USF Libraries
USF Digital Collections

Modeling three-dimensional shape of sand grains using Discrete Element Method

MISSING IMAGE

Material Information

Title:
Modeling three-dimensional shape of sand grains using Discrete Element Method
Physical Description:
Book
Language:
English
Creator:
Das, Nivedita
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Fourier transform
Spherical harmonics
Shape descriptors
Skeletonization
Angularity
Roundness
Liquefaction
Overlapping discrete element cluster
Dissertations, Academic -- Civil Engineering -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The study of particle morphology plays an important role in understanding the micromechanical behavior of cohesionless soil. Shear strength and liquefaction characteristics of granular soil depend on various morphological characteristics of soil grains such as their particle size, shape and surface texture. Therefore, accurate characterization and quantification of particle shape is necessary to study the effect of grain shape on mechanical behavior of granular assembly. However, the theoretical and practical developments of quantification of particle morphology and its influence on the mechanical response of granular assemblies has been very limited due to the lack of quantitative information about particle geometries, the experimental and numerical difficulties in characterizing and modeling irregular particle morphology.Motivated by the practical relevance of these challenges, this research presents a comprehensive approach to model irregular particle shape accurately both in two and three dimensions. To facilitate the research goal, a variety of natural and processed sand samples is collected from various locations around the world. A series of experimental and analytical studies are performed following the sample collection effort to characterize and quantify particle shapes of various sand samples by using Fourier shape descriptors. As part of the particle shape quantification and modeling, a methodology is developed to determine an optimum sample size for each sand sample used in the analysis. Recently, Discrete Element Method (DEM) has gained attention to model irregular particle morphology in two and three dimensions.In order to generate and reconstruct particle assemblies of highly irregular geometric shapes of a particular sand sample in the DEM environment, the relationship between grain size and shape is explored and no relationship is found between grain size and shape for the sand samples analyzed. A skeletonization algorithm is developed in this study in order to automate the Overlapping Discrete Element Cluster (ODEC) technique for modeling irregular particle shape in two and three dimensions. Finally, the two-dimensional and three-dimensional particle shapes are implemented within discrete element modeling software, PFC2D and PFC3D, to evaluate the influence of grain shape on shear strength behavior of granular soil by using discrete simulation of direct shear test.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Nivedita Das.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 133 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001919512
oclc - 184843132
usfldc doi - E14-SFE0002072
usfldc handle - e14.2072
System ID:
SFS0026390:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Modeling Three-Dimensional Shape of Sand Grains Using Discrete Element Method by Nivedita Das A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Civil and Environmental Engineering College of Engineering University of South Florida Co-Major Professor: Alaa K. Ashmawy, Ph.D. Co-Major Professor: Sudeep Sarkar, Ph.D. Manjriker Gunaratne, Ph.D. Beena Sukumaran, Ph.D. Abla M. Zayed, Ph.D. Date of Approval: May 4, 2007 Keywords: Fourier transform, spherical harmonics, shape descriptors, skeletonization, angularity, roundness, liquefaction, overlapping discrete element cluster Copyright 2007, Nivedita Das

PAGE 2

DEDICATION To my parents.....

PAGE 3

ACKNOWLEDGEMENTS First of all, I would like to thank my doctoral committee members Dr. Alaa Ashmawy, Dr. Sudeep Sarkar, Dr. Manjriker Gunaratne, Dr. Abla Zayed and Dr. Beena Sukumaran for their insights and suggestions that have immensely contributed in improving the quality of this dissertation. I would also like to thank Dr. Chris Ferekides for serving as the chair for my doctoral dissertation defense. In particular, I wish to express my sincere gratitude to my advisor, Dr. Alaa Ashmawy and co-advisor, Dr. Sudeep Sarkar for their invaluable guidance and continued support for this research work. Without their generous help and support, successful completion of this dissertation would not have been possible. Sincere thanks to Dr. Beena Sukumaran, Dr. Shreekanth Mandayam and the graduate students at the Rowan University for making three-dimensional particle database available online. I am also very thankful to my colleagues, in particular, former USF graduate students Mr. Delfin Carreon and Mr. Jorge Rivas for helping me greatly in characterizing two-dimensional particle shapes. I am indebted to the Department of Civil and Environmental Engineering at USF for providing excellent research environment and facilities. Finally, I would like to extend my heartfelt thanks to my parents and my husband, Amlan for their unconditional sacrifice and inspiration during the completion of this dissertation.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES.............................................................................................................iv LIST OF FIGURES.............................................................................................................v ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION.........................................................................................1 1.1 Problem Statement..........................................................................................1 1.2 Particle Shape Modeling in Two and Three Dimensions...............................2 1.3 Research Objectives........................................................................................4 1.4 Outline of the Dissertation..............................................................................5 CHAPTER 2 STATE OF THE ART IN PARTICLE SHAPE QUANTIFICATION AND MODELING....................................................6 2.1 Existing Methods of Quantifying Particle Shape............................................6 2.1.1 Shape Descriptors in Two Dimensions...............................................6 2.1.1.1 Shape Factor (SF)................................................................8 2.1.1.2 Angularity Factor (AF)........................................................9 2.1.1.3 Fractal Based Shape Measures...........................................10 2.1.1.4 Fourier Shape Descriptors..................................................16 2.1.2 Shape Descriptors in Three Dimensions...........................................17 2.1.2.1 Representing Grain Shape in Spherical Coordinates.........19 2.2 Relationship Between Grain Size and Shape................................................21 2.3 Modeling Particle Shape in Two Dimensions..............................................24 2.3.1 Existing Methods of Modeling Irregular Particle Shape..................24 2.3.2 Modeling Angular Particles as Clusters............................................26 2.3.3 Overlapping Discrete Element Clusters............................................26 2.4 Modeling Particle Shape in Three Dimensions............................................29 2.5 Effect of Particle Shape on Shear Strength Behavior of Cohesionless Soil..........................................................................................31 CHAPTER 3 MATERIALS..............................................................................................35 3.1 Sand Samples Collected for the Present Study.............................................35 3.1.1 Sample Selection Procedure.............................................................35 3.2 Data Sets and Sample Characteristics...........................................................40

PAGE 5

ii CHAPTER 4 PARTICLE SHAPE CHARACTERIZATION AND QUANTIFICATION..................................................................................45 4.1 Characterizing Particle Shape in Two Dimensions......................................45 4.2 Characterizing Particle Shape in Three Dimensions....................................46 4.3 Quantification of Particle Shape...................................................................48 4.3.1 Particle Shape Quantification in Two Dimensions...........................48 CHAPTER 5 RELATIONSHIP BETW EEN GRAIN SIZE AND SHAPE......................52 5.1 Introduction...................................................................................................52 5.2 Methodology to Determine Sample Size......................................................52 5.3 Relationship Between Grai n Size and Grain Shape......................................58 5.3.1 Data Sets...........................................................................................58 5.3.2 Fourier Shape Descriptors.................................................................59 5.3.3 Grain Size Grain Shape Relationship............................................61 5.3.4 Summary and Discussion..................................................................65 CHAPTER 6 SKELETONIZATION AND OVERLAPPING DISCRETE ELEMENT CLUSTER ALGORITHM.....................................................67 6.1 Skeletonization of Grain Shape....................................................................67 6.1.1 Skeletonization Algorithm in Two Dimensions...............................71 6.1.2 Skeletonization Algorit hm in Three Dimensions.............................75 6.2 Overlapping Discrete Element Cluster.........................................................81 6.2.1 ODEC Algorithm in Two Dimensions.............................................81 6.2.2 ODEC Algorithm in Three Dimensions...........................................83 CHAPTER 7 IMPLEMENTATION OF PARTICLE SHAPE WITHIN DISCRETE ELEMENT MODELING SIMULATION.............................90 7.1 Introduction...................................................................................................90 7.2 Two-Dimensional Discrete Element Simulation..........................................90 7.2.1 Model Set-Up....................................................................................91 7.2.2 Numerical Simulation.......................................................................92 7.3 Three-Dimensional Discrete Element Simulation........................................97 7.3.1 Model Set-Up....................................................................................98 7.3.2 Numerical Simulation.......................................................................98 7.4 Particle Shape Library.................................................................................104 CHAPTER 8 SUMMARY AND CONCLUSIONS........................................................110 8.1 Summary and Conclusions.........................................................................110 8.2 Methodological Contributions....................................................................113 8.3 Practical Contributions................................................................................114 8.4 Future Recommendations...........................................................................115 REFERENCES................................................................................................................117

PAGE 6

iii APPENDICES.................................................................................................................128 Appendix A: Data Sets.........................................................................................129 ABOUT THE AUTHOR.......................................................................................End Page

PAGE 7

iv LIST OF TABLES Table 2.1 Relationship Between Number of Miniature Pieces and Fractal Dimensions................................................................................................11 Table 3.1 Sand Samples Collected for the Study.......................................................41 Table 5.1 Values of Variances of Different Shape Parameters..................................55 Table 5.2 Estimated Sample Size for Sand Samples Used in the Study ...................57 Table 5.3 Relevant Propertie s of Granular Materials................................................58 Table 5.4 Average Values of Fourier De scriptors for Different Sand Samples........60 Table 5.5 Regression Analysis Results for Different Shape Descriptors..................65 Table 6.1 Number of Discs Require d for Daytona Beach Sand Sample...................81 Table 6.2 Number of Spheres Requir ed for Michigan Dune and Daytona Beach Sand Grains ....................................................................................83 Table 7.1 Values of Maximum Shear Stresses for Different Vertical Stresses (Two-Dimensional Simulation).................................................................96 Table 7.2 Values of Maximum Shear Stresses and Internal Friction Angles for Different Particle Arrangements (Three-Dimensional Simulation)...............................................................................................102 Table 7.3 Comparison of Internal Fric tion Angles Obtained from 2-D and 3D Simulations...........................................................................................103 Table 7.4 Geomaterial Database..............................................................................105 Table 7.5 ODEC Data (2-D) for Dayt ona Beach Sand Grains (98% Area Coverage).................................................................................................106 Table 7.6 ODEC Data (2-D) for Michig an Dune Sand Grains (98% Area Coverage).................................................................................................107 Table 7.7 ODEC Data (3-D) for Michig an Dune Sand Grain (85% Volume Coverage).................................................................................................108 Table 7.8 ODEC Data (3-D) for Dayt ona Beach Sand Grain (85% Volume Coverage).................................................................................................109

PAGE 8

v LIST OF FIGURES Figure 2.1 Ideal Geometric Shape Used to Define the Shape and Angularity Factors..........................................................................................................8 Figure 2.2 Perimeter of an Aggregate...........................................................................9 Figure 2.3 Convex Perimeter.........................................................................................9 Figure 2.4 Self-Similar Figures: (A) Li ne Segments, (B) Square, (C) Cube..............10 Figure 2.5 Sierpinski Triangle.....................................................................................11 Figure 2.6 The Koch Curve.........................................................................................12 Figure 2.7 Single Fractal Element Overall Represented by D T ...................................13 Figure 2.8 Multiple Fractal Elements Represented by D 1 and D 2 ...............................14 Figure 2.9 M-R Plot for a Sedimentary Particle..........................................................15 Figure 2.10 Fourier Analysis in Closed Form...............................................................17 Figure 2.11 Spherical Harmonic Transform..................................................................20 Figure 2.12 Mean Roundness Values of Eight Samples Plotted Against MidPoints of Size Grades.................................................................................22 Figure 2.13 Relationship Between Partic le Angularity and Particle Size.....................22 Figure 2.14 (A) Outline of Sand Particle, (B) DEM Disc Element Superimposed Over Sand Particle (C) DEM Disc Particles are Joined Together in a Rigid Conf iguration (Cluster), (D) Several Possible Combination of Discs to Form Clusters......................................27 Figure 2.15 Disc Elements Inscribed within a Particle Outline to Capture the Shape..........................................................................................................27 Figure 2.16 Random Assemblies of Eight Circular Particles (Left) and the Transformed Equivalent A ngular Particles (Right)...................................29 Figure 2.17 Virtual Force Acting on the Elements.......................................................30 Figure 3.1 Variation of Minimum and Maximum Void Ratio (Group # 1)................36 Figure 3.2 Variation of Minimum and Maximum Void Ratio (Group # 2)................36 Figure 3.3 Variation of Minimum and Maximum Void Ratio (Group # 3)................37

PAGE 9

vi Figure 3.4 Variation of Minimum and Maximum Void Ratio (Group # 4)................37 Figure 3.5 Variation of Maximum Void Ratio with e max e min (Group # 1)...............38 Figure 3.6 Variation of Maximum Void Ratio with e max e min (Group # 2)...............38 Figure 3.7 Variation of Maximum Void Ratio with e max e min (Group # 3)..............39 Figure 3.8 Variation of Maximum Void Ratio with e max e min (Group # 4)...............39 Figure 3.9 Particle Size Distri bution (Sand Samples, Group # 1)...............................43 Figure 3.10 Particle Size Distri bution (Sand Samples, Group # 2)...............................43 Figure 3.11 Particle Size Distri bution (Sand Samples, Group # 3)...............................44 Figure 3.12 Particle Size Distri bution (Sand Samples, Group # 4)...............................44 Figure 4.1 Motic SMZ-168 Stereo Zoom Microscope................................................46 Figure 4.2 Motic AE-31 Inverted Microscope............................................................46 Figure 4.3 SkyScan 1072 X-Ray CT System..............................................................47 Figure 4.4 Fourier Transform on Particle Boundary...................................................48 Figure 4.5 Mean Amplitude Spectra for Tecate River Sand.......................................50 Figure 4.6 Mean Amplitude Spectra for Daytona Beach Sand...................................50 Figure 4.7 Variation of Harmonic Amp litude with Descriptor Number for Different Sand Samples.............................................................................51 Figure 5.1 Standard Normal Distribution....................................................................54 Figure 5.2 Variation of Error w ith Sample Size for Toyoura Sand............................55 Figure 5.3 Variation of Change of Erro r Per Unit Sample with Sample Size.............56 Figure 5.4 Variation of Error as Pe rcent of Mean with Sample Size..........................56 Figure 5.5 Two-Dimensional Images of Sand Samples..............................................59 Figure 5.6 Fourier Amplitude Spectra for Toyoura Sand...........................................60 Figure 5.7 Fourier Amplitude Spectra for Michigan Dune Sand................................61 Figure 5.8 Frequency Distributions of Shape Parameters: (A) Diameter, (B) Elongation, (C) Triangularity and (D) Squareness....................................62 Figure 5.9 Variation of Ci rcularity with Diameter......................................................63 Figure 5.10 Variation of Elongation with Diameter......................................................63 Figure 5.11 Variation of Triangularity with Diameter..................................................64 Figure 5.12 Variation of Squareness with Diameter.....................................................64 Figure 6.1 Thinning Algorithm: (A) A Set of Structuring Elements, (B) Successive Steps of Thinning....................................................................68

PAGE 10

vii Figure 6.2 Skeleton in Two Dimensions : (A) L-Shaped Object, (B) Fraser River Sand Grain........................................................................................69 Figure 6.3 Neighborhood Arrangement Used by the Thinning Algorithm.................70 Figure 6.4 Skeleton in Two Dimensions: (A) T-Shaped Object, (B) Michigan Dune Sand Grain........................................................................................70 Figure 6.5 Edge Detection...........................................................................................71 Figure 6.6 First Pixel to Start the Skeleton.................................................................72 Figure 6.7 First Skeleton Point Obtained....................................................................73 Figure 6.8 Next Skeleton Point...................................................................................74 Figure 6.9 Skeleton Continued Following the First Branch........................................74 Figure 6.10 Final Skeleton Obta ined Through the Algorithm......................................75 Figure 6.11 A Daytona Beach Sand Particle (Left) and the Skeleton (Right)..............75 Figure 6.12 Non-Adjace nt Boundary Voxels................................................................76 Figure 6.13 Surface Protrusion Marked as Red Voxels................................................76 Figure 6.14 Original Particle of Daytona Beach Sand (DB #1)....................................77 Figure 6.15 Three-Dimensional Skelet on of Daytona Beach Sand Grain (DB #1)......................................................................................................77 Figure 6.16 Original Particle of Daytona Beach Sand (DB #2)....................................78 Figure 6.17 Three-Dimensional Skelet on of Daytona Beach Sand Grain (DB #2)......................................................................................................78 Figure 6.18 Original Particle of Michigan Dune Sand (MD #1)...................................79 Figure 6.19 Three-Dimensional Skelet on of Michigan Dune Sand Grain (MD #1).....................................................................................................79 Figure 6.20 Original Particle of Michigan Dune Sand (MD #2)...................................80 Figure 6.21 Three-Dimensional Skelet on of Michigan Dune Sand Grain (MD #2).....................................................................................................80 Figure 6.22 Overlapping Discrete Element Cluster for a Daytona Beach Sand Grain..........................................................................................................82 Figure 6.23 Original Particle of Daytona Beach Sand in Two Dimensions..................82 Figure 6.24 Particle Shape Ob tained Through ODEC Technique................................82 Figure 6.25 Overlapping Discrete Elem ent Cluster (MD #1, Volume Covered =85%).........................................................................................................84

PAGE 11

viii Figure 6.26 Overlapping Discrete Elem ent Cluster (MD #1, Volume Covered =90%).........................................................................................................84 Figure 6.27 Overlapping Discrete Elem ent Cluster (MD #1, Volume Covered =95%).........................................................................................................85 Figure 6.28 Overlapping Discrete Elem ent Cluster (MD #2, Volume Covered =85%).........................................................................................................85 Figure 6.29 Overlapping Discrete Elem ent Cluster (MD #2, Volume Covered =90%).........................................................................................................86 Figure 6.30 Overlapping Discrete Elem ent Cluster (MD #2, Volume Covered =95%).........................................................................................................86 Figure 6.31 Overlapping Discrete Elem ent Cluster (DB #1, Volume Covered =85%).........................................................................................................87 Figure 6.32 Overlapping Discrete Elem ent Cluster (DB #1, Volume Covered =90%).........................................................................................................87 Figure 6.33 Overlapping Discrete Elem ent Cluster (DB #1, Volume Covered =95%).........................................................................................................88 Figure 6.34 Overlapping Discrete Elem ent Cluster (DB #2, Volume Covered =85%).........................................................................................................88 Figure 6.35 Overlapping Discrete Elem ent Cluster (DB #2, Volume Covered =90%).........................................................................................................89 Figure 6.36 Overlapping Discrete Elem ent Cluster (DB #2, Volume Covered =95%).........................................................................................................89 Figure 7.1 Two-Dimensional Direct Sh ear Test Simulation with Daytona Beach Sand Sample....................................................................................92 Figure 7.2 History of Servo-Wall Stress with Time....................................................93 Figure 7.3 Movement of Shear Box with Daytona Beach Sand (2-D Simulation).................................................................................................93 Figure 7.4 Variation of Shear Stress with Shear Displacement for Daytona Beach Sand Sample (Two-Dimensional Simulation)................................94 Figure 7.5 Variation of Vertical Disp lacement with Shear Displacement for Daytona Beach Sand Sample (Two-Dimensional Simulation)..................94 Figure 7.6 Variation of Shear Stress with Shear Displacement for Circular Particles (Two-Dimensional Simulation)..................................................95 Figure 7.7 Variation of Vertical Disp lacement with Shear Displacement for Circular Particles (Two-Dimensional Simulation)....................................95 Figure 7.8 Variation of Shear Stress with Normal Stress (Two-Dimensional Simulation).................................................................................................97

PAGE 12

ix Figure 7.9 Three-Dimensional Direct Sh ear Test Simulation with Spherical Particles......................................................................................................98 Figure 7.10 Three-Dimensional Direct Shear Test Simulation with Daytona Beach Sand Sample....................................................................................99 Figure 7.11 Variation of Shear Stress with Shear Displacement for Daytona Beach Sand Sample (Three-Dimensional Simulation)............................100 Figure 7.12 Variation of Vertical Disp lacement with Shear Displacement for Daytona Beach Sand Sample (Three-Dimensional Simulation)..............100 Figure 7.13 Variation of Shear Stress with Shear Displacement for Spherical Particles (Three-Dimensional Simulation)..............................................101 Figure 7.14 Variation of Vertical Disp lacement with Shear Displacement for Spherical Particles (Three -Dimensional Simulation)..............................101 Figure 7.15 Variation of Shear Stress w ith Normal Stress for Daytona Beach Sand (Three-Dimensional Simulation)....................................................102 Figure 7.16 Variation of Shear Stress with Normal Stress for Spherical Particles (Three-Dimensional Simulation)..............................................103 Figure A.1 Sand Samples Collected for the Present Study........................................128

PAGE 13

x MODELING THREE-DIMENSIONAL SHAPE OF SAND GRAINS USING DISCRETE ELEMENT METHOD Nivedita Das ABSTRACT The study of particle morphology plays an important role in understanding the micromechanical behavior of cohesionle ss soil. Shear streng th and liquefaction characteristics of granular soil depend on various morphological ch aracteristics of soil grains such as their particle size, shape and surface texture. Therefore, accurate characterization and quantificati on of particle shape is necessary to study the effect of grain shape on mechanical beha vior of granular assembly. However, the theoretical and practical developments of quantification of particle morphology and its influence on the mechanical response of granular assemblies ha s been very limited due to the lack of quantitative information about particle ge ometries, the experimental and numerical difficulties in characterizing and modeling irregular particle morphology. Motivated by the practical relevance of these challenges, this research presents a comprehensive approach to model irregular pa rticle shape accurately both in two and three dimensions. To facilitate the research goal, a variety of natural and processed sand samples is collected from various locations around th e world. A series of experimental and analytical studies are performed following the sa mple collection effort to characterize and quantify particle shapes of various sand samples by using F ourier shape descriptors. As part of the particle shape quantificati on and modeling, a methodology is developed to determine an optimum sample size for each sand sample used in the analysis. Recently, Discrete Element Method (DEM) has gained attention to model irregular particle morphology in two and three dimensions. In order to generate and reconstruct particle

PAGE 14

xi assemblies of highly irregular geometric shapes of a particular sand sample in the DEM environment, the relationship between grain size and shape is explored and no relationship is found between grain size and shape for the sand samples analyzed. A skeletonization algorithm is developed in this study in order to automate the Overlapping Discrete Element Cluster (ODE C) technique for modeling irregular particle shape in two and three dimensions. Finally, the two-dimensi onal and three-dimensional particle shapes are implemented within discrete element modeling software, PFC 2D and PFC 3D to evaluate the influence of grain shape on shear strength behavior of granular soil by using discrete simulation of direct shear test.

