portion of this top run was cracked in both the vertical and radial 
directions.   
 
78 
 
The second continuous sample run was performed from 6.1 to 7.62 m below the ground 
surface.  The top and middle portions of this sample were severely cracked in both the vertical 
and radial direction.  The bottom third of the sample was uncracked; however, during sampling 
the sample tube was bowed and stretched.  It was noted during sampling that the stiffness of the 
Demopolis Chalk formation made advancing the sample tubes increasingly difficult with depth.  
The torque necessary to advance the continuous sampler in the formation was great enough to 
distort the continuous sample tube and increase the outside diameter from 64 mm to 
approximately 74 mm.  This increase in diameter of the sample tubes made this section of the 
Demopolis Chalk formation unable to be tested and advanced in the EFA.  Due to the difficulty 
in sampling the stiff Demopolis Chalk formation only two continuous sample runs were 
performed, resulting in approximately 30 cm of EFA testable sample. 
4.4.2 EFA Testing 
  A total of five Erosion Function Apparatus tests were performed on the Demopolis Chalk 
formation.  The Demopolis Chalk Formation was unable to be automatically advanced into the 
EFA during testing due to the stiffness of the sample.  Therefore, prior to each EFA test the 
sample was manually advanced approximately one millimeter using a hydraulic extruder, and 
then placed into the EFA at the testing height.  The height of the test specimen protruding from 
the flume could be calculated by subtracting the know flume height from the corrected distances 
measured by the ultrasonic sensor.  This changed the shape of the erosion rate versus elapsed 
time figures generated by the ultrasonic sensor and data acquisition system.  Instead of a figure 
showing a starting location at 0, a push advancing the sample approximately one millimeter, and 
subsequent scour movement, the figure showed a starting location at 0 and subsequent scour 
movement from that datum.   
 
79 
 
  The first test performed on the Demopolis Chalk formation was titled 
?DemopolisChalk19.0_1?, and was started at a flume velocity of 1.2 m/s.  Using scales attached 
to the outside of the EFA viewing window it appeared that the sample protruded from the flume 
approximately one millimeter.  After 40 minutes of not witnessing any visually or by ultrasonic 
readings, the flume velocity was increased to 2.0 m/s.  Another 20 minutes passed without any 
evidence of scour, and the velocity was increased to the maximum test velocity of 3.0 m/s.  The 
test was allowed to run for one hour at the maximum velocity and no scour was witnessed.  It 
was determined that three more tests should be performed at 3.0 m/s to determine if the 
Demopolis Chalk formation was scour resistant according to the parameters of this study.   
  The three tests mentioned above were titled ?DemopolisChalk19.0_2?, 
?DemopolisChalk19.0_3?, and ?DemopolisChalk19.0_4?.  These three tests averaged a 
protruded distance of 1.13 mm above the flume with values of 0.3, 1.5, and 1.6 mm respectively.  
All three of these tests were performed for at least on hour, and all three did not show any signs 
of major scour.  The only scour witnessed was a very minor rounding of the sharp edges of the 
test samples.  It is believed that this is due to stress concentrations that are not observed in the 
field with a continuous stream bed.  This slight rounding of edges was visually observed and was 
not significant enough to be recorded by the ultrasonic sensor.  The ?DemopolisChalk19.0_3? 
and ?DemopolisChalk19.0_4? tests both showed a slight raise in height in the ultrasonic output, 
averaging 0.2 mm over the duration of the tests, hinting that some minor swelling behavior is 
possible in the formation.   
  The final test performed on the Demopolis Chalk formation was titled 
?DemopolisChalk19.0_5?.  This test was intended to mirror recurring storm events on a small 
scale.  The first hour of the test the flume velocity was 3.0 meters per second, dropping to 1.0 
 
80 
 
m/s for the next 30 minutes, and then increased to 3.0 m/s for another hour.  It was calculated 
that the test specimen started at an average height of 1.7 mm above the base of the flume.  No 
major scour was observed during the 2.5 hour test.  Again only a slight rounding off of the front 
edge of the sample was noticed as viewed below in Figure 4-16.  In summary a total of five tests 
were performed on the Demopolis Chalk formation and all five did not show any signs of 
measurable scour. 
  The results of all five tests may be viewed in Appendix C.  This was the first sample 
tested in this study that did not exhibit any scour behavior with the set testing regiment.  In case 
future testing of the Demopolis Chalk formation is necessary, the remaining testable section of 
the formation was sealed and placed into a curing chamber to preserve field moisture conditions. 
 
Figure 4-16.  ?DemopolisChalk19.0_5? After Testing 
4.4.3 Geotechnical Testing 
Three initial moisture contents were taken to obtain the insitu moisture content of the 
Demopolis Chalk formation prior to EFA testing.  These tests yielded moisture contents of 22.5, 
21.5, and 21.4 percent, with an average of 21.8 percent.  Using soil processed by following 
ASTM D421, a full grain size distribution was performed on the formation following ASTM 
D422.  The results of the grain size analysis can be viewed below in Figure 4-17.  The grain size 
 
81 
 
analysis on the Demopolis Chalk showed that the formation consisted of very fine-grained 
particles with approximately 97 percent of the sample passing the number 200 sieve.  From the 
grain size analysis the mean particle diameter was calculated to be 0.021 mm.   
 
Figure 4-17.  Demopolis Chalk Formation Grain Size Distribution 
 
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Demopolis Chalk formation.  The liquid limit was determined to be approximately 
37, and the plastic limit was determined to be approximately 27, resulting in a plasticity index of 
10.  The calculated plasticity index validated the earlier observation during EFA testing, that this 
formation has minor swelling potential.  Using these values along with the information gathered 
from the grain size distribution, the sample was classified as a lean clay (USCS Classification 
CL).  
 
0
10
20
30
40
50
60
70
80
90
100
0.0010.010.1110
%
 Pa
ssi
ng
Particle Diameter (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.0021 mm
 
82 
 
4.5 Mooreville Chalk 
4.5.1 Sampling 
The Mooreville Chalk formation was sampled on April 30, 2012 in Dallas County just 
west of Selma, Alabama.  An ALDOT geologist classified the Mooreville Chalk formation as a 
stiff gray chalk.  The ALDOT geologist identified the formation in drill cuttings and split spoon 
samples at approximately 5 m below the ground surface.  A Standard Penetration Test was 
performed at the top of the formation yielding an N value of 60 blows.  The ALDOT drill crew 
performed three CME continuous sample runs on the Mooreville Chalk formation.  
The sampling of the Mooreville Chalk formation occurred in two drill holes, the Standard 
Penetration Test and first sample run were performed in the first drill hole.  The Standard 
Penetration Test was taken from 6.1 to 6.55 m.  The CME continuous sample was taken from 
6.85 to 8.38 m below the ground surface.  During sampling it was determined that a plug had 
developed in the drill hole and the continuous sampler had to be retracted.  After the sample 
tubes were recovered, the sample tube from 7.62 to 8.38 m showed approximately 30 cm of 
testable section.  However, the sample tubes were extremely bowed approaching an outside 
diameter of 75 mm. 
The second drill hole was located approximately 3 meters from the first, and the 
formation was encountered at a similar depth.  Two continuous sample runs were performed in 
this hole.  The first run was taken from 5.33 to 6.85 m below the ground surface.  A small 
testable section, approximately 15 cm long, was located in the top half of the sample.  Another 
testable section with a similar length was located at the bottom of the sample run.  The rest of the 
run was severely cracked in both the vertical and radial directions.   The second CME continuous 
sample run was performed from 6.85 to 8.38 m below the ground surface.  The top half of the 
 
83 
 
sample run was severely cracked and unusable for EFA testing.  A testable section with a length 
of approximately 30 cm was located at the base of the sample run.  However, 20 cm of this 
section was bowed out similar to the sample tube in the first sample hole.  The Mooreville Chalk 
formation similar to the Demopolis Chalk formation proved difficult to sample with the CME 
continuous sampling system, but sufficient testable sections were recovered.    
4.5.2 EFA Testing 
Ten Erosion Function Apparatus tests were performed on the Mooreville Chalk formation 
resulting in varying results.  The section used to test the Mooreville Chalk was from the second 
drill hole mentioned above at a depth of 6.7 m.  This section was the testable section with the 
least bowed sample tube.  The sample tube was cut according the methods provided in Chapter 3, 
but also had to be sanded in order to fit into the flume of the EFA.  Once the sample tube was 
altered to fit into the base of the flume it was determined that the automatic advancing motor on 
the EFA could not advance the sample.  The sample was manually advanced in a similar manner 
to that of the Demopolis Chalk formation.   
The first test on the Mooreville Chalk formation was performed to determine the base 
velocity for testing the formation.  The test titled ?MoorevilleChalk22.0_2? was started at 1.0 
m/s and quickly showed measurable scour.  It was determined that this test was not indicative of 
the scour parameters of the formation as it was calculated that the sample protruded from the 
base of the flume two millimeters which was double the distance specified by Briaud.  The next 
test performed on the Mooreville Chalk formation was titled ?MoorevilleChalk22.0_3?, was 
started at a flume velocity of 0.3 m/s, and was able to be automatically advanced into the EFA 
flume.  This test showed no signs of scour at 0.3 m/s.  Some large masses of material scoured in 
chunks at 0.6 m/s, and scour continued once the flume velocity was increased to 1.0 m/s.  From 
 
84 
 
this test it was determined that testing should start at 0.6 m/s.  It was noted that the sample 
section used to obtain the results from these two tests could have been exposed to varying stress 
states, not primarily associated to EFA testing.  It appeared that the test sample could have been 
exposed to differing shear and torsional stress states, by being manually advanced with an 
extruder through a bowed sample tube.   
Two EFA tests were performed at a flume velocity of 0.6 m/s.  The first titled 
?MoorevilleChalk22.0_4? resulted in a scour rate of 1.4 mm/hr.  However, all the scour recorded 
was measured from two large mass scours in which a large chunk of material accounted for the 
measured scour.  The second test performed at a flume rate of 0.6 m/s did not show any signs of 
scour visually or ultrasonically.  These two tests did not demonstrate the behavior as the previous 
two with regards to varying stress states, as both were easily automatically advanced.  An 
additional test was performed at a flume velocity of 1.0 m/s on the Mooreville Chalk formation 
and no measurable scour was recorded.   
Two EFA tests were performed at a flume rate of 1.5 m/s on the Mooreville Chalk 
formation.  The first of these two tests was titled ?MoorevilleChalk22.0_7? and resulted in one 
third of the volume of the protrusion being scoured away by one large chunk in the upstream 
portion of the sample.  The final result of this test can be viewed below in Figure 4-18.  The 
second test performed at a flume velocity of 1.5 m/s resulted in no measurable amount of scour.  
The only scour was that occurred during this test was witnessed visually and consisted of the 
rounding off of the front edges of the sample similar to that witnessed in testing the Demopolis 
Chalk formation.   
 
