IDENTIFICATION AND ASSESSMENT OF THE DOMINANT AGGREGATE SIZE
RANGE OF ASPHALT MIXTURE
By
SUNGHO KIM
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2006
Copyright 2006
by
Sungho Kim
To my wife, Heejoo Moon, and parents, Byungtae Kim and Sangyeoun Lee.
ACKNOWLEDGMENTS
It is a great pleasure for me to thank and acknowledge the many individuals who
assisted and supported me during the course of my doctoral program. First of all, I would
like to express my sincere appreciation to my committee chairman, Dr. Reynaldo Roque,
and my committee cochairman, Dr. Bjorn Birgisson, for their invaluable guidance and
support throughout my studies at the University of Florida. I would have not been able to
reach this milestone if it was not for their advice and understanding. I would also like to
express my gratitude to the other committee members, Dr. Mang Tia, Dr. Byron E. Ruth,
and Dr. Bhavani V. Sankar, for their support in accomplishing my work. I could not ask
for a better committee group. They were all great advisors and mentors.
I would like to thank George Lopp , Alvaro Guarin, and Avraham A. Chileuitt for
their adivise and assistance in performing testing and analysis.
I would like to thank Mr. Gregory A. Sholar, Howie Moseley, and Mrs. Shanna
Johnson of the Florida Department of Transportation for their assistance in performing
testing.
I would like to thank Dr. Christos Drakos, Tanya Riedhammer, and others in
Materials group for their help and friendship.
I would also like to thank all Korean students in our department and member of
Gainesville Korean Catholic Community for sharing a lot of time together. I won’t forget
every moment we had.
iv
TABLE OF CONTENTS
page
ACKNOWLEDGMENTS ................................................................................................. iv
LIST OF TABLES........................................................................................................... viii
LIST OF FIGURES .............................................................................................................x
ABSTRACT.......................................................................................................................xv
CHAPTER
1
INTRODUCTION ...................................................................................................1
1.1
1.2
1.3
2
LITERATURE REVIEW ........................................................................................4
2.1
2.2
2.3
2.4
2.5
3
Problem ........................................................................................................1
Objectives ....................................................................................................2
Scope............................................................................................................2
Shear Resistance and Rutting Potential .......................................................4
Criteria Associated with VMA and Restricted Zone ...................................4
Gradation Parameters: n and a .....................................................................5
Bailey Method..............................................................................................6
Summary ....................................................................................................10
NEW DEVELOPMENT........................................................................................12
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
Porosity as a Criterion for Interlocking .....................................................12
Application to Asphalt Mixture .................................................................13
Porosity of Individual Particle Sizes..........................................................15
Theoretical Developments .........................................................................18
3.4.1 Dominant Aggregate Size Range (DASR) ....................................18
3.4.2 Interstitial Volume (IV) .................................................................20
3.4.3 Interstitial Surface (IS)...................................................................20
Particle Spacing on the IS ..........................................................................22
Determination of DASR ............................................................................22
Spacing Analysis and Interaction Diagram................................................23
Interaction Diagrams and DASR Porosity.................................................29
v
4
EVALUATION AND REFINEMENT..................................................................35
4.1
4.2
4.3
4.4
4.5
4.6
5
FURTHER TESTING............................................................................................87
5.1
5.2
5.3
5.4
5.5
5.6
5.7
6
Introduction................................................................................................35
Field Performance: Superpave Monitoring Projects 1 to 8.......................36
Laboratory Performance: Superpave Monitoring Projects 8 to 12 ...........46
WesTrack Test Sections.............................................................................57
4.4.1 General Description .......................................................................57
4.4.2 Experiment Design and Performance History ...............................58
4.4.3 Interaction Diagrams......................................................................61
4.4.4 Summary ........................................................................................71
NCAT Test Sections ..................................................................................72
4.5.1 Interaction Diagrams: Coarse Mixtures ........................................73
4.5.2 Interaction Diagrams: Fine Mixtures ............................................75
4.5.3 Interaction Diagrams: Dense-Coarse Mixtures.............................78
4.5.4 Interaction Diagrams: SMA Mixtures ..........................................80
4.5.4 Summary ........................................................................................81
Additional Observations ............................................................................83
4.6.1 Excessively Low DASR Porosity ..................................................84
Introduction................................................................................................87
Materials ....................................................................................................87
Gradations ..................................................................................................88
5.3.1 Georgia Granite..............................................................................88
5.3.2 Rinker South Florida Limestone....................................................90
Mix Design.................................................................................................94
APA Test....................................................................................................94
ServoPac Test.............................................................................................99
Summary ..................................................................................................101
CLOSURE ...........................................................................................................103
6.1
6.2
6.3
Summary of Findings...............................................................................103
Conclusions..............................................................................................105
Recommendations....................................................................................106
APPENDIX
A
GRADATIONS FOR SUPERPAVE MONITORING PROJECT ......................107
B
POROSITY RESULTS FOR SUPERPAVE PROJECTS...................................116
C
TRAFFIC AND RUT DEPTH DATA FOR SUPERPAVE MONITORING
PROJECT.............................................................................................................119
D
THEORETICAL CALCULATION FOR SURFACE AREA.............................123
vi
E
LABORATORY MIXTURES INFOMATION ..................................................131
LIST OF REFERENCES.................................................................................................139
BIOGRAPHICAL SKETCH ...........................................................................................142
vii
LIST OF TABLES
Table
page
2-1
Recommended Aggregate Ratios for Coarse Mixtures..............................................9
4-1
Original Experimental Factors .................................................................................58
4-2
Experiment Design ...................................................................................................59
4-3
Materials...................................................................................................................59
4-4
Rut depth for Original Coarse Mixtures...................................................................66
4-5
Rut Depth for Coarse Replacement Sections (36, 37, 55, 56) .................................66
4-6
Field Rut Depth for Coarse Replacement Sections (35, 38, 39, 54) ........................69
4-7
Rut Depth for Fine Mixtures ....................................................................................69
4-8
Rut Depth for Fine plus Mixtures ............................................................................70
4-9
Reference Figures and Tables for NCAT.................................................................73
4-10 Field Rut Depth for Sections E2, E3, and E4...........................................................75
4-11 Field Rut Depth for Sections E8, E9, and E10.........................................................77
4-12 Field Rut Depth for Sections N5, N6, N7 and N8....................................................80