PAGE 15

1 CHAPTER 1 INTRODUCTION 1.1 Problem Statement Particle morphology is an important paramete r to study the micromechanical behavior of granular media. Accurate modeling of particle shape is necessary to study the effect of grain shape on mechanical beha vior of granular soil such as inter-particle contacts, dilation and liquefaction. The shear strength of granular soil is greatly influenced by its dilative and contractive beha vior which in turn depends on various intrinsic soil properties such as grain size, size distri bution, shape (angularity or roundness) and surface roughness of soil grains. Substantial research on particle morphol ogy has been conducted to evaluate the effect of grain shape on the mechanical re sponse of granular soil by theoretical and experimental investigation and numerical m odeling. However, in spite of significant progress in particle shape characterizati on and reconstruction th rough digital imaging, majority of the available resources descri bing the particle shape modeling technique remains limited to two-dimensional mode ling and characterization, with minimal progress in three-dimensional domain due to lack of quantitative information about particle geometries, and the experimental and numerical difficulties associated with characterizing and modeling irregular partic le shape. However, the microstructural phenomena (soil fabric, particle to particle in teraction) in granular assemblies cannot be represented by two-dimensional simulation wh ere the degrees of freedom are limited by enforcing restrictions in the third dimensions. These limitations warrant accurate modeling and quantitative characterization of three-dimensiona l particle shape to better understand the mechanical processes that cont rol natural phenomena such as liquefaction susceptibility and shear flow. Proper charac terization of particle shape is also

PAGE 16

2 relevant to several other geotechnical applications such as studying soil-structure interaction, strength and de formation characteristics of pavement, soil erosion, and geotechnical interface design. In the recent past, Discrete Element Method (DEM) has gained momentum in the field of micromechanical modeling. With the advancement of DEM, characterization of particle morphology in two and three dimensions has become increasingly relevant. Discrete element method is a numerical technique which allows modeling of a system of discontinuous material as an assembly of disc rete elements interacting with each other. The main advantage of DEM lies in its abil ity to capture the mechanical interaction between different discrete bodies that cannot be solved by traditional continuum-based techniques such as the Finite Elemen t Method. Although the two-dimensional DEM framework has seen considerable devel opment in the area of particle shape characterization, however, three dimensi onal characterization and modeling of nonspherical particles is still in the early stage of development. Considering the current need to advance the existing research methodologi es, the current research aims to model irregular particle geometries to evaluate the influence of three-dimensional particle shape on the shear strength behavior of granular soil using DEM. 1.2 Particle Shape Modeling in Two and Three Dimensions Traditional approaches in DEM have modeled soil samples as an assembly of twodimensional discs or three-dimensional spheres. But considering each particle as a disc or sphere would be too idealized and it does not ca pture the real behavior of the system as the shapes of soil grains are highly irregular in reality. Moreover, the circular or spherical particles have a higher tendency to rotate compared to the act ual particle. Hence the angle of internal shearing resistance of the material comprising of circular or spherical particles would be much less than that of actual ma terial. To overcome the limitation of this methodology, various formulations have been suggested in the literature to model noncircular particle outline such as approxima ting the particle shape by ellipse, polygon and combining several circular outlin es into a cluster. However, the irregular particle shape could not be simulated accurately using th ese approaches. Notably, Ashmawy et al.

PAGE 17

3 (2003) proposed a more accurate clumping t echnique (ODEC) in two dimensions to model angular particle shapes using real sand particles. In the ODEC (Overlapping Discrete Element Cluster) method, two-dimensional particle shape was modeled by clumping a number of overlapping disc elements within the particle boundary so that the resulting outline resembles the outline of the actual particle. The advantage of the ODEC method is that the built-in clump logic allows bonding betw een several disc elements without detecting contacts between disc elements belonging to the same clump (Ashmawy et al., 2003). The ability of the OD EC technique to model the behavior of irregular particle sh ape was verified numerically and experimentally by Sallam (2004). Good agreement was observed between experi mental results and numerical simulations. However, the procedure primarily relied on several manual operations which necessarily warranted the development of a computer-bas ed technique. The current research intends to take the next step to automate the ODE C technique in two dime nsions by developing an algorithm (ODEC2D). In th is method, overlapping disc elem ents are inscribed within the particle outline until a reasonable percentage of the grain area is covered. The ODEC2D algorithm is found to be capable of capturing angular particle shape accurately in two dimensions. As it can be conceivabl e that two-dimensional simulation limits the accuracy of modeling soil behavior from quantitative standpoint, which further necessitates extension of the ODEC technique in three dime nsions to better understand the mechanical response of granular media in its entirety. Literature describing particle shape mode ling technique in three dimensions is scarce due to many research constraints su ch as difficulties in image capturing, threedimensional reconstruction and handling large volume of data sets. However, some of the studies which are worth mentioning include ellipsoid-based three-dimensional DEM code, ELLIPSE3D (Lin and Ng, 1997), polyhedron-based approach (Ghaboussi and Barbosa, 1990) and three-dimensional imagebased discrete element modeling procedure considering a virtual attraction between the grain surface and a number of primitive elements (Matsushima, 2004). The current st udy suggests another sphere-based approach by developing an algorithm (ODEC3D) to m odel three-dimensional irregular particle shape. Skeletonization is an efficient techni que to inscribe sphere within a particle

PAGE 18

4 boundary because skeleton of a region is the lo cus of centers of a ll maximally inscribed spheres. After generating skeleton of a gi ven shape of sand particle, the volume is covered by clumping a number of overlapping spheres within the particle surface until the desired level of accuracy is achieved. In order to generate and reconstruct pa rticle assemblies for discrete element modeling simulation, relationship between grain size and shape needs to be explored. Therefore, this research conducts a detailed st atistical analysis to explore the existence of any relationship between grain size and shape using different natura l and processed sand samples. If no relationship exis ts between grain size and shape, then shapes belonging to a particular sand sample can be selected randomly from the particle shape library irrespective of sizes, for discre te element modeling simulation. However, if any particular sand sample is found to exhibit such relationshi p, then it would be necessary to separate the particles into different groups or bins based on their size. Optimum sample size selection is consid ered as an important aspect of any experimental design. In the context of th is study, it is hypothesi zed that the model behavior would be highly sensitive to the vari ability of size and shape of particles within a particular sand sample. Therefore, this study offers a statistical procedure to determine the optimum sample size for different materi als used in the analysis. These tasks render proper characterization and qua ntification of particle mo rphology. In the current study, the grain shapes are quantified using Fourier shape descriptor s and the first four Fourier descriptors are used to inve stigate the relationship between grain size and shape. 1.3 Research Objectives Considering the emerging need of evaluating micromechanical behavior of granular soil, it is important to accurately model highly irregular particle sh ape in two and three dimensions and incorporate those shapes within discrete element modeling simulation. To fulfill these research goals, the main objectives of this research can be broadly classified into the following categories:

PAGE 19

5 Compilation of particle morphology data for sand samples collected from locations around the world and additional information obtained from various sources in the literature. Characterization of two-dimensional and three-dimensional shapes of granular materials. Quantification of particle shape usin g Fourier Shape Descriptors in two dimensions. Verification of any existi ng relationship between grai n size and grain shape. Development of an algorithm for skeletoniza tion of irregular pa rticle shapes in two and three dimensions. Automation of the ODEC (Ove rlapping Discrete Element Cluster) method in two and three dimensions. Implementation of two-dimensional and thre e-dimensional particle shapes within DEM simulations to study of influence of particle shapes on the shear strength behavior of soil. Development of an online data base for particle morphology. 1.4 Outline of the Dissertation The state-of-the-art practices in quantifyi ng and modeling angular particle shape are discussed in the next chapter. The third chapter describes the properties of various materials collected from different loca tions around the world. Particle shape characterization and quantificati on technique will be presented in the fourth chapter. The fifth chapter offers a methodology to determ ine the sample size for sand samples and explores relationship between grain size and grain shape. A detailed skeletonization algorithm and a procedure of automating the ODEC technique in two and three dimensions will be discussed in the sixth chapter. Chapter 7 describes the implementation of particle shape within DEM simulation and influence of particle shape on shear strength behavior of granular soil. Information a bout particle shape library will also be documented in chapter 7. Finally, a concludi ng discussion, research summary and future recommendation will be included in the eighth chapter.

PAGE 20

6 CHAPTER 2 STATE OF THE ART IN PARTICLE SHAPE QUANTIFICATION AND MODELING The previous chapter has introduced the impor tance of modeling part icle shape in two and three dimensions. Having the research probl em stated, the need for accurate particle shape modeling approaches is acknowledged and then research objectives are outlined. The current chapter discusses about the past and recent developments in particle shape quantification and modeling. 2.1 Existing Methods of Qu antifying Particle Shape Many research studies have been conducted to characterize and quantify the particle shape in two and three dimensions. Conventional methods available to quantify particle shape do not provide any quantitative information. For example, the comparison charts developed by Krumbein (1941) and the verbal descriptors assigned by Powers (1953), are based on qualitative visual assessment. Th e shape descriptors commonly used in the literature to quantify particle morphology are described next. 2.1.1 Shape Descriptors in Two Dimensions A shape can be quantitatively described by a set of numbers which are often called descriptors. The three main featur es used to describe a shape are form roundness and surface texture (Barrett, 1980). Form is the first order morphological descriptor, used to describe the gross shape of a particle. Form is related to the three principal axes, usually quantified in terms of sphe ricity (Diepenbroek, et al, 1992) and is independent of angularity and surface roughness (Sukumaran & Ashmawy, 2001). Roundness and angularity the second order descriptors, reflect the variations in corners, edges and faces and are related to surface texture Roundness was defined as the

PAGE 21

ratio of the curvature of corners and edges of the particle to that of the overall particle (Wadell, 1932). The defining equation is as follows: Degree of Roundness of a particle in one plane = arithmetic mean of the roundness of individual corners in that plane = N Rr r is the radius of curvature of a corner and R is the radius of the largest inscribed circle within the shape, Rr is the sum of the roundness values of the corners,is the number of corners of the particle in the given plane (Wadell, 1932). Roundness is sensitive to abrasion during transportation and can be used as a measure of distance of transport. N Surface texture, the third order descriptor, reflects the roughness along the particle surface and on corners (Sukumaran & Ashmawy, 2001). This property is usually used to describe the small-scale details along the particle surface. Other two-dimensional shape descriptors include Aspect ratio, elongation, circularity, shape factor and angularity factor. Aspect ratio is the ratio between major axis and minor axis of ellipse equivalent to the shape. Elongation is defined as the ratio of the length of the longest chord of the shape to the longest chord perpendicular to it. Wadell (1933) described circularity as the ratio of circumference of a circle of the same area as the shape, to the actual circumference of the shape. The standard equation to calculate circularity is: 2)4(perimeterareayCircularit (2.1) Sukumaran & Ashmawy (2001) proposed a shape factor which is defined as the sum of the deviation of global particle outline from a circle and an angularity factor which is defined in terms of the number and sharpness of corners. In this method the particle shape was approximated by an equivalent polygon and compared to an ideal shape (circle) as shown in Figure 2.1. 7

PAGE 22

Figure 2.1 Ideal Geometric Shape Used to Define the Shape and Angularity Factors [Source: Sukumaran (1996)] 2.1.1.1 Shape Factor (SF) Sukumaran and Ashmawy (2001) defined the particle shape in terms of the deviation of the particle outline from a circle and distortion diagram was used to determine the shape factor. Distortion diagram (Sukumaran, 1996) is a plot of distortion angles ( i ) against the cumulative sampling interval. The distortion diagram is a mapping technique from which the particle shape can be fully reconstructed (Sukumaran and Ashmawy, 2001). The normalized shape factor was obtained by using equation 2.2. SF = %1004501 N NiiParticle (2.2) Where, the numerator is the sum of the absolute values of i for a given shape and the denominator represents the sum of the distortion angles for a flat particle. The shape factor is zero for a circle and one for a flat particle. 8

PAGE 23

2.1.1.2 Angularity Factor (AF) Sukumaran and Ashmawy (2001) defined the angularity of a particle in terms of the number and sharpness of the corners and the angularity factor was estimated by using the following equation: AF = %100)/360()180(3)/360()180(22122NNNiiParticle (2.3) Based on the above equation, the angularity factor of a sphere will be zero. Though reliable, the method is difficult to implement for three-dimensional shapes. Figure 2.2 Perimeter of an Aggregate Figure 2.3 Convex Perimeter [Source: Janoo, 1998] Janoo (1998) proposed a roundness/angularity index and roughness of particles. The angularity index was obtained by using equation 2.4: 24pAnR (2.4) where, is the roundness index, nR A is the area of aggregate and is the perimeter of aggregate (Figure 2.2). An object with irregular surface will have a smaller value than the circular one. p The roughness of a grain was defined as the ratio of perimeter to the convex perimeter (Figure 2.3) and it was calculated using equation 2.5. cppr (2.5) 9

PAGE 24

Where, r is the roughness, p is the perimeter of aggregate and is the convex perimeter of aggregate (Janoo, 1998). For a smooth particle, the roughness factor is 1.00 and the roughness factor increases with the roughness of the particle (Uthus et al.,2005). cp 2.1.1.3 Fractal Based Shape Measures Fractal Analysis is another approach to quantify particle shape. The 20th century mathematician, Benoit Mandelbrot (1977) first developed the concept of fractal dimension in order to quantify the complexity of nature in relatively simplistic ways. A point is a dimensionless object, a line has one dimension, a plane has two dimensions and space has three dimensions. Fractals can have fractional dimensions. (A) (B) (C) Figure 2.4 Self-Similar Figures: (A) Line Segments, (B) Square, (C) Cube All of the above figures are self-similar. In figure 2.4 (A) the line is divided into two similar pieces. When magnified by a factor of 2, each of the two pieces will look exactly like the original line. In figure 2.4 (B) each of the four small squares will be identical to the original large square when magnified by a factor of 2. Similarly, in figure 2.4 (C) each of the eight small cubes needs to be magnified by a factor of 2 to generate the original large cube. 10

PAGE 25

Table 2.1 Relationship Between Number of Miniature Pieces and Fractal Dimensions Figure Dimension No. of miniature pieces Line 1 2 Square 2 4 Cube 3 8 So, there exists a relationship between number of miniature pieces (N) and the dimension (D) of the object which is as follows: DMN (2.6) where, M is the scaling factor/magnification factor. Sierpinski Triangle (Figure 2.5) is another self-similar figure. Doubling the sides generates three similar copies. So, considering the above relationship, 3 = 2 D ; ln (3) = D ln (2); D = ln (3) / ln (2) = 1.585, so fractal is a geometric figure that can have fractional dimension. This relationship [D = ln (N) / ln (M)] is used to compute the fractal dimension (D) of any self-similar fractals. Figure 2.5 Sierpinski Triangle [Source: http://math.rice.edu/~lanius/fractals ] 11

PAGE 26

Figure 2.6 The Koch Curve [Source: www.jimloy.com/fractals/koch.html] One example of such fractal is Van-Koch snowflake (Figure 2.6). The Van-Koch curve starts with a straight line and replaces it with four lines; each of those is one-third the length of the original line. From Figure 2.7, N = 4, M = 3, D = ln (4) / ln (3) = 1.262. So, a fractal can be defined as an irregular geometric object with an infinite nesting of structure. Fractals are self-similar copies of themselves. In the existing literatures there are some examples of using fractal technique for particle shape analysis. In fractal based shape measurement, the shape was approximated by a series of equilateral polygons (Kennedy & Lin, 1992). With the increase in number of polygon sides, there was a resulting increase in polygon perimeter and decrease in step length (length of each polygon side), so irregularities in the original shape were more closely observed. A relationship was developed by Mandelbrot (1967) between the perimeter estimate (P) and the step length (S). A series of log P versus log S pairs yielded a line (M-R plot) as shown in Figure 2.7, and the slope of the line was used as a measure of the irregularity of the shape. The information about the shape of the particle can be obtained from the slope of the line, greater slope indicates more irregularity in the shape. The relationship between the slope of the line (b) and the fractal dimension (FD) is as follows: FD = 1 b (2.7) 12

PAGE 27

This relationship is valid for self-similar (same degree of irregularity at all scales, Kennedy & Lin, 1992). The shapes which are not self-similar (sedimentary particles) can not be represented by a single fractal dimension since the M-R plot for these particles is not a single straight line. In case of non self-similar particles, two linear elements can appear in the M-R plot. These two linear elements represent two separate fractal dimensions, D 1 and D 2 (Figure 2.8). D 1 is related to the smallest step length and referred to as the textural fractal (Kaye, 1978; Flook, 1979) and the second element (D 2 ) is defined as the structural fractal (Flook, 1979) which represents the gross shape features of the particle. T r marks the boundary between the textural and the structural fractal. Two different fractal elements represent two separate self-similar scales of particle geometry (Orford and Whalley, 1983). The occasional existence of the third element was considered as an artifact of the algorithm proposed by Schwarz and Exner (1980). ln Step (S)ln Perimeter (P) b = 1.0 DT Single Fractal Element Figure 2.7 Single Fractal Element Overall Represented by D T [Source: Orford and Whalley (1983)] 13

PAGE 28

ln Step (S)ln Perimeter (P) Tr Structural Fractal b' = 1.0 D1 b''' = 1.0 D3 Tr Textureal Fractal Multiple Fractal Elements b'' = 1.0 D2 Figure 2.8 Multiple Fractal Elements Represented by D1 and D2 [Source: Orford and Whalley (1983)] Orford & Whalley (1983) used fractal dimension to quantify the morphology of irregular-shape particles. They defined the boundary of the two fractal models as T r (where T r =S/H max S is the step length at the boundary and H max is the particles A-axis length). Three types of fractal combination were observed from the relative steepness of fractal element slopes. Type I was the standard single fractal element which usually refers to the gross shape of particle outline. Type II exhibited more irregular fractal structure and no prominent edge texture was observed in type II. Type III showed a concave fractal assemblage with marked irregular edge texture. According to many researchers, the problem of having two different fractal elements in the M-R plot can be solved by segmenting the M-R plot into two components, one of them represents represent the gross shape of the particle and the other represents small scale details or surface texture. Kennedy & Lin (1992) mentioned that there are two problems with this approach, first is to identify the hinge point separating the two components. The inflection point occurs at different step lengths for each object. Therefore, selection of a single hinge point cannot represent some fractal components properly. The second problem is that it may be necessary to consider more than two fractal components in case of complex plot. 14

PAGE 29

These problems can be overcome by segmenting the M-R plot into a larger number of fractal components, each of which is associated with the information about particle shape at a specific scale (Figure 2.9). The plot was divided into 10 straight line segments, the first representing the gross shape of the particle and the last representing finer scale features and the fractal dimension of each line segment was calculated. So each grain in a sample was represented by 10 fractal components that are analogous to the 24 Fourier harmonics and the dimension of each fractal component was analogous to the amplitude of the Fourier series (Kennedy & Lin, 1986). From their study, it was concluded that the fractal-based approach can be considered for the shape characterization of sedimentary particles. At the same time, research needs to be done to better understand the usefulness of this technique in discriminating the particles of vastly different shapes. -0.027-0.0010.0250.051-2.11-1.75-1.39-1.03-0.67ln (S)ln (P) 2 3 4 5 6 7 8 9 10 1 Figure 2.9 M-R Plot for a Sedimentary Particle [Source: Kennedy and Lin (1992)] 15

PAGE 30

The studies suggest that the irregular particle shape can be characterized by the use of fractal analysis and the power of this technique improves as the particle outline becomes more complex and irregular (Orford and Whalley, 1983). However, Fractal dimension is a measure of surface texture rather than the overall particle morphology, hence, the large scale surface features (gross shape) cannot be quantified by the use of Fractal dimension only. Another limitation of this approach is the difficulty in defining the range of fractal length (Dodds, J., 2003). 2.1.1.4 Fourier Shape Descriptors Fourier transform is a mathematical technique that provides an alternative approach to characterize two-dimensional particle shape by fitting a Fourier series on the unrolled particle outline. The grain shape can be analyzed by ( ,R ) Fourier method in closed form (Ehrlich & Weinberg, 1970), where the outline of a two-dimensional soil particle is traced out as shown in Figure 2.10. The equation of the profile is: NnnnnBnARR10)]sin()cos([)( (2.8) where, )( R is the radius at angle is the total number of harmonics, is the harmonic number, and are the coefficient, giving the magnitude for each harmonic (Bowman et al., 2001). One drawback with this approach is the possibility of re-entrant angles where the radius intersects the particle outline twice resulting in two possible values of radius N n nA nB R at a particular angle (Figure 2.10). Fourier descriptor method (Clark, 1981) is a possible alternative to the above method. In this method the particle outline is traversed in the complex plane at constant speed. The complex function used in the analysis is: 2/12/2sin2cos)(NNnnnmmMnmiMnmibaiyx (2.9) where, ( y x ) are the coordinates of the particle outline; is the total number of descriptors; n is the descriptor number; N M is the total number of points describing the particle; is the index number of a point on the particle; a and b are the coefficients for m 16

PAGE 31

each descriptor; and i is the imaginary unit number (Bowman et al., 2001). This method doesnt suffer from the re-entrant angle problem. In this research, particle shape is characterized by the Fourier mathematical technique. R1() y x R2() Figure 2.10 Fourier Analysis in Closed Form [Source: Bowman et al., 2001] 2.1.2 Shape Descriptors in Three Dimensions Three-dimensional descriptors commonly used to characterize particle shape include the length ratios of orthogonal axes (Krumbein, 1941; Yudhbir & Abedinzadeh, 1991), sphericity (Wadell, 1932; Krumbein, 1941). Sphericity is the ratio of surface area of a sphere of the same volume as the shape, to actual surface area of the shape (Wadell, 1932). Krumbein (1941) defined the sphericity as the ratio of particle volume to that of the smallest circumscribing sphere. Sphericity and roundness are two different morphological properties. Sphericity is related to the form and elongation, while roundness is related to angularity and surface roughness. Therefore, a spherical particle may have low roundness value if the surface is rough (Bowman et al., 2001). Conversely, a non-spherical particle can be perfectly round in shape (ellipse-shaped particle) and equidimensional particles (cube or hexahedron) can be very angular (Cho et al., 2006). Lees (1964) described the shape of the aggregates by a shape factor (F) and the sphericity () and these descriptors were defined in terms of the flatness and elongation ratio. The flatness ratio (p) is the ratio of the short length (thickness) to the intermediate 17