85 
 
 
Figure 4-18.  ?MoorevilleChalk22.0_7? After Testing 
Since there was a limited amount of testable sample obtained from the Mooreville Chalk 
formation it was imperative that testing continued regardless of slightly varying results.  The 
next EFA test performed on the Mooreville Chalk formation was performed at a flume velocity 
of 2.0 m/s and no scour was recorded.  The flume velocity was increased to 3.0 m/s for the next 
EFA test and no scour was recorded.  The final test performed on the Mooreville Chalk 
formation was titled ?MoorevilleChalk22.0_11? and was a repeat even test similar to that 
performed on the Demopolis Chalk formation.  This test was intended to be performed with a 
flume velocity of 3.0 m/s for one hour, 1.0 m/s for thirty minutes, and 3.0 meters per second for 
another hour.  This test showed no signs of scour through the first hour and a half of the test.  
However, soon after the velocity was increased from 1.0 to 3.0 m/s two masses of material 
scoured from the sample.  The smaller of the two masses came from the far side portion of the 
sample, while the larger mass scoured from the near side of the sample.  After these two masses 
scoured away no other scour was noticed or measured throughout the duration of the test.  The 
 
86 
 
volume of the scour measured from these two masses was approximately one third of the total 
protrusion. 
In summary, the behavior of the Mooreville Chalk formation varied, but it was 
recognized that the formation could be scour resistant.  Testing was limited to the amount of 
testable sections recovered during sampling.  Once the sample was able to be automatically 
advanced in the EFA, the formation showed minimal scour rates with no tests recording scour 
rates greater than 1.0 mm/hr.  Finally, it was determined that the Mooreville Chalk formation did 
not scour uniformly in a particle by particle or even flake by flake fashion as observed in sands 
and earlier tested clays.  The scour observed in the Mooreville Chalk formation consisted of 
large mass chunks of the material scouring at once.  All EFA test data for the Mooreville Chalk 
formation can be viewed in Appendix D. 
4.5.3 Geotechnical Testing 
Two initial moisture contents were taken to obtain the insitu moisture content of the 
Mooreville Chalk formation.  These tests yielded moisture contents of 21.9 and 24.9 percent, 
with an average of 23.4 percent.  Using soil processed by following ASTM D421, a full grain 
size distribution was performed on the formation following ASTM D422.  The results of the 
grain size analysis can be viewed below in Figure 4-19.  The grain size analysis on the 
Mooreville Chalk showed that the formation consisted of very fine-grained particles with 
approximately 92 percent of the sample passing the number 200 sieve.  From the grain size 
analysis the mean particle diameter (D50) was calculated to be 0.024 mm.  The mean particle 
diameter and percent of the Mooreville Chalk sample passing the number 200 sieve mirrored that 
of the Demopolis Chalk formation.  Also after 24 hours in the Hydrometer both formations had 
at least thirty percent of the sample still in solution.      
 
87 
 
 
Figure 4-19.  Mooreville Chalk Formation Grain Size Distribution 
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Mooreville Chalk formation.  The liquid limit was determined to be approximately 
52, and the plastic limit was determined to be approximately 25, resulting in a plasticity index of 
27.  The calculated plasticity index is similar to that observed in the Bucatunna Clay formation.  
Using these values along with the information gathered from the grain size distribution, the 
sample was classified as a fat clay (USCS Classification CH).  
4.6 Prairie Bluff Chalk 
4.6.1 Sampling 
The Prairie Bluff Chalk formation was sampled on May 1, 2012 in Marengo County west 
of Demopolis, Alabama.  An ALDOT geologist classified the Prairie Bluff Chalk formation as a 
stiff gray chalk.  Outcrops of the formation were viewed in a stream bed beside the drilling 
location.  The ALDOT geologist identified the formation in drill cuttings and split spoon samples 
0
10
20
30
40
50
60
70
80
90
100
0.00010.0010.010.1110
%
 Pa
ssi
ng
Particle Diameter (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.0024 mm
 
88 
 
at approximately 3.5 m below the ground surface.  A Standard Penetration Test was performed at 
the top of the formation yielding an N value of 86 blows.  The ALDOT drill crew performed 
three CME continuous sample runs on the Prairie Bluff Chalk formation.  
The first sample run was taken from 4.57 to 6.10 m below the ground surface.  The top 
half of the sample consisted of a transitioning layer from the Standard Penetration Test.  The 
bottom half of the sample was uncracked and acceptable for EFA testing.  The second 
continuous sample run was taken from 6.10 to 7.62 m below the ground surface.  Similar to the 
previously sample chalk formations the Prairie Bluff Chalk was very stiff and difficult to sample.  
The second sample run was to be very difficult to advance, and upon recovery it was discovered 
that the acrylic sample tubes melted during sampling.  The sections of the tubes that did not melt 
were severely cracked and could not be used for EFA testing.  The third CME continuous sample 
run was taken from 7.62 to 9.14 m below the ground surface.  This run was also very difficult to 
advance as the sampler bowed out during drilling and became wedged the hollow stem auger.  
Once the sample was recovered, it was discovered that most of the sample tube was melted due 
to a plug that developed at the bottom of the drill hole.  Upon recommendation from the ALDOT 
drill crew, sampling of the Prairie Bluff Chalk was abandoned due to difficult sampling 
conditions.  A total of approximately 30 cm of the Prairie Bluff Chalk formation was collected 
for EFA testing.   
4.6.2 EFA Testing 
Seven Erosion Function Apparatus tests were performed on the Prairie Bluff Chalk 
formation to determine erosion rates.  Once the samples were brought back to Auburn University 
it was discovered that the amount of testable sample was very limited due to bowed sample 
tubes.  A 15 cm section of the sample was cut and prepared for testing.  Similar to the Mooreville 
 
89 
 
Chalk formation, the exterior of the sample tube had to be sanded down to fit into the flume of 
the EFA.  Once the sample was prepared, it was determined that the automatic motor on the EFA 
could not advance the sample into the flume.  Therefore, the sample was manually advanced 
using an extruder approximately 1 mm and placed into the EFA at the testing height.  The height 
of each protrusion could be calculated by the values provided from the ultrasonic sensor.   
The first test on the Prairie Bluff Chalk formation started at a flume velocity of 1.0 m/s.  
The protrusion into the flume was calculated to be 1.51 mm or fifty percent higher than the 
height specified by Briaud.  After fifty minutes a large mass chunk scoured away, the volume of 
the scoured section was approximately equal to five times the volume of the protrusion.  It was 
noted during testing that the sample seemed dry and loose in some areas, and that the large 
scoured volume started from the loose areas of the sample.  Another EFA test was performed at a 
flume velocity of 1.0 m/s, with a total protrusion of 1.85 mm.  No scour was observed during this 
test, suggesting that the flume velocity should be increased for future tests.  The idea that the test 
samples that had altered tubes and been manually advanced were exposed to differing stress 
states than initially intended during EFA testing was confirmed during testing of the Prairie Bluff 
Chalk formation.  It was thought that these samples were exposed to shear or torsional forces 
during extruding.  This idea was reinforced by cracking on the edges of the advanced sample as 
viewed in Figure 4-20 and Figure 4-21.  These shear and torsional forced could have been 
developed through an eccentricity between the bowed sample tube and the extruder advancing 
the sample.  It is unknown if these forces negatively influenced EFA testing.  It was noted that 
these cracks that developed in the EFA sample created shear planes for scouring to occur that 
would not naturally exist in a stream bed.   
 
90 
 
 
Figure 4-20.  Prairie Bluff Chalk Formation Cracking Example 1 
 
Figure 4-21.  Prairie Bluff Chalk Formation Cracking Example 2 
An EFA test was performed at a flume velocity of 1.5 meters per second, and a protrusion 
into the flume of 0.8 mm.  This test showed one millimeter of erosion after thirty-four minutes 
resulting in a scour rate of 1.76 mm/hr.  The scour started from dry loose areas noticed in the 
middle of the sample during sample preparation.  The next EFA test performed on the Prairie 
Bluff formation was performed at a flume velocity of 2.0 m/s.  The total protrusion into the 
flume from this test was calculated to be 1.33 mm.  After eighty minutes no scour was observed 
visually or ultrasonically. 
 
91 
 
Three tests were performed at a flume velocity of 3.0 m/s, and all three tests showed no 
measurable scour.  The protrusions from all three of these tests were greater than 1.0 mm and in 
one case the protrusion was double the specified height.  It was noted that these samples did not 
seem to be as dry and loose as samples noted above.  Also the cracks due to manually extruding 
these samples were smaller in size than those noticed during earlier tests that exhibited scour.  
Due to the lack of testable sample material, a repeat event test was not performed on the Prairie 
Bluff Chalk formation.  It was also recognized that the samples tested that did not seem to be 
disturbed during sampling or extruding exhibited a resistance to scour at flume velocities up to 
3.0 m/s.  All EFA tests performed on the Prairie Bluff Chalk formation can be viewed in 
Appendix E. 
4.6.3 Geotechnical Testing 
Three initial moisture contents were taken to obtain the insitu moisture content of the 
Prairie Bluff Chalk formation.  These tests provided moisture contents of 16, 17.1, and 20.1 
percent with an average of 17.7 percent.  As noted during EFA testing, these moisture contents 
were approximately five percent less than those observed from the other two chalk formations.  
Using soil processed by following ASTM D421, a full grain size distribution was performed on 
the formation following ASTM D422.  The results of the grain size analysis can be viewed 
below in Figure 4-22.  The grain size analysis on the Prairie Bluff Chalk showed that the 
formation consisted of fine-grained particles with approximately 82 percent of the sample 
passing the number 200 sieve.  From the grain size analysis the mean particle diameter (D50) was 
calculated to be 0.028 mm.  The mean particle diameter and percent of the Prairie Bluff Chalk 
sample passing the number 200 sieve were similar to the other two chalk formations tested.  The 
 
92 
 
fine sand evident from the grain size distribution explains some of the difficulty in preparing the 
EFA test samples with some loose portions.    
 