4-13 Field Rut Depth Sections W3 lower, W3 upper, W4 lower, and W4 upper ............82
5-1
Aggregate Sources....................................................................................................87
5-2
Gradation IDs for Testing ........................................................................................88
5-3
Summary of the DASR Porosity and Interaction Diagram for GA Granite
Gradations ................................................................................................................91
5-4
Summary of the DASR Porosity and Interaction Diagram for Rinker South FL
Limestone Gradations ..............................................................................................93
5-5
Summary for Test Matrix .........................................................................................93
viii
5-6
Designed volumetric information.............................................................................94
E-1 Blending Percent for GA Granite Gradations ........................................................131
E-2 Blending Percent for Rinker South FL Limestone.................................................131
E-3 JMF for GA Granite Gradation ..............................................................................132
E-4 JMF for Rinker South FL Limestone .....................................................................132
E-5 Batch Weight for Granite Gradations.....................................................................133
E-6 Batch Weight for Limestone Gradations................................................................135
E-7 DRD, ARD, and Area Change Results from APA.................................................137
E-8 Student T-test Results for DRD .............................................................................138
E-9 Student T-test Results for ARD .............................................................................138
ix
LIST OF FIGURES
Figure
page
2-1
Determination of Mix Type........................................................................................7
2-2
Four Main Principles of Bailey Method for Coarse Mixtures ...................................8
2-3
Example of Coarse Gradation Mix.............................................................................9
3-1
Relationship among Soil Phases ..............................................................................12
3-2
Mixture Components for Porosity Calculation ........................................................14
3-3
Example Gradations .................................................................................................16
3-4
Individual Porosity Results ......................................................................................17
3-5
Dominant Aggregates and Interstitial Volume.........................................................19
3-6
The Failure Surface from a Broken IDT Sample .....................................................21
3-7
Hexagonal Pattern Distribution and Spacing Calculation for Each Size .................25
3-8
Modified Hexagonal Area from Outer Solid to Dotted Line (nt=2).........................26
3-9
The Representative Areas Based on Hexagonal Patterns for Each Step..................27
3-10 Spacing Result for the Binary Mixture with 9.5, 4.75mm .......................................28
3-11 Slope (spacing change) for the Binary Mixture .......................................................30
3-12 Interaction Diagram..................................................................................................31
3-13 Porosity Result after Considering Interaction ..........................................................33
4-1
Interaction Diagram for Field Mixtures of Project 1 and 2......................................37
4-2
Interaction Diagram for Field Mixtures of Project 3 Layer A .................................38
4-3
Interaction Diagram for Field Mixtures of Project 3 Layer B..................................38
4-4
Interaction Diagram for Field Mixtures of Project 4 Layer A .................................39
x
4-5
Interaction Diagram for Field Mixtures of Project 4 Layer B..................................39
4-6
Interaction Diagram for Field Mixtures of Project 5 Layer A .................................40
4-7
Interaction Diagram for Field Mixtures of Project 5 Layer B..................................40
4-8
Interaction Diagram for Field Mixtures of Project 6................................................41
4-9
Interaction Diagram for Field Mixtures of Project 7................................................41
4-10 Interaction Diagram for Field Mixtures of Project 8................................................42
4-11 Porosity Result for Field Mixtures ...........................................................................42
4-12 Rut Depth/ESALs from Field Measurement ............................................................44
4-13 Average Rut Depth/ESALs for Different Porosity Groups (Round I) .....................44
4-14 Average Rut Depth/ESALs for Different Porosity Groups (Round II)....................45
4-15 Interaction Diagram for Plant Mixtures of Project 8................................................47
4-16 Interaction Diagram for Plant Mixtures of Project 9................................................47
4-17 Interaction Diagram for Plant Mixtures of Project 10..............................................48
4-18 Interaction Diagram for Plant Mixtures of Project 11..............................................48
4-19 Interaction Diagram for Plant Mixtures of Project 12..............................................49
4-20 Porosity Result for Plant Mixtures ...........................................................................49
4-21 APA Test Result (Rib) for Plant Mix Gradations ....................................................50
4-22 Concepts for DRD and ARD....................................................................................51
4-23 Area Change Interpretation ......................................................................................51
4-24 Absolute Rut Depth for Different Porosity Groups (APA)......................................53
4-25 Area Change for Different Porosity Groups (APA).................................................53
4-26 Servopac Result for Plant Mix Gradations...............................................................54
4-27 Servopac Result for Different Porosity Groups .......................................................56
4-28 WesTrack - Layout of Test Track (not to scale) ......................................................57
4-29 JMF Mixtures Gradations.........................................................................................61
xi
4-30 Interaction Diagram for JMF Coarse and JMF Coarse Replacement ......................62
4-31 Interaction Diagram for JMF Fine and JMF Fine plus.............................................62
4-32 Gradation of Coarse Replacement Sections (36, 37, 55, 56) ...................................63
4-33 Interaction Diagram for Sections 36 and 37.............................................................64
4-34 Interaction Diagram for Sections 55 and 56.............................................................64
4-35 DASR Porosity (ηDASR) of Coarse Replacement Sections (36, 37, 55, 56)..............65
4-36 Gradation of Coarse Replacement Sections (35, 38, 39, 54) ...................................67
4-37 Interaction Diagram for Sections 35 and 38.............................................................67
4-38 Interaction Diagram for Sections 39 and 54.............................................................68
4-39 DASR Porosity (ηDASR) of Coarse Replacement Sections (35, 38, 39, 54)..............68
4-40 Maximum Rut Depth for Fine and Fine plus Mixtures............................................70
4-41 NCAT - Layout of Test Track (not to scale) ............................................................72
4-42 Gradation of Sections E2, E3, and E4 ......................................................................74
4-43 Interaction Diagram for Sections E2, E3, and E4 ....................................................74
4-44 DASR Porosity of Sections E2, E3, and E4 .............................................................75
4-45 Gradation of Sections E8, E9, and E10 ....................................................................76
4-46 Interaction Diagram for Sections E8, E9, and E10 ..................................................76
4-47 DASR Porosity of Sections E8, E9, and E10 ...........................................................77
4-48 Gradation of Sections N5, N6, N7, and N8..............................................................78
4-49 Interaction Diagram for Sections N5, N6, N7, and N8 ............................................79
4-50 DASR Porosity of Sections N5, N6, N7, and N8.....................................................79
4-51 Gradation of Sections W3 lower, W3 upper, W4 lower, and W4 upper..................80
4-52 Interaction Diagram for Sections W3 lower, W3 upper, W4 lower, and W4
upper.........................................................................................................................81
4-53 DASR Porosity of Sections W3 lower, W3 upper, W4 lower, and W4 upper.........82
4-54 Finite Element Model of Aggregate and Interstitial Volume...................................85
xii
4-55 Interstitial Spacing (Volume) vs Local Stress..........................................................86
5-1
Gradations for GA Granite .......................................................................................89
5-2
Interaction Diagram for GA Granite Gradations......................................................89
5-3