PAGE 32

length (width) and the elongation ratio (q) is the ratio of the intermediate length to the longest length (length). The shape factor was defined by equation 2.10 which is the ratio of the elongation ratio and the flatness ratio (Uthus et al.,2005). F = p/q (2.10) The sphericity of a particle was also expressed by the flatness and elongation ratios as shown in Equation 2.11. )21(216)1(1)32(8.12qpqpqp (2.11) Rao et al. (2002) described a three-dimensional descriptor, called Angularity Index by obtaining an angularity value for each of the three 2-D images acquired from the three views using the image analysis procedure. Then the angularity was calculated as a weighted average of all three views. Ang. (front). Area (front) + Ang. (top). Area (top) + Ang. (side).Area(side) AI Particle = Area(front) + Area(top) + Area(side) (2.12) Where, AIParticle is the Angularity Index of the particle. The unit for AI is degree. Masad et al. (1999) determined the Surface Texture (ST) by fine aggregates by erosion-dilation technique. Erosion is a morphological operation which causes an object shrink by one pixel along the boundary. Dilation is the reverse of erosion, where the boundary of an object is dilated or grown by a layer of pixels. ST is defined by the area lost due to the erosion-dilation operation as a percentage of the total area of the original image. ST = [(A 1 A 2 ) 100] / A 1 (2.13) Where, A 1 and A 2 are the areas of the two-dimensional image before and after erosion-dilation respectively. The surface texture of an object was calculated as a weighted average of all three views. 18

PAGE 33

ST (front) Area (front) + ST (top) Area (top) + ST (side) Area(side) ST Particle = Area (front) + Area (top) + Area (side) (2.14) Where, ST Particle is the Surface Texture of the particle, ST is the Surface Texture for one view and Area is the area for one view. 2.1.2.1 Representing Grain Shape in Spherical Coordinates Three-dimensional surface of any particle can be characterized by ),( RR in closed form (Schwarcz & Shane, 1969) where are angles measured from two perpendicular axes intersecting at the centroid of the particle ( 20 ; 0 ) and R is the radial distance from the centroid to a point on the surface. For such surfaces ),( R can be represented by a set of spherical harmonics of the form: ],),(0),([),(lmmlYmlBemlYmlAR (2.15) where and are the spherical harmonics based on the Legendre function (Morse & Feschbach, 1953). The coefficients and are obtained as follows: )(cos)cos(mlemlPmY )(cos)sin(0mlmlPmY mlP mlA mlB 200sin),()!()!(4)12(dmlYRdmlmlmn (2.16) 0mlforYmlBmlforYmlA 10 2n (=1, 2, ) n The spherical harmonics are the angular portion of the solution to Laplaces equation in spherical coordinates. Spherical coordinates are a system of curvilinear coordinates, describing positions on a sphere (Figure 2.11). 19

PAGE 34

Figure 2.11 Spherical Harmonic Transform [Source: http://mathworld.wolfram.com/SphericalCoordinates ] The equations of transformation between Cartesian and spherical coordinates are as follows: cossinRx (2.17-a) 2/1222)(zyxR sinsinRy 2/12221)(coszyxz (2.17-b) cosRz 2/12212/1221)(cos)(sinyxxyxy (2.17-c) Three-dimensional particle surface can be quantified using Spherical Harmonic Transform (3-D equivalent of 2-D Fourier transform). Cartesian coordinates (x, y, z) of the each voxel are transformed into spherical coordinates ),,( R using equations (2.17-a) to (2.17-c). In this case the object surface is scanned by the variation of two parameters ( and ) where 0 and 20 A suitable sampling interval should be chosen for proper characterization of the particle shape. An efficient algorithm of spherical harmonic series has been developed by Garboczi (2002) and implemented by Masad et al. (2005) on various three-dimensional particle shapes. 20

PAGE 35

21 2.2 Relationship Between Grain Size and Shape Particle morphology plays a very important role in understanding the micromechanical behavior of cohesionless soils. Grain shape de pends on various factors such as source of material, mineralogical composition, distance of transport and environmental conditions affecting formation of deposit. In order to ge nerate and reconstruct particle assemblies of highly irregular geometric shapes of a par ticular sand sample, the relationship between grain size and shape needs to be evaluate d. For example, size-shape relationships are necessary to generate represen tative assemblies of angular pa rticles for discrete element modeling simulations. Various research studies have been documented in the literature describing the relationship between particle si ze and particle shape. The de pendence of part icle shape on particle size was investigated by Russell and Taylor (1937), Pollack (1961), Ramez and Mosalamy (1969), Wadell (1935), Pettij ohn and Lundahl (1934), McCarthy (1933), Inman (1953), Inman et al. (1966) and Conolly (1965) and these studies demonstrated a decrease in roundness with a de crease in particle size for in tertidal sands (Balazs, 1972). Banerjee (1964) conducted a study to evaluate size-s hape relation and observed that the finer grains are more rounded than the coarser ones (Figure 2.12) and it was concluded that the negative correlation between grain size and roundness of the grains is due to two different sources of sands (Pe ttijohn, 1957). The study also suggested that the finer rounded particles were generated from a mature pre-existing sedimentary rock whereas the coarser angular particles were originated from nearby freshly-weathered igneous and metamorphic rocks (Banerjee, 1964). A reverse relationship was found in a study (Yudhbir and Abedinzadeh, 1991) where a relationship was established between av erage value of particle angularity and the grain size for each sieve fraction (Figure 2.1 3). The study proposed that the particle angularity decreases (or roundne ss increases) with size which was also suggested by Twenhofell (1950), Folk (1978), Khalaf and Gharib (1985).

PAGE 36

0.250.2750.30.3250.350.3750.40.42500.511.5Grain Size (mm)Mean Roundness (visual) 2 Figure 2.12 Mean Roundness Values of Eight Samples Plotted Against Mid-Points of Size Grades [Source: Banerjee, 1964] 051015202500.20.40.60.81Grain Size (mm)Angularity Ganga Kalpi San Fernando 5 San Fernando 6 San Fernando 7 Lagunillas (94A+94B) Lagunillas (94A-M19) Figure 2.13 Relationship Between Particle Angularity and Particle Size [Source: Yudhbir and Abedinzadeh (1991)] 22

PAGE 37

23 Another study was conducted by Goudie an d Watson (1981) to investigate the roundness of quartz grains from different dune areas around the world and the study revealed that the majority of the sample s were sub-rounded and more angularity was observed in smaller grains compared to larg er ones and the grain roundness varied from one dune location to the next. In addition, it is suggested that the shape of dune sand particles depends on the transport and sediment ation conditions, as well as the nature and origin of the material (Thomas, 1987). In a study conducted by Mazzullo et al. (1992) grains were divided into thre e bins based on the grain size distribution and for each bins higher order harmonics were used to study th e effect of grain size on grain roundness for increasing distance of transport. In that st udy, no major variation in grain roundness was observed among the three bins. The available literature exploring the re lationship between grain size and grain shape sometimes fails to demonstrate consistent results. In various research studies it has been observed that roundness of sand grains is extremely susceptible to abrasion and wear to which particles are subjected duri ng transportation by wind or water (Krumbein, 1941). An increase in roundness of very coarse sand (Plumley, 1948) and a slight decrease in roundness of fine sand (Russell a nd Taylor, 1937) with distance of transport by fluvial action were documented in the li terature, whereas Pollack (1961) reported negligible changes in roundness in the direction of transport in the South Canadian River (Balazs, 1972). The shapes of individual part icle in a composite soil sample also depend on the extent of gradation. Likewise, an increase in roundness was observed when sand particles are separated from grav el during segregation by tidal currents since the presence of gravel results in de crease in roundness of sand (B alazs, 1972; Anderson, 1926; Russell, 1939; Twenhofel, 1946). Abrasion of sand grains also depends on the environment in which they are being transported. It has been observed in literat ure (Kuenen, 1960) that abrasion (in terms of weight loss) of quartz grains in aeolian e nvironment is 100 to 1000 times more than that in fluvial environment over the same distance of transport because sand grains have to resist more friction in air than in water (Sorby, 1877). Another fact or controlling grain shape is the mineralogical composition of the individual grain. Hunt (1887) indicated that

PAGE 38

24 quartz and feldspar are highly resistant to abrasion. Moreover prolonged transport results in abrasion in feldspar and hence feldspar is expected to be less rounded than quartz (Balazs, 1972). Mazzullo et al. (1986) suggested that the shapes of quartz grains can be influenced by different mechanical (abr asion, fracturing, grindi ng and sorting) and chemical (silica dissolution and precipitation) processes and the variability in resulting grain shapes is likel y to be dependent on the variation in mechanical and chemical processes to which the grains are subjected. In this researc h, Fourier descriptors are used to verify any existing relations hip between grain size and shape. 2.3 Modeling Particle Shape in Two Dimensions Discrete element method (DEM) is being cons idered as a significant achievement in the area of micromechanical modeling. The recent development of DEM has made it possible to model particle morphology in two and th ree dimensions and examine soil behavior from micromechanical standpoint. The Discre te Element Method was first developed to model rock slopes (Cundall, 1971) and the mechanical behavior of 2-D assemblies of circular discs (Strack and Cundall, 1978). Later, the method was extended to three dimensions to model 3-D assemblies of s pheres (Cundall & Strack, 1979) by the program TRUBAL. Extensive research ha s been conducted to study the constitutive behavior of coarse grain soil using the modified versions of TRUBAL. In DEM, material is modeled as a random assembly of discrete elements interacting with each other through contact forces. 2.3.1 Existing Methods of Modeling Irregular Particle Shape The DEM tool has been adopted by numer ous researchers to study the mechanical response of granular soil from macroscopic to microscopic level (Mustoe et al., 1989; Williams & Mustoe, 1993). Traditional approaches in DEM modeled soil mass as an assembly of discs or spheres (Cundall a nd Strack, 1979). The first DEM code, BALL was introduced by Strack and Cundall (1978), wh ere a two-dimensional system was modeled as an assembly of discs. Following this procedure, various DEM formulations were

PAGE 39

25 proposed during the last two decades using ci rcular and spherical particles, such as TRUBAL (Cundall and Strack, 1979), C ONBAL (Ng, 1989; Ng and Dobry, 1991), GLUE (Bathurst and Rothenburg, 1989), DISC (Ting et al., 1989), DMC (Taylor and Preece, 1989) and others. The circular or s pherical particles have a higher tendency to rotate compared to the actual particle. Hence the angle of internal shearing resistance of the material comprising of circular or spherical particles will be much less than that of actual material and the use of spherical par ticles in discrete element modeling simulation would be too idealized to study the micros copic behavior of soil mass (Lin and Ng, 1997). To overcome these limitations and to bett er understand the soil behavior through numerical simulation, different modeling t echniques were proposed in many research studies where the non-circular particle outlines were approximated by various mathematical functions, such as ellipses (Ting et al, 1993; Ng, 1994), super-quadratics (Williams and Pentland, 1992; Cleary, 2000), and continuous circular segments (Potapov and Campbell, 1998) to model highly irregular particle shape. Barbosa and Ghaboussi (1992) and Matuttis et al (2000) suggested polygonshaped particles which is a more realistic representation of modeli ng irregular particle shape. However, the contact detection algorithm was very time consuming for polygonshaped particles (Jensen et al., 1999) and the method is computationally intensive as the particle outline becomes more complex and ir regular, especially in three dimensions. Ting et al. (1993) developed an algorithm for DEM simulation using twodimensional ellipse-shaped particles to comput e particle-to-particle and particle-to-wall contacts and good agreement was observed between the numerical simulation and the behavior of real soil. Though ellipse-shaped particles have fewer tendencies to rotate compared to circular particles, the shape of irregular particle c ould not be represented accurately by ellipse. Potapov and Campbell (1998) used oval-shap ed particles in order to generate representative assemblies in DEM environm ent where ellipse was approximated by oval shape whose boundary was determined by four circular arches of two different radii that are joined together in a continuous way. More complex shapes can be reproduced by

PAGE 40

26 changing the radii of the arches. Though the procedure was computationally efficient, the application of the method in th ee dimensions was not verified. Favier et al (1999) modeled axisymmetri cal particles as multi-sphere discrete elements by using overlapping spheres with fixed rigidity. The method was capable of modeling any axisymmetrical sh ape, however highly angular particles cannot be modeled properly using the procedure. 2.3.2 Modeling Angular Particles as Clusters Jensen et al. (1999) proposed a new cluste ring technique where a number of circular discrete elements were clumped together in a semi-rigid configuration to capture the shape of irregular particle (Figure 2.14). The main concept behind this clustering technique is that each cluster rotates and translates as a rigid body. The relative translation and rotation among the discs within a cluster can also be prevented by enforcing kinematics restrictions on discs fo rming the cluster (Thomas and Bray, 1999). In both methods, only non-overlapping elements were used within each cluster and the number of discs within a cluster was limited to three or four to decrease computation time. Therefore, the simulated particle outlines did not resemble that of actual particles. 2.3.3 Overlapping Discrete Element Clusters Ashmawy et al (2003) proposed the Overla pping Discrete Element Cluster (ODEC) technique to model angular particle shap es accurately in two dimensions by using discrete element modeling code PFC 2D and Itascas software-specific programming language, Fish In the ODEC method, two-dimensi onal particle shape was modeled by clumping a number of overlapping discs within the particle boundary so that the resulting outline resembles the outline of the actual particle (Figure 2.15). The ODEC method is computationally efficient, because the bui lt-in clump logic cannot detect contacts between disc elements belonging to the sa me clump. The number of overlapping discs needed to accurately model the irregular particle shape depends on the degree of nonuniformity in the original particle shape and angularity, the desire d level of geometric accuracy and the required computation time limit (Ashmawy et al., 2003). It was

PAGE 41

Figure 2.14 (A) Outline of Sand Particle, (B) DEM Disc Element Superimposed Over Sand Particle, (C) DEM Disc Particles are Joined Together in a Rigid Configuration (Cluster), (D) Several Possible Combination of Discs to Form Clusters [Source: Jensen et al. (1999)] Figure 2.15 Disc Elements Inscribed within a Particle Outline to Capture the Shape [Source: Ashmawy et al. (2003)] 27

PAGE 42

observed that ten to fifteen discs are sufficient to capture the shape of a particle accurately. Due to overlapping, the density scaling should be necessary for each disc belonging to a particular clump so that the mass of the particle remains proportional to the area. An approximate method was proposed by Ashmawy et al. (2003) to scale the density of overlapping discrete elements as follows: pdpdAA (2.17) where d is the density of the discs, is the area of the particle, is the sum of the areas of the disc elements and pA dA p is the density of the particle. Equation (2.11) does not guarantee the moment of inertia and the center of mass of the model particle to be identical to those of the actual particle. Sallam (2004) introduced a modification to the ODEC method where the compatibility of the particle centroid and inertia was satisfied after generating all the discs inside the particle. The ODEC method was implemented within PFC 2D by means of a series of Fish functions that convert a particle assembly of discs into their corresponding angular particle as follows: Particles were first generated as circular discrete elements within the desired range of grain sizes using the built-in particle generation techniques. Each circular particle was then transformed into its angular equivalent by using the shape conversion algorithm that replaced each circular outline with a corresponding set of circular discrete element cluster, selected randomly from the particle shape library. A random rotation between 0 and 360 was applied to each transformed particle to ensure uniform particle orientations within the assembly. (Ashmawy et al., 2003). Figure 2.16 shows a random assembly of circular particles generated in PFC 2D and transformed to their equivalent angular shapes using the ODEC technique. 28

PAGE 43

Figure 2.16 Random Assemblies of Eight Circular Particles (Left) and the Transformed Equivalent Angular Particles (Right) [Source: Ashmawy et al. (2003)] Sallam (2004) experimentally verified the ability of DEM using the ODEC technique developed by Ashmawy et al. (2003) to model the behavior of irregular particle shapes. An experimental set-up was built to study the translations and rotations of particles and inter-particle contact resulting from external disturbance. Good agreement was observed between experimental results and numerical simulations. In the current research the two-dimensional particle shapes are modeled using the ODEC technique. An algorithm is developed to automate the ODEC technique. 2.4 Modeling Particle Shape in Three Dimensions Most of the available literature on modeling irregular particle shape is limited to two dimensions with minimal progress in the three-dimensional domain. Lin and Ng (1997) developed a three-dimensional DEM code, ELLIPSE3D where the three-dimensional irregular particle shape was approximated by ellipsoid. Numerical simulations were performed to study the mechanical behavior of mono-sized particle arrays using the ELLIPSE3D program. The use of non-spherical particles in discrete element modeling showed an improvement in the results of numerical simulations. However, the highly irregular three-dimensional particle shape cannot be modeled accurately using ellipsoid. Another DEM formulation was proposed by Ghaboussi and Barbosa (1990) where three-dimensional granular particles were modeled as polyhedron. None of the methods used the shape of real sand particles for DEM simulation. 29

PAGE 44

Matsushima (2004) first suggested a 3-D image-based method to model irregular particle shapes in three dimensions. 3-D images of Toyoura sand were obtained with a micro X-ray CT and they were directly converted into Discrete Element models (Figure 2.17). A number of primitive elements of different sizes were placed within the particle surface and a virtual attraction was assumed between each point on the grain surface and the element closest to the point. Due to this attraction, elements moved from their initial position and increased or decreased in size to reduce the distance from the surface point and the procedure was continued until an optimized converged solution was obtained (Matsushima, 2004). The accuracy of the model was estimated and an average error of 5.8% was found in terms of grain radius and the volumes of most of the modeled grains were 10 to 15% less than that of actual grains. The volumes of some grains were even bigger than the original volumes of the corresponding grains and these errors were considered as an inaccuracy of the modeling technique. Figure 2.17 Virtual Force Acting on the Elements [Source: Matsushim a 2004] In this research, the ODEC m e thod (d eveloped by Ashm awy et al., 2003) is extended to three dim e nsions. A number of ove rlapp i ng sph e ric a l discre te elem ents is clum ped together within the particle volum e. The num b er of spheres necessary to cover the particle volum e depends on the overall sh ape and angularity of the particle. Threedim e nsional shapes are implem e nted within 3-D DEM code, PFC 3D by us ing Itascas software-specific programming language, Fish that converts a particle assem b ly of spheres into their equivale nt angular particles. 30

PAGE 45

2.5 Effect of Particle Shape on Shear Strength Behavior of Cohesionless Soil Shear force in cohesionless soil is derived from the frictional resistance of soil which depends on the inter-particle friction, particle interlocking, packing density, grain crushing, rearrangement and dilation during shearing. These factors can also be influenced by inherent soil properties, such as particle size, shape and surface roughness. Grain shape is one of the major contributing factors that affect the mechanical behavior of granular assembly. Shear strength and liquefaction characteristics of granular soil depend on particle size, grain size distributions, shape and surface texture of the individual grains. Granular packing which is governed by the void ratio of the assembly is another important factor influencing the shear strength behavior of soil (Holtz and Kovacs, 1981). The maximum (loosest state) and minimum (densest state) void ratio of a soil mass depend on the grain shape and grain size distribution of the assembly of grains. Early research found an increase of maximum (e max ) and minimum (e min ) void ratio and void ratio difference (e max e min ) with increasing particle angularity or decreasing roundness and sphericity (Youd, 1973; Cho et al., 2006; Fraser, 1935; Shimobe and Moroto, 1995; Miura et al., 1998; Cubrinovski and Ishihara, 2002; Dyskin et al. 2001; Jia and Williams, 2001; and Nakata et al., 2001). Based on several experiments published in the literature, the angle of internal friction ( ) decreases with an increase in void ratio (Zelasko et al., 1975; Shinohara et al., 2000). Therefore, the shear strength of soil also decreases since is a measure of shear strength of cohesionless soil. The angle of shearing resistance of soil can also be influenced by the angularity (or roundness) and the surface texture of the individual grains. The increase in angularity and surface roughness of the soil particles results in an increase of (Zelasko et al., 1975; Alshibi et al., 2004). A reverse relationship was documented in literature where the void ratio was increased with increasing particle angularity (Jensen et al, 2001). Angular particles cannot produce dense packing since the grains are separated by sharp corners (Dodds, J., 2003). Conflicting knowledge is available in the literature describing the relationship between particle size and the angle of internal friction. Koerner (1970) observed a decrease in with increasing mean grain size. In another study, the authors demonstrated a decrease in void ratio with increasing grain size whereas the friction angle reduces or 31

PAGE 46

remains almost constant for each sand sample. Though the larger grains show greater initial interlocking, it is compensated by the greater degree of grain crushing and fracturing due to the greater force per contact of larger grains (Lambe & Whitman, 1969). Norris (1976) and Zelasko et al. (1975) also suggested that the angle of internal friction of soil is not influenced by particle size considering the void ratio, angularity and roughness remaining constant (Jensen et al., 2001), rather the shear strength of soil is greatly influenced by gradation or particle size distribution (Zelasko et al. 1975). The relationship between grain size and angle of internal friction can be explained by the phenomenon of interlocking. For example, the angle of internal friction of well graded soil is higher than that of poorly graded sand, because the smaller size particles fill the void spaces between larger size particles; hence the void ratio is reduced resulting in an increase in the strength of soil mass. The effect of particle shape on void ratio was investigated by Zelasko et al. (1975) and the study demonstrated an increase in shear strength and value with a decrease in particle roundness. More interlocking is observed between angular grains and hence the angular particles are found to exhibit more shearing resistance than do rounded particles. Sukumaran and Ashmawy (2001) conducted a study to evaluate the relationship between shear strength and shape and angularity factor and found that the large-strain drained friction angle increases with an increase in shape and angularity factor. An increase in large-strain angle of shearing resistance with increasing surface roughness was also observed by Santamarina and Cascante (1998). In general, dense specimen with angular particles will provide more resistance to shearing than rounded particles (Shinohara et al., 2000) due to increase in interlocking effect. The shearing resistance of soil is developed due to particle rotation and translation (rolling and sliding). Frictional resistance will increase if the particles are frustrated from rotating (Santamarina and Cascante, 1998). If the density of soil is low, particles are free to rotate which results in lower frictional resistance to shearing. In densely packed soil (low void ratio), higher density and higher coordination number (number of contacts per particle) hinder rotation, causing slippage at particle contacts and this will results in dilation and thus an increase in the shearing resistance of soil (Santamarina and Cascante, 1998). The two major components influencing the shearing resistance of granular soil are 32

PAGE 47

dilatancy (developed from particle rearrangement and interlocking) and interparticle sliding resistance (Taylor, 1948; Santamarina and Cascante, 1998; Alshibi et al., 2004). Dilatancy of granular soils is defined as the change in soil volume during shear and dilatancy of granular soil is greatly influenced by the angularity of the grains, void ratio and confining pressure (Chen et al., 2003). Dilation is usually represented by the angle of dilation ( ) which is defined as the ratio of volumetric strain rate to shear strain rate. Dense sand under undrained condition exhibits strain hardening behavior. After an initial tendency to contract, dilation starts and causes the pore pressure to decrease and effective stress to increase. Liquefaction susceptibility of granular soil also depends on particle size, shape and size distribution. Poorly-graded sands with rounded particles are more susceptible to liquefaction than well-graded sands with angular particles since the shearing resistance of angular particles is higher due to high coordination number and thus the particle interlocking is stronger compared to rounded particles. Liquefaction is a phenomenon that may take place during earthquake shaking and is one of the major causes of ground failure in earthquakes. Loose, saturated, uniformly-graded, fine grain sands are very susceptible to liquefaction. Liquefaction takes place when seismic shear waves pass through a saturated granular soil layer, distorting its particle arrangements and breaking the inter-particle contacts. During earthquake loading, the shearing stage is so rapid that the pore water pressure cannot get enough time to dissipate, resulting in rapid increases of pore water pressure and accompanying very low effective stress such that the shear strength of the soil can no longer sustain the weight of the overlying structures and the soil flows like a viscous fluid. To mitigate the post earthquake hazards, areas susceptible to liquefaction should be identified. The effect of particle shape and angularity on shear strength, dilation and liquefaction characteristics of granular media in two dimensions has already been investigated by several researchers and the findings are documented in literature (Sallam, 2004, Ashmawy et al., 2003). Not much progress has been made in evaluating the dilation angle of granular soils in three dimensions and influence of three-dimensional particle shape on liquefaction behavior of cohesionless soil. To simulate the real 33

PAGE 48

34 micromechanical behavior of granular medi a, accurate characterization and modeling of particle shape in three dimensions are neces sary. The current study presents a detailed description of particle shape modeling techni que both in two and three dimensions using Discrete Element Method to evaluate the influence of particle shape on the shear strength behavior of granul ar assembly.