Figure 4-22.  Prairie Bluff Chalk Formation Grain Size Distribution 
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Prairie Bluff Chalk formation.  The liquid limit was determined to be 
approximately 32, and the plastic limit was determined to be approximately 19, resulting in a 
plasticity index of 13.  Using these values along with the information gathered from the grain 
size distribution, the sample was classified as a lean clay with sand (USCS Classification CL).   
4.7 Porter?s Creek Clay 
4.7.1 Sampling 
The Porter?s Creek Clay formation was sampled on May 1, 2012 in Sumter County, 
Alabama.  An ALDOT geologist classified the Porter?s Creek Clay formation as a stiff brown 
clay.  The ALDOT geologist identified the formation in drill cuttings and split spoon samples at 
approximately 2.5 m below the ground surface.  A Standard Penetration Test was performed at 
0
10
20
30
40
50
60
70
80
90
100
0.00010.0010.010.1110
%
 Pa
ssi
ng
Particle Diameter (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.0028 mm
 
93 
 
the top of the formation yielding an N value of 30 blows.  The ALDOT drill crew performed 
three CME continuous sample runs on the Porter?s Creek Clay formation. 
The first continuous sample run was performed between 3.05 and 4.57 m below the 
ground surface.  Unlike the previous chalk samples, the Porter?s Creek Clay did not have any 
trouble advancing the sampler.  Most of the sample run was unusable with cracks in the vertical 
and radial directions.  A testable section was located in the bottom of this run and was 
approximately 15 cm in length.  The second continuous sample run was performed between 4.57 
and 6.1 m below the ground surface.  A layer of sand was located in the top half of this run and 
was not used for EFA testing.  The bottom half of the run consisted of a 30 cm EFA testable 
section.  It was noticed during sampling that some cracks due to weathering separated the 
testable sections.  The third continuous sample run was performed between 6.1 and 7.62 m below 
the ground surface.  The top half of the sample was determined to be unusable for EFA testing 
due to heavy cracking and the presence of a wet seam.  The bottom half of the sample was 
uncracked and could be used for EFA testing.  The ALDOT geologist also noticed that the 
testable sections were divided by cracks representing weathered areas. 
4.7.2 EFA Testing 
A total of 20 Erosion Function Apparatus tests were performed on the Porter?s Creek 
Clay formation.  Three different test samples were cut and prepared for EFA testing ranging 
from 5.7 to 6.1 m below ground surface.  The first test performed was started at a flume velocity 
of 0.3 m/s and did not scour after one hour, and swelling similar to that of the Bucatunna Clay 
was observed.  The flume velocity was increased to 0.6 m/s for the next EFA test and the sample 
exhibited one millimeter of scour after 15 minutes, resulting in a scour rate of 4.0 mm/hr.  Two 
more EFA tests were performed at a flume velocity of resulting in scour rates of 3.33 and 4.62 
 
94 
 
mm/hr.  The average scour rate for the three EFA tests performed at a flume velocity of 0.6 m/s 
was 3.98 mm/hr.  The results of the EFA tests performed at a flume velocity of 0.6 meters per 
second are shown in Table 4-9 below. 
Table 4-9:  Porter?s Creek Clay Results at 0.6 m/s 
 
Three EFA tests were performed on the Porter?s Creek Clay formation at velocity of 1.0 
m/s.  During these tests, the advancement of the sample occurred one minute after the flume and 
ultrasonic sensor was started.  This time period was typically 3 to 5 minutes, allowing for a clear 
baseline to be created during testing.  However, significant surface roughening had occurred 
during prior tests before the sample was advanced.  The ultrasonic sensor and data reduction 
software only required one minute to set an adequate base line for readings.  The three tests 
performed at a flume velocity of 1.0 m/s resulted in an average scour rate of 9.17 mm/hr.  Table 
4-10 below shows the results of these three tests. 
Table 4-10:  Porter?s Creek Clay Results at 1.0 m/s 
 
 
The Porter?s Creek Clay formation was tested four times with a flume velocity of 1.5 m/s.  
However, one of these tests scoured immediately due to a weathered surface.  The results of 
these tests are shown in Table 4-11.   
Porter's Creek Clay 19.5_2 15 15 4.00 4.00
Porter's Creek Clay 19.5_3 18 18 3.33 3.33
Porter's Creek Clay 19.5_4 13 13 4.62 4.62
Scour Rate 
Visual 
(mm/hr)
Scour Rate 
Ultrasonic 
(mm/hr)
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Porter's Creek Clay 19.5_5 6 6 10.00 10.00
Porter's Creek Clay 19.5_7 6 6 10.00 10.00
Porter's Creek Clay 19.0_1 8 8 7.50 7.50
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
95 
 
Table 4-11:  Porter?s Creek Clay Results at 1.5 m/s 
 
The average scour rate of these tests was calculated to be 10.19 mm/hr.  The scour 
observed during these tests mirrored that of the Bucatunna Clay formation and not that of the 
three chalk formations previously tested.  The scour did not typically involve large mass chunks, 
but small flakes that consistently scoured away the protrusion into the flume over time.  This 
scour did occur rather quickly as elapsed times for scour at a flume velocity of 1.5 m/s was 5, 7, 
and 6 minutes respectively.   
Three EFA tests were also performed at a velocity of 2.0 m/s.  The mechanism for scour 
was similar to those tests performed at a velocity of 1.5 m/s.  The average scour rate for these 
three tests was 10.86 mm/hr.  The results of these tests are shown below in Figure 4-12.  The 
average scour rates for the EFA tests performed for flume velocities of 1.0, 1.5, and 2.0 m/s were 
similar with a difference of only 1.69 mm/hr. 
Table 4-12:  Porter?s Creek Clay Results at 2.0 m/s 
 
The final set of EFA tests, used to determine scour rates, performed on the Porter?s Creek 
Clay formation were executed at a flume velocity of 3.0 m/s.  The average scour rate on these 
three tests jumped to 15.67 mm/hr.  The results of these tests are shown below in Table 4-13. 
Porter's Creek Clay 19.0_2 5 5 12.00 12.00
Porter's Creek Clay 19.0_3 7 7 8.57 8.57
Porter's Creek Clay 19.0_5 6 6 10.00 10.00
*Porter's Creek Clay 19.0_4 Instant Scour (Weathered Sample)
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
Porter's Creek Clay 19.0_6 5 5 12.00 12.00
Porter's Creek Clay 19.0_7 7 7 8.57 8.57
Porter's Creek Clay 19.0_8 5 5 12.00 12.00
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
Scour Rate 
Visual 
(mm/hr)
 
96 
 
Table 4-13:  Porter?s Creek Clay Results at 3.0 m/s 
 
During testing of the Porter?s Creek Clay formation it was noted that weathered planes 
existed within the tested samples.  It was also noted that scour typically started on or around 
these weathered planes, which appeared to be planes of weakness with respect to scour 
resistance.  As a result of these weathered lines and planes in the prepared sample four EFA tests 
were abandoned immediately following advancement into the flume.  All four of these tests had 
large mass scouring occur at the location of the weathered planes immediately following the 
push.  The volume of scour was not measured by the ultrasonic sensor as scour occurred very 
quickly before a scan from the sensor could be performed.  Visually the volume of scour in these 
four instances was greater than the one millimeter protrusion into the flume. An image of these 
weathered lines and planes can be viewed below in Figure 4-23.  When the weathered lines were 
noticed in split spoon samples during sampling the ALDOT geologist confirmed that this 
weathering pattern was quite common in samples from the Porter?s Creek Clay formation. 
  
Porter's Creek Clay 19.0_9 5 5 12.00 12.00
Porter's Creek Clay 19.0_12 4 4 15.00 15.00
Porter's Creek Clay 19.0_14 3 3 20.00 20.00
Scour Rate 
Visual 
(mm/hr)
Sample:
Elapsed Time 
Ultrasonic 
(min)
Elapsed 
Time Visual 
(min)
Scour Rate 
Ultrasonic 
(mm/hr)
 
97 
 
 
Figure 4-23.  Porter?s Creek Clay Formation Scoured along Weathered Planes 
The last test performed on the Porter?s Creek Clay formation was intended to determine 
the threshold velocity, which scour first occurs.  From previous tests it was determined that the 
threshold velocity, or critical shear velocity, was between 0.3 and 0.6 m/s.  The velocity during 
the first test performed on the Porter?s Creek Clay formation typically ranged from 0.3 to 0.35 
m/s.  The velocity for the final test on the formation, titled ?Porter?s Creek Clay 18.75_1?, was 
started at 0.4 meters per second.  At first surface scouring began to occur, at a reduced rate of 
that observed in the samples tested at 0.6 m/s.  After about ten minutes small flakes began to 
scour away from the specimen, while the flume velocity ranged from 0.4 to 0.42 m/s.  This 
scouring was substantial enough for the threshold velocity to be established at 0.4 m/s.  It should 
be noted that the threshold velocity obtained is not an exact number.  Previous work at Auburn 
University involved the calibration of the flow meter located in the EFA.  This work showed at 
velocities less than 0.5 m/s the flow meter could have a margin of error of approximately ten 
percent (Mobley, 2009).  The total results for the testing of the Porter?s Creek Clay formation 
can be viewed in Figure 4-24, with the averaged results in Figure 4-25. 
 
98 
 
 
Figure 4-24.  EFA Test Results, Porter?s Creek Clay Formation 
 
Figure 4-25.  Porter?s Creek Clay Velocity versus Scour Rate 
0
5
10
15
20
25
0 0.5 1 1.5 2 2.5 3 3.5
Sc
ou
r Ra
te 
(m
m/
hr
)
Velocity (m/s)
0.6 m/s
1.0 m/s
1.5 m/s
2 m/s
3.0 m/s
0
2
4
6
8
10
12
14
16
18
0 0.5 1 1.5 2 2.5 3 3.5
Sc
ou
r R
ate
 (m
m/
hr
)
Velocity (m/s)
AVG
 
99 
 
 
In summary, the Porter?s Creek Clay formation behaved similarly to the Bucatunna Clay 
formation during EFA testing.  The major concern in testing the Porter?s Creek Clay formation 
was the scour behavior of the formation with respect to the weathered lines and planes within the 
formation.  As stated above these weathered lines created planes of weakness during EFA testing 
increasing the observed scour rates.  All EFA test results for the Porter?s Creek Clay formation 
can be viewed in Appendix F. 
4.7.3 Geotechnical Testing 
Three initial moisture contents were taken to obtain the insitu moisture content of the 
Porter?s Creek Clay formation.  These tests provided moisture contents of 36.7, 33.6, and 36.9 
percent with an average of 35.7 percent.  Using soil processed by following ASTM D421, a full 
grain size distribution was performed on the formation following ASTM D422.  The results of 
the grain size analysis can be viewed below in Figure 4-25.  The grain size analysis on the 
Porter?s Creek Clay showed that the formation consisted of fine-grained particles with 
approximately 90 percent of the sample passing the number 200 sieve.  From the grain size 
analysis the mean particle diameter (D50) was calculated to be 0.082 mm.  The percent of the 
Porter?s Creek Clay sample passing the number 200 sieve was similar to the three chalk 
formations that were tested.   
Atterberg limits were performed to complete the soil classification and geotechnical 
testing of the Porter?s Creek Clay formation.  The liquid limit was determined to be 
approximately 62, and the plastic limit was determined to be approximately 53, resulting in a 
plasticity index of 9.  Using these values along with the information gathered from the grain size 
distribution, the sample was classified as an elastic silt (USCS Classification MH).  
 