DASR Porosity for GA Granite Gradations .............................................................90
5-4
Gradations for Rinker South FL Limestone .............................................................91
5-5
Interaction Diagram for Rinker South FL Limestone Gradations............................92
5-6
DASR Porosity for Rinker South FL Limestone Gradations...................................93
5-7
APA Results by System Measurement.....................................................................95
5-8
Differential Rut Depth and Absolute Rut Depth Results from APA .......................96
5-9
Area Change Results from APA ..............................................................................97
5-10 Relationship between Hill Height (DRD-ARD) and, DRD or ARD .......................98
5-11 Relationship between Hill Height (DRD-ARD) and Area Change..........................98
5-12 Results of The Maximum Hill Height (DRD-ARD) ................................................99
5-13 ServoPac Test Results ............................................................................................100
5-14 Relationship between the Failure Strain and the Rut Depth ..................................100
5-15 Pictures for Bad Performance Samples after APA Test ..........................................101
5-16 Pictures for Good Performance Samples after APA Test........................................101
A-1 Gradations for Project 1 and 2................................................................................108
A-2 Gradations for Project 3 Layer A ...........................................................................108
A-3 Gradations for Project 3 Layer B ...........................................................................109
A-4 Gradations for Project 4 Layer A ...........................................................................109
A-5 Gradations for Project 4 Layer B ...........................................................................110
A-6 Gradations for Project 5 Layer A ...........................................................................110
A-7 Gradations for Project 5 Layer B ...........................................................................111
A-8 Gradations for Project 6 .........................................................................................111
xiii
A-9 Gradations for Project 7 Layer A ...........................................................................112
A-10 Gradations for Project 8 Layer A ...........................................................................112
A-11 Gradations for Project 8 Layer B ...........................................................................113
A-12 Gradations for Project 8 Plant Mixture ..................................................................113
A-13 Gradations for Project 9 .........................................................................................114
A-14 Gradations for Project 10 .......................................................................................114
A-15 Gradations for Project 11 .......................................................................................115
A-16 Gradations for Project 12 .......................................................................................115
B-1 Porosity Results for Group 1 (Field Gradations for Projects 3, 4, 5, 7, and 8, and
Plant-Mix Gradations for Project ...........................................................................117
B-2 Porosity Results for Group 2 (Field Gradation for Projects 6, and Plant-Mix
Gradations for Projects 8, and 12)..........................................................................118
C-1 Cumulative Average Rut Depth for Each Round...................................................120
C-2 Cumulative ESALs for Each Round ......................................................................121
C-3 Total Rut Depth and ESALs...................................................................................122
D-1 Mixture Cut Through by an Arbitrary Interstitial Plane ........................................124
D-2 Particles on an Interstitial Plane .............................................................................124
D-3 Maximum Protrusion Area (Hemisphere)..............................................................125
D-4 m Times Cuts for a Hemisphere.............................................................................125
D-5 The Case with Protruded and Embedded Spheres on the Plane.............................125
D-6 The Case with Only Protruded Spheres on the Plane.............................................126
D-7 Surface Area of the Spherical Cap .........................................................................126
D-8 Surface Area for m Cut with r = 3..........................................................................127
D-9 Example of the Protruded or Embedded Depth of Particles ..................................128
D-10 Different Types of Prolate Spheroid ......................................................................129
D-11 Surface Area for m Cut with a = 1, b = 2 ...............................................................130
xiv
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
IDENTIFICATION AND ASSESSMENT OF THE DOMINANT AGGREGATE SIZE
RANGE OF ASPHALT MIXTURE
By
Sungho Kim
May 2006
Chair: Reynaldo Roque
Cochair: Bjorn Birgisson
Major Department: Civil and Coastal Engineering
The importance of aggregate structure on asphalt mixture performance has been
well established on the basis of experience and is well documented in the literature.
Furthermore, coarse aggregate structure is most important for resistance to rutting, and
recent work has shown that it can also play a significant role in resistance to damage and
fracture. Therefore, large enough aggregates should engage dominantly in the structure
for good mixture performance. This study focused on the development of a conceptual
and theoretical approach to evaluate coarse aggregate structure based on gradation.
According to a well-known fact in soil mechanics, the porosity of an assemblage
of granular particles (e.g., the aggregate within an asphalt mixture) must be no greater
than 50% for the particles to be in contact with each other. This also implies that one can
use porosity as a criterion to assure contact between large enough particles within the
mixture to provide suitable resistance to deformation and fracture.
xv
A theoretical analysis procedure was developed to calculate the center to center
spacing between specific size particles within a compacted assemblage of particles of
known gradation. Thus, the 70/30 proportion can be used to determine whether particles
on contiguous Superpave sieves can form an interactive network of particles in
continuous contact with each other. The range of particle sizes determined to be
interactive was referred to as the dominant aggregate size range (DASR) and its porosity
must be no more than 50% for the particles to be in contact with each other.
It was concluded through the extensive analysis with existing database and lab tests
that porosity of the DASR may provide a good criterion for determining the suitability of
gradation for dense-graded asphalt mixture. The approach should be further developed
and evaluated for use in mixture design and analysis.
xvi
CHAPTER 1
INTRODUCTION
1.1 Problem
The performance of Hot-Mix Asphalt (HMA) is related to particle size distribution,
which affects the most important properties of the mix, such as cracking resistance,
rutting resistance, durability, permeability, and workability. Therefore, having an
adequate aggregate particle distribution is a very important factor in order to have good
field performance. Typically the selection of aggregate gradation is made based on
specification bands within control points, but the main question is how to choose the best
possible blend to achieve better performance (Asphalt Institute and the Heritage Group,
2005).
The Superpave mix design method requires that gradation should pass within
control points and avoid a specified restricted zone (Asphalt Institute, 2001). However,
many HMA mixtures that pass through the restricted zone have been found to perform
well. On the other hand, many mixtures, which meet Superpave criteria, have not
exhibited good performance. Additionally, Superpave gradation specifications have not
considered aggregate structure fully.
This study focused on the development of a conceptual and theoretical approach to
evaluate coarse aggregate structure based on gradation. The goal was to develop a
system to help evaluate and, if necessary, modify gradations to ensure that mixtures will
have sufficient aggregate interlock to resist deformation and cracking. It is recognized
that this alone would obviously not ensure good mixture performance, which will also
1
2
depend on the characteristics and properties of the finer components of the mixture,
including the binder, but it would help to eliminate mixtures that will not perform well,
regardless of the quality of these other components. The study also led to concepts that
may lead to the development of other useful criteria associated with these other
components.
1.2 Objectives
As mentioned earlier, this study focused on the development of a conceptual and
theoretical approach to evaluate coarse aggregate structure based on gradation. The main
purpose was to develop an approach to analyze mixture gradation to determine whether
the coarse aggregate will interlock sufficiently to provide necessary resistance to
deformation and fracture (i.e., the condition commonly referred to as stone-on-stone
contact). Detailed objectives may be summarized as follows:
•
Develop a numerical approach to describe the aggregate structural characteristics
based on gradation.