PAGE 49

35 CHAPTER 3 MATERIALS The previous chapter has offered a review of the conceptual developments in particle shape quantification and modeling techniques. Th e current chapter presents the properties of various granular materials collected from different geographical locations and those obtained from various sour ces in the literature. 3.1 Sand Samples Collected for the Present Study A wide variety of natural and processed sand samples having different roundness and angularity are collected from various locati ons around the world in conjunction with the current study. The intent was to obtain materials from as wide a geographical coverage as possible so that they would be more likely to be different in mean grain size, size distribution, and morphology due to differences in the deposition process. The sand samples collected for the current study encompass natural sands from beaches, rivers, dunes and manufactured crushed sands. The tw o-dimensional projection images of the sand samples collected for the study are documented in Appendix A. 3.1.1 Sample Selection Procedure A detailed description of different sand samples obtained from various sources in the literature and their engineering properties such as gradation (mean grain size, uniformity coefficient, coefficient of curvature), p acking (minimum and maximum void ratio and dry density) are documented in the form of a spreadsheet. These parameters can be a useful source for selecti ng materials for the present study. Figure 3.1 through Figure 3.8 present the variation of mi nimum and maximum void ratio for different types of sand samples.

PAGE 50

0.30.50.70.91.11.31.51.70.50.70.91.11.31.51.71.92.1e maxe min Kogyuk Erksak Syncrude TS Toyoura Leighton Buzzard Massey Tunnel Monterey Alaska Nerlerk Ticino Blasting sand Glass beads Cambria Hokksund Chattahoochee River Medium grained Silica Chiba Mersey River Figure 3.1 Variation of Minimum and Maximum Void Ratio (Group # 1) 0.30.50.70.91.11.31.51.70.50.70.91.11.31.51.71.92.1e maxe min Crushed Silica Brasted River Portland River Chonan Silty Kiyosu Lagunillas Lornex Tia Juana Silty Hostun RF Brenda Kizugawa Echigawa Abashiri Tottori Sado Sumaura Nevada Sydney Ottawa Santa Monica sand Likan sand Figure 3.2 Variation of Minimum and Maximum Void Ratio (Group # 2) 36

PAGE 51

0.30.50.70.91.11.31.51.70.50.70.91.11.31.51.71.92.1e maxe min Douglas Lake Calcareous, Gujarat Ganga sand Kalpi sand Glazier Way sand Mortar sand Agsco sand Jebba sand Colorado sand Quiou sand Cositas dam sand Till sand LSF dam sand Lytle sand, colorado Enewetak coral sand Sacramento river sand Hokksund sand Karlsruhe sand Yatesville sand Fontainebleau sand Daytona Beach sand Michigan Dune sand Figure 3.3 Variation of Minimum and Maximum Void Ratio (Group # 3) 0.30.50.70.91.11.31.51.70.50.70.91.11.31.51.71.92.1e maxe min Yurakucho sand Kenya sand Catania sand Dog's Bay sand Kingfish sand Halibut sand Ballyconneely sand Bombay Mix sand Amami sand Mol sand Berlin sand Chengde sand Chiibishi sand Sao Paulo sand Oklahoma sand Banding sand Ham river sand Fraser river sand Loire River sand Mailiao sand 37 Figure 3.4 Variation of Minimum and Maximum Void Ratio (Group # 4)

PAGE 52

0.50.70.91.11.31.51.71.92.100.20.40.60.811.2e max e mine max Kogyuk Syncrude TS Toyoura Massey Tunnel Monterey Sacramento River Nerlerk Ticino Blasting Sand Glass Beads Cambria Sand Hokksund Banding Chattahoochee River Medium grained silica Chiba Portland River Ottawa Figure 3.5 Variation of Maximum Void Ratio with e max e min (Group # 1) 0.50.70.91.11.31.51.71.92.100.20.40.60.811.2e max e mine max Nevada Crushed Silica Chonan Silty Fraser River Kiyosu Lagunillas Lornex Tia Juana Silty Hostun RF Brenda Kizugawa Echigawa Tottori Sado Sumaura Alaska Leighton Buzzard Erksak san d 38 Figure 3.6 Variation of Maximum Void Ratio with e max e min (Group # 2)

PAGE 53

0.50.70.91.11.31.51.71.92.100.20.40.60.811.2e max e mine max Douglas Lake Calcareous, Gujarat Ganga sand Kalpi sand Glazier Way sand Mortar sand Agsco sand Jebba sand Colorado sand Quiou sand Cositas dam sand Till sand LSF dam sand Lytle sand Enewetak coral sand Mersey river Sydney sand Ham river sand Abashiri sand Fontainebleau sand Santa Monica sand Likan sand Daytona Beach sand Michigan Dune sand Figure 3.7 Variation of Maximum Void Ratio with e max e min (Group # 3) 0.50.70.91.11.31.51.71.92.100.20.40.60.811.2e max e mine max Yurakucho sand Kenya sand Catania sand Karlsruhe sand Dog's Bay sand Kingfish sand Halibut sand Ballyconneely sand Bombay Mix sand Amami sand Mol sand Berlin sand Chengde sand Chiibishi sand Sao Paulo sand Oklahoma sand Brasted river sand Yatesville sand Loire River sand Mailiao sand Figure 3.8 Variation of Maximum Void Ratio with e max e min (Group # 4) 39

PAGE 54

Granular packing is represented by the void ratio of the assembly and the shear strength behavior of soil is influenced by the packing density of the granular mass (Holtz and Kovacs, 1981). The maximum and minimum void ratios of a soil mass depend on the shape of the individual grain and grain size distribution. Research had been documented in the literature describing the relationship between void ratio and angle of internal friction of cohesionless soil and the studies suggested an increase of maximum (e max ) and minimum (e min ) void ratio and void ratio difference (e max e min ) with increasing particle angularity or decreasing roundness and sphericity (Youd, 1973; Cho et al., 2006; Fraser, 1935). Based on several experiments published in the literature, a decrease in void ratio results in an increase in angle of shearing resistance ( ) and therefore an increase in shear strength of granular soil mass (Zelasko et al., 1975; Shinohara et al., 2000). Materials can be selected based on the above figures. For example, Chiba, Alaska, medium grained Silica, Lagunillas, Kizugawa, crushed Silica, Quiou, Jebba, Colorado, Till, Ganga, Glazier Way, Dogs Bay, Kenya, Ballyconneely, Kingfish, Sao Paulo, Oklahoma, Erksak, Leighton Buzzard sands can be some of the interesting materials to study since their void ratios fall a way above or a way below on the plot. Though it was intended to acquire these materials for the purpose of the present study, it was difficult to obtain all these materials from different parts of the world. Instead, the materials collected for the current study are easily available, but still cover a wide range of geographic locations and have various degrees of angularity and roundness. Therefore, it is expected that they are likely to be different in grain size, shape, mineralogical composition and other engineering properties. The properties of sand samples collected for this study are presented next. 3.2 Data Sets and Sample Characteristics Total 26 types of different sand samples are collected for the present study. Their locations and engineering properties are presented in table 3.1. Figure 3.9 through Figure 3.11 presents the particle size distribution for some of the materials collected. For some samples, sieve analysis is not performed because of insufficient amount of sample. 40

PAGE 55

41 Instead, the particle size distributions for those materials are obtained through image analysis. Table 3.1 Sand Samples Collected for the Study Gradation Type of Sand D 50 D 10 C u Location US-Silica, #1 Dry 0.270 0.130 2.462 Newport, NJ US-Silica, Std. Melt 0.280 0.140 2.286 Newport, NJ Daytona Beach 0.130 0.080 1.875 Daytona Beach, FL Rhode Island 0.250 0.190 1.421 Rhode Island Nice 0.200 0.073 3.014 Var River bed, Nice, France Fontainebleau 0.200 0.110 1.909 Fontainebleau, France Loire River 0.730 0.520 1.500 Loire River bed, Orlans, France Hostun 0.580 0.410 1.659 Hostun, France Toyoura Beach 0.200 0.170 1.235 Toyoura Town, Japan Indian Rocks Beach 0.220 0.160 1.500 Clearwater, FL Belle Air Beach 0.420 0.182 2.802 Clearwater, FL Clearwater Beach 0.180 0.095 1.947 Clearwater, FL Gulf Beach 0.220 0.170 1.353 Clearwater, FL

PAGE 56

42 Table 3.1 (Continued) Gradation Type of Sand D 50 D 10 C u Location Madeira Beach 0.200 0.105 1.952 Clearwater, FL Redington Shores 0.215 0.140 1.643 Clearwater, FL Belmont Pier 0.245 0.170 1.559 Long Beach, CA Boca Grande Beach 0.200 0.160 1.313 Cartagena, Columbia Tecate River 0.950 0.760 1.316 Tecate, Mexico Oxnard 0.800 0.525 1.600 Oxnard, CA Arroyo Alamar 0.900 0.600 1.550 Alamar River, a tributary of Tijuana River in Tijuana Rincon Beach 0.680 0.460 1.609 Beaches of Rincon, Puerto Rico Panama Malibu Beach 0.250 0.180 1.472 Malibu Beach, Gorgona, Panama Michigan Dune 0.345 0.260 1.385 Michigan Ala Wai Surfers Beach 1.050 1.010 1.054 Ala Wai Surfers Beach, Oahu, Hawaii Kahala Beach 0.700 0.500 1.520 Kahala Beach, Oahu, Hawaii Red Sea Dune 0.900 0.600 1.550 Suez, Egypt

PAGE 57

0204060801000.010.1110Grain Size (mm)Percent Finer #1 Dry Daytona Toyoura Michigan Oxnard Figure 3.9 Particle Size Distributions (Sand Samples, Group # 1) 0204060801000.010.1110Grain Size (mm)Percent Finer Rhode Island Hostun Loire River Clearwater Fontainebleau Figure 3.10 Particle Size Distributions (Sand Samples, Group # 2) 43

PAGE 58

0204060801000.010.1110Grain Size (mm)Percent Finer Std. Melt Kahala Tecate Red Sea Dune Long Beach Figure 3.11 Particle Size Distributions (Sand Samples, Group # 3) 0204060801000.010.1110Grain Size (mm)Percent Finer Ala Wai Rincon Panama Malibu Arroyo Alamar Boca Grande Figure 3.12 Particle Size Distributions (Sand Samples, Group # 4) 44

PAGE 59

45 CHAPTER 4 PARTICLE SHAPE CHARACTERIZ ATION AND QUANTIFICATION The previous chapter has discussed about the material selection procedure and engineering properties and locat ions of sand samples collect ed for the present study. The current chapter includes a detailed descripti on of microscope system, image analysis software and X-ray CT as we ll as the particle shape characterization and quantification techniques in two and three dimensions. 4.1 Characterizing Particle Shape in Two Dimensions An automated procedure is established to ch aracterize particle shape in two dimensions using photo microscopy and an image pro cessing software package from Media Cybernetics, Inc: Image-Pro Plus 5.1 and the add-ins Scope-Pro 5.0, Sharp Stack 5.0 and 3D Constructor 5.0. Different illumination sy stems and filtering techniques is also suggested (Rivas, 2005). A semi-automated routin e is implemented within the software to capture a large number of images of sand particles at a time. The microscopes used in this study are Motic SMZ-168 Stereo Microscope with Magnification of 0.75X to 5X (Figure 4.1) and Motic AE31 Inverted Microscope with Magnification of 4X to 40X (Figure 4.2) manufactured by Motic Instruments, Inc. Va rious sand samples collected for the present study are analyzed using the optical micr oscopes and digital camera system. Twodimensional images of part icle outline are obtained by orthogonal projection. The captured images of the sand particles are processed through successive steps of erosion, dilation and contrast enhancement technique using the image-processing software and different shape parameters in terms of asp ect ratio, roundness, diam eter, perimeter etc. are obtained.

PAGE 60

46 Figure 4.1 Motic SMZ-168 Figure 4.2 Motic AE-31 Inverted Stereo Zoom Microscope Microscope 4.2 Characterizing Particle Shape in Three Dimensions Characterizing the particle surface accurately in three dimensions is a challenging task. Serial Sectioning has been found to be a very accurate method for three-dimensional reconstruction of particle surface. In this destructive technique, the grains are embedded in epoxy-resin matrix and successive thin layers are removed using a polishing machine. The first slice of a particular thickness is removed and the remaining portion is observed in a microscope to acquire the image. The same process can be repeated to remove successive layers of same thickness and each time the images of the remaining portion of the grains are captured. Then all the stacks of images are combined to reconstruct three-dimensional shapes of the sand grains. Though reliable, the experimental procedure, especially the sample preparation is very time consuming and laborious. X-ray Computed Tomography is an alternative technique for three-dimensional reconstruction of particle shape where no prior sample preparation is necessary. X-ray CT is a completely nondestructive technique for visualizing internal structure of solid objects and obtaining digital information about their three-dimensional geometries and properties. The fundamental principle behind computed tomography is to acquire multiple views of an

PAGE 61

object over a range of angular orientations. Computed Tomography is based on the x-ray principle: as x-rays pass through the object, they are absorbed or attenuated, creating a profile of x-ray beams of different strength. This x-ray profile is captured by a detector within the CT to generate an image. Figure 4.3 SkyScan 1072 X-Ray CT System [ www.rowan.edu/colleges/engineering/clinics/shreek/database.html ] In the present study the three-dimensional particle shape is characterized using X-ray CT since it allows visualizing and measuring a complete 3D object without any special sample preparation. The machine is located at the Rowan University, NJ (Figure 4.3). Two-dimensional grayscale image stacks are obtained from the online geomaterial database developed at the Rowan University. The grayscale image stacks are then processed through successive steps of erosion and dilation to obtain binary image stacks. Three-dimensional coordinates (x, y, z) of boundary and internal voxels are extracted from the two-dimensional image stacks using Matlab 7.2. 47

PAGE 62

4.3 Quantification of Particle Shape In various research studies the particle shapes are quantified using different shape descriptors. A detailed description of shape descriptors can be found in Chapter 2. In the current study, particle shape is quantified using Fourier shape descriptors in two dimensions. 4.3.1 Particle Shape Quantification in Two Dimensions Fourier analysis is a mathematical approach to quantify grain shape in two-dimensions. In this research, two-dimensional signature descriptors of each shape are obtained by measuring the radial distance between the centroid and particle boundary at constant sampling interval, and Fourier transforms are used to obtain the spectral information of each shape. A starting point on the boundary is selected and periphery radii (R) are measured from the centroid to the boundary of the particle at a uniform sampling interval (Figure 4.4). This periodic function is then expanded in a Fourier series to obtain a set of harmonic amplitudes, each at a different frequency. The first harmonic indicates the deviation from the center of gravity. It can be represented as circularity. The second, third and fourth and fifth harmonics add a component of elongation, triangularity, sqaureness and pentagonality respectively. The nth harmonic contributes a sinusoidal wave of frequency n to the (Diepenbroek, et al., 1992). These shape descriptors can be a measure of angularity or roundness of grains. 0R R3R2R1132XY Figure 4.4 Fourier Transform on Particle Boundary 48

PAGE 63

The equation of the profile is: NnnnnBnARR10)]sin()cos([)( (4.1) Where, 200)(21dRR (4.2-a) 20)cos()(1dnRAn (4.2-b) 20)sin()(1dnRBn (4.2-c) N total number of discrete samples taken = 256; total sampling angle = 2 ; sampling interval = 02454.02562 N radian = 1.40625 degree; sf sampling frequency = 74367.401N cycles/radian; f frequency increment = 211 0.15916 cycles/radian = fundamental frequency of the Fourier series. In this study, total 256 data points are taken along the periphery of the grain at a sampling interval of 1.40625 degree and the first 128 shape descriptors (in terms of amplitude) are used to quantify the shape. Figure 4.5 and Figure 4.6 show the mean amplitude spectra for Tecate River sand and Daytona Beach sand respectively for the first 24 harmonics. FFT is performed on each grain of six different sand samples to quantify the grain shapes in terms of Fourier shape descriptors. These shape descriptors are used as parameters to determine the sample size of different sand samples and to explore the relationship between grain size and grain shape. 49

PAGE 64

00.040.080.120.160.21357911131517192123Harmonic NumberMean Amplitude Figure 4.5 Mean Amplitude Spectra for Tecate River Sand 00.040.080.120.160.21357911131517192123Harmonic NumberMean Amplitude Figure 4.6 Mean Amplitude Spectra for Daytona Beach Sand 50

PAGE 65

Meloy (1977) found a straight line relationship between log (descriptor number) and log (amplitude) above order 8. These higher order descriptors give an idea about the surface texture along the periphery of the grain. Surface texture is a measure of roughness of the particle. Figure 4.7 follows the same pattern as obtained by Bowman et al. (2001) where the surface roughness was quantified by slope and intercept. As explained by the authors, a higher intercept indicates a greater degree of roughness and a steeper gradient suggests a greater decay of roughness toward the finer scale. Therefore, this plot can be used to measure the surface roughness of sands. In Figure 4.7, the gradients are almost identical for all the sands but the Tecate River sand shows the highest intercept and Michigan Dune sand shows the lowest intercept. Therefore, Tecate River sand should possess highest surface roughness and Michigan Dune sand should have lowest surface roughness. -12-11-10-9-8-7-6-5-42.83.33.84.34.85.35.8Log base2 (Descriptor No.)Log base2 (Amplitude) Daytona Beach Kahala Beach Toyoura Beach Michigan Dune #1 Dry Tecate Figure 4.7 Variation of Harmonic Amplitude with Descriptor Number for Different Sand Samples 51

PAGE 66

52 CHAPTER 5 RELATIONSHIP BETWEEN GRAIN SIZE AND SHAPE The previous chapter has described particle shape characterization and quantification techniques in two and three-dimensions. The current chapter explores the relationship between grain size and shape of different na tural and processed sand samples. It also offers a methodology to determine the sample si ze that will be the representative of the population for each type of sand sa mple collected for the study. 5.1 Introduction In order to generate and r econstruct particle assemblies of highly irregular geometric shapes of a particular sand sample, the rela tionship between grain size and shape needs to be evaluated. For example, size-shape re lationships are necessary to generate representative assemblies of angular particles for discrete element modeling simulations. The present study develops a methodology to determine an optimum sample size for a given sand sample. Determination of sample size is important to verify any existing relationship between grain size and shape since size-shape relationship depends on the number of particles used in the analysis. Mo reover, design of an optimum sample size can save significant amount of resources. 5.2 Methodology to Determine Sample Size The present study focuses on determining samp le size (optimum number of particles within a particular sand sample) for different natural and processed sand samples using different shape descriptors such aspect ra tio, elongation, triangul arity and squareness. Aspect ratio is found to have greater influe nce on sample size determination than Fourier

PAGE 67

shape descriptors. The minimum number of particles needed to be to be analyzed for a particular sand sample depends on the variability of particle size and shape within that sand sample. Statistical analysis is performed to find out the sample size for sand samples from different locations based on relevant shape parameters like aspect ratio, roundness, elongation, triangularity and squareness. Sample size depends on the following factors: Confidence Level is the estimated probability that a population estimate lies within a given margin of error. Margin of error (E) measures the precision with which an estimate from a single sample approximates the population value. The margin of error is same as the confidence interval which is a range of values. For example, if x is the sample mean, x confidence will be the range of population mean. The margin of error is the maximum difference between the sample mean x and the population mean ExEx The value of population standard deviation. The formula for calculating the sample size for a simple random sample is: 22/Ezn (5.1) Where, is the standard normal variate; 2/z is the risk of rejecting a true hypothesis and it is called the significance level. For =0.05, confidence level is 95%. ( = 1.645 for 90% confidence level, 1.96 for 95% confidence level, and 2.575 for 99% confidence level). 2/z is the population standard deviation; is the sample size. The sample standard deviation, n 1)(2 nxxsi is the consistent estimator of the population standard deviation, so s can be replaced by s in equation (5.1). 53

PAGE 68

Figure 5.1 Standard Normal Distribution Table 5.1 presents the values of the variances of different shape descriptors for the sand samples. In the present study, sample size is calculated based on aspect ratio of the particle shape, because (1) more variation is observed in aspect ratio compared to other shape parameters for each sand sample; (2) the gross shape of the particles can be defined by elongation and in various research studies, the two-dimensional particle shapes were approximated by ellipse. Elongation is closely related to aspect ratio of a given shape. The higher the aspect ratio, the more elongated is the shape of the particle; (3) the magnitude of error in terms of aspect ratio is stabilized at a magnitude of 0.04 or below for all the sand samples tested whereas no specific error threshold can be established for elongation, triangularity and squareness since the magnitude of error for each of these parameters varies from one sand sample to the next. The magnitude of error is plotted against sample size for Toyoura sand in Figure 5.2 and Figure 5.3 presents the variation in change of error per unit sample with sample sizes for the same sand. It can be found from Figure 5.2, that the variation in errors is marginal at a magnitude of 0.04 and beyond and the same trend was observed for all other sand samples. Corresponding to the error magnitude of 0.04, the sample size is 90. The threshold for change of error per unit sample is chosen as .001or less (Figure 5.3) and the corresponding sample size is 70 and after that the change of error per unit sample is negligible and gradually approaching to zero. 54 Error as percent of mean (e p ) is also considered as a criterion for the determination of sample size (Figure 5.4) and a value of 5% is selected as a reasonable threshold for this experimental design and based on the value of e p a sample size of 50 is