100 
 
 
Figure 4-26.  Porter?s Creek Clay Formation Grain Size Distribution 
4.8 Clayton Formation 
4.8.1 Sampling 
Three other clay formations were sampled to complete this study, however geotechnical 
and EFA testing was not performed on these formations.  The testing of these formations will 
continue in the future.  The Clayton Clay formation was sampled on June 21, 2012 in Barbour 
County, Alabama.  An ALDOT geologist classified the Clayton formation as a light brown clay.  
The Clayton formation was difficult to locate as the formation is traditionally thin in parts of 
southern Alabama.  Sampling the Clayton formation was made additionally difficult as the 
formation often underlies varying layers of sand.  The ALDOT geologist identified the formation 
in drill cuttings and split spoon samples at approximately 8.7 m below the ground surface.  A 
Standard Penetration Test was performed at the top of the formation yielding an N value of 23 
0
10
20
30
40
50
60
70
80
90
100
0.0010.010.1110
% 
Pa
ssi
ng
Particle Diameter (mm)
No. 10
2 mm
No. 200
0.075 mm
HydrometerFine SieveCoarse Sieve
D50: 0.0082 mm
 
101 
 
blows.  After the Standard Penetration Test was performed on the Clayton formation, the drill 
hole caved in due to sand in the upper part of the hole.  The drill crew started another drill hole 
approximately 5 maway from the first and the formation was located approximately 7.3 m below 
the ground surface.  The ALDOT drill crew performed three CME continuous sample runs on the 
Clayton formation. 
The first of the continuous sample runs was performed between 7.32 and 8.84 meters 
below the ground surface.  The top acrylic sample tube in this run was completely empty with no 
recovery.  The bottom half of the sample had a small testable section about 30 cm in length.  The 
second continuous sample run was performed between 8.84 and 10.36 m below the ground 
surface.  This sample run included many small testable sections between cracks in the formation, 
but none of the sections were greater than 15 cm in length.  The third continuous sample run was 
performed between 10.36 and 11.89 m below the ground surface.  The sample tube hit the top of 
bedrock during sample resulting in very little recovered sample.  Among this recovered sample 
there were no EFA testable sections.    
4.9 Nanafalia Clay Formation 
4.9.1 Sampling 
The Nanafalia Clay formation was sampled on June 6, 2012 in Coffee County, Alabama.  
An ALDOT geologist classified the Nanafalia Clay formation as a plastic brown clay.  The 
ALDOT geologist identified the formation in drill cuttings and split spoon samples at 
approximately 2.5 m below the ground surface.  A Standard Penetration Test was performed at 
the top of the formation yielding an N value of 13 blows.  The ALDOT drill crew performed 
three CME continuous sample runs on the Nanafalia Clay formation.   
 
102 
 
The first CME sampler run was performed from 2.74 to 4.26 m below the ground surface.  
The top half of the sample was not usable for EFA testing and included some transition from the 
Standard Penetration Test.  A testable section was located in the top 15 centimeters of the bottom 
acrylic sampling tube.  The second continuous sample run was performed from 4.26 to 5.79 m 
below the ground surface.  Most of the top half of the sample was not recovered, and some small 
usable sections were located in the bottom half of the sample.  The third and final continuous 
sample run on the Nanafalia Clay formation was performed from 5.79 to 7.31 m below the 
ground surface.   The top half of this sample run was mostly unusable for EFA testing, but could 
be used for future geotechnical testing.  The bottom acrylic sampling tube was almost entirely 
uncracked, and could be used for EFA testing.  The bottom half of this continuous sample was 
the first sample obtained that could have an unconfined compression test performed without the 
presence of multiple cracks. 
4.10 Naheola Clay Formation 
4.10.1 Sampling 
The Naheola Clay formation was sampled on June 7, 2012 in Marengo County, Alabama.  
An ALDOT geologist classified the Naheola Clay formation as a gray clay.  The ALDOT 
geologist identified the formation in drill cuttings and split spoon samples at approximately 3.9 
m below the ground surface.  A Standard Penetration Test was performed at the top of the 
formation yielding an N value of 16 blows.  The ALDOT drill crew performed three CME 
continuous sample runs on the Naheola Clay formation.   
The first continuous sample run on the Naheola Clay formation as performed between 
3.96 and 5.49 m below the ground surface.  The top half of the sample was disturbed by the 
Standard Penetration test.  The bottom half of the sample was mostly usable for EFA testing.  
 
103 
 
The color of the formation changed from gray to dark yellow in the bottom half of the sample.  
The ALDOT geologist observed the sample and determined that the colored change was not 
unordinary for the Naheola Clay formation.  The second CME sample run was performed 
between 5.49 and 7.01 m below the ground surface.  The color of the formation continued to 
change intermittingly between gray and yellow in the second sample.  The top half of the sample 
run was severely cracked and not able to be used for EFA testing.  The bottom half of the sample 
contained approximately 40 cm of EFA testable sample.  The third continuous sample run was 
performed from 7.01 to 8.53 m below the ground surface.  The top half of the sample did not 
recover any of the Naheola Clay formation.  The bottom half of sample contained several small 
rock segments.  A 25 centimeter testable section was located at the bottom of the sample.  With 
approximately 80 cm of uncracked Naheola Clay formation collected, unconfined compression 
tests can be performed on the formation in the future.  
 
104 
 
Chapter 5 Results Discussion 
5.1 Sampling Observations 
As stated in earlier chapters, nine clay and chalk formations from the coastal plains in 
southern Alabama were sampled using the Central Mining Equipment continuous sample 
tube system.  The quality and consistency of the acquired samples varied between 
formations.  Ideally the samples gathered would serve three purposes and be used for EFA 
testing, shear strength testing, and geotechnical index testing.   
All samples were successfully used for EFA testing.  However, the clay formations 
tested in the EFA were easier to test than the chalk formations.  It appeared the very stiff 
chalk formations created a high amount of skin friction between the sample and the acrylic 
tube.  At times this skin friction was too large for the sample to be automatically advanced 
using the stepping motor on the Erosion Function Apparatus.  In these cases EFA testing was 
not ideal because the protrusion into the flume not consistent with the calibrated one 
millimeter protrusion set forth in Briaud?s (1999) procedure.  The clay samples acquired, the 
Bucatunna Clay, Yazoo Clay, and Porter?s Creek Clay formations, were able to be EFA 
tested without any difficulties. 
The samples acquired, using the CME continuous sampler, were successfully used for 
geotechnical index testing.  Samples were extruded from the sample tubes and processed for 
testing.  However, the samples acquired were not conducive for shear strength testing.  Given 
the cohesive material used for this study, the unconfined compression test is ideal for shear 
strength testing.  With the diameter of the continuous sample being 5.7 cm, an acceptable 
 
105 
 
unconfined compression sample would need to be at least 11.4 cm.  Unfortunately the 
sampling method did not yield uncracked samples with a minimum length of 114 mm.  
Therefore, it was not possible for an unconfined compression test to be performed on any of 
the acquired samples.   
During sampling it was noted that most, if not all, of the uncracked sections 
acceptable for use in EFA testing were located in the bottom half of the acrylic sampling 
tubes.  It is thought that this was due to friction being created between the cohesive surface of 
the sample and the acrylic sample tube as the sampler was advanced during sampling.  It 
appeared that after a certain length the sampled material began to fail in friction and crack 
vertically within the sampling tube.  This would explain the bottom of the samples being 
uncracked, as this segment of sample would be exposed to small or negligible frictional 
forces along the edges of the sample.  This cracking was nominally improved by spraying a 
lubricant inside the sample tubes, although the effectiveness of the lubricant varied 
depending on the stiffness of the formation being sampled.  Another issue associated with the 
cracking of sampled material is that the bearing assembly was not always successful in 
keeping the sample tube isolated from the rotation of the drill rig.  This could expose the 
sample to torsional and shear forces during sample, which would be evident by the cracks in 
the recovered sample.  This also could compromise the undisturbed state of the sample 
specimen.  It is important to note this was not always the case and was only observed when 
sampling very stiff formations. 
Another issue observed during sampling of the chalk formations was the bowing of 
the sample tube assembly.  In these cases it appeared as if the acrylic sampling tubes were 
heated, due to the friction of material, allowing the tube to yield and expand radially.  This 
 
106 
 
too could question the undisturbed state of the obtained samples.  Also the bowing of the 
sample tube made it difficult to test these samples in the EFA, as the opening in the flume of 
the EFA is equal to the diameter of the sample tube.      
5.2 EFA Testing Observations 
Six of the sampled formations were tested in the EFA to determine scour parameters.  
Of the six formations tested, three were clay formations and three were chalk formations.  
The trends in EFA testing appear to illustrate a clear difference in the scour behavior of clay 
and chalk formations.  The three clay formations tested were the Bucatunna Clay formation, 
the Yazoo Clay formation, and the Porter?s Creek Clay formation.   
The Bucatunna Clay formation scoured with scour rates ranging from 4 to 11 mm/hr 
after scour initially occurred at a flume velocity of 0.45 m/s.  The Bucatunna Clay formation 
was easy to prepare for EFA tests and was easily advanced into the flume.  Scour typically 
occurred in small flakes resulting in nearly linear erosion, until the entire protrusion was 
eroded.  At velocities close to or lower than the threshold velocity, the formation appeared to 
swell during testing.  At times the magnitude of this swelling was approximately five times 
the initial protrusion.  Swelling did not occur at velocities above the threshold velocity 
because the formation was not able to absorb moisture and swell while scouring.   
 The Yazoo Clay formation also scoured with highly variable scour rates which caused 
doubt in establishing a threshold velocity.  The Yazoo Clay proved difficult to prepare as 
larger particle sizes made it difficult to create a level surface.  EFA testing was eventually 
suspended due to high variability.  Scour rates observed in the Yazoo Clay formation were 
higher at a flume velocity of 1.0 m/s than at a flume velocity of 1.5 m/s.  Likewise, two 
consecutive tests completely failed before the ultrasonic sensor was even allowed to start.  It 
 
107 
 
appeared that the sample was inconsistent with respect to depth and contained a high sand 
content.  The sand content was excessive in this sample, upwards of fifty-five percent, which 
is uncharacteristic of the formation. 
The Porter?s Creek Clay formation was also scoured at the tested velocities.  The 
scour rates for the Porter?s Creek Clay formation ranged from 4 to 15 mm/hr.  The Porter?s 
Creek Clay formation behaved similarly to the Bucatunna Clay during EFA testing.  The 
Porter?s Creek Clay formation was also similar to the Bucatunna Clay formation in regards to 
ease of sample preparation.  The sample generally scoured in a flake by flake pattern with 
some larger chucks eroding at times.  The Porter?s Creek Clay formation sample contained 
many weathered lines and planes.  Scour was observed along these planes, and these planes 
became shear planes during testing. 
The observations presented above are consistent with the results presented by Crim 
(2003) and Mobley (2009).  Both Crim and Mobley tested samples classified as silts and 
clays.  Crim (2003) and Mobley (2009) observed scour rates ranging from 0 mm/hour 
upwards of 100 mm/hour.  Typical erosion rates observed by Crim and Mobley were 
approximately 20 mm/hr.  These results are similar to those observed in the Bucatunna and 
Porter?s Creek formations. 
The three chalk formations tested were the Demopolis Chalk, the Mooreville Chalk, 
and the Prairie Bluff Chalk.  All three of these formations were scour resistant according to 
the testing regiment established in this study.  Mobley (2009) also tested the Mooreville 
Chalk formation in the EFA.  Mobley observed that the formation was resistant to scour in 
velocities upwards of 6 m/s. 
 