•
Identify and develop an approach to determine the range of interactive coarse
aggregate particles for a specified gradation (i.e., the particle size or sizes that make
up the primary structure or "skeleton" of the mixture).
•
Identify a criterion to assess whether the range of interactive coarse aggregate
particles are sufficiently dense within the asphalt mixture to actually be in contact
and provide the interlock necessary to resist deformation and fracture.
•
Evaluate the approach and the criterion developed using mixtures of known
performance.
•
Evaluate and confirm the detailed criterion by laboratory tests.
1.3 Scope
The approach developed in this study was based on packing theory of spherical
particles of multiple sizes. Consequently, the criteria developed are probably most
applicable to aggregates that are not excessively elongated or cubicle in shape. However,
3
the authors see no reason why it would not be possible to extend the concepts and
theoretical calculations developed to particles that are not spherical. It is also recognized
that aggregate angularity and texture can affect the quality of aggregate interlock and
these factors were not dealt with in this study. However, the concepts and criteria
developed should be valid for aggregate of any angularity and texture. In other words,
gradations that result in better interlock are beneficial regardless of the aggregate
angularity or texture. That being said, further research and evaluation in the future may
allow for modified criteria based on measurable characterization of angularity and texture.
CHAPTER 2
LITERATURE REVIEW
2.1 Shear Resistance and Rutting Potential
Roque et al. (1997) found that the gradation characteristics of the coarse aggregate
fraction had the strongest effect on mixture shear resistance for the mixtures evaluated.
Eighteen mixtures were prepared with different coarse aggregate (> 2.0 mm) gradations
ranging from Stone Matrix Asphalt (SMA) to those corresponding to the maximum
density line. They found that asphalt mixture shear resistance appeared to be most
strongly related to the gradation characteristics of the coarse aggregate fraction (> 2.0
mm) of the mixtures. Coarseness of the aggregate and the shape (curvature) and position
of the coarse aggregate gradation curve relative to the maximum density line were all
found to influence mixture shear resistance. In addition, aggregate voids in mineral
aggregate (VMA), which is a function of the denseness of the aggregate structure, was
not found to be related to mixture shear resistance.
2.2 Criteria Associated with VMA and Restricted Zone
The SuperpaveTM specifications have certain guidelines for gradations through the
use of control points and restricted zone on a 0.45 power gradation chart. Control points
limit the percent of material retained or passing some selected sieve sizes depending on
the nominal maximum aggregate size (NMAS) to help ensure continuous gradations,
whereas a restricted zone was proposed to prevent the production of tender mixes.
Kandhal et al. (2001) showed that potentially good mixes have been rejected because
their gradations pass through the restricted zone. Chowdhury et al. (2001) found that
4
5
there is no relationship between the restricted zone and permanent deformation when
crushed aggregates are used in the mixture design. Kandhal and Cooley (2002) found
that there was no significant difference between coarse and fine-graded mixture based on
limited tests.
Nukunya et al. (2001) suggested that the effective film thickness was more useful
than VMA, and showed that VMA reqirements based on NMAS does not account for the
effect of mixture gradation, and is therefore insufficient to correctly differentiate goodperforming mixtures from bad-performing ones. Kandhal et al. (1998) also
recommended using a minimum average film thickness instead of the minimum VMA
requirement to ensure mixture durability. Coree and Hislop (2000) found that the
specified VMA values provided by Superpave did not appear to be adequate for
identifying mixture performance. They suggested the volume percentage of effective
binder, Vbe, was relatively insensitive to the level of compaction and appeared to be a
critical parameter.
2.3 Gradation Parameters: n and a
Birgisson and Ruth (2001) developed Power law parameters (n, a) to evaluate and
classify gradation curves according to performance. Gradations were initially analyzed
using power law regression that characterized the coarse aggregate gradation (retained on
the 4.75 mm) and the fine aggregate gradation (from the 2.36 mm down to 0.15 mm),
according to the following power relationship:
P = a (d ) n
where,
P = percent passing
d = sieve size opening, mm
(2-1)
6
a = constant
n = exponent
The key characteristics that tend to define the desired gradations for coarse or finegraded mixtures are primarily a continuous, well-balanced, coarse aggregate gradation
from the 1.18, 2.36, or 4.75 mm sizes, a reasonable reduction or increase in the amount of
fine aggregate, and mineral filler content less than 6 %.
The study of these parameters was expanded by Ruth et al. (2002). The results
presented the concepts and guidelines for the selection of coarse or fine-graded aggregate
blends using gradation characterization factors based on power law constants (aCA, aFA)
and exponents (nCA, nFA).
2.4 Bailey Method
Typically the selection of aggregates gradation is made based on specification
bands (coarse, medium, or fine gradation), but the main question is how to choose the
best possible blend to achieve good workability and field performance. The Bailey
method is a more systematic way to find a starting point (Vavrik et al., 2001, 2002, and
Asphalt Institute and the Heritage Group, 2005).
The Bailey method was developed by Bob Bailey in the early 1980s; the main
purpose of this approach is to control the mix properties--volumetric properties,
workability, segregation, and compactibility--during construction.
The focus of the Bailey method is aggregate packing based on Voids in the Mineral
Aggregate (VMA). The method determines coarse fraction as those particles that create
voids and fine fraction as those particles that fit into the voids crated by coarse
aggregates.
7
The Bailey method also defines three types of mixes (coarse, SMA, or fine) based
on the volume of the coarse fraction, as shown in Figure 2-1.
< LUW
LUW
Fine-Graded
< 90%
Coarse-Graded
95~105%
RUW
SMA
110~125%
Figure 2-1. Determination of Mix Type
There are four main principles to the Bailey method.
•
•
•
•
Principle No. 1:
Principle No. 2:
Principle No. 3:
Principle No. 4:
Definition of coarse fraction and fine fraction.
Coarse fraction analysis.
Coarse part of fine fraction evaluation.
Fine part of fine fraction analysis.
These four principles are related not only to compactibility and segregation
susceptibility of the mix in the field but also to the expected change in VMA or voids
from one design trial to the next, or from one QC sample to the next. Figures 2-2 and 2-3
shows how to determine four principal sieve sizes.
The Bailey method utilizes the Nominal Maximum Particle Size (NMPS) to
estimate the void size within the coarse fraction. The break between coarse and fine
fractions is defined as the Primary Control Sieve (PCS) which is estimated as the closest
sieve to the result of 0.22×NMPS.