PAGE 69

obtained. The maximum of these three values is selected as the required sample size for the particular sand sample. Therefore, a sample size of 90 was chosen for Toyoura sand. Table 5.2 provides the sample sizes estimated for different sand samples used in the analysis. Table 5.1 Values of Variances of Different Shape Parameters Type of Sand Aspect Ratio Elongation Triangularity Squareness Daytona Beach 0.077 0.006 0.002 0.0008 Toyoura Beach 0.045 0.006 0.003 0.0006 Tecate River 0.088 0.007 0.002 0.0009 Michigan Dune 0.054 0.006 0.001 0.0006 US-Silica #1 Dry 0.101 0.008 0.002 0.0008 Kahala Beach 0.112 0.010 0.001 0.001 00.040.080.120.160.204080120160200Sample SizeError Figure 5.2 Variation of Error with Sample Size for Toyoura Sand 55

PAGE 70

-0.01-0.008-0.006-0.004-0.00200.0020.00410-2030-4050-6070-8090-100110-120130-140150-160170-180Sample sizeChange of error per unit sample Figure 5.3 Variation of Change of Error Per Unit Sample with Sample Size 05101520050100150200Sample SizeError as percent of mean Figure 5.4 Variation of Error as Percent of Mean with Sample Size 56

PAGE 71

57 Table 5.2 Estimated Sample Size fo r Sand Samples Used in the Study Material Sample Size Hostun Sand 110 US-Silica Std. Melt 200 Boca Grande Beach Sand 230 US-Silica #1 Dry 150 Daytona Beach Sand 130 Loire River Sand 130 Toyoura Beach Sand 90 Oxnard Beach Sand 120 Kahala Beach Sand 170 Ala Wai Beach Sand 200 Michigan Dune Sand 100 Rincon Beach Sand 110 Arroyo Alamar River Sand 140 Rhode Island Sand 150 Tecate River Sand 150 Clearwater Beach Sand 150 Red Sea Dune Sand 90 Fontainebleau Sand 110 Long Beach Sand 130 Panama Malibu Beach Sand 70

PAGE 72

58 5.3 Relationship Between Grain Size and Grain Shape The relationship between particle size and sh ape for a given sand plays an important role in generating and reconstruc ting particle assemblies for micromechanical modeling. However, limited and conflicting knowledge is available in the literature describing the relationship between particle size and particle shape. The present study focuses on verifying any existing relationship between grain size and grain shape using different natural and processed sand samples. The findings can provide important information in understanding the nature of vari ation or lack thereof of shape parameters such as elongation, triangularity, squa reness with grain size. 5.3.1 Data Sets Total 6 sand samples are used to find out th e size-shape relationship and they encompass natural sands from beaches (Daytona, T oyoura, Kahala), rivers (Tecate), dunes (Michigan) and a manufactured crushed sand (#1 Dry). Figure 5.5 shows the twodimensional projection images of the sand samples and Table 5.3 summarizes the relevant engineering properties a nd locations of various materials. Table 5.3 Relevant Properties of Granular Materials Material No. of particles D 50 (mm) C u Location Daytona Beach 100 0.13 1.87 Florida Toyoura Beach 90 0.20 1.24 Japan Tecate River 130 0.95 1.32 Mexico Michigan Dune 100 0.33 1.5 Michigan US Silica #1 Dry 150 0.27 2.46 New Jersey Kahala Beach 170 0.70 1.52 Hawaii

PAGE 73

Daytona Beach Sand Toyoura Sand Tecate River Sand Michigan Dune Sand US-Silica #1 Dry Kahala Beach Figure 5.5 Two-Dimensional Images of Sand Samples 5.3.2 Fourier Shape Descriptors Fourier transform is performed to characterize grain shape in two dimensions by fitting a Fourier series to the unrolled particle outline. In the present study, grain shapes are quantified by ( ,R ) Fourier method in closed form (Ehrlich & Weinberg, 1970). Table 5.4 presents the average values of the first four shape descriptors for different sand samples and Figure 5.6 and Figure 5.7 show the mean amplitude spectra for Toyoura Beach sand and Michigan Dune sand respectively for the first 24 harmonics. 59

PAGE 74

Table 5.4 Average Values of Fourier Descriptors for Different Sand Samples Type of Sand Harmonic1 (Circularity) Harmonic2 (Elongation) Harmonic3 (Triangularity) Harmonic4 (Squareness) Daytona Beach 0.016378 0.168487 0.078809 0.052109 Toyoura Beach 0.016165 0.139724 0.100307 0.0518 Tecate River 0.017707 0.168221 0.082782 0.057527 Michigan Dune 0.010951 0.140109 0.063223 0.04587 US-Silica #1 Dry 0.017826 0.167989 0.088005 0.054268 Kahala Beach 0.015493 0.167344 0.080716 0.048941 00.040.080.120.161357911131517192123Harmonic NumberMean Amplitude Figure 5.6 Fourier Amplitude Spectra for Toyoura Sand 60

PAGE 75

00.040.080.120.161357911131517192123Harmonic NumberMean Amplitude Figure 5.7 Fourier Amplitude Spectra for Michigan Dune Sand 5.3.3 Grain Size Grain Shape Relationship The study aims to explore the relationship between grain size (mean diameter) and grain shape (shape descriptors) for a given sand sample. Figure 5.8 presents the frequency distribution of the shape parameters such as diameter, elongation, triangularity and squareness respectively. The distributions are very close to standard normal distribution. Figure 5.9 through Figure 5.12 shows the variation of shape descriptors with grain size for Tecate River sand and Table 5.5 summarizes the R-square values obtained from regression analysis. 61

PAGE 76

140012001000800 Diameter (Microns) 2520151050 Frequency 0.500.400.300.200.100.00 Elongation 2520151050 Frequency (A) Mean = 980.34, Std. Dev. = 146.57 (B) Mean = 0.168, Std. Dev. = 0.086 0.250.200.150.100.050.00 Triangularity 20151050 Frequency 0.200.150.100.050.00 Squareness 2520151050 Frequency (C) Mean = 0.083, Std. Dev. = 0.044 (D) Mean = 0.058, Std. Dev. = 0.03 Figure 5.8 Frequency Distributions of Shape Parameters: (A) Diameter, (B) Elongation, (C) Triangularity and (D) Squareness 62

PAGE 77

00.020.040.060.080.1600800100012001400Diameter (Microns)Circularity Figure 5.9 Variation of Circularity with Diameter 00.080.160.240.320.40.480.56600800100012001400Diameter (Microns)Elongation Figure 5.10 Variation of Elongation with Diameter 63

PAGE 78

00.040.080.120.160.20.246008001000120014001600Diameter (Microns)Triangularity Figure 5.11 Variation of Triangularity with Diameter 00.040.080.120.160.26008001000120014001600Diameter (Microns)Squareness Figure 5.12 Variation of Squareness with Diameter 64

PAGE 79

Table 5.5 Regression Analysis Results for Different Shape Descriptors Circularity Elongation Regression R-square Linear 0.0081 Logarithmic 0.0074 Polynomial (2 nd degree) 0.0097 Exponential 0.0301 Regression R-square Linear 0.0804 Logarithmic 0.0786 Polynomial (2 nd degree) 0.0813 Exponential 0.0727 Triangularity Squareness Regression R-square Linear 0.0001 Logarithmic 0.0002 Polynomial (2 nd degree) 0.0024 Exponential 0.0016 Regression R-square Linear 0.0093 Logarithmic 0.0104 Polynomial (2 nd degree) 0.0128 Exponential 0.00004 5.3.4 Summary and Discussion The current study focuses on the verification of any existing relationship between grain size and grain shape using six different natural and processed sand samples. To investigate the relationship, the first four Fourier descriptors are used to quantify the particle shapes. Though in previous literature a strong relationship has been documented between grain size and grain shape, it is evident from the current research that there is no relationship between grain size and grain shape in terms of circularity, elongation, triangularity or squareness for the sand samples analyzed. Considering the dependency of grain shape on various factors, such as origin and mineralogical compositions, distance of transport, mechanical and chemical processes, it is unlikely to obtain a universal relationship between grain size and grain shape for all sands. Even though the sands in the current study came from different sources and underwent considerably different depositional processes, no relationship was found between shape and size. The current findings have significant implications in terms of constructing discrete element model 65

PAGE 80

66 assemblies of angular particles (e.g., Ashm awy et al., 2003). Because no relationship exists between particle size and shape, shap es belonging to a particular sand sample can be randomly selected from the particle shape library, and resized according to the desired grain size distribution, irrespec tive of size. However, if a particular sand is found to exhibit a correlation between shape and size, it would be necessary to separate the particles into different groups or bins ba sed on their size in order to generate a representative assemblies of angular particles for discrete element simulation.

PAGE 81

67 CHAPTER 6 SKELETONIZATION AND OVERLAPPING DISCRETE ELEMENT CLUSTER ALGORITHM The previous chapter has presented a methodology to determine the sample size of various natural and processed sand samples. Th e analysis suggested that aspect ratio has greater influence on sample size determination compared to other shape descriptors. The relationship between grain size and shape has also been explored and no relationship was found for the sand samples analyzed. The current chapter explains th e skeletonization and Overlapping Discrete Element Cluster (ODEC) algorithms to model irregular particle shape. 6.1 Skeletonization of Grain Shape In the present study particle shape is modeled using th e ODEC technique proposed by Ashmawy et al. (2003) where the irregular particle shape is covered by inscribing a number of overlapping circular or spherical discrete elements to generate representative assemblies of angular particles for di screte element modeling simulations. To make the ODEC technique computationally efficient, skeletonization of the shape is necessary. Skeletonization is often called Medial Axis Transformation which is defined as the locus of centers of maximally inscribe d discs. In 3-D, the medial surface is the locus of centers of all maximally inscribed spheres. One of the methods to generate skeleton is by thinning the object with a set of structuring elements. Figure 6.1 shows an example of successive thinning of a set A by structuring element B. The thinning procedure is performed on A by one pass with B 1 Then the thinning is continued on the resulting image by one pass of B 2 and so on, until A is thinned with one pass of B n The process is then repeated until A can not be thinned further. Figure 6.1 (A) shows a set of structuring elements commonly used for thi nning and Figure 6.1(B) shows the successive

PAGE 82

1B 2B 3B 4B 5B 6B 7B 8B (A) A 1BA 2BA 3BA 4BA 5BA 6BA 8,7BA 3,2,1BA 3,2,1,8,7,6,5,4BA (B) Figure 6.1 Thinning Algorithm: (A) A Set of Structuring Elements, (B) Successive Steps of Thinning [Source: Gonzalez & Woods, 2003] steps of thinning with passes of different structuring element. Finally the thinned set is converted to m-connectivity to eliminate multiple paths. Figure 6.2 presents the skeleton obtained using this method. Although the method is accurate for generating the skeleton of a regular L shaped object, it can not generate accurate skeleton for irregular object, such as Fraser River sand grain. It can be seen from Figure 6.2(B) that part of the medial axis (marked in red) is not equidistant from at least two non-consecutive points on the outline. 68

PAGE 83

(A) (B) Figure 6.2 Skeleton in Two Dimensions: (A) L-Shaped Object, (B) Fraser River Sand Grain Contour thinning is another method to make skeleton of a binary region. In this method, the region points are assumed to have a value of 1 and the background points to have a value of 0. The procedure is based on two basic steps. The neighborhood arrangement is shown in Figure 6.3. Step 1: Contour point P 1 is marked for deletion if the following conditions are satisfied: (a) 2<=N(P 1 )<=6; (b) T(P 1 ) = 1; (c) P 2 P 4 P 6 = 0; (d) P 4 P 6 P 8 = 0 where, N(P 1 ) is the number of non-zero neighbors of P 1 i.e. N(P 1 ) = P 2 + P 3 + + P 8 + P 9 and T(P 1 ) is the number of 0-1 transitions in the ordered sequence P 2 P 3 P 4 P 5 P 6 P 7 P 8 P 9 P 2 Step 2: Conditions (a) and (b) remain the same, but condition (c) and (d) are changed to (c) P 2 P 4 P 8 = 0; and (d) P 2 P 6 P 8 = 0 Step 1 is first applied to each border pixel and if all conditions are satisfied, the point is marked for deletion. After processing all border points, the marked points are deleted (changed to zero). Step 2 is then applied to the resulting image in the same manner as step 1. The entire process is repeated until no further points are deleted. The resulting image will be the skeleton of the region. Figure 6.4 (A) and Figure 6.4 (B) show the skeletons obtained by using this method. Though the skeleton was accurate for regular T shaped object, the method could not generate accurate skeleton for irregular shape, for example Michigan Dune sand grain because the red portion in the skeleton is not equidistant from at least two non-contiguous points on the outline. The skeleton obtained 69

PAGE 84

through thinning algorithm may not be accurate since it is obtained by erosion of pixels, not by measuring the Cartesian distances from any internal point to the boundary. P 9 P 2 P 3 P 8 P 1 P 4 P 7 P 6 P 5 Figure 6.3 Neighborhood Arrangement Used by the Thinning Algorithm (A) (B) Figure 6.4 Skeleton in Two Dimensions: (A) T-Shaped Object, (B) Michigan Dune Sand Grain 70

PAGE 85

6.1.1 Skeletonization Algorithm in Two Dimensions In the present study an algorithm is developed in Matlab to extract the skeleton of 2-D particle outline and it is capable of generating skeleton of any irregular shape. First the two-dimensional silhouette of the particle is obtained using edge detection algorithm. The edge pixel is a pixel for which one or more of the four neighbors is on the background. The edge of the particle is marked in red in Figure 6.5. The edge pixels are then numbered consecutively, starting from a particular point and then traversing either clockwise or anti-clockwise direction. The next step is to define any point on the skeleton. For this purpose, the centroid of the particle, which coincides with, say, the black pixel (Figure 6.6), is chosen as the first pixel to start the skeleton. The Cartesian distances (D) from the black pixel (centroid) to all the edge pixels are calculated and the minimum distance (D min ) from the centroid to the edge pixels is obtained. Now the edge pixels that are within a distance of D min and D min + 1 are marked in yellow/red dots. If all marked edge pixels are adjacent (i.e., their order is sequential, for example, 5, 6, 7 or Figure 6.5 Edge Detection 71

PAGE 86

Figure 6.6 First Pixel to Start the Skeleton separated by less than a previously defined threshold value, then the black pixel is not a skeleton point. In this study, a threshold of 10 pixels is considered. It should be noted that the first pixel and the last pixel on the boundary are adjacent. Since the black point is not a skeleton pixel based on the criteria, it is then marked in the same color as the internal pixels. The next pixel to be checked is the pixel that is one pixel along the D min line away from the edge pixel as shown by blue arrow in Figure 6.6 and is marked in black (Figure 6.7). At the new marked (black) point, the same process is repeated until the first skeleton point is obtained. First, the Cartesian distances (D) from the new black pixel to all the edge pixels are calculated. Now the edge pixels that are within a distance of D min and D min + 1 are marked in yellow/red dots. In this case, the marked edge pixels are not all adjacent (i.e., their order is not sequential). Therefore, the black point is a skeleton point (Figure 6.7). 72

PAGE 87

Figure 6.7 First Skeleton Point Obtained The same process is then repeated on all 8 neighbors of the current skeleton point to check if they are the skeleton points or not. The distance between each of them and all the edge pixels are obtained. In this case, the yellow pixels satisfy the conditions required for a skeleton pixel (Figure 6.8) and the shaded yellow pixels do not. To move along the skeleton, anyone of those yellow pixels is chosen. This point is the next point on the skeleton and is then marked in black. It is noted that the yellow point at the bottom is also a skeleton point. This branch of the skeleton will be tracked later. From the current point, the procedure is repeated, marking all 8-neighbors that satisfy the skeleton pixel condition (Figure 6.9) and any of the 8-neighbors that have been previously marked in black as a skeleton pixel and shaded yellow as a non-skeleton pixel are not considered. When one branch of the skeleton reaches the edge pixel or being a certain threshold distance from the edge pixel, another branch is followed to continue the skeleton. The whole procedure is continued until the whole skeleton is defined (Figure 6.10). The output of the skeletonization algorithm is shown in Figure 6.11. 73

PAGE 88

Figure 6.8 Next Skeleton Point Figure 6.9 Skeleton Continued Following the First Branch 74

PAGE 89

Figure 6.10 Final Skeleton Obtained Through the Algorithm Figure 6.11 A Daytona Beach Sand Particle (Left) and the Skeleton (Right) 6.1.2 Skeletonization Algorithm in Three Dimensions Different approaches have been suggested in early studies to extract skeleton of three-dimensional shape (Lien and Amato, 2005; Chuang et al., 2000; Wu et al., 2003). In the present study, three-dimensional skeleton of granular particles are extracted using the same method as developed in two-dimensions. After characterizing the grain shape, three-dimensional coordinates of each and every voxel of the particle are obtained. The centroid was chosen as the first voxel to start the skeleton. The distances from this voxel to all surface voxels are calculated and the minimum distance (D min ) from the centroid to the surface voxels is obtained. Now the surface voxels which are at a distance of D min and 75

PAGE 90

D min + 1 are marked. If the marked voxels are non-adjacent (Figure 6.12), then the voxel at the centroid is a skeleton point. If not, this procedure should be continued on all of its 26 neighbors until the first skeleton point is obtained. The process is then repeated on all 26 neighbors of the current skeleton point to check if they are the skeleton points or not. When one branch of the skeleton reaches the boundary or being a certain threshold distance from the boundary, another branch is followed to continue the skeleton. The procedure is continued until the whole skeleton is defined. One problem that may arise during this skeleton generation procedure is explained below. Lets say the black voxels in Figure 6.13 are the marked surface voxels that are within a distance of D min and D min + 1 from a particular internal voxel. This kind of situation may occur if any particular region on the surface (as shown in red color) is protruded out of the neighboring region. Even though the marked black voxels are contiguous, the internal voxel should still be a skeleton point. The three-dimensional skeletons obtained by using the algorithm along with the original particles are shown in Figure 6.14 through Figure 6.21 for Daytona Beach sand grains (1 voxel = 0.003 mm) and Michigan Dune sand grains (1 voxel = 0.006 mm). The average computational time of this algorithm was 96 hours for Daytona Beach sand grains and 1 month for Michigan Dune sand grains. Figure 6.12 Non-Adjacent Boundary Voxels Figure 6.13 Surface Protrusion Marked as Red Voxels 76

PAGE 91

Figure 6.14 Original Particle of Daytona Beach Sand (DB #1) Figure 6.15 Three-Dimensional Skeleton of Daytona Beach Sand Grain (DB #1) 77

PAGE 92

Figure 6.16 Original Particle of Daytona Beach Sand (DB #2) Figure 6.17 Three-Dimensional Skeleton of Daytona Beach Sand Grain (DB #2) 78

PAGE 93

Figure 6.18 Original Particle of Michigan Dune Sand (MD #1) Figure 6.19 Three-Dimensional Skeleton of Michigan Dune Sand Grain (MD #1) 79

PAGE 94

Figure 6.20 Original Particle of Michigan Dune Sand (MD #2) Figure 6.21 Three-Dimensional Skeleton of Michigan Dune Sand Grain (MD #2) 80

PAGE 95

81 6.2 Overlapping Discrete Element Cluster After extracting the skeleton of two-dimensional particle outline and three-dimensional particle surface, the irregul ar shapes of sand grains are modeled using the ODEC technique for implementation in di screte element modeling code PFC 2D and PFC 3D In the current research, a fully automated procedur e is developed to model two-dimensional particle shape using the ODEC technique proposed by As hmawy et al. (2003). The procedure is also extended to three dimensions to model hi ghly irregular particle surface for implementation in three-dimensional DEM simulation. 6.2.1 ODEC Algorithm in Two Dimensions In the ODEC technique the two-dimensional pa rticle outline is covered by inscribing a number of overlapping circular discrete elements so that the simulated particle outline resembles the outline of actua l particle. To reduce the numbe r of overlapping discs within the clump, the biggest disc covering maximum area is inscribed first. Th e next disc to be added is the disc that covers the maximum uncovered area and so on. The disc generation procedure is continued until at least a reasonabl e percentage of the grain area is covered. The output of the ODEC algorithm is shown in Figure 6.22 for Daytona Beach sand grain. Figure 6.23 and Figure 6.24 show the orig inal particle of a Daytona Beach sand grain and the modeled part icle obtained through the ODEC technique. Table 6.1 summarizes the number of discs required to capture the shapes of various grains of Daytona Beach sand sample. Table 6.1 Number of Discs Required for Daytona Beach Sand Sample Area Covered (in percent) Number of Discs 98 % 9 14 95 % 4 8 90 % 3 6 85 % 2 4

PAGE 96

Figure 6.22 Overlapping Discrete Element Cluster for a Daytona Beach Sand Grain Figure 6.23 Original Particle of Daytona Figure 6.24 Particle Shape Obtained Beach Sand in Two Dimensions Through ODEC Technique A significant difference between the number of discs is observed (Table 6.1) if the area coverage is increased from 95% to 98%, in comparison to the difference between the number of discs, when the area coverage is increased from 90% to 95%. Therefore, an area coverage of 95% was considered as a reasonable threshold for ODEC technique because a small increase in the number of discs can cover 5% more area if the threshold is increased from 90% to 95%. 82

PAGE 97

83 6.2.2 ODEC Algorithm in Three Dimensions In this case three-dimensional particle su rface is covered by inscribing a number of overlapping spherical discrete elements so that the simulated particle surface resembles the surface of actual particle. To reduce th e number of overlapping spheres within the clump, the biggest sphere covering maximum volume is inscribed first. The next sphere to be added is the sphere that covers the maximum uncovered volume and so on. The procedure is continued until at least a reasona ble percent of the grain volume is covered. The output of the ODEC algorithm is s hown in Figure 6.25 through Figure 6.36 for Michigan Dune sand (1 voxel = 0.006 mm) and Daytona Beach sand grains (1 voxel = 0.003 mm) respectively for different percentage of covered volume. The original particles for the corresponding Daytona Beach sand gr ains (DB #1 and DB #2) and Michigan Dune sand grains (MD #1 and MD#2) are presented in Figure 6.14, Figure 6.16, Figure 6.18 and Figure 6.20 respectively. Table 6.2 summa rizes the number of spheres required to capture the shape of Michigan Dune sand and Daytona Beach sand grains. The average computational time of this algorithm was 96 hours for Daytona Beach sand grains and 56 hours for Michigan Dune sand grains with an error of 5% in grain volume. Table 6.2 Number of Spheres Required for Michigan Dune and Daytona Beach Sand Grains Number of Spheres Volume Covered (in percent) Michigan Dune Daytona Beach 98 % 217 297 251 296 95 % 98 126 123 133 90 % 49 57 61 62 85 % 32 35 39 It is evident from the figures that that a threshold of 95% can model irregular particle shape accurately in three dimensions. Howeve r, a big difference between the number of spheres is observed if the volume coverage is increased from 90% to 95%, in comparison with the number of discs between 85% to 90% volume coverage. Due to large volume of data sets, only two particles are processed for Daytona Beach and Michigan Dune sand

PAGE 98

samples, therefore it is not possible to set a threshold for percent volume covered. Sufficient number of grains needs to be analyzed to verify the threshold. Figure 6.25 Overlapping Discrete Element Cluster (MD #1, Volume Covered =85%) Figure 6.26 Overlapping Discrete Element Cluster (MD #1, Volume Covered =90%) 84

PAGE 99

Figure 6.27 Overlapping Discrete Element Cluster (MD #1, Volume Covered =95%) Voxel Voxel Voxel Figure 6.28 Overlapping Discrete Element Cluster (MD #2, Volume Covered =85%) 85

PAGE 100

Voxel Voxel Voxel Figure 6.29 Overlapping Discrete Element Cluster (MD #2, Volume Covered =90%) Voxel Voxel Voxel Figure 6.30 Overlapping Discrete Element Cluster (MD #2, Volume Covered =95%) 86

PAGE 101

Voxel Voxel Voxel Figure 6.31 Overlapping Discrete Element Cluster (DB #1, Volume Covered =85%) Voxel Voxel Voxel Figure 6.32 Overlapping Discrete Element Cluster (DB #1, Volume Covered =90%) 87

PAGE 102

Voxel Voxel Voxel Figure 6.33 Overlapping Discrete Element Cluster (DB #1, Volume Covered =95%) Voxel Voxel Voxel Figure 6.34 Overlapping Discrete Element Cluster (DB #2, Volume Covered =85%) 88

PAGE 103

Figure 6.35 Overlapping Discrete Element Cluster (DB #2, Volume Covered =90%) Figure 6.36 Overlapping Discrete Element Cluster (DB #2, Volume Covered = 95%) 89

PAGE 104

90 CHAPTER 7 IMPLEMENTATION OF PARTICLE SHAPE WITHIN DISCRETE ELEMENT MODELING SIMULATION The previous chapter presented the skel etonization and ODEC algorithms and it was observed that the algorithms are capable to modeling the irregular particle shape accurately. The current chapter describes the implementation of particle shape within DEM software and the effect of grain shape on the shear strength behavior of granular media. 7.1 Introduction In recent years Discrete Element Method (DEM) has become an emerging technique to characterize irregular particle shapes in two and three dimensions. In the current research, the effect of particle shape on the micromech anical behavior of gr anular materials is studied using Discrete Element Meth od. The output of the ODEC algorithm is implemented within the DEM software, PFC 2D and PFC 3D (Itasca, 1999) to model angular particles by means of a series of Fish functions that will convert a particle assembly of discs and sphe res into their corre sponding angular particles using the technique described in chapter 2. 7.2 Two-Dimensional Discret e Element Simulation Direct shear test is perfor med to evaluate the influence of grain shape on the shear strength response of cohesionless soil. Angul ar materials dilate and rounded materials contract when subjected to shear. In this study, Daytona B each sand (angular materials) and circular particles (rounded ma terials) were selected to si mulate the direct shear test numerically.