108 
 
However, the Demopolis Chalk, Mooreville Chalk, and Prairie Bluff Chalk samples 
were difficult to test as all three samples had to be manually extruded during EFA testing.  
This manual extrusion resulted in a loss in calibration of the protrusion height according to 
the procedure set forth by Briaud.  Typically the manual protrusion height was greater than 
one millimeter meaning that the shear stress acting on the sample was actually greater than 
the intended shear stress. Although these samples were exposed to higher shear stresses scour 
was not observed.  After the samples were manually extruded small cracks were noticed on 
the surface of the samples.  These cracks were attributed to possible shear and torsional 
forces exerted on the sample during sampling and extruding.  Although these cracks would 
eventually affect the scour performance of the formations, scour was not observed in any of 
the three chalk formations.  These formations were also difficult to level during sample 
preparation due to the stiffness of the chalk formations. 
5.3 Geotechnical Testing Observations 
The geotechnical testing component of this study helped reinforce the observations 
made during EFA testing.  Similarly the geotechnical testing program assigned additional 
properties to the formations tested in the EFA, to find trends between geotechnical and scour 
parameters.  One example showing the validity of the geotechnical testing program in 
providing explanation for observations during EFA testing is the Bucatunna Clay formation.  
As previously stated, the Bucatunna Clay formation swelled during EFA testing.  This 
swelling behavior was noted and quantified during EFA testing.  During geotechnical testing 
this observation was validated as the formation had a plasticity index of 29 which is well 
above the limit for a swelling clay. 
 
109 
 
Another example of the results in geotechnical testing providing validity to 
observations made in EFA testing was the Yazoo Clay formation.  The Yazoo Clay 
formation was highly variable with large particles noticed during some EFA tests.  The 
geotechnical testing it was determined the sample acquired did not classify as a clay, but as a 
silty sand.  This evidence along with the variability and observations obtained during EFA 
testing hint that the obtained sample was not a pure sample of the Yazoo Clay formation, and 
the results obtained were not indicative of the entire formation.   
Similarly the geotechnical testing provided some clarity in the testing of the three 
chalk formations.  The Demopolis Chalk, Mooreville Chalk, and Prairie Bluff Chalk 
formations were all determined to be scour resistant during EFA testing.  These formations 
also had similar geotechnical index properties, which explained some of the similarity in 
EFA testing.  All three formations had similar mean particle diameters within a range of 
0.007 millimeters.  Also the percent passing the number 200 sieve was within 15 percent for 
the three formations with a minimum of 82 percent and a maximum of 97 percent. 
5.4 Correlations in Testing 
Several general scour trends were noticed while correlating the scour results to the 
geotechnical index properties obtained for the six formations tested.  Generally speaking, 
scour-resistant chalk formations had higher Standard Penetration Test N values, lower insitu 
moisture contents, higher percentage of percent passing the number 200 sieve, and a smaller 
mean particle diameter.  These trends correlate to previous research that states the critical 
shear stress increases when plasticity index increases, when undrained shear strength 
increases, and with the percent passing the number 200 sieve (Briaud 2001b).  The trend 
stating that plasticity index increases as the critical shear strength increases was not affirmed 
 
110 
 
in this study.  There were no correlations made between scour parameters and plasticity 
index, as the plasticity index for all six formations was highly variable.  The Standard 
Penetration Test N value can roughly represent the undrained shear strength. Briaud (2001b) 
stated that the critical shear stress increases with the shear strength of the soil.  This theory 
was confirmed in this study as the formations with higher SPT N values were scour resistant. 
Although the samples tested hinted that the insitu moisture content influences the critical 
shear stress and scour rate of a formation, this value is highly variable both between different 
formations and within one formation.   
Both the evidence of this study and prior research tend to show that the percentage of 
particles passing the number 200 sieve influences the critical shear stress of the formation.  
The three chalk formations that were proven to be scour resistant all had a high percentage of 
particles passing the number 200 sieve.  The Bucatunna Clay and Yazoo Clay formations, 
which both showed scour at the tested velocities, had lower percentages of particles passing 
the number 200 sieve than the three scour resistant formations.  The outlier by this measure 
was the Porter?s Creek Clay, which showed scour at the tested velocities, but had 90 percent 
passing the number 200 sieve.  An explanation for the Porter?s Creek Clay scouring with a 
high percentage passing the number 200 sieve could be the weathered planes that were 
observed during EFA testing.  It was noted that scour routinely occurred along these planes, 
which could explain why the formation does not abide by this correlation.  Previous research 
stated that the rate of scour was increased when the percentage of clay particles decreased 
(Briaud, 2001b).  This trend is indirectly similar to the trend stated above that the critical 
shear stress increases with the percentage passing the number 200 sieve.  Again all the 
formations tested except the Porter?s Creek Clay formation follow this qualitative trend.   
 
111 
 
An observation was made with the scour test results and the geotechnical index 
parameters using the moisture content and SPT N value.  In Figure 5-1 below, a horizontal 
and vertical line are drawn, and approximately intersect at a moisture content of 23 percent 
and an SPT N value of 60.  All of the formations falling down and to the right of this 
intersection, the Bucatunna Clay formation, the Yazoo Clay formation, and the Porter?s 
Creek Clay formation all scoured during testing.  The formations up and to the left of the 
intersection, the Demopolis Chalk formation, the Mooreville Chalk formation, and the Prairie 
Bluff Chalk formation were scour resistant.  This observation combined all three aspects of 
this study in sampling, EFA testing, and geotechnical testing.  The SPT N value was obtained 
during sampling, the scour data was obtained during EFA testing, and the moisture content 
was collected during geotechnical testing.  Additionally, one data point from Mobley (2009) 
was added to this data set and also follows this observation. 
 
Figure 5-1.  Scour Observation using Moisture Content and SPT N Value 
 
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80
SP
T N
 Va
lue
Moisture Content
Bucatunna Clay
Yazoo Clay
Demopolis Chalk
Moreville Chalk
Prairie Bluff Chalk
Porter's Creek Clay
Mobley (2009)
ScourResistant
Scourable
 
112 
 
Similarly, an observation was also made between SPT N Value and the percent of 
particles passing the number 200 sieve.  Figure 5-2 shows this observation, as all of the 
formations falling below the line scoured during testing and all the formations plotted above 
the line were scour resistant.  Similar to Figure 5-1, data from Mobley (2009) was added to 
this plot.  Mobley (2009) observed scour in this sample with a SPT N value of 39 and 98 
percent of the sample passing the #200 sieve, which follows the observation in Figure 5-2.  
This observation between scour parameters, SPT N value, and grain size is further described 
by Figure 5-3.   
 
Figure 5-2.  Scour Observation using Percent Passing #200 and SPT N Value 
Figure 5-3 shows the mean particle size (D50) on a logarithmic scale plotted with the 
SPT N value.  In Figure 5-3, a horizontal and vertical line are placed approximately 
intersecting at a mean grain size of 0.0082 mm and an SPT N value of 60.  All the formations 
plotted above and to the left of the intersection line were scour resistant and the formations 
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80 90 100
SP
T N
 Va
lue
Percent Passing #200
Bucatunna Clay
Yazoo Clay
Demopolis Chalk
Moreville Chalk
Prairie Bluff Chalk
Porter's Creek Clay
Mobley (2009)
ScourResistant
Scourable
 
113 
 
plotted below and to the right of the intersection scoured during EFA testing.  Six additional 
samples tested by Crim (2003) are plotted on Figure 5-3 and follow the same observation, as 
all six samples fall below and to the right of the intersection and exhibited scour behavior 
during testing. 
 
Figure 5-3.  Scour Observation using Mean Grain Size and SPT N Value 
 
In summary, all six of the formations tested proved to be unique.  Even the chalk 
formations that generally performed the same in geotechnical and EFA testing, were each 
unique.  This uniqueness made creating and proving quantitative correlations between scour 
behavior and geotechnical behavior more difficult. 
 