8
Coarse
Fraction
Fine
Fraction
Half Sieve = 0.5 x NMPS
2
CA Ratio
PCS = 0.22 x NMPS
1
% CA LUW
SCS = 0.22 x PCS
TCS = 0.22 x SCS
3
FAc Ratio
4
FAf Ratio
Figure 2-2. Four Main Principles of Bailey Method for Coarse Mixtures
The calculation of the Coarse Aggregate ratio (CA), Fine Coarse Aggregate ratio
(FAc), and Fine fine aggregate ratio (FAf) can be made by using the following equations:
CA Ratio =
% passing half sieve − %passing PCS
100 − % passing half sieve
(2-2)
FA c Ratio =
% passing SCS
%passing PCS
(2-3)
FA c Ratio =
% passing TCS
%passing SCS
(2-4)
The use of the four principles and admissible values for the different ratios depend
upon the type of gradation (coarse, fine or SMA). Table 2-1 shows the recommended
values of the different ratios for coarse mixes.
9
100
90
80
70
% Passing
60
Sieve
A
B
C
D
E
F
G
H
I
J
K
% Passing
100
97
76
63
39
25
17
11
7
5
4.2
2
1
50
40
30
20
3
10
Fine fraction
Coarse fraction
0
K J
I
H
G
F
E
D
C
B
A
Sieve Size (mm) ^ 0.45
Figure 2-3. Example of Coarse Gradation Mix
Table 2-1. Recommended Aggregate Ratios for Coarse Mixtures
NMPS
37.5mm
25.0mm
19.0mm
12.5mm
9.5mm
4.75mm
CA ratio
0.80-0.95
0.70-0.85
0.60-0.75
0.50-0.65
0.40-0.55
0.30-0.45
FAc ratio
0.35-0.50
FAf ratio
0.35-0.50
In conclusion, the Bailey method is a pretty good tool for evaluating volumetrics
and compactibility of the mix, but further research is required to find the optimum
aggregate gradation based on mixture performance, for example, rutting, fatigue
cracking, and thermal cracking resistance.
10
2.5 Summary
Some recent studies have focused on evaluating the effects of aggregate
characteristics and structure to determine which gradations are most resistant to cracking
and rutting in Superpave mixtures.
The Bailey method of mix design provided a better understanding of relationships
between aggregate gradation and mixture voids, and offers a means to design and analyze
the aggregate structure in an asphalt mixture. The method defined gradation parameters
(CA, FAc, FAf ratios) that were related to air voids and VMA. In addition, the design
approach attempts to achieve a suitable coarse aggregate structure by requiring the
density of the coarse aggregate in the compacted mixture to be between 95% and 105%
of the loose density of the coarse aggregate as determined in the laboratory.
The developers of the Bailey method clearly recognized the need to have large
enough particles in contact with each other for suitable mixture performance. However,
achieving a specified coarse aggregate density may not necessarily ensure a suitable
coarse aggregate structure. For example, the coarse aggregates may be proportioned in
such a way that the range of different sized particles is not in continuous contact. Finer
coarse aggregate particles may simply be filling voids between relatively few coarser
aggregate particles, or coarser aggregate particles may just be floating in a matrix of finer
coarse aggregate particles. In either case, particles within the coarse aggregate range may
be acting independently of each other and not providing a suitable network for resistance
to deformation and fracture.
Therefore, it would be useful to have a system to determine whether different size
coarse aggregate particles from a specified gradation are proportioned properly so that
they can result in an interactive network of particles in continuous contact. In addition, it
11
would also be of benefit to have a criterion to assess whether the range of interactive
coarse aggregate particles are sufficiently dense within the asphalt mixture to actually be
in contact and provide the interlock necessary to resist deformation and fracture. It
would be particularly beneficial if the criterion did not require laboratory testing.
CHAPTER 3
NEW DEVELOPMENT
3.1 Porosity as a Criterion for Interlocking
Porosity has been used extensively in fields like soil mechanics as a dimensionless
parameter that describes the relative proportion of voids to total volume. In soil
mechanics, a typical element of soil contains three distinct phases: solid (mineral
particles), gas, and liquid (usually water). Figure 3-1 is a phase diagram illustrating the
Wg
≈0
Vs
Solid
Volumes
Ws
V
W
Liquid
Ww
Vg
Gas
Vw
Vv
three phases separately. Porosity (n) is the ratio of void volume (VV) to total volume (V).
Weights
Figure 3-1. Relationship among Soil Phases
Porosity, n =
VVoid VV
=
VTotal
V
12
(3-1)
13
It is a well-known fact in soil mechanics that the porosity of granular materials in
the loose state is approximately constant between 45% and 50%, regardless of particle
size or distribution (Lambe and Whitman, 1969, and Freeze and Cherry, 1979). This
implies that the porosity of an assemblage of granular particles (e.g., the aggregate within
an asphalt mixture) must be no greater than 50% for the particles to be in contact with
each other. This also implies that one can use porosity as a criterion to assure contact
between large enough particles within the mixture to provide suitable resistance to
deformation and fracture. As mentioned earlier the Bailey Method of mix design takes a
very similar approach by requiring the density of the coarse aggregate in the compacted
mixture to be between 95% and 105% of the loose density of the coarse aggregate as
determined in the laboratory. Use of a porosity criterion would preclude the need for
laboratory compaction of coarse aggregate.
Therefore, a maximum porosity of 50% was selected as a starting point for
evaluation as a criterion for asphalt mixture, which is essentially a granular material with
asphalt and fines between the granular particles. The basic principles associated with the
calculation of porosity of different components within the asphalt mixture are presented
below.
3.2 Application to Asphalt Mixture
VMA in asphalt mixtures, which is the volume of available space between
aggregates in a compacted mixture, is analogous to void volume in soil.
VMA = V − V AGG
(3-2)
By assuming that a mixture has a certain effective asphalt content and air voids for
a given gradation (i.e., VMA), porosity can be calculated for each aggregate particle size.
14
For example, the porosity of particles retained on the 4.75mm sieve and passing the
9.5mm sieve is calculated by subtracting the volume of larger aggregates (i.e., those
retained on the 9.5mm sieve) from the total volume of mixture (V) as shown in Figure 32..
VT( 4 .75−9.5 ) = VTM − V AGG( ≥9.5 )
(3-3)
where,
VT(4.75-9.5)
VTM
VAGG(≥9.5)
= Total volume available for particles retained on the 4.75mm
sieve and passing the 9.5mm sieve
= Total volume of mixture
= Volume of particles retained on the 9.5mm sieve
Figure 3-2. Mixture Components for Porosity Calculation
The volume of voids within VT(4.75-9.5) includes the volume of aggregates passing
the 4.75mm sieve, in addition to the volume of effective asphalt plus the volume of air
(i.e., the VMA of the mixture).