PAGE 105

To implement particle shape within DEM simulation, the circular particles are first generated within the desired range of grain sizes using discrete element modeling code PFC 2D The shape conversion algorithm (Ashmawy et al., 2003) is then invoked to transform each circular particle into its angular equivalent through replacement with a corresponding set of circular discrete element cluster, selected randomly from the particle shape library. 7.2.1 Model Set-Up To simulate direct shear test numerically, the size of direct shear box is chosen as 5 mm in height and 12 mm in width, which can be split into two halves. Even though from practical standpoint, it is not possible to prepare such a small sample, numerical results can still be compared with the real-world experiments with respect to the nature of variation in stress-strain and volume change behavior between angular and rounded materials. Since the purpose of the current study is to compare the shear induced dilative or contractive response of otherwise similar materials of different particle shapes, the grain size distribution and particle density are kept constant in all simulations, with a mean grain size of 0.375 mm. The inter-particle friction coefficient is taken as 0.5. The shear and normal stiffness of the wall forming the box are set to 1 10 7 N/m. Total 416 circular particles are generated within the box with a void ratio of 0.3. The density, shear and normal stiffness of the particles are set to 2500 kg/m 3 1 10 7 N/m respectively. The linear-stiffness contact model is used in the analysis. Figure 7.1 shows the simulated model for direct shear box with Daytona Beach sand grains that is generated in PFC 2D The circular particles are considered in the analysis because of their perfectly round shape. The direct shear simulation of the sample with circular particles also started with the same setup including the same properties as the Daytona Beach sand sample. 91

PAGE 106

Figure 7.1 Two-Dimensional Direct Shear Test Simulation with Daytona Beach Sand Sample 7.2.2 Numerical Simulation Soil specimen within the box is subjected to a vertical compressive load (F) and simultaneously sheared by applying a gradually increasing lateral load until the sample fails or the shear displacement reaches a certain value. The test is started by vertically moving the top wall of the box until a specific vertical stress ( ) is reached. The vertical force (F) is kept constant throughout the simulation by using Itascas built-in servo-controlled fish functions that controls the movement of the top wall of the shear box. Figure 7.2 presents the history of servo-wall stress with time after application of normal load for Daytona Beach sand. After the vertical stress reaches a constant value, shearing stage is started by moving the upper half of the shear box laterally with a constant velocity of 510 -3 m/step (V). The test is continued until the horizontal displacements are equal to 3.0 mm for Daytona Beach sand, and 2.8 mm for circular particles. Figure 7.3 shows the movement of the shear box after application of normal load and horizontal velocity after a certain time interval for Daytona Beach sand. 92

PAGE 107

0501001502002500.2070.20720.20740.20760.20780.2080.20820.20840.2086Accumulated Time (Sec)Servo-Wall Stress (kPa) Figure 7.2 History of Servo-Wall Stress with Time Figure 7.3 Movement of Shear Box with Daytona Beach Sand (2-D Simulation) 93

PAGE 108

015304560759010500.511.522.5Shear Displacement (mm)Shear Stress (kPa) 3 Figure 7.4 Variation of Shear Stress with Shear Displacement for Daytona Beach Sand Sample (Two-Dimensional Simulation) -0.0200.020.040.060.080.10.1200.511.522.5Shear Displacement (mm)Vertical Displacement (mm) 3 Figure 7.5 Variation of Vertical Displacement with Shear Displacement for Daytona Beach Sand Sample (Two-Dimensional Simulation) 94

PAGE 109

0153045607500.40.81.21.622.42.8Shear Displacement (mm)Shear Stress (kPa) Figure 7.6 Variation of Shear Stress with Shear Displacement for Circular Particles (Two-Dimensional Simulation) -0.2-0.16-0.12-0.08-0.04000.40.81.21.622.42.8Shear Displacement (mm)Vertical Displacement (mm) Figure 7.7 Variation of Vertical Displacement with Shear Displacement for Circular Particles (Two-Dimensional Simulation) 95

PAGE 110

The variation of shear stress and vertical displacement with shear displacement are shown in Figure 7.4 through Figure 7.7 for Daytona Beach sand, and circular particles ( = 200 kPa). From the stress-displacement plots, it is observed that the Daytona Beach sand sample possesses highest shear strength compared to the other material. Due to particle breakage and rearrangement, there are some spike drops in the stress-strain plot. From the displacements plots, it can be seen that, after initial contraction, the Daytona Beach sand exhibits dilation throughout the simulation due to the angularity of grains. On the other hand, the circular particles are predominantly contractive in nature which results from roundness of grains. The test is performed for four different normal stresses and the corresponding peak shear stresses are obtained. Table 7.1 presents the value of peak shear stresses for different applied normal stresses. Table 7.1 Values of Maximum Shear Stresses for Different Vertical Stresses (TwoDimensional Simulation) Maximum Shear Stress (kPa) Materials = 50 kPa = 100 kPa = 150 kPa = 200 kPa tan Daytona Beach Sand 27.82 50.29 78.10 101.50 0.51 27 0 Circular Particles 15.07 33.89 45.91 59.89 0.31 17.2 0 The coefficients of internal shearing resistance (tan ) are evaluated by plotting the shear stress values against normal stresses in Figure 7.8. The angles of internal friction for Daytona Beach sand sample and circular particles are 27 0 and 17.2 0 respectively. It is evident that the Daytona Beach sand sample possesses higher value of internal friction coefficient which can be explained by the angularity of the grains. 96

PAGE 111

050100150200250050100150200250Normal Stress (kPa)Shear Stress (kPa) Daytona Beach Sand Circular Particles Figure 7.8 Variation of Shear Stress with Normal Stress (Two-Dimensional Simulation) 7.3 Three-Dimensional Discrete Element Simulation From quantitative standpoint, three-dimensional discrete simulation is necessary to compare the test results with the experiments. Therefore, three-dimensional numerical simulation of direct shear test was performed to understand the real micromechanical behavior of granular media. The same materials are used for three-dimensional simulation as were used for two-dimensional simulation except the circular particles are replaced by spherical ones for three-dimensional simulation. To implement the particle shape within 3-D DEM simulation, discrete spherical particles are first generated within the desired range of grain sizes (as shown in Figure 7.9) using discrete element modeling code PFC 3D (Itasca, 1999). The shape conversion algorithm (described in Chapter 2) is then invoked to transform each spherical particle into its angular equivalent (as shown in Figure 7.10) following the same procedure as used for two-dimensional simulation. 97

PAGE 112

7.3.1 Model Set-Up The size of direct shear box is chosen as 2 mm in height, 4 mm in length and 4 mm in width. The grain size distribution and particle density are kept constant in all simulations, with a mean grain size of 0.375 mm. The inter-particle friction coefficient is taken as 0.5. The shear and normal stiffness of the wall forming the box are set to 110 3 N/m. Total 800 spherical particles are generated within the box with a void ratio of 0.45. The density, shear and normal stiffness of the particles are set to 2500 kg/m 3 1 10 3 N/m respectively. Figure 7.9 and Figure 7.10 show the simulated model of direct shear box with spherical particles generated in PFC 3D and corresponding equivalent angular particles of Daytona Beach sand sample respectively. Figure 7.9 Three-Dimensional Direct Shear Test Simulation with Spherical Particles 7.3.2 Numerical Simulation After generating the assembly of particles in PFC 3D a vertical load (F) is applied on the top wall of the shear box and the force is kept constant throughout the simulation. After the vertical stress ( ) reaches a constant value, shearing stage is started by moving the upper half of the shear box to the right with a constant velocity of 510 -3 m/step. The test 98

PAGE 113

is continued until the horizontal displacements are equal to 2.1 mm for Daytona Beach sand sample and 1.6 mm for the materials with spherical particles (sample #2). The variation of shear stress and vertical displacement with shear displacement are shown in Figure 7.11 through Figure 7.14 for Daytona Beach sand sample ( = 250 kPa) and spherical particles ( = 200 kPa) with different particle arrangements. The simulations are performed for three different particle assemblies. No significant variation is observed in stress-displacement and volumetric behaviors of spherical particles for different particle orientations whereas small variation is observed for Daytona Beach sand due to difference in fabric, coordination numbers and contact force chain generated for each assembly. From the stress-displacement plots, it is observed that the Daytona Beach sand (sample #1) possesses highest shear strength compared to the spherical particles. The same trend is observed in volumetric behavior for both the samples as was observed in two-dimensional simulation. After initial contraction, the Daytona Beach sand dilates due to angularity of the grains whereas the spherical particles contracts due to the roundness of the particles. The test is performed for three different normal stresses and the corresponding peak shear stresses are obtained. Table 7.2 presents the values of peak shear stresses and internal friction angles of Daytona Beach sand and spherical particles with three different particle assembly. Figure 7.10 Three-Dimensional Direct Shear Test Simulation with Daytona Beach Sand Sample 99

PAGE 114

05010015020025000.30.60.91.21.51.82.1Shear Displacement (mm)Shear Stress (kPa) Assembly 1 Assembly 2 Assembly 3 Figure 7.11 Variation of Shear Stress with Shear Displacement for Daytona Beach Sand Sample (Three-Dimensional Simulation) -0.06-0.04-0.0200.020.040.0600.30.60.91.21.51.82.1Shear Displacement (mm)Vertical Displacement (mm) Assembly 1 Assembly 2 Assembly 3 Figure 7.12 Variation of Vertical Displacement with Shear Displacement for Daytona Beach Sand Sample (Three-Dimensional Simulation) 100

PAGE 115

02040608010012000.20.40.60.811.21.41.61.8Shear Displacement (mm)Shear Stress (kPa) Assembly 1 Assembly 2 Assembly 3 Figure 7.13 Variation of Shear Stress with Shear Displacement for Spherical Particles (Three-Dimensional Simulation) -0.12-0.1-0.08-0.06-0.04-0.02000.20.40.60.811.21.41.61.8Shear Displacement (mm)Vertical Displacement (mm) Assembly 1 Assembly 2 Assembly 3 Figure 7.14 Variation of Vertical Displacement with Shear Displacement for Spherical Particles (Three-Dimensional Simulation) 101

PAGE 116

Table 7.2 Values of Maximum Shear Stresses and Internal Friction Angles for Different Particle Arrangements (Three-Dimensional Simulation) Maximum Shear Stress (kPa) Materials Particle Orientation = 150 kPa = 200 kPa = 250 kPa tan Assembly 1 118.04 166.60 199.93 0.813 39.1 0 Assembly 2 138.80 197.10 223.80 0.929 42.9 0 Daytona Beach Sand Assembly 3 139.00 168.20 218.70 0.875 41.2 0 Assembly 1 82.95 100.04 111.80 0.50 26.6 0 Assembly 2 82.09 97.76 104.40 0.474 25.3 0 Spherical Particles Assembly 3 79.35 92.86 107.90 0.467 25.0 0 102 050100150200250300050100 1 50200250300Normal Stress (kPa)Shear Stress (kPa) Numerical Simulation (Assembly1) Numerical Simulation (Assembly 2) Numerical Simulation (Assembly 3) Experimental Results Figure 7.15 Variation of Shear Stress with Normal Stress for Daytona Beach Sand (Three-Dimensional Simulation)

PAGE 117

103 050100150200250300050100150200250300Normal Stress (kPa)Shear Stress (kPa) Numerical Simulation (Assembly 1) Numerical Simulation (Assembly 2) Numerical Simulation (Assembly 3) Numerical Simulation (Assembly 4) Figure 7.16 Variation of Shear Stress with Normal Stress for Spherical Particles (Three-Dimensional Simulation) Table 7.3 Comparison of Internal Friction Angles Obtained from 2-D and 3-D Simulations Material 2-D Simulation 3-D Simulation Experiments Daytona Beach Sand 27.0 0 39.1 0 42.9 0 37.4 0 Rounded Particles 17.2 0 25 0 26.6 0 24.4 0 27 0 Figure 7.15 and Figure 7.16 present the variation of shear stress with normal stress for different particle assemblies of Daytona Beach sand and spherical particles respectively. For comparison purpose, experimental data of direct shear test are collected from Rowan University (NJ) and plotted in Figure 7.15 for Daytona Beach sand sample

PAGE 118

with normal stresses of 50 kPa, 100 kPa and 150 kPa. Even though three different numerical simulations produced significantly different particle arrangements, the angle of shearing resistance obtained from different particle assemblage are very close to each other for both materials. It is evident from the figures that the Daytona Beach sand sample possesses higher value of internal shearing resistance than the spherical particles due to particles angularity. Table 7.3 summarizes the values of internal friction angle obtained from two-dimensional and three-dimensional simulations. By comparing experimental results with numerical simulations, it is found that the two-dimensional simulations underestimate the results for both materials. The angle of shearing resistance ( ) of Daytona Beach sand obtained from three-dimensional simulations varies from of 39.1 0 42.9 0 which is close to the experimental value (37.4 0 ) whereas for rounded materials, the angle of shearing resistance varies from 25 0 26.6 0 which is within the range of internal friction angle of 24.4 0 27 0 reported in literature (OSullivan et al., 2004, Phillips et al., 2006). Although two-dimensional simulation allows simplistic ways to capture and reconstruct particle images, three-dimensional simulation still remains more practical from quantitative standpoint. 7.4 Particle Shape Library A particle shape library is developed in MS Excel format and will be made available online to the geotechnical community. The database is comprised of a wide variety of natural and processed sands, their morphological characteristics, size distributions, shape descriptors and other shape parameters. The database includes (1) general information on geomaterials regarding geographic location, particle size distribution, sample size; (2) pixel and voxel coordinate data of each particle; (3) quantitative descriptors such as Fourier Descriptors and (4) centers and radii of circular and spherical discrete elements used to reconstruct the angular particles. The geomaterial database is summarized in Table 7.4 and the Two-dimensional projection images of the sand samples are presented in Appendix A. 104

PAGE 119

105 Table 7.4 Geomaterial Database Gradation Shape Descriptors Type of Sand D 50 (mm) D 10 (mm) C u Location Aspect Ratio Elongation Triangularity Squareness Sample Size Daytona 0.13 0.08 1.87 Florida 1.423 0.168487 0.078809 0.052109 130 Toyoura 0.20 0.17 1.24 Japan 1.333 0.139724 0.100307 0.0518 90 Michigan 0.34 0.26 1.39 Michigan 1.341 0.140109 0.063223 0.04587 100 Tecate 0.95 0.76 1.32 Mexico 1.421 0.168221 0. 082782 0.057527 150 #1 Dry 0.27 0.13 2.46 New Jersey 1.420 0.167989 0.088005 0.054268 150 Kahala 0.70 0.50 1.52 Hawaii 1.426 0.167344 0.080716 0.048941 170 Hostun 0.58 0.41 1.66 France 1.377 110 Oxnard 0.8 0.53 1.60 California 1.401 120 Ala Wai 1.05 1.01 1.05 Hawaii 1.500 200 Std. Melt 0.28 0.14 2.29 New Jersey 1.501 200 Loire 0.73 0.52 1.5 France 1.403 130 Rincon 0.68 0.46 1.61 Puerto Rico 1.369 110 Long Beach 0.25 0.17 1.56 California 1.444 130 Fontainebleau 0.20 0.11 1.91 France 1.386 110 Arroyo Alamar 0.90 0.60 1.55 Tijuana 1.438 140 Clearwater 0.18 0.10 1.95 Florida 1.460 150 Rhode Island 0.25 0.19 1.42 Rhode Island 1.474 150 Boca Grande 0.20 0.16 1.31 Columbia 1.425 230 Red Sea 0.44 0.28 1.79 Egypt 1.369 90 Panama Malibu 0.25 0.18 1.47 Panama 1.289 70 Table 7.5 and 7.6 present the centers and radii of circular discrete elements of four particles for Daytona Beach and Michigan D une sand samples and these data were used to reconstruct the assemblies of two-dimensional angular particles for discrete element modeling. The whole dataset cannot be shown in its entirety due to its large volume. Only a portion of the database is presented here. However, comprehensive information about the database will be compiled in the form of an online particle shape library.

PAGE 120

106 Table 7.5 ODEC Data (2-D) for Daytona Be ach Sand Grains (98% Area Coverage) Type of Sand Particle Index # Disc # X c Y c Radius 1 0.044911 0.071133 0.837580 2 -0.11790 -0.476535 0.666082 3 -0.029098 0.603999 0.554822 4 0.163326 -0.713364 0.384278 5 -0.029098 0.367170 0.710642 6 -0.014296 0.826027 0.397726 7 0.533372 -0.121291 0.380554 8 0.548174 0.248755 0.370046 9 -0.088305 -0.269309 0.740092 10 0.000506 0.011926 0.828109 1 11 -.043900 0.485585 0.632335 1 -0.042798 0.049540 0.765721 2 0.005206 -0.366491 0.744529 3 -0.026796 0.417568 0.734140 4 0.389235 -0.638512 0.416031 5 0.181219 0.737592 0.448034 6 -0.154806 -0.622511 0.519238 7 -0.202810 0.561578 0.534066 2 8 0.213222 -0.510502 0.576043 1 0.086271 0.140719 0.781997 2 -0.632913 -0.379744 0.511787 3 0.777067 0.055552 0.461458 4 -0.244932 -0.162096 0.649161 5 -0.793783 -0.493299 0.420223 6 -0.074599 0.689570 0.336966 7 0.341771 0.140719 0.674795 8 1.051492 0.197496 0.243292 9 0.956863 -0.143170 0.243292 10 -0.093525 -0.020152 0.719433 3 11 -0.973579 -0.682558 0.199621 1 0.137521 0.140768 0.782373 2 -0.336156 -0.486854 0.550490 3 0.137521 0.543393 0.595991 4 -0.691413 -0.996056 0.221225 5 0.587513 0.081558 0.444979 6 0.137521 1.005227 0.272364 7 0.042785 -0.072387 0.720121 8 -0.478259 -0.664482 0.442133 9 -0.265104 -0.392118 0.598456 10 0.090153 0.211819 0.746040 11 -0.786149 -1.102633 0.153945 12 0.8954036 0.06971 0.192772 13 0.5993559 0.27102 0.364800 Daytona Beach Sand 4 14 -0.383524 0.318396 0.285191

PAGE 121

107 Table 7.6 ODEC Data (2-D) for Michigan Dune Sand Grains (98% Area Coverage) Type of Sand Particle Index # Disc # X c Y c Radius 1 0.069201 0.134643 0.805323 2 -0.189597 -0.522703 0.623633 3 0.131312 0.357210 0.778547 4 -0.329348 -0.926428 0.331019 5 -0.163717 -0.326017 0.705470 6 0.457398 0.403793 0.492369 7 -0.065374 -0.118979 0.748154 1 8 0.012265 0.538368 0.580169 1 -0.111929 -0.006224 0.846130 2 0.227153 0.117078 0.798189 3 -0.370865 -0.345307 0.625967 4 0.498419 0.283537 0.565885 5 -0.611306 -0.499436 0.411543 6 0.603227 -0.148023 0.409878 7 0.264144 0.708932 0.265960 8 -0.438682 0.437666 0.344642 9 -0.284553 -0.332977 0.642625 10 0.036034 0.055427 0.838187 11 0.812842 0.394510 0.270846 2 12 -0.216737 0.006106 0.754115 1 0.062274 0.366943 0.747018 2 -0.075913 -0.349117 0.662960 3 -0.031944 -1.096583 0.319109 4 0.062274 0.775223 0.506954 5 -0.057069 -0.079024 0.720619 6 -0.069632 -0.694585 0.497448 7 0.018306 -1.291302 0.212543 8 0.074837 0.429756 0.737771 9 0.018306 0.247600 0.745458 3 10 0.049712 0.530255 0.679564 1 0.259312 -0.047216 0.708469 2 -0.789453 0.098815 0.498936 3 0.591200 -0.080404 0.652664 4 -0.278346 0.098815 0.638468 5 -0.988585 0.085539 0.449215 6 1.049205 -0.491945 0.209589 7 0.033629 0.025800 0.697090 8 0.777057 -0.219797 0.493386 9 0.982827 0.112090 0.267576 10 -0.497392 0.112090 0.577332 Michigan Dune Sand 4 11 -1.161167 -0.000751 0.283565