0
10
20
30
40
50
60
70
80
90
100
0.001 0.01 0.1 1
SP
T N
 Va
lue
Mean Grain Size (mm)
Bucatunna Clay
Yazoo Clay
Demopolis Chalk
Moreville Chalk
Prairie Bluff Chalk
Porter's Creek Clay
Crim (2003)
ScourResistant
Scourable
 
114 
 
Chapter 6 Summary, Conclusions, and Recommendations 
6.1 Summary 
The ability to determine scour parameters in order to accurately predict scour depth is 
imperative to designing safe, economic, and efficient bridge foundations.  As previously 
stated, scour behavior is well understood for cohesionless soils; however, much research has 
been performed in an effort to better understand scour behavior of cohesive soils.  In this 
study the research objectives were to adapt the EFA to measure scour rates automatically 
with the aid of an ultrasonic sensor, create an EFA testing procedure and regimen that 
incorporates the updated sensor, and conduct EFA tests on cohesive soil samples that were 
collected from the coastal plain of Alabama.  Additionally, common geotechnical parameters 
such as Atterberg limits and grain size distribution were measured to determine if any 
correlations exists with scour parameters.   
The Erosion Function Apparatus was updated with a 16 transducer ultrasonic sensor.  
Using the ultrasonic sensor and the updated EFA, a new testing regimen was created for 
determining scour parameters of cohesive soils.  To obtain these samples, cooperation with 
the Alabama Department of Transportation was necessary.  With the advice of ALDOT 
drilling coordinators, the sampling method used for this research was the Central Mining 
Equipment continuous sample tube system.  ALDOT collected a total of nine formations, 
using the continuous sample tube system, in the coastal plains of Alabama.  Six of the 
collected formations were tested in the EFA with a wide variety of results.  These formations 
were then tested for traditional geotechnical index properties including a full grain size 
 
115 
 
distribution and Atterberg limits.  Using the results from the EFA and geotechnical testing, 
qualitative correlations were determined.  
The six formations sampled, erosion tested, and index tested include the Bucatunna 
Clay, Yazoo Clay, Demopolis Chalk, Mooreville Chalk, Prairie Bluff Chalk, and Porter?s 
Creek Clay formations.  The Demopolis Chalk, Mooreville Chalk, and Prairie Bluff Chalk 
formations were all scour resistant according to the testing program previously established.  
The Bucatunna Clay, Yazoo Clay, and Porter?s Creek Clay formations all showed signs of 
measurable scour during testing.  EFA testing of the Yazoo Clay formation was suspended 
due to high variability and inconsistencies in testing.  Weathered planes in the Porter?s Creek 
Clay formation heavily altered the EFA testing of the formation.   
6.2 Conclusions 
Determining correlations between erosion testing and geotechnical testing is difficult 
given the limited number of tests performed in this study.  At best the correlations 
determined in this study should be expounded upon in future work.  Figure 5-2 and Figure 5-
3 indirectly validate the correlation between critical shear stress and particle size presented 
by Briaud (1999).  Similarly, Figure 5-2 and Figure 5-3 confirm a correlation between scour 
rates and particle size.  Figure 5-1 shows an observation between moisture content and SPT 
N values.  EFA testing is highly variable and unique to each individual formation.  This 
makes it increasingly difficult to quantify the results with standardized geotechnical 
properties.  Overall the qualitative correlations confirmed or created throughout this study 
should be further studied to create more quantitative correlations.  A summary of the research 
results is provided below in Table 6-1.  Table 6-1 allows for easy correlations between test 
results and visual observations.  The ?Pushable? column in Table 6-1 states if a sample was 
 
116 
 
able to be automatically advanced in the EFA.  The ?Swelling Witnessed? column in Table 
6-1 notates if swelling was witnessed during EFA testing.  Table 6-1 illustrates all formations 
that required manual advancement into the EFA did not scour.  Similarly all formations that 
swelled during testing exhibited scour.  From the geotechnical index testing, formations 
containing a SPT N value greater than 60, a percent of particles passing the #200 sieve 
greater than 80, and insitu moisture content less than 25, and a mean particle size less than 
0.003 mm will not exhibit scour.  Again these correlations are qualitative but can be used to 
compare future results and observations. 
Table 6-1:  Summary of Results and Observations 
 
 
6.3 Recommendations 
Generally, the means and methods used in this study to collect and test EFA samples 
were successful.  However, below are a few recommendations meant to improve the 
interpretation of test results.  The three formations sampled and not tested?the Clayton 
formation, the Nanafalia Clay formation, and the Naheola Clay formation?should all be 
tested in the EFA.  Also these three formations should be tested for geotechnical index 
properties to determine if the data obtained continues to follow the same correlations listed 
above.  The Yazoo Clay formation should be resampled and tested to ensure that the correct 
Formation: Scourable LL PL PI D50 (mm)
% 
Passing 
#200
USCS 
Classification
SPT N 
Value
Average 
Moisture Pushable
Swelling 
Witnessed
Bucatunna 
Clay Yes 68 39 29 0.033 64
MH (Sandy 
Elastic Silt) 9 47.9 Yes Yes
Yazoo Clay Yes 57 NP NP 0.088 44 SM (Non-Plastic Silty Sand) 15 59.9 Yes No
Demopolis 
Chalk No 37 27 10 0.0021 97 CL (Lean Clay) 92 21.8 No No
Mooreville 
Chalk No 52 25 27 0.0024 92 CH (Fat Clay) 60 23.4 Yes/No No
Prairie Bluff 
Chalk No 32 19 13 0.0028 82
CL (Lean Clay 
with Sand) 86 17.7 No No
Porter's 
Creek Clay Yes 62 53 9 0.0082 90 MH (Elastic Silt) 30 35.7 Yes Yes
 
117 
 
sample is gathered.  The Yazoo Clay sample acquired should classify as a clay with more 
than fifty percent of the sample passing a number 200 sieve.   
Chalk formations should be sampled with rock cores.  The chalk samples were unable 
to be automatically advanced from the continuous sample tubes do to the stiffness of the 
samples.  Chalk formations can be tested in the EFA by using traditional rock cores, without 
sacrificing resolution from the ultrasonic sensor.  Future sampling of clay formations should 
include at least one Shelby tube sample, if possible, to accompany the continuous samples.  
These changes to the sampling procedure ensure every formation sampled would have a 
sufficient sample for shear strength testing.  An unconfined compression test can be 
performed on a prepared rock core sample to determine the undrained shear strength.  
Likewise an unconfined compression test can be performed from the Shelby tube sample 
gathered from clay formations.  This simple change in sampling technique allows future 
studies to obtain shear strength parameters for each of the tested formations which would add 
validity to correlations made with strength parameters.  By obtaining shear strength 
parameters, the critical shear stress of the soil from EFA testing can be compared to the 
ultimate shear stress in compression. 
The testing program should be expanded to include EFA tests with a longer duration 
and a higher velocity.  The testing regiment should be expanded to velocities upwards of 6.0 
meters per second in an attempt to determine the critical shear stress for the stiffer 
formations.  Similarly a longer duration test should be introduced to determine if scour 
occurs at a lower velocity after multiple hours of exposure to relatively low shear stresses.  
This longer duration test can be altered to fit the format of the multiple events test performed 
on the Demopolis and Mooreville Chalk formations.   
 
118 
 
Finally, the testing program should be expanded beyond the nine formations sampled.  
This additional testing would create more data from which correlations can be drawn, with 
greater accuracy.  This data can be compared with correlations initially created by Briaud to 
statistically advance the legitimacy of any quantitative correlations.  In conclusion, to 
completely understand the scour behavior at an individual stream crossing, an extensive 
unique EFA study should be performed.  However, with extensive EFA testing, using the 
ultrasonic sensor for higher resolution, there is potential for creating statistically adequate 
correlations for estimating scour behavior of a specific formation. 
    
 
119 
 
References 
Briaud, J.L., Ting, F.C.K., Chen, H.C., Gudavalli, R., Peregu, S., and Wei, G. (1999). 
?SRICOS: prediction of scour rate in cohesive soils at bridge piers,? Journal of 
Geotechnical and Geoenvironmental Engineering, Vol. 125, No.4, April, pp. 237-246, 
American Society of Civil Engineers, Reston, Virginia, USA. 
 
Briaud, J.L., Chen, H.C., Kwak, K.W., Han, S.W., and Ting, F.C.K. (2001a).  ?Multiflood 
and multilayer method for scour rate prediction at bridge piers,? Journal of Geotechnical 
and Geoenvironmental Engineering, Vol. 127, No.2, February, pp. 114-125, American 
Society of Civil Engineers, Reston, Virginia, USA. 
 
Briaud, J.L., Ting, F.C.K., Chen, H.C., Cao, Y., Han, S.W., and Kwak, K.W. (2001b). 
?Erosion function apparatus for scour rate predictions,? Journal of Geotechnical and 
Geoenvironmental Engineering, Vol. 127, No. 2, February, pp. 105-113, American 
Society of Civil Engineers, Reston, Virginia, USA. 
 
Briaud, J.L., Chen, H.C., Li, Y., Nurtjahyo, P., and Wang, J. (2004). ?Pier and contraction 
scour in cohesive Soils,? NCHRP Report 516, Transportation Research Board, 
Washington, D.C. 
 
Central Mining Equipment Company (2012). Central Mining Equipment Soil Sampling 
Catalogue, St. Louis, Missouri. 
 
Crim Jr. S. (2003).  ?Erosion functions of cohesive soils,? M.S. Thesis, Draughon Library, 
Auburn University, Auburn, Alabama. 
 
Jette, C. (2010).  ?SeaTek 5 MHz ultrasonic ranging system,? SeaTek Instrumentation and 
Engineering, Gainesville, Florida. 
 
Lagasse, P., Clopper, P., Zevenbergen, L. and Girard, L., (2007)., ?Countermeasures to 
protect bridge piers from scour,?: NCHRP Report 593, Transportation Research Board, 
National Academy of Science, Washington, D.C. 
 
Lagasse, P., Schall, J., and Richardson, E., (2001). ?Evaluating scour at bridges,? FHWA 
NHI 01-002 HEC-20, U.S. Federal Highway Administration, Washington, D.C. 
 
120 
 
 
Mobley, T. (2009). ?Erodibility testing of cohesive soils,? M.S. Thesis, Draughon Library, 
Auburn University, Auburn, Alabama. 
 
Murillo, J.A. (1987). ?The scourge of scour,? Civil Engineering, Vol. 57, No.7, July, pp. 66-
69, ASCE, Reston, Virginia, USA. 
 
Navarro, H. (2004). ?Flume measurements of erosion characteristics of soils at bridge 
foundations in Georgia,? M.S. Thesis, Georgia Institute of Technology, Atlanta, Georgia 
 
Richardson, E. and Davis, S., (2001). ?Evaluating scour at bridges,? FHWA NHI 01-001 
HEC-18, U.S. Federal Highway Administration, Washington, D.C. 
 
Sturm, T., Sotiropoulos, F. , Landers, M., Gotvald, T., Lee, S., Ge, L., Navarro, R., and 
Escauriaza, C., (2004). ?Laboratory and 3D numerical modeling with field monitoring of 
regional bridge scour in Georgia,? Final Report , GADOT Research Report No. FHWA-
GA-04-2002. 
 
Sheppard,D., Bloomquist, D., Henderson, M., Kerr, K., Trammell, M., Marin, J., and Slagle, 
P., (2005). ?Design and construction of apparatus for measuring scour rate of water 
erosion of sediments,? Final Report FDOT Project BC354 RPWO #12. 
 
Sheppard,D., Bloomquist, D., and Slagle, P., (2006). ?Rate erosion properties of rock and 
clay,? Final Report FDOT Project BD545 RPWO #3. 
 
Wardhana, K. and Hadipriono, F.C. (2003). ?Analysis of recent bridge failures in the United 
States,? Journal of Performance of Constructed Facilities, 17(3), August, pp. 144-150, 
American Society of Civil Engineers, Reston, Virginia. 
 