VV ( 4.75−9.5) = V AGG ( <4.75) + VMA
where,
VV(4.75-9.5)
= Volume of voids within VT(4.75-9.5)
(3-4)
15
VAGG(<4.75)
= Volume of particles passing the 4.75mm sieve
The porosity of this aggregate particle size is then calculated as follows.
n( 4.75−9.5) =
VV ( 4.75−9.5)
VT ( 4.75−9.5)
=
V AGG ( < 4.75) + VMA
VTM − V AGG ( ≥9.5)
⎛ VTM − V AGG ( ≥ 4.75)
=⎜
⎜ V −V
AGG ( ≥9.5 )
⎝ TM
⎞
⎟
⎟
⎠
(3-5)
where,
VAGG(≥4.75)
= Volume of particles retained on the 4.75mm sieve
Similar calculations can be performed for any other particle size or range of particle
sizes within the mixture.
3.3 Porosity of Individual Particle Sizes
Some typical mixture gradations are shown in Figure 3-3, which includes coarsegraded, fine-graded, and SMA mixtures. Porosity analysis was applied to check the
coarse aggregate structure in these mixtures. Figure 3-4 shows the porosity of each
individual particle size for the three gradations presented in Figure 3-3. As shown in the
figure, the only single aggregate size with porosity less than 50% was the aggregate
retained on the 9.5 mm sieve for the SMA mixture. This finding was expected, since
SMA mixtures are designed specifically to achieve stone-on-stone contact with a singlesize aggregate. The finding also seems to indicate that the 50% porosity criterion is
reasonable.
None of the individual particle sizes met the 50% porosity criterion for either the
coarse-graded or the fine-graded mixtures. However, both of these dense-graded
Superpave mixtures are known to have good resistance to deformation and fracture, so it
is not logical that the coarse aggregate in these mixtures exists in a state where the
particles are not in contact with each other as reflected by the porosity being much
greater than 50%. Therefore, it seems clear that there must be a range of contiguous
100
90
80
% passing
70
60
MDL
50
C1
40
F1
SMA
16
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
Sieve size, ^0.45
Figure 3-3. Example Gradations
⅜" ½"
¾"
110
100
90
Porosity, %
80
70
C1
60
F1
50
SMA
40
MDL
Limit
17
30
20
10
Sieve size, mm
Figure 3-4. Individual Porosity Results
0
0.075
0.15
0.3
0.6
1.18
2.36
4.75
9.5
12.5
19
0
18
coarse aggregate particle sizes that form a network of interactive particles with a porosity
of less than 50%. The challenge was to develop an approach to objectively determine
what specific particle sizes, if any, are interacting such that they should be considered to
be a single unit in the determination of porosity. Here again it is important to emphasize
that the 50% porosity criterion is independent of particle size or distribution, so it is
equally applicable to a range of interactive particle sizes as to single size particles.
A theoretical analysis procedure was developed to determine whether a given
proportion of contiguous particle sizes are interacting to form a continuous network. The
development and results of the analysis are presented in the following section.
3.4 Theoretical Developments
Several important concepts were employed in the theoretical development of a
system to determine whether different size particles are interacting in space. Perhaps the
most important one involves the physical model used to describe an asphalt mixture,
which can be viewed as being composed of the following elements:
3.4.1 Dominant Aggregate Size Range (DASR)
For all asphalt mixtures, this is the interactive range of particle sizes that forms the
primary structural network of aggregates. It was hypothesized that the DASR must be
composed of coarse enough particles and its porosity must be no greater than 50% for a
mixture to effectively resist deformation and cracking. Particle sizes smaller than the
DASR will serve to fill the void space between the DASR (the interstitial volume
described below) along with binder and fines. Particles larger than the DASR will simply
float in the DASR matrix and will not play a major role in the aggregate structure. These
concepts are illustrated in Figure 3-5, which shows the DASR for three different types of
mixtures.
Dominant
Aggregate
IC, IV
19
(a) SMA
Figure 3-5. Dominant Aggregates and Interstitial Volume
(b) Coarse dense
(c) Fine dense
20
3.4.2 Interstitial Volume (IV)
This is the volume of material (asphalt, aggregate and air voids) that exists within the
interstices of the DASR. The components within this volume are referred to as the
interstitial Components (IC). This volume serves to hold together the DASR, and its
characteristics, as well as the properties of the IC will strongly influence the durability
and fracture resistance of mixtures. Excessively low stiffness and/or excessive IV can
lead to excessive creep rate, which is related to rate of damage development. Conversely,
excessively high stiffness and/or insufficient IV can make a mixture brittle and have low
dissipated creep strain energy to failure (DCSEf), which defines a mixture’s tolerance to
damage.
3.4.3 Interstitial Surface (IS)
This surface is defined by an approximately straight plane taken through the
interstitial volume. It can be most easily visualized as a failure surface of an asphalt
mixture pulled apart in tension, as shown in Figure 3-6. The characteristics of this
surface, including its roughness, protrusion of different size aggregate particles, and
presence of asphalt and fines, will strongly influence the mixture's resistance to
deformation and fracture, and particularly shear deformation associated with rutting.
Rougher interstitial surfaces with larger particle protrusions will result in mixtures with
greater shear resistance. Shear resistance will be further enhanced if particles on this
surface are arranged in such a way as to form an interlocking network of particles.
Therefore, determination of the characteristics of this interstitial surface, which are
controlled by gradation, should provide useful parameters for mixture evaluation and
design.
21
Figure 3-6. The Failure Surface from a Broken IDT Sample
22
For example, one can determine whether or not particles are interacting with each
other by determining their center-to-center spacing on the interstitial surface. A
theoretical procedure was developed to determine this spacing for specified gradations
and thus determine which particles within the gradation interact to form the DASR. The
development and results of this procedure are presented below.
3.5 Particle Spacing on the IS
For a given particle size distribution (gradation) compacted to a specified density,
one can easily calculate the number of particles of any given size that will be present
within a specified representative volume. Furthermore, one also can calculate how many
particles of each size will be present within a representative cross-sectional area (i.e., the
interstitial surface) taken through the representative volume. The spacing between each
particle size on the IS can also be calculated if certain characteristics regarding the
distribution between the different particle sizes in the area are known or assumed. For
asphalt mixtures it is reasonable to assume that particles are generally uniformly
distributed within the representative volume or area. In addition, if the mixture is not
segregated, the largest particles will be uniformly distributed over the entire volume or
area, while smaller particles will be uniformly distributed within the remaining volume or
area (i.e., the volume or area between the larger aggregate particles). In other words,
smaller particles are uniformly distributed locally but not globally over the entire volume
or area. These were the basic assumptions made in making the theoretical spacing
calculations.
3.6 Determination of DASR
As explained earlier, the DASR may be composed of one size or multiple sizes.
Particle sizes interacting with each other to form the primary network that carries load
23
induced stresses have to be determined. Open-graded or uniform gradations such as
SMA have a very distinct DASR, because only one size aggregate makes up most of the
mixture volume. However, determination of the DASR is less clear for dense-graded
mixtures. Therefore, a system is needed to determine which contiguous sizes are
interacting as a unit to make up the DASR. To do this, an interaction diagram was
developed based on the spacing analysis between particles on the interstitial surface.