PAGE 122

108 In the similar fashion, three-dimensional da ta for sand samples are complied in the database and these include the coordinates (X c Y c Z c ) of the center a nd radii of the spheres needed to capture the shape. Tabl e 7.7 and Table 7.8 summarize the information about spherical discrete elements used to model a three-dimensional grain of Michigan Dune sand and Daytona Beach sand for 85% volume coverage which corresponds to 32 and 39 spheres for those tw o sand grains respectively. Table 7.7 ODEC Data (3-D) for Michigan Dune Sand Grain (85% Volume Coverage) Material Particle Index # Disc # X c Y c Z c Radius 1 0 -0.177794 0.400037 0.594681 2 -0.222243 0.133346 0 0.556939 3 0.133346 0.62228 -0.400037 0.440018 4 -0.177794 -0.577832 0.577832 0.464057 5 0.31114 0.400037 0.177794 0.400037 6 0 0.355589 -0.533383 0.440018 7 0.177794 -0.177794 -1.066766 0.28461 8 -0.088897 -0.31114 -0.488934 0.291469 9 0.355589 0.755626 -0.222243 0.384936 10 -0.266691 -0.444486 0.93342 0.341416 11 0.177794 -0.044449 0.400037 0.537074 12 -0.133346 -0.222243 0.400037 0.581241 13 0.088897 -0.266691 -0.222243 0.314299 14 0 -0.400037 -0.844523 0.239363 15 0.133346 0.444486 -0.711177 0.358356 16 0.31114 1.066766 -0.222243 0.273999 17 -0.355589 -0.177794 -0.044449 0.39756 18 -0.177794 0.62228 -0.400037 0.332623 19 0.133346 0.088897 -0.93342 0.255338 20 0.222243 0.044449 0.62228 0.400037 21 -0.355589 0 0.444486 0.39756 22 -0.177794 -0.711177 0.62228 0.400037 23 -0.400037 -0.666729 1.022317 0.239363 24 0.31114 -0.444486 0.355589 0.314299 25 0.400037 0.666729 0.222243 0.27037 26 -0.133346 -0.31114 0.666729 0.480785 27 0.266691 0.577832 -0.666729 0.323591 28 -0.044449 0.93342 -0.400037 0.239363 29 -0.355589 -0.444486 0.177794 0.332623 30 -0.533383 0.177794 -0.044449 0.332623 31 0 -0.400037 -1.155663 0.183266 Michigan Dune Sand 1 32 0.266691 -0.044449 -1.200112 0.208482

PAGE 123

109 Table 7.8 ODEC Data (3-D) for Daytona Beach Sand Grain (85% Volume Coverage) Material Particle Index # Disc # X c Y c Z c Radius 1 0.371605 0.222963 0.371605 0.532057 2 -0.07432 -0.03716 -0.22296 0.501323 3 -0.11148 0.074321 0.891852 0.442819 4 0.557408 0.297284 -0.07432 0.445926 5 -0.29728 -0.26012 0.334445 0.395022 6 -0.26012 -0.07432 -0.66889 0.400231 7 0.222963 0.297284 0.780371 0.450547 8 0.520247 0.111482 -0.59457 0.308678 9 -0.44593 -0.44593 -0.14864 0.317499 10 -0.37161 -0.1858 0.74321 0.378964 11 0 -0.14864 0.408766 0.462645 12 -0.33445 -0.33445 -1.07766 0.262764 13 0.780371 0.334445 0.222963 0.360284 14 -0.52025 -0.48309 -0.66889 0.252035 15 0.185803 0.074321 -0.11148 0.49856 16 0.74321 0.334445 -0.26012 0.340582 17 -0.14864 -0.03716 1.151976 0.319667 18 0.074321 0 -0.70605 0.290233 19 -0.55741 -0.52025 0.074321 0.260124 20 0.148642 0.594568 0.631729 0.283006 21 -0.66889 -0.44593 0.594568 0.200115 22 -0.37161 0 -0.96617 0.24928 23 0.222963 0.222963 1.114815 0.270533 24 0.891852 0.334445 0.037161 0.315317 25 0.260124 0.557408 0.260124 0.292602 26 -0.44593 -0.22296 1.040494 0.240828 27 -0.37161 -0.29728 -0.29728 0.365989 28 0.445926 0.371605 0.74321 0.350571 29 -0.33445 -0.40877 -1.26346 0.213471 30 -0.59457 -0.63173 -0.33445 0.200115 31 -0.22296 -0.22296 0.037161 0.430164 32 0.408766 0.074321 -0.85469 0.203536 33 0.631729 0.148642 -0.70605 0.24928 34 0.222963 0.520247 -0.03716 0.260124 35 -0.26012 0.260124 -0.33445 0.24928 36 -0.59457 -0.33445 0.780371 0.246495 37 -0.48309 -0.59457 -0.92901 0.161979 38 0 0.408766 0.817531 0.315317 Daytona Beach Sand 1 39 0.260124 0.074321 0.445926 0.48877

PAGE 124

110 CHAPTER 8 SUMMARY AND CONCLUSIONS This chapter provides a summar y and draws conclusions of th e research effort presented in the previous chapters. A comprehensiv e research has been conducted on threedimensional characterization of particle morphology and discrete element modeling to understand the effect of particle shap e on the behavior of geomaterials. 8.1 Summary and Conclusions A wide variety of sand samples having differe nt angularity and roundness were collected for the purpose of the current study. Two-dime nsional images of the sand grains were acquired using the microscope system and image analysis software and the acquired images were processed through successive steps of erosion, d ilation and contrast enhancement techniques to obtain the final bi nary images of the particles. In three dimensions, particle shapes were character ized using X-ray CT, which generated stacks of two-dimensional images. These image stacks were then combined together to reconstruct three-dimensional shape of the particles and to obtain three-dimensional surface points in voxel format. After shape characterization, particle shape quantificati on was performed in order to design an optimum sample size and to inve stigate the relationshi p between grain size and shape. The grain shapes were quantified using Fourier shape desc riptors and the first four Fourier descriptors were used to inve stigate the relationship between grain size and shape. Considering the importance of optimum samp le size selection to any experimental design, a suitable methodology was established to obtain sample size for various sand samples. The optimum sample size of a particular sand sample was determined based on

PAGE 125

111 the variability in size and shape of partic les within that sample. In the current study, sample size was found to vary from 70 (for Panama Malibu Beach sand) to 230 (for Boca Grande Beach sand). Selection of an optimum sample size is important to increase accuracy of numerical simulation and it can save significant amount of resources. The study verified any existing relations hip between grain size and grain shape using six different natural and processed sand samples. Though in the previous literature strong relationships were documented, the curr ent research verified that there is no relationship between grain si ze and grain shape in term s of circularity, elongation, triangularity and squareness for the sand sample s analyzed. It should be noted that the grain shapes are highly variab le in nature and they are li kely to be influenced by the origin and mineralogical compositions, distan ce of transport, mechanical and chemical processes, velocity of the transporting medium and the depositional environment (aeolian, fluvial, glacial). Considering all these factors, it is unlikely to identify a universal relationship between grain size and grain shape for all sands. Even though the sands in the current study came from diffe rent sources and underwent considerably different depositional processes, no relations hip was found between shape and size. As such, discrete element models of angular particle assemblies were constructed by selecting the shapes belonging to a particular sand sample randomly from its particle shape library, irrespective of size. However, all the sand samples used in the study to evaluate the size-shape relationship were unifo rmly-graded. Therefore, it is suggested to perform the analysis on well-graded sand samp les to confirm the ve racity of the above statement. This research presented a detailed particle shape modeling procedure in two and three dimensions using the ODEC technique. A skeletonization algorithm was developed to automate the ODEC technique and the algo rithm is capable of generating skeleton of any irregular shape. The ODEC2D and ODEC3D algorithms were found to model irregular particle shape accurately in tw o and three dimensions respectively. It was observed that 8 to 16 discs are sufficient to capture the shape accurately in two dimensions, resulting an uncovered area of 2% (error). However, based on the analysis, an area coverage of 95% was considered as a reasonable threshold for this experimental

PAGE 126

112 design and the corresponding number of discs varied from 4 to 8 for Daytona Beach sand sample. Three-dimensional shapes were modeled using ODEC3D algorithm with an error of 5% in grain volume which corresponds to 123-133 spherical elements for Daytona Beach sand and 98-126 spherical elements for Mi chigan Dune sand. No specific value of error threshold could be estab lished for volume coverage, because very limited number of particles were modeled using the ODEC tec hnique in three dimensions due to large volume of data sets. After modeling partic le shape using the ODE C technique, shapes were then implemented within disc rete element modeling software, PFC 2D and PFC 3D by using shape conversion algorithm. Numerical Simulation of direct shear test was performed to study the effect of grain shape on shear strength characteristics of granular soil. Daytona Beach Sand was chosen to study the effect of particle shape on the shear strength behavior of the granular soils. To compare the behavior of angular Daytona Beach sand with rounded materials, circular and spherical partic les were used in the study fo r two-dimensional and threedimensional simulations respectively. It was observed that the stress-strain and volumetric behavior of simulated materials followed typical soil behavior of angular and rounded particles. Angularity of grains and interlocking effect cont ributed more shearing resistance and higher shear strength in Dayt ona Beach sand in comparison with spherical particles. The three-dimensional simulation wa s repeated three times to generate three different particle assemblies and to study the effect of particle or ientation on the shear strength response of granular soil. No significant variation was observed between spherical assemblies whereas small varia tion was observed between assemblies of Daytona Beach sand sample due to the differe nce in coordination numbers of the three assemblies. A very small-size model with limited number of particles was generated in PFC to reduce the computation time. The incr ease in the dimension of shear box, total number of particles within the assembly and number of spheres within each particle can increase the accuracy of the model estimati on results. Angle of shearing resistance was computed from numerical simulation and the results obtained from three-dimensional simulations are in good agreement with the va lues reported in literature and obtained through experiments. A direct comparison wa s made between two-dimensional and three-

PAGE 127

dimensional peak angle of internal friction ( ) and higher value of and hence higher shear strength were obtained from three-dimensional simulation. From practical standpoint it would not be appropriate to compare the results between two and three-dimensional simulations because only two different types of shapes of Daytona Beach sand sample were characterized due to large number of surface data points and the number of spherical discrete elements within each clump was limited to 15 to reduce the computation time in three-dimensional simulation, still three-dimensional characterization and modeling is more practical in order to simulate soil behavior accurately from computational standpoint. 8.2 Methodological Contributions The automation of ODEC2D algorithm and development of ODEC3D algorithm in the current study can be considered as a valuable contribution in the field of geotechnical engineering, particularly in three-dimensional discrete element modeling. This method would allow modeling any number of particles of arbitrary shape with a desired level of accuracy. Grain shape can be modeled more realistically using the ODEC method compared to other ellipsoid or polyhedron based approach (Lin and Ng, 1997; Ghaboussi and Barbosa, 1990) and the percent error in grain volume was found to be less than that obtained from another sphere-based approach proposed by Matsushima (2004). The main contribution of this research effort is to automate the entire multi-phase procedure starting from image capturing, reconstruction to three-dimensional particle shape modeling and finally linking the process to discrete element modeling framework by implementing three-dimensional particle shape into DEM environment which will allow researchers to simulate particulate behavior of cohesionless soil numerically. The relationship between grain size and shape was explored and documented in the literature, but the concept has never been applied in the area of discrete element modeling. For example, size-shape relationship is important for generating and reconstructing particle assemblies in DEM environment. If any such relationship exists, then it would be necessary to divide the particle size into different bins in order to generate representative assemblies of angular particles for discrete element modeling 113

PAGE 128

114 simulation. In other situations, where there is no relationship between grain size and shape such as in case of poorly-graded sand sa mples as evidenced from the present study, shapes belonging to a particular sand sample can be selected randomly from its particle shape library irrespective of size for DEM simulation. Therefore, understanding such relationship would certainly improve th e accuracy of numerical simulation. The study also develops a methodology to se lect an optimum sample size for each sand sample. The idea of using optimum sample size has never been taken into account in early research efforts in micromechanical m odeling. The study explic itly considers that the behavior of a soil mass is greatly influenced by the variability of the particle shapes (sample size) within the samp le. The use of small sample size may lead to inaccurate results while too large sample size can result inefficient use of resources. 8.3 Practical Contributions The development of an online particle shape library will be an extremely valuable contribution for future research on granular particle morphology. In this study, a particle shape library was developed in excel format and the images of different natural and processed sand samples, their lo cations, grain size distributio ns, sample size, quantitative shape descriptors, pixel and voxel coordinate s of each particle, cen ters and radii of circular and spherical discrete elements us ed to reconstruct th e two-dimensional and three-dimensional angular particles were comp iled in the database. It would be the most diverse particle shape library in the field of discrete element modeling. This database will be made available online and would provide researchers an instant access to a diverse range of particle morphology data to simula te different aspects of micromechanical behaviors of granular soil and such arra ngement would potentially reduce the sample collection, image capturing and reco nstruction effort. The research methodology presented in this study can have significant contribution to many areas in the field of geotechnical engineering. For example, this research can be applied to pavement engine ering to study the strength and deformations characteristics of base, sub-base layers of granular substructures. The study can also be applied in numerical simula tion of stress-strain relati onship and stress-paths to

PAGE 129

115 understand the dilative/contractive behavior a nd deformation characteristics of granular material. Another potential a pplication of the current res earch could be to study the influence of particle shape on liquefaction behavior of cohesionl ess soil in earthquakeprone regions. Other practical applications of this resear ch could be studying the effect of grain shape on the flow rate of materials flowi ng from hopper, modeling dynamic behavior of debris during rockfall which is important for cliff stabilization, designing geotechnical interfaces, selecting suitable materials fo r constructions, studying surface roughness of sand grains, soil-structure interaction and others. 8.4 Future Recommendations This section addresses some future reco mmendations and shed light on some potential future research problems for further exte nsion of the current research effort. Applications of advanced mathematical techniques are rapidly emerging in the field of micromechanical modeling. Such tech niques have been contributing in gaining computational efficiency and solving many unsolved complex problems in geotechnical engineering. The application of spherical harmonics in qua ntitative characterization of particle surface by means of three-dimensiona l shape descriptors is one of the finest examples that hold enormous promise in the future of three-dimensional quantification of particle morphology. The spherical harmonic series algorithm can be used to obtain surface information in terms of three-dimensi onal shape descriptors and to reconstruct the particle surface accurately. It was concluded from the current study th at there is no relationship between grain size and shape for the sand samples analyzed. However, it is necessary to perform the analysis on well-graded sand samples to verify the relationship. Use of three-dimensional shape descriptors to explore the size-shape re lationship would be an interesting future research topic. As previously mentioned, the threedimensional modeling was performed on very limited number of different particle shapes, which may be considered as a limitation of the present analysis. Therefore, it is s uggested that sufficient number of different

PAGE 130

116 particle shapes as much as the optimum sample size should be used to perform the threedimensional simulation to closely capture the ac tual behavior of soil. Very small size of specimen was used in the study which warrant s a bigger specimen size with more number of sand particles for numerical simulation to compare it with experimental results. Evaluating the effect of three-dimensional grain shape on the liquefaction susceptibility of granular soil by using the ODEC technique will be a worth while extension of the current research effort and a valuable contribution in the field of geotechnical engineering. This research methodology can also be extended to study the surface roughness of particles, grain crushing duri ng shear, the influence of thre e-dimensional particle shape on the flow rate of granular materials flowing from hoppers. The behavior of the interface between two materials (soil-foundation or so il-geosynthetic interface) can also be investigated.

PAGE 131

117 REFERENCES Alarcon-Guzman, A., Leonards, G.A., and Chameau, J.L. (1988). Undrained monotonic and cyclic strength of sands. Journal of Geotechnical Engineering, 114(10), 10891109. Alshibi, K.A., and Alsaleh, M.I. (2004). Ch aracterizing surface roughness and shape of sands using digital microscopy. Journal of Computing in Civil Engineering, 18(1), 36-45. Anderson, G.E. (1926). Experiments on the rate s of wear of sand grains. Journal of Geology, 34, 144-158. Arulanandan, K., Seed, H. B., Yogachandran, C., Muraleetharan, K. K., and Seed, R. B. (1993). Centrifuge study on volume changes and dynamic stability of earth dams. Journal of Geotechnical E ngineering, 119(11), 1717-1731. Ashmawy, A.K., Sukumaran, B., and Hoang, A. V. (2003). Evaluating the influence of particle shape on liquefaction behavi or using Discrete Element Method. Proceedings of the thirteenth intern ational offshore and polar engineering conference (ISOPE 2003), Honolulu, Hawaii, May 2003. Balazs, R.J., and Klein, G.D. (1972). R oundness-mineralogical relations of some intertidal sands. Jour nal of Sedimentary Petrology, 42(2), 425-433. Banerjee, I. (1964). Size-r oundness relation in the Baraka r sandstones of the south Karanpur Coalfield, Indi a. Sedimentology, 3, 22-28. Barbosa, R., and Ghaboussi, J. (1992). Di screte finite element method. Engineering Computations, 9, 253-266. Barret, P.J. (1980). The shape of rock part icles, a critical review. Sedimentology, 27, 291-303. Bathurst, R.J., and Rothenburg, L. (1989). Inve stigation of micromechanical features of idealized granular assemblies using DEM. Proceedings of the 1 st US Conference on Discrete Element Methods, CSM, Golden, CO (1989); Engineering Computations, 9, 199-210 (1992).

PAGE 132

118 Been, K., and Jefferies, M. G. (1985). A st ate parameter for sands. Geotechnique, 35(2), 99-112. Been, K., Jefferies, M. G., and Hachey, J. (1991). The critical state of sands. Geotechnique, 41(3), 365-381. Bolton, M.D. (1986). The strength and dilata ncy of sands. Geotechnique, 36(2), 65-78. Borgefors, G., Nystrm, I., and Baja, G.S.D. (1999). Computing skeletons in three dimensions. Pattern Recognition, 32(7), 1225-1236. Bowman, E.T., Soga, K., and Drummond, W. (2001). Particle shape characterization using Fourier descriptor analys is. Geotechnique, 51(6), 545-554. Chen, J.W., Lin, C.Y., and Lee, W.F. (2003). Dilative Behavior of Granular Materials. International Journal of Offshor e and Polar Engineering, 13(4). Cho, G., Dodds, J, and Santamarina, J.C. (2006). Particle shape effects on packing density, stiffness, and strength: natural and crushed sands. Journal of Geotechnical and Geoenvironmenta l Engineering, 132(5), 591-602. Chu, J., and Lo, S. C. R. (1993). On the measur ement of critical state parameters of dense granular soils. Geotechnical Testing Journal, 16(1), 27-35. Chuang, J.-H., Tsai, C.-H., and Ko, M.-C (2000) Skeletonization of three-dimensional object using generalized poten tial field. IEEE Transactions on Pattern Analysis and Machine Intelligence, 22(11), 1241-1251. Clark, M.W. (1981). Quantitative shape analys is: a review. Mathematical Geology, Vol. 13(4), 303-320. Cleary, P.W. (2000). DEM simulation of industrial particle flows: case studies of dragline excavators, mixing in tumblers and centrifugal mills. Powder Technology, 109(1), 83-104. Conolly, J.R. (1965). The occurrence of pol ycrystallinity and undulatory extinction in quartz in sandstones. Journal of Sedimentary Petrology, 35, 116-135. Cubrinovski, M., and Ishihara, K. (2002). Maximum and minimum void ratio characteristics of sands. Soils and Foundations, 42(6), 65-78. Cundall, P. A. (1971). A computer model for simulating progressive large scale movements in block rock systems. Pr oceedings of the Symposium of the International Society of Rock Mech anics, Nancy, II, Article 8.

PAGE 133

119 Cundall, P. A., and Strack O. D. L. (1979) A discrete numerical model for granular Assemblies. Geotechnique, 29, 47-65. Cundall, P. A., and Strack O. D. L. (1979). The distinct element method as a tool for research in granular media, Part II. Report to NSF, Department of Civil and Mineral engineering, University of Minnesota. Minneapolis, MN. Das, N. and Ashmawy, A.K. (2007). Relationshi p between grain size and shape of natural and crushed sand. Geo-Denver 2006 Conference, Denver, CO. Diepenbroek, M., Bartholoma, A., and Ibbe ken, H. (1992). How r ound is round? A new approach to the topic roundness by Four ier grain shape analysis. Sedimentology, 39, 411-422. Dinesh, S.V., and Sitharam, T.G. (2003). discrete element simulation of liquefaction behavior of granular so ils under cyclic loading. CPFTEGE. Dodds, J. (2003). Particle shape and stiffness effects on soil behavior. M.S.C.E. Thesis, Georgia Institute of Technology, GA. Dowdeswell, J.A. (1982). Scanning electron micrographs of quartz sands from cold environments examined using Fourier Shape Analysis. J. Sediment. Petrol., 52(4), 1315-1323. Dyskin, A. V., Estrin, Y., Kanel-Belov, A. J., and Pasternak, E. (2001). Toughening by fragmentation How topology help s. Adv. Eng. Mater., 3(1), 885-888. Ehrlich, R., Brown, J., Yarus, J. M., and Przygocki, R. S. (1980). The origin of shape frequency distributions and the relationship between size and shape. J. Sediment. Petrol., 50(2), 475-484. Ehrlich, R., and Weinberg, B. (1970). An exact method for characterization of grain shape. J. Sediment. Petrol., 40(1), 205-212. Favier, J., Abbaspour-Fard, M, Kremmer, M., and Raji, A. (1999). Shape representation of axi-symmetrical, non-spherical particle s in discrete element simulation using multi-element model particles. Engineering Computations, 16, 467-480. Flook, A.G. (1979). The characterization of te xtural and structural profiles by the automated measurement of their fractal dimensions. 2 nd European Symposium on Particle Characterization, 591-599. Folk, R.L. (1978). Angularity and silica co atings of Simpson Desert sand grains, Northern Territory Australia. Journa l of Sedimentary Petrology, 48, 611-624.