121 
 
APPENDIX A 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
BUCATUNNA CLAY FORMATION 
 
TEST VELOCITY: 0.6 METERS PER SECOND 
 
Figure A-1.  ?Bucatunna 27.0_3? Test Results 
-1
-0.5
0
0.5
1
1.5
2
0 5 10 15 20 25 30
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.0_3
Veloctiy: 0.6 (m/s)
Scour Rate: 4.29 (mm/hr)
 
122 
 
 
Figure A-2. ?Bucatunna 27.0_4? Test Results 
 
Figure A-3. ?Bucatunna 27.0_5? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation:  Bucatunna Clay
Test: Bucatunna 27.0_4
Velocity: 0.6 (m/s)
Scour Rate: 3.53 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatanna 27.0_5
Velocity: 0.6 (m/s)
Scour Rate: 5.00 (mm/hr)
 
123 
 
 
Figure A-4. ?Bucatunna 27.0_6? Test Results 
 
Figure A-5. ?Bucatunna 27.0_7? Test Results 
 
 
-1
-0.5
0
0.5
1
1.5
2
0 5 10 15 20 25 30 35
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.0_6
Velocity: 0.6 (m/s)
Scour Rate: 2.73 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 2 4 6 8 10 12 14 16 18 20
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.0_7
Velocity: 0.6 (m/s)
Scour Rate: 4.00 (mm/hr)
 
124 
 
TEST VELOCITY: 1.0 METERS PER SECOND 
 
 
Figure A-6. ?Bucatunna 27.5_2? Test Results 
 
Figure A-7. ?Bucatunna 27.5_3? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_2
Velocity: 1.0 (m/s)
Scour Rate: 4.62 (mm/hr)
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0 2 4 6 8 10 12 14 16
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_3
Velocity: 1.0 (m/s)
Scour Rate: 8.57 (mm/hr)
 
125 
 
 
Figure A-8.  ?Bucatunna 27.5_4? Test Results 
 
Figure A-9. ?Bucatunna 27.5_5? Test Results 
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25 30 35 40 45 50
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_4
Velocity 1.0 (m/s)
Scour Rate: 1.58 (mm/hr)
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_5
Velocity: 1.0 (m/s)
Scour Rate: 4.0 (mm/hr)
 
126 
 
 
Figure A-10.  ?Bucatunna Clay 27.5_6? Test Results 
 
Figure A-11. ?Bucatunna 27.5_7? Test Results 
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25
Ave
rag
e H
eigh
t C
ha
nge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna Clay 27.5_6
Velocity: 1.0 (m/s)
Scour Rate: 3.53 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25 30
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_7
Velocity: 1.0 (m/s)
Scour Rate: 3.33 (mm/hr)
 
127 
 
 
Figure A-12. ?Bucatunna Clay 27.5_8? Test Results 
 
TEST VELOCITY: 1.5 METERS PER SECOND 
 
 
Figure A-13.  ?Bucatunna Clay 27.5_9? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25 30 35
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_8
Velocity: 1.0 (m/s)
Scour Rate: 2.86 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25 30
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 27.5_9
Velocity: 1.5 (m/s)
Scour Rate: 4.29
 
128 
 
 
Figure A-14. ?Bucatunna 26.5_2? Test Results 
 
Figure A-15. ?Bucatunna 26.5_3? Test Results 
-2
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14 16 18 20
Ave
rag
e H
eigh
t C
ha
nge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_2
Velocity: 1.5 (m/s)
Scour Rate: 5.00 (mm/hr)
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 2 4 6 8 10 12 14
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_3
Velocity: 1.5 (m/s)
Scour Rate: 8.57 (mm/hr)
 
129 
 
 
Figure A-16.  ?Bucatunna 26.5_4? Test Results 
 
Figure A-17. ?Bucatunna 26.5_5? Test Results 
 
 
 
 
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 5 10 15 20 25
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_4
Velocity: 1.5 (m/s)
Scour Rate: 4.62 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_5
Velocity: 1.5 (m/s)
Scour Rate: 5.45 (mm/hr)
 
130 
 
TEST VELOCITY: 2.0 METERS PER SECOND 
 
 
Figure A-18. ?Bucatunna 26.5_6? Test Results 
 
Figure A-19. ?Bucatunna 26.5_7? Test Results 
-2
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14 16 18
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_6
Velocity: 2.0 (m/s)
Scour Rate: 6.0 (mm/hr)
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 5 10 15 20 25 30 35
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_7
Velocity: 2.0 (m/s)
Scour Rate: 2.73 (mm/hr)
 
131 
 
 
Figure A-20. ?Bucatunna 26.5_8? Test Results 
 
Figure A-21. ?Bucatunna 26.5_9? Test Results 
-1.5
-1
-0.5
0
0.5
1
1.5
0 5 10 15 20 25
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_8
Velocity: 2.0 (m/s)
Scour Rate: 6.00 (mm/hr)
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_9
Velocity: 2.0 (m/s)
Scour Rate: 8.57 (mm/hr)
 
132 
 
 
Figure A-22. ?Bucatunna 26.5_10? Test Results 
 
TEST VELOCITY: 3.0 METERS PER SECOND 
 
 
Figure A-23. ?Bucatunna 26.5_11? Test Results 
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_10
Velocity: 2.0 (m/s)
Scour Rate: 10.00 (mm/hr)
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12 14 16 18
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_11
Velocity: 3.0 (m/s)
Scour Rate: 7.5 (mm/hr)
 
133 
 
 
Figure A-24. ?Bucatunna 26.5_12? Test Results 
 
Figure A-25. ?Bucatunna 23.0_1? Test Results 
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14 16
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 26.5_12
Velocity: 3.0 (m/s)
Scour Rate: 8.57 (mm/hr)
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 23.0_1
Velocity: 3.0 (m/s)
Scour Rate: 12.00 (mm/hr)
 
134 
 
 
Figure A-26. ?Bucatunna 23.0_2? Test Results 
 
Figure A-27. ?Bucatunna 23.0_3? Test Results 
 
 
-2.5
-2
-1.5
-1
-0.5
0
0.5
0 2 4 6 8 10 12
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 23.0_2
Velocity: 3.0 (m/s)
Scour Rate: 12.00 (mm/hr)
-2
-1.8
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0 1 2 3 4 5 6 7 8 9 10
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Bucatunna Clay
Test: Bucatunna 23.0_3
Velocity: 3.0 (m/s)
Scour Rate: 15.00 (mm/hr)
 
135 
 
THRESHOLD VELOCITY TEST 
 
 
Figure A-28. ?Bucatunna 23.0_4? Test Results 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
0 5 10 15 20 25 30 35 40 45
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
? 0.35 m/s
? 0.45 m/s
Push1 mm
 
136 
 
GEOTECHNICAL TESTING RESULTS 
 
 
Figure A-29. Bucatunna Clay Liquid Limits Results 
20
30
40
50
60
70
80
90
100
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 68
 
137 
 
APPENDIX B 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
YAZOO CLAY FORMATION 
 
INITIAL SCOUR TEST 
 
Figure B-1.  ?Yazoo Clay 18.5_2? Test Results 
 
 
 
 
 
 
 
 
-3
-2
-1
0
1
2
3
4
5
0 20 40 60 80 100 120 140 160 180
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
0.45 m/s
0.6 m/s 1.0 m/s
Formation: Yazoo Clay
Test: Initial Scour Test
 
138 
 
TEST VELOCITY: 1.0 METERS PER SECOND 
 
 
Figure B-2. ?Yazoo Clay 18.5_3? Test Results 
 
Figure B-3. ?Yazoo Clay 18.5_4? Test Results 
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12 14
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_3
Velocity: 1.0 (m/s)
Scour Rate: 12.00 (mm/hr)
-1.5
-1
-0.5
0
0.5
1
1.5
0 2 4 6 8 10 12 14 16 18
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_4
Velocity: 1.0 (m/s)
Scour Rate: 7.50 (mm/hr)
 
139 
 
 
Figure B-4.  ?Yazoo Clay 18.5_5? Test Results 
 
Figure B-5. ?Yazoo Clay 18.5_6? Test Results 
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 2 4 6 8 10 12 14
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_5
Velocity: 1.0 (m/s)
Scour Rate: 8.57 (mm/hr)
-1
-0.5
0
0.5
1
1.5
2
0 2 4 6 8 10 12 14 16
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_6
Velocity: 1.0 (m/s)
Scour Rate: 8.57 (mm/hr)
 
140 
 
 
Figure B-6.  ?Yazoo Clay 18.5_7? Test Results 
 
TEST VELOCITY: 1.5 METERS PER SECOND 
 
 
Figure B-7. ?Yazoo Clay 18.5_8? Test Results 
-1.5
-1
-0.5
0
0.5
1
1.5
2
0 20 40 60 80 100 120 140
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_7
Velocity: 1.0 (m/s)
Scour Rate: n/a
-4
-3
-2
-1
0
1
2
3
4
0 10 20 30 40 50 60
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_8
Velocity: 1.5 (m/s)
Scour Rate: 1.40 (mm/hr)
 
141 
 
 
Figure B-8. ?Yazoo Clay 18.5_9? Test Results 
 
TEST VELOCITY: 2 METERS PER SECOND 
 
 
Figure B-9.  ?Yazoo Clay 18.5_10? Test Results 
0
0.5
1
1.5
2
2.5
0 20 40 60 80 100 120
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_9
Velocity: 1.5 (m/s)
Scour Rate: n/a
-12
-10
-8
-6
-4
-2
0
2
4
0 2 4 6 8 10 12 14 16 18 20
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Yazoo Clay
Test: Yazoo Clay 18.5_10
Velocity: 2.0 (m/s)
Scour Rate: 4.29 (mm/hr)
 
142 
 
GEOTECHNICAL TESTING RESULTS 
 
 
Figure B-10.  Yazoo Clay Liquid Limits Results 
20
30
40
50
60
70
80
90
100
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 57
 
143 
 
APPENDIX C 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
DEMOPOLIS CHALK FORMATION 
 
INITIAL SCOUR TEST 
 
Figure C-1.  ?Demopolis Chalk 19.0_1? Test Results 
 
 
 
 
 
 
 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100 120
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Demopolis Chalk
Test: Demopolis Chalk 19.0_1
Initial Scour Test
1.0 m/s
2.0 m/s 3.0 m/s
 
144 
 
TEST VELOCITY: 3.0 METERS PER SECOND 
 
 
Figure C-2. ?Demopolis Chalk 19.0_2? Test Results 
 
Figure C-3. ?Demopolis Chalk 19.0_3? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Demopolis Chalk
Test: Demopolis Chalk 19.0_2
Velocity: 3.0 (m/s)
Scour Rate: n/a
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Demopolis Chalk
Test: Demopolis Chalk 19.0_3
Velocity: 3.0 (m/s)
Scour Rate: n/a
 