3.7 Spacing Analysis and Interaction Diagram
As mentioned above, spacing between particles for each size in the representative
volume can be calculated to check whether there is interaction between contiguous sizes
for specified gradations. The spacing calculations assumed that particles are distributed
according to a hexagonal pattern within the available area, which results in a uniform
particle distribution. The center-to-center spacing among the same sizes of particles was
calculated in order, from the biggest size to smallest size in order to account for the fact
that smaller particles are only locally uniformly distributed between the larger aggregate
particles. At first, the biggest particles are distributed within the total representative area,
then the next smallest particles are distributed with the same pattern within the available
area, which is the remaining area after subtracting the area taken up by the biggest
particles from the total representative area. The 3-D distribution of particles in space can
be determined and defined on a plane for a given density, VMA, etc. The number
particles (n’) of each size that intersect a plane in space is determined as follows:
n′ = total no. of spheres ×
diameter of the sphere
height of the representative volume
24
∴ n′ =
nD 2nr
=
h
h
(3-6)
where,
n = the number of particles of each size in the total volume
h = height of the representative volume
D
= diameter of particles
r = radius of particles
Figure 3-7 shows the patterns used to perform the spacing calculations for each
size. The triangular number (nt) is the number of hexagonal layers required to
accommodate the number of particles (n’), which is equal to 3 in Figure 3-7. The
centered hexagonal number (Hn) is the number of particles that we want to distribute
(note: Hn equals n’ for our problem), which is equal to 37 in Figure 3-7. The general
equation to calculate Hn is:
2
H n = 3nt + 3nt + 1
(3-7)
The spacing between particles is “s” and “a” is the side length of the outermost
hexagon that encompasses the total available area for a specified particle size. Therefore,
the area of the hexagon (Ah) is determined as follows:
Ah =
3
3
3 a2 =
3 ( nt × s ) 2
2
2
(3-8)
However, as shown in Figure 3-7, the total area occupied by the n’ particles, is
greater than Ah, which is the area inside the hexagon associated with the centerline of the
outermost particles. The modified area (A’h), which is the area of the dotted line hexagon
in Figure 3-8, is calculated as follows:
25
nt = 3
nt = 2
nt = 1
s
a
Figure 3-7. Hexagonal Pattern Distribution and Spacing Calculation for Each Size
Ah′ =
3
s
3
s
3 (a + ) 2 =
3 ( nt × s + ) 2
2
2
2
2
∴ Ah′ =
3
3 ((nt + 0.5) s) 2
2
(3-9)
The triangular number (nt) can be determined by rearranging Equation 3-7.
nt =
4H n − 1 1
− =
12
2
4n ′ − 1 1
−
12
2
(3-10)
One can solve for the particle spacing (s) by rearranging Equation 3-9.
s=(
2 Ah′
2
)
2 nt + 1 3 3
(3-11)
Therefore, the spacing within the hexagonal pattern is easily determined if the
number of particles (n’) and the total area (A’h) are known. This procedure was repeated
for all particle sizes within the gradation.
26
nt = 2
nt = 1
s
a
a+s/2
Figure 3-8. Modified Hexagonal Area from Outer Solid to Dotted Line (nt=2)
Figure 3-9 shows the basic principles employed in these calculations. The spacing
among the biggest particles within the total area is calculated with the hexagonal pattern
distribution. If the biggest particles take 20% of the total area, the remaining area, 80%,
will be the representative total area for the next size. The next smallest particles were
distributed with the same pattern within this remaining available area.
Figure 3-10 shows results of spacings calculated for a binary mixture with 9.5 and
4.75 mm size particles. As the proportion of larger/smaller particles decreases, the larger
particle spacing increases. In other words, as the number of larger particles decreases,
their spacing increases. The smaller particles (4.75 mm) obviously show a reverse trend.
Figure 3-10 shows that for each size particle the spacing starts to increase dramatically
once the relative proportions of different sized aggregates reaches a certain level. In
order to more precisely determine the relative proportion at which the particle spacing
27
(a) The biggest particles distribution
(b) The 2nd size particles distribution
solid particles = **20%
shaded rest area = **80%
*
*, representative area for the next step
**, percentage of the initial total area
Figure 3-9. The Representative Areas Based on Hexagonal Patterns for Each Step
solid particles = 30% x 80% = **24%
*
rest area
= 70% x 80% = **56%
4
Spacing, cm
3
2
Large
Small
1
Large/Small Particle Proportion
Figure 3-10. Spacing Result for the Binary Mixture with 9.5, 4.75mm
0/100
5/95
10-90
15/85
20/80
30/70
40/60
50/50
60/40
70/30
80/20
85/15
90/10
95/5
100/0
28
0
29
starts to change rapidly, the rate of change of the slope of the spacing diagram presented
in Figure 3-10 was plotted in Figure 3-11. These results indicate that the particle spacing
for either particle size begins to increase more rapidly once the relative proportion of the
different size aggregate is about 70/30. It should be noted that this result would be the
same for any two particle sizes having a size ratio of 2:1, which is generally the size ratio
used between contiguous size sieves in asphalt mixture design.
This finding implies that one particle size will significantly disrupt the ability of
another particle size to interact once the relative proportions of the particle sizes is about
70/30. In other words, once the proportions exceed this value, the spacing of the particles
with the smaller proportion increases so much that these particles are simply floating in
the matrix and are no longer an effective part of the aggregate structure. That is, the
particles are not part of the DASR. Conversely, at proportions less than 70/30 (e.g.,
40/60, 50/50, 60/40), as shown in Figures 3-9 and 3-10, each particle size maintains a
fairly stable spacing, so both are part of the DASR. All contiguous particle sizes
determined to be interactive are considered part of the DASR, and are considered to act
as a unit for determination of porosity.
3.8 Interaction Diagrams and DASR Porosity
For any given gradation, the criteria described above can be used to determine
which contiguous sizes are interacting. One simply needs to determine the relative
proportion of the contiguous sizes and determine whether or not it is less than 70/30.