PAGE 134

120 Fraser, H. J. (1935). Experimental study of the porosity and permeability of elastic sediments. Journal of Geology, 13(8), 910-1010. Garboczi, E.J. (2002). Three-dimensional Mathem atical Analysis of Particle Shape using x-ray Tomography and Spherical Harmonics: Application to Aggregates used in Concrete. Cement and Concre te Research, 32(10), 1621-1638. Ghaboussi, J., and Barbosa, R. (1990). Three-dimensional discrete element method for granular materials. International Journal for Numerical and Analytical Methods in Geomechanics, 14, 451-472. Gonzalez, R.C., and Woods, R.E. (2003). Digital image processing. Prentice Hall. Goudie, A.S., and Watson, A. (1981). The shape of desert sand grains. Journal of Arid Environment, 4, 185-190. Holtz, R.D., and Kovacs, W.D. (1981). An Introduction to Geotechnical Engineering. Prentice-Hall, Inc., Englewood, Cliffs, NJ. Houlsby, G.T. (1991). How the dilatancy of soils affects their behavior. European Conference on Soil Mechanics and Foundation Engineering, 4, 1189-1202. Hunt, A.R. (1887). The evidence of the Ske rries Shoal on the wearing of fine sand by waves. Devonshire Association for the A dvancement of Science, Literature and Art, Report and Transactions, 19, 498-515. Inman, D.L. (1953). Areal and seasonal variat ions in beach and nearshore sediments at LaJolla, California. U.S. Army Corp. of Engineers, Beach Erosion Board, Tech. Memo 39, 134. Inman, D.L., Ewing, G.C., and Corliss, J.B. (1966). Coastal sand dunes of Guerrero Negro, Baja, California, Me xico. Geological Society of America Bulletin, 77, 787-802. Ishihara, K. (1985). Stability of natura l deposits during earthquakes. Proc. 11 th Int. Conf. Soil Mechanics and Foundation Engi neering, San Francisco, CA, 1, 321-376. Itasca Consulting Group (1999). Particle Fl ow Code in Two Dimensions. Software Version 3.1. Itasca Consulting Group (1999). Particle Fl ow Code in Three Dimensions. Software Version 3.1.

PAGE 135

121 Iyer, N., Jayanti, S., Lou, K., Kalyan araman, Y., and Ramani, K. (2005). Threedimensional shape searching: state-of-the-a rt review and future trends. ComputerAided Design, 37, 509. Janoo, V. (1998). Quantification of shape, angularity, and surface texture of base course materials. Special Report 98-1, Cold Regions Research & Engineering Laboratory, US Army Corps of Engineers. Jensen, R., Bosscher, P., Plesha, M., and Ed il, T. (1999). DEM simulation of granular media-structure interface: effect of surface roughness and particle shape. International Journal for Numerical and Analytical Methods in Geomechanics, 23, 531-547. Jensen, R.P., Edil, T.B., Bosscher, P.J., Plesha M.E., and Ben Kahla, N. B. (2001). Effect of particle shape on interface behavior of DEM-simulated granular materials. The International Journal of Ge omechanics, 1(1), 1-19. Jia, X., and Williams, R. A. (2001). A packing algorithm for particles of arbitrary shapes. Powder Technology, 120, 175-186. Kaye, B.H. (1978). Specification of the rugge dness and/or texture of a fine particle profile by its fractal dimens ion. Powder Technology, 21, 207-213. Kazhdan, M., Funkhouser, T. and Rusinkiewicz, S. (2003). Rotation invariant spherical harmonic representation of 3D shape de scriptors. Eurographics Symposium on Geometry Processing. Kennedy, S.K., and Lin, W.H. (1986). FRACTa FORTRAN subroutine to calculate the variables necessary to determine the fractal dimension of closed forms. Computers and Geoscience, 12, 705-727. Kennedy, S.K., and Lin, W.H. (1992). A comparis on of Fourier and fractal techniques in the analysis of closed forms. Journa l of sedimentary petrology, 62(5), 842-848. Khalaf, F.I., and Gharib, I.M. (1985). Roundne ss parameters of quartz grains of recent aeolian sand deposits in Kuwa it. Sedimentary Geology, 45, 147-158. Koerner, R.M. (1970). Effects of Particle Ch aracteristics on Soil Strength. Journal of Soil Mechanics and Foundation Div., ASCE, 96(4), 1221-1233. Kolar, J. (2004). Global indexing of 3-D vector geographic features. In Proceedings of International Society for Photogrammetr y and Remote Sensing, 20th Congress, 4, 669-672.

PAGE 136

122 Konrad, J. M., and Watts, B. D. (1995). U ndrained shear strength for liquefaction flow failure analysis. Canadian Geotechnical Journal, 32, 783-794. Krumbein, W.C. (1941). Measurement a nd geological significance of shape and roundness of sedimentary particles. Jour nal of Sedimentary Petrology, 11(2), 6472. Kuenen, P.H. (1960). Experimental abra sion, 4. Aeolian action on sand. Journal of Geology, 68, 427-449. Lambe, T.W. and Whitman, R.V. (1969). Soil Mechanics, John Wiley & Sons, Inc., NY, USA. Lanius, C. (1997). Fractal Dimension. http://math.rice.edu/~lanius/fractals/dim.html (Last accessed: 04/24/2007). Lees, G. (1964). The measurement of particle shape and its influence in engineering materials. Journal of the British Granite and Whinestone Federation, London, 122. Li, J. (2002). Three-dimensional shape m odeling: segmentation, reconstruction and registration. Ph.D. Thesis, Electrical E ngineering, University of Michigan. Lien, J.-M., and Amato, N.M. (2005). Simultaneous Shape Decomposition and Skeletonization Using Approximate Convex Decomposition. Technical Report, TR05-004, Parasol Lab. Department of Computer Science, Texas A&M University. Lin, X., and NG, T.T. (1997). A three-dimensi onal discrete element model using arrays of ellipsoids. Geotechnique, 47(2), 319-329. Lipson, H., and Shpitalni, M. (2002). Correla tion-based reconstruction of a 3D object from a single freehand sketch. Proceedings of AAAI Spring Symposium Series Sketch understanding, 99-104, http://citeseer. ist.psu.edu/509787.html Loy, J. (2002). The Koch Curve. http://www.jimloy.com/fractals/koch.htm (Last accessed: 04/24/2007). Luaa, V. (1996). The Spherical Harmonics gallery page, Universidad de Oviedo, Oviedo, Spain http://web.uniovi.es/qcg/harmonics/harmonics.html (Last accessed: 04/24/2007). MacCarthy, G.R. (1933). The rounding of beach sands. American Journal of Science, 25, 205-224.

PAGE 137

123 Mandelbrot, B. B. (1967). How long is the coas t of Britain? Statistical self-similarity and fractal dimension. Science, 156(3775), 636-638. Mandelbrot, B. B. (1977). Fractals: Forms, chance and Dimensions. W. H. Freeman, San Francisco, 361. Manzari, M.T., and Nour, M.A. (2000). Signifi cance of soil dilatanc y in slope stability analysis. Journal of Geotechnical and Geoenvironmental Engineering, 126(1), 7580. Masad, E., and Button, J. W. (2000). Unified imaging approach for measuring aggregate angularity. The International Journal of Computer-Aided Civil and Infrastructure Engineering, 15(4), 273-280. Masad, E., Saadeh, S., Al-Rousan, T., Garboczi E., and Little, D. (2005). Computations of particle surface characteristics using optical and X-ray CT images. Computational Materials Science, 34, 406-424. Matsushima, T. (2004). 3-D image-based di screte element mode ling for irregularlyshaped grains. Numerical Modeling in Micromechanics via Particle Methods 2004 Shimizu, Hart & Cundall (eds.) 2004 Taylor & Francis Group, London, 421-427. Matuttis, H.G., Luding, S., and Herrmann, H.J. (2000). Discrete element simulations of dense packings and heaps made of spheri cal and non-spherical particles. Powder Technology, 109(1), 278-292. Mazzullo, J., Alexander, A., Tieh, T., a nd Menglin, D. (1992). The effects of wind transport on the shapes of quartz silt gr ains. Journal of Sedimentary Petrology, 62(6), 961-971. Mazzullo, J., Sims, D., and Cunningham, D. ( 1986). The effect of aeolian sorting and abrasion upon the shapes of fine quartz sand grains. Journal of Sedimentary Petrology, 56(1), 45-56. Meloy, T.P. (1977). Fast Fourier transform applied to shape analysis of particle silhouettes to obtain morphological data. Powder Technology, 17, 27-35. Miura, K., Maeda, K., Furukawa, M., and Toki S. (1998). Mechanical characteristics of sands with different primary proper ties. Soils and Foundations, 38, 159-172. Montgomery, D.C., and Runger, G.C. (2003). Applied statistics and probability for engineers. John Wiley & Sons, Inc., USA.

PAGE 138

124 Morse, P.M., and Feschbach, H. (1953). Met hods of theoretical phys ics. McGraw Hill, New York, NY, 1978. Mustoe, G.G.W., Henriksen, M., and Huttelm aier, H.P. (1989). Proceedings of the 1 st US Conference on Discrete Element Methods, CSM, Golden, CO. Nakata, Y., Kato, Y., Hyodo, M., Hyde, A. F. L., and Murata, H. (2001). Onedimensional compression behavior of uniformly graded sand related to single particle crushing strength. Soils and Foundations, 41(2), 39-51. Ng, T.T. (1989). Numerical simulation of granular soils under monotonic and cyclic loading: A particulate mechanics approach. Ph.D. Dissertation, Rensselaer Polytechnic Institute, Troy, NY. Ng, T.T. (1994). Numerical simulations of gr anular soil using elli ptical particles. Computers and Geotech nics, 16(2), 153-169. Ng, T.T., and Dobry, R. (1991). CONBAL Si mulated granular material using quartz spheres with the Discrete Element Method. Report to NSF, Rensselaer Polytechnic Institute, Troy, NY. Norris, G. (1976). The effect of particle size and the natural variation in particle shape and surface roughness on the stress-strain a nd strength behavior of uniform quartz sands. Ph.D. thesis, University of California, Berkeley, CA. Orford, J.D., and Whalley, W.B. (1983). The use of fractal dimension to quantify the morphology of irregular-shaped particles. Sedimentology, 30, 655-668. OSullivan, C., Cui, L., and Bray, J.D. ( 2004). Three-dimensional discrete element simulations of direct shear test. Nu merical Modeling in Micromechanics via Particulate Methods 2004 Shimizu, Ha rt and Cundall (eds.), 2004 Taylor and Francis Group, London. Pettijohn, F.J. (1957). Sedimentary Rock s. Harper & Bros, New York, NY, 718. Pettijohn, F.J., and Lundahl, A.C. (1934). Shape and roundness of Lake Erie beach sand. Journal of Sedimentary Petrology, 14, 69-78. Phillips, J.C., Hogg, A.J., Kerswell, R.R., and Thomas, N.H. (2006). Enhanced mobility of granular mixtures of fi ne and coarse particles. Earth and Planetary Science Letters, 246, 466. Plumley, W.J. (1948). Black hills terrace gravel s: a study in sediment transport. Journal of Geology, 56, 526-577.

PAGE 139

125 Podczeck, F. (1997). A shape factor to asses th e shape of particles using image analysis. Powder Technology, 93, 47-53. Pollack, J.M. (1961). Significance of compos itional and textural properties of South Canadian River Channel sands, New Mexico, Texas and Oklahoma. Journal of Sedimentary Petrology, 31, 15-37. Potapov, A., and Campbell, C. (1998). A fast model for the simulation of non-round particles. Granular Matter, 1, 9-14. Powers, M.C. (1953). A new roundness scale fo r sedimentary particles. Journal of Sedimentary Petrology, 23, 117-119. Ramez, M.R., and Mosalamy, F.H. (1969). The deformed nature of various size fractions in some clastic sands. Journal of Sedimentary Petrology, 39, 1182-1187. Rao, C., Tutumluer, E., and Kim, I. T. (2002). Quantification of coarse aggregate angularity based on image analysis. Transportation Research Record, 1787, Transportation Research Board, Nationa l Research Council, Washington DC. Riester, D.D., Shipp, R.C., and Ehrlich, R. (1982). Patterns of quartz sand grain shape variation, Long Island Littoral and shelf. J. Sediment. Petrol., 52(4), 1307-1314. Rivas, J.A. (2005). Three-dimensional dig ital image processing a nd reconstruction of granular particles. M.S.C.E. Thesis, Univ ersity of South Florida, Tampa, FL. Russell, R.D. (1939). Recent marine sediment s. Society of Economic Paleontologists and Mineralogists Special Publication, 4, 736. Russell, R.D., and Taylor, R.E. (1937). Roundne ss and shape of Mississippi River sands. Journal of Geology, 45, 225-267. Sallam, A.M. (2004). Studies on Modeling Angu lar Soil Particles Using the Discrete Element Method. Ph.D. Thesis, Universi ty of South Florida, Tampa, FL. Santamarina, J.C., and Cascante, G. (1998). Effect of surface roughness on wave propagation parameters. Ge otechnique, 48(1), 129-136. Santamarina, J. C., and Cho, G. C. (2001). Dete rmination of critical state parameters in sandy soils-simple procedure. Geotec hnical Testing Journal, 24(2), 185-192. Santamarina, J.C., and Cho, G.C. (2004). Soil Behavior: The Role of Particle Shape. Proc. Skempton Conf., March, London.

PAGE 140

126 Sasitharan, S., Robertson, P. K., Sego, D. C., and Morgenstern, N. R. (1994). Stateboundary surface for very loose sand and its practical implications. Canadian Geotechnical Journal, 31, 321-334. Schanz, T., and Vermeer, P.A. (1996). A ngle of friction and dilatancy of sand. Geotechnique, 46(1), 145-151. Schwarcz, H.P., and Shane, K.C. (1969). Measurement of particle shape by Fourier analysis. Sedimentology, 13, 179-212. Schwarz, H.B., and Exner, H.E. (1980). The implementation of the concept of fractal dimension in a semi-automatic image analyzer. Powder Technology, 27, 207-213. Shimobe, S., and Moroto, N. (1995). A new classification chart for sand liquefaction. Earthquake geotechnical engi neering, K. Ishihara, ed., Balkema, Rotterdam, The Netherlands, 315-320. Shinohara, K., Oida, M., and Golman, B. (2000). Effect of particle shape on angle of internal friction by triaxial compression test. Powder Technology, 107, 131-136. Sorby, H.C. (1877). The application of the microscope to geology. Monthly Microscopical Journal, 17, 113-136. Strack O. D. L. and Cundall, P. A. (1978). The distinct element method as a tool for research in granular media, Part I. Report to NSF, Department of Civil and Mineral engineering, University of Minnesota. Minneapolis, MN. Sukumaran, B. (1996). Study of the effect of pa rticle characteristics on the flow behavior and strength properties of pa rticulate materials. Ph.D. Thesis, Purdue University, West Lafayette, IN. Sukumaran, B., and Ashmawy, A.K. (2001). Quantitative characterization of the geometry of discrete partic les. Geotechnique, 51(7), 619-627. Sukumaran, B., and Ashmawy, A.K. (2001). Influence of inherent particle characteristics on the strength properties of Particulate materials. Annual International Society of Offshore and Polar Engineering Conference, Oslo, Norway. Taylor, D.W. (1948). Fundamentals of soil m echanics. John Wiley and Sons, New York. Taylor, L.M., and Preece, D.S. (1989). Simulatio n of blasting induced rock motions using spherical element models. Proceedings of the 1 st US Conference on Discrete Element Methods, CSM, Golden, CO.

PAGE 141

127 Thomas, D.S.G. (1987). The roundness of aeo lian quartz sand grains. Sedimentary Geology, 52, 149-153. Thomas, M.C., Wiltshire, R.J., and Williams, A. T. (1995). The use of Fourier descriptors in the classification of particle shape. Sedimentology, 42, 635-645. Thomas, P.A., and Bray, J.D. (1999). Capturing nonspherical shape of granular media with disk clusters. Journal of Geotec hnical and Geoenvironmental Engineering, 125(3), 169-178. Ting, J., Khawaja, M., Meachum, L., and Rowe ll, J. (1993). An ellipse-based discrete element model for granular materials. International Journal for Numerical and Analytical Methods in Geomechanics, 17, 603-623. Ting, J.M., Corkum, B.T., Kauffman, C.R., and Greco, C. (1989). Discrete numerical model for soil mechanics. Journal of Geotechnical Engineering, ASCE, (3), 379398. Twenhofel, W.H. (1946). Beach and river sa nds on the coastal region of southwest Oregon with particular reference to black sands. American Journal of Science, 244, 114-139, 200-214. Twenhofel, W.H. (1950). Principles of sedi mentation. McGraw-Hill Book Co. Inc., New York, 302-311. Uthus, L, Hoff, I., and Horvli, I. (2005). Evaluation of grain shape characterization methods for unbound aggregates. 7th Intern ational Conference on the bearing capacity of roads, railways and airfields 2005, BCRA. Trondheim, Norway. Wadell, H. (1932). Volume, shape & roundness of rock particles. J. of Geology, 40, 443451. Wadell, H. (1933). Sphericity and roundness of rock particles. J. of Geology, 41, 310331. Wadell, H. (1935). Volume, shape & roundness of quartz particles. Journal of Geology, 43, 250-286. Williams, J., and Mustoe, G.G.W. (1993). Procee dings of the International Conference on the Discrete Element Met hods, MIT, Cambridge, MA. Williams, J.R., and Pentland, A.P. (1992). Superquadrics and modal dynamics for discrete elements in interactive desi gn. Engineering Comput ations, 9(2), 115-127.

PAGE 142

128 Wu, F.-C., Ma, W.-C., Liou, P.-C., Laing, R.-H. and Ouhyoung, M. (2003). Skeleton extraction of 3d objects with visible repul sive force. In Computer Graphics Workshop 2003, Hua-Lien, Taiwan. Yang, J., and Li, X.S. (2004). State-Dependent Strength of Sands from the Perspective of Unified Modeling. J. Geotech. and Geoenvir. Eng., 130(2), 186-198. Youd, T.L. (1973). Factors controlling ma ximum and minimum densities of sands, evaluation of relative density and its role in geotechnical projects involving cohesionless soils, ASTM STP 523, 98-112. Yudhbir and Abedinzadeh, R. (1991). Quantifi cation of particle shape and angularity using image analyzer. Geotechnical Testing Journal, 14(3), 296-308. Zahn, C.T., and Roskies, R.Z. (1972). Fourier descriptors for plane closed curves. IEEE Transactions on Com puters, C-21(3), 269-282. Zelasko, J.S., Krizek, R.J., and Edil, T.B. ( 1975). Shear Behavior of Sands as a Function of Grain Characteristics. Istanbul Conference on Soil Mech. Found. Eng., 1, 5564. Zwillinger, D. (1995). Spherical coordi nates in space, .9.3 in CRC standard mathematical tables and formulae. Boca Raton, FL: CRC Press, 297-298.

PAGE 143

129 APPENDICES

PAGE 144

Appendix A: Data Sets Total 26 types of different sand samples were collected from various locations. Their two-dimensional projection images are shown below. US Silica #1 Dry US Silica Std.Melt Daytona Beach Sand Rhode Island Sand Nice Sand Fontainebleau Sand Figure A.1 Sand Samples Collected for the Present Study 130

PAGE 145

Appendix A: (Continued) Loire River Sand Hostun Sand Toyoura Sand Michigan Dune Sand Tecate River Sand Kahala Beach Sand Indian Rocks Beach Sand Belle Air Beach Sand 131 Figure A.1 (Continued)

PAGE 146

Appendix A: (Continued) Clearwater Beach Sand Gulf Beach Sand Long Beach Sand Boca Grande Beach Sand Oxnard Beach Sand Arroyo Alamar River Sand Figure A.1 (Continued) 132

PAGE 147

Appendix A: (Continued) Rincon Beach Sand Panama Malibu Beach Sand Ala Wai Beach Sand Red Sea Dune Sand Madeira Beach Sand Redington Shores Sand Figure A.1 (Continued) 133

PAGE 148

ABOUT THE AUTHOR Ms. Nivedita Das received a bachelors De gree in Civil Engineering from Jadavpur University, India in 2000 and Masters Degr ee in Geotechnical Engi neering from Bengal Engineering College, India in 2002. She started working with an engineering consultancy firm until she joined the Ph.D. program in Geotechnical Engineering in the department of Civil and Environmental Engineering at th e University of South Florida in 2003 under the supervision of Dr. Alaa Ashmawy. Her pr imary areas of doctoral research include three-dimensional characterization particle morphology and discrete element modeling. She served as a secretary/treasurer for th e USF Geotechnical Society Student Chapter. She coauthored two conference publications a nd three of her research papers are under review in referred journals.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001919512
003 fts
005 20071218131746.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071218s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002072
035
(OCoLC)184843132
040
FHM
c FHM
049
FHMM
090
TA145 (ONLINE)
1 100
Das, Nivedita.
0 245
Modeling three-dimensional shape of sand grains using Discrete Element Method
h [electronic resource] /
by Nivedita Das.
260
[Tampa, Fla.] :
b University of South Florida,
2007.
520
ABSTRACT: The study of particle morphology plays an important role in understanding the micromechanical behavior of cohesionless soil. Shear strength and liquefaction characteristics of granular soil depend on various morphological characteristics of soil grains such as their particle size, shape and surface texture. Therefore, accurate characterization and quantification of particle shape is necessary to study the effect of grain shape on mechanical behavior of granular assembly. However, the theoretical and practical developments of quantification of particle morphology and its influence on the mechanical response of granular assemblies has been very limited due to the lack of quantitative information about particle geometries, the experimental and numerical difficulties in characterizing and modeling irregular particle morphology.Motivated by the practical relevance of these challenges, this research presents a comprehensive approach to model irregular particle shape accurately both in two and three dimensions. To facilitate the research goal, a variety of natural and processed sand samples is collected from various locations around the world. A series of experimental and analytical studies are performed following the sample collection effort to characterize and quantify particle shapes of various sand samples by using Fourier shape descriptors. As part of the particle shape quantification and modeling, a methodology is developed to determine an optimum sample size for each sand sample used in the analysis. Recently, Discrete Element Method (DEM) has gained attention to model irregular particle morphology in two and three dimensions.In order to generate and reconstruct particle assemblies of highly irregular geometric shapes of a particular sand sample in the DEM environment, the relationship between grain size and shape is explored and no relationship is found between grain size and shape for the sand samples analyzed. A skeletonization algorithm is developed in this study in order to automate the Overlapping Discrete Element Cluster (ODEC) technique for modeling irregular particle shape in two and three dimensions. Finally, the two-dimensional and three-dimensional particle shapes are implemented within discrete element modeling software, PFC2D and PFC3D, to evaluate the influence of grain shape on shear strength behavior of granular soil by using discrete simulation of direct shear test.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 133 pages.
Includes vita.
590
Advisor: Alaa K. Ashmawy, Ph.D.
653
Fourier transform.
Spherical harmonics.
Shape descriptors.
Skeletonization.
Angularity.
Roundness.
Liquefaction.
Overlapping discrete element cluster.
690
Dissertations, Academic
z USF
x Civil Engineering
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2072