145 
 
 
Figure C-4. ?Demopolis Chalk 19.0_4? Test Results 
 
REPEAT EVENT TEST 
 
 
Figure C-5. ?Demopolis Chalk 19.0_5? Test Results 
 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Demopolis Chalk
Test: Demopolis Chalk 19.0_4
Velocity: 3.0 (m/s)
Scour Rate: n/a
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100 120 140 160 180
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Demopolis Chalk
Test: Demopolis Chalk 19.0_5
Velocity: 3.0 (m/s)
Scour Rate: n/a3.0 m/s
1.0 m/s 3.0 m/s
 
146 
 
GEOTECHNICAL TESTING RESULTS 
 
 
Figure C-6. Demopolis Chalk Liquid Limit Results 
 
0
5
10
15
20
25
30
35
40
45
50
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 37
 
147 
 
APPENDIX D 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
MOOREVILLE CHALK FORMATION 
 
INITIAL SCOUR TEST 
 
Figure D-1.  ?Mooreville Chalk 22.0_3? Test Results 
 
 
 
 
 
 
 
-1
-0.5
0
0.5
1
1.5
2
0 20 40 60 80 100 120
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
0.3 m/s
0.6 m/s
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_3
Initial Scour Test
 
148 
 
TEST VELOCITY: 0.6 METERS PER SECOND 
 
 
Figure D-2. ?Mooreville Chalk 22.0_4? Test Results 
 
Figure D-3. ?Mooreville Chalk 22.0_5? Test Results 
 
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 10 20 30 40 50 60
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_4
Velocity: 0.6 (m/s)
Scour Rate: 1.4 (mm/hr)
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 20 40 60 80 100 120 140
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_5
Velocity: 
Scour Rate: n/a
 
149 
 
TEST VELOCITY: 1.0 METERS PER SECOND 
 
 
Figure D-4. ?Mooreville Chalk 22.0_6? Test Results 
 
TEST VELOCITY: 1.5 METERS PER SECOND 
 
 
Figure D-5. ?Mooreville Chalk 22.0_7? Test Results 
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_6
Velocity: 1.0 (m/s)
Scour Rate: n/a
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70 80A
ve
ra
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_7
Velocity: 1.5 (m/s)
Scour Rate: n/a
 
150 
 
 
Figure D-6. ?Mooreville Chalk 22.0_8? Test Results 
 
TEST VELOCITY: 2.0 METERS PER SECOND 
 
 
Figure D-7. ?Mooreville Chalk 22.0_9? Test Results 
 
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_8
Velocity: 1.5 (m/s)
Scour Rate: n/a
-0.5
0
0.5
1
1.5
2
0 10 20 30 40 50 60 70A
ve
ra
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_9
Velocity: 2.0 (m/s)
Scour Rate: n/a
 
151 
 
TEST VELOCITY: 3.0 METERS PER SECOND 
 
 
Figure D-8. ?Mooreville Chalk 22.0_10? Test Results 
 
REPEAT EVENT TEST 
 
 
Figure D-9. ?Mooreville Chalk 22.0_11? Test Results 
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70 80
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_10
Velocity: 3.0 (m/s)
Scour Rate: n/a
-0.3
-0.1
0.1
0.3
0.5
0.7
0.9
1.1
1.3
1.5
0 20 40 60 80 100 120 140 160 180
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
3.0 m/s
1.0 m/s
3.0 m/s
Formation: Mooreville Chalk
Test: Mooreville Chalk 22.0_11
Repeat Event Test
 
152 
 
GEOTECHNICAL TESTING RESULTS 
 
 
Figure D-10. Mooreville Chalk Liquid Limits Results 
20
30
40
50
60
70
80
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 52
 
153 
 
APPENDIX E 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
PRAIRIE BLUFF CHALK FORMATION 
 
TEST VELOCITY: 1.0 METERS PER SECOND 
 
Figure E-1. ?Prairie Bluff Chalk 18.5_2? Test Results 
-8
-6
-4
-2
0
2
4
6
0 10 20 30 40 50 60 70 80 90 100
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_2
Velocity: 1.0 (m/s)
Scour Rate: 1.2 (mm/hr)
 
154 
 
 
Figure E-2. ?Prairie Bluff Chalk 18.5_3? Test Results 
 
TEST VELOCITY: 1.5 METERS PER SECOND 
 
 
Figure E-3. ?Prairie Bluff Chalk 18.5_4? Test Results 
 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_3
Velocity: 1.0 (m/s)
Scour Rate: n/a
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0 5 10 15 20 25 30 35 40
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_4
Velocity: 1.5 (m/s)
Scour Rate: 1.76 (mm/hr)
 
155 
 
TEST VELOCITY: 2.0 METERS PER SECOND 
 
 
Figure E-4. ?Prairie Bluff Chalk 18.5_5? Test Results 
 
TEST VELOCITY: 3.0 METERS PER SECOND 
 
 
Figure E-5. ?Prairie Bluff Chalk 18.5_6? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70 80 90
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_5
Velocity: 2.0 (m/s)
Scour Rate: n/a
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Ave
rage
 H
eigh
t C
han
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_6
Velocity: 3.0 (m/s)
Scour Rate: n/a
 
156 
 
 
Figure E-6. ?Prairie Bluff Chalk 18.5_7? Test Results 
 
Figure E-7. ?Prairie Bluff Chalk 18.5_8? Test Results 
 
 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_7
Velocity: 3.0 (m/s)
Scour Rate: n/a
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70
Av
era
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Prairie Bluff Chalk
Test: Prairie Bluff Chalk 18.5_8
Velocity: 3.0 (m/s)
Scour Rate: n/a
 
157 
 
GEOTECHNICAL TESTING RESULTS 
 
 
Figure E-8. Prairie Bluff Chalk Liquid Limits Results 
0
5
10
15
20
25
30
35
40
45
50
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 32
 
158 
 
APPENDIX F 
 
EFA AND GEOTECHNICAL TEST RESULTS: 
PORTER?S CREEK CLAY FORMATION 
 
TEST VELOCITY: 0.3 METERS PER SECOND 
 
 
Figure F-1. ?Porter?s Creek 19.5_1? Test Results 
 
 
 
 
 
 
 
0
2
4
6
8
10
0 10 20 30 40 50 60 70
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_1
Velocity: 0.3 (m/s)
Scour Rate: N/A
 
159 
 
TEST VELOCITY: 0.6 METERS PER SECOND 
 
 
Figure F-2. ?Porter?s Creek 19.5_2? Test Results 
 
Figure F-3. ?Porter?s Creek 19.5_3? Test Results 
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0 5 10 15 20 25
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_2
Velocity: 0.6 (m/s)
Scour Rate: 4.00 (mm/hr)
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25Ave
ra
ge
 H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_3
Velocity: .06 (m/s)
Scour Rate: 3.33 (mm/hr)
 
160 
 
 
Figure F-4. ?Porter?s Creek 19.5_4? Test Results 
 
TEST VELOCITY: 1.0 METERS PER SECOND 
 
 
Figure F-5. ?Porter?s Creek 19.5_5? Test Results 
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12 14 16 18 20
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_4
Velocity: .06 (m/s)
Scour Rate: 4.62 (mm/hr)
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_5
Velocity: 1.0 (m/s)
Scour Rate: 10.00 (mm/hr)
 
161 
 
 
Figure F-6. ?Porter?s Creek 19.5_7? Test Results 
 
Figure F-7. ?Porter?s Creek 19.0_1? Test Results 
 
 
-1.5
-1
-0.5
0
0.5
1
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.5_7
Velocity: 1.0 (m/s)
Scour Rate: 10.00 (mm/hr)
-1.8
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0 2 4 6 8 10 12 14
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_1
Velocity: 1.0 (m/s)
Scour Rate: 7.50 (mm/hr)
 
162 
 
TEST VELOCITY: 1.5 METERS PER SECOND 
 
 
Figure F-8. ?Porter?s Creek 19.0_2? Test Results 
 
Figure F-9. ?Porter?s Creek 19.0_3? Test Results 
-2.5
-2
-1.5
-1
-0.5
0
0.5
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_2
Velocity: 1.5 (m/s)
Scour Rate: 12.00 (mm/hr)
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12 14
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_3
Velocity: 1.5 (m/s)
Scour Rate: 8.57 (mm/hr)
 
163 
 
 
Figure F-10. ?Porter?s Creek 19.0_5? Test Results 
 
TEST VELOCITY: 2.0 METERS PER SECOND 
 
 
Figure F-11. ?Porter?s Creek 19.0_6? Test Results 
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_5
Velocity: 1.5 (m/s)
Scour Rate: 10.00 (mm/hr)
-1
-0.5
0
0.5
1
1.5
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_6
Velocity: 2.0 (m/s)
Scour Rate: 12.00 (mm/hr)
 
164 
 
 
Figure F-12. ?Porter?s Creek 19.0_7? Test Results 
 
Figure F-13. ?Porter?s Creek 19.0_8? Test Results 
 
 
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 2 4 6 8 10 12
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_7
Velocity: 2.0 (m/s)
Scour Rate: 8.57 (mm/hr)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 1 2 3 4 5 6 7 8 9
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_8
Velocity: 2.0 (m/s)
Scour Rate: 12.00 (mm/hr)
 
165 
 
TEST VELOCITY: 3.0 METERS PER SECOND 
 
 
Figure F-14. ?Porter?s Creek 19.0_9? Test Results 
 
Figure F-15. ?Porter?s Creek 19.0_12? Test Results 
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6 7 8 9
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_9
Velocity: 3.0 (m/s)
Scour Rate: 12.00 (mm/hr)
-1.5
-1
-0.5
0
0.5
1
0 1 2 3 4 5 6 7 8
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_12
Velocity: 3.0 (m/s)
Scour Rate: 15.00 (mm/hr)
 
166 
 
 
Figure F-16. ?Porter?s Creek 19.0_14? Test Results 
 
GEOTECHNICAL TEST RESULTS 
 
 
Figure F-17. Porter?s Creek Clay Liquid Limits Results 
 
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 1 2 3 4 5 6 7
Av
er
ag
e H
eig
ht 
Ch
an
ge
 (m
illi
me
ter
s)
Elapsed Time (minutes)
Formation: Porter's Creek
Test: Porter's Creek 19.0_14
Velocity: 3.0 (m/s)
Scour Rate: 20.00 (mm/hr)
30
40
50
60
70
80
90
1 10 100
M
ois
tur
e C
on
ten
t
Number of Blows
N = 25
LL ? 62