Figure 3-12 presents an interaction diagram, showing the relative proportion of all
contiguous sizes for the three gradations presented in Figure 3-3. For purposes of
illustration, the interaction diagram is shown for all aggregate sizes. However, only the
interaction and porosity of the coarser aggregate is relevant for this evaluation, which is
0.25
Slope
0.20
0.15
Large
Small
0.10
30
0.05
0.00
0
10
20
30
40
50
60
70
% passing for sections
Figure 3-11. Slope (spacing change) for the Binary Mixture
80
90
100
Large/Small Particle Proportion
100/0
90/10
IC
80/20
70/30
60/40
C1
50/50
F1
40/60
SMA
30/70
Limit
20/80
10/90
31
Contiguous sieve sizes, mm
Figure 3-12. Interaction Diagram
0.075~0
0.15~0.075
0.3~0.15
0.6~0.3
1.18~0.6
2.36~1.18
4.75~2.36
9.5~4.75
12.5~9.5
0/100
32
intended to determine whether the range of interactive coarse aggregate particles are
sufficiently dense within the asphalt mixture to actually be in contact and provide the
interlock necessary to resist deformation and fracture. For this purpose, the particle size
passing the 2.36 mm sieve, but retained on the 1.18 mm sieve, was selected as the
smallest particle coarse enough to contribute to aggregate interlocking. This selection
was based on existing definitions of coarse and fine aggregates for asphalt mixture, which
generally separate coarse and fine aggregate at the 2.36 mm sieve, and knowledge of soil
mechanics indicating that particles finer than this have little internal friction. The Bailey
method defined coarse aggregate as particles large enough to create voids of a certain size
when placed in a unit volume. The primary control sieve (PCS) separates coarse and fine
aggregate in the Bailey method. For a nominal maximum particle size (NMPS) of
12.5mm, the PCS is 2.36mm, based on a packing factor of 0.22. However, the packing
factor can vary from 0.18~0.28 in the Bailey method. Therefore, selection of particles
passing the 2.36 mm and retained on the 1.18 mm sieve for the intended purpose is also
consistent with the Bailey approach.
As shown in Figure 3-12, in the coarse aggregate range, both the SMA and the
coarse-graded mixture exhibit interaction between the 4.75/2.36 mm sizes and the
2.36/1.18 mm sizes. The fine-graded mixture exhibited interaction at three levels:
9.5/4.75 mm, 4.75/2.36 mm, and 2.36/1.18 mm. Therefore, the interaction diagram
indicates that several potential DASR ranges need to be checked for these mixtures. The
actual DASR of each mixture is the set of interactive (or single) particles that result in the
lowest porosity for the mixture.
33
It is interesting to note that all contiguous particle sizes exhibit strong interaction
for the gradation associated with the maximum density line (MDL). This result, of course,
was anticipated, and lends credence to the interaction criterion established on the basis of
spacing.
For example, the SMA has three potential DASRs: the aggregate retained on the
12.5 mm sieve, the aggregate passing the 12.5 mm sieve but retained on the 9.5 mm sieve,
and the aggregates passing the 4.75 mm sieve and retained on the 1.18 mm sieve, which
includes two interactive sizes. Because of the large amount of material retained on 9.5
mm sieve, this single aggregate size was the DASR for the SMA, even though two other
sizes are interactive. As shown in Figure 3-13, the resulting DASR porosity for the SMA
mixture was 42%, whether or not interaction was considered.
100
90
80
Porosity, %
70
60
50
40
30
20
10
0
Individual
Interaction
Coarse
65
36
Fine
74
46
SMA
42
42
Figure 3-13. Porosity Result after Considering Interaction
34
For both the coarse-graded and fine-graded mixtures, the interactive aggregate was
the DASR, and as shown in Figure 3-13, the interaction made a dramatic difference in the
determination of DASR porosity. Whereas the lowest porosity for individual coarse
aggregate particles (i.e., no interaction) was 65% for the coarse-graded mixture and 74%
for the fine-graded mixture, the resulting DASR porosities were 36% and 46%,
respectively, once interaction was considered. Both mixtures met the proposed porosity
criterion of 50%, which indicates that these gradations will result in good resistance to
deformation and fracture. As indicated earlier, these mixtures are both known to be good
performers in the state of Florida.
CHAPTER 4
EVALUATION AND REFINEMENT
4.1 Introduction
Mixtures for which gradation has been well determined and documented, and for
which rutting performance has been determined either from field measurements,
laboratory rut tests, test track measurements, or measurements from accelerated pavement
testing facilities (APT’s) were used to evaluate the gradation evaluation system
developed and presented in chapter 3. Five excellent sources of data were identified and
obtained for this purpose:
•
Field rutting performance measurements from the first eight projects associated
with the comprehensive Superpave monitoring project being conducted by FDOT.
•
Laboratory rutting performance determined from asphalt pavement analyzer (APA)
and Servopac results on plant mixtures obtained from Projects 8 through 12 of the
Superpave monitoring project (reliable field rut measurements were not yet
available for these recently placed sections).
•
Rutting performance of mixtures placed and tested at FHWA’s WesTrack road test
facility in Nevada.
•
Rutting performance of mixtures placed and tested at NCAT’s test track in
Alabama.
For each data set, the gradation of each mixture evaluated was analyzed using the
approach developed. Interaction diagrams were developed from the gradation data to
identify the dominant aggregate size range (DASR) and the porosity of the DASR.
Mixtures were separated into one of the following three groups based on the interaction
diagram characteristics and porosity of DASR:
35
36
•
Group I: mixtures with DASR porosity less than 50% and having a clearly
interactive DASR range. These mixtures were expected to perform well.
•
Group II: mixtures with DASR porosity greater than 50%. These mixtures were
expected to exhibit greater rutting than those with porosity less than 50%.
•
Group III: mixtures with marginal interaction between aggregate sizes in the
DASR (i.e., the relative proportion of larger to smaller aggregate sizes was very
close to 70/30), and with DASR porosity less than 50% if interaction was
considered, but greater than 50% if interaction was not considered. These mixtures
were expected to exhibit marginal to poor performance and sensitivity to changes
in asphalt content or gradation.
The rutting performance of each group was determined and compared to evaluate
whether or not these criteria distinguished between mixtures exhibiting different rutting
performance. Results of the evaluations are presented in the following sections.
4.2 Field Performance: Superpave Monitoring Projects 1 to 8
A comprehensive monitoring project was initiated by FDOT with the intention of
studying construction and performance data of Superpave mixtures in the state of Florida
to establish appropriate and realistic performance-based specifications. Twelve projects
from throughout the state of Florida constructed with Superpave mixtures were
monitored during and after construction. Extensive sampling was done by taking field
cores from projects already constructed (Projects 1 to 7), and plant mix and field cores for
Projects 8 to 12. Field performance has been continually monitored and an extensive
laboratory testing program has been conducted on both field cores and plant mixtures
obtained from the projects. Projects 1 to 8 have been subjected to over three years of
traffic now, so valuable field rutting performance data is available for evaluation.
Interaction diagrams for mixture gradations associated with these projects are
presented in Figures 4-1 to 4-10. It is emphasized that these gradations are in-place
gradations as determined from field cores, and not simply job-mix-formula gradations
37
which may or may not be representative of the final result in the field. Resulting DASR
porosity of each project mixture is presented in Figure 4-11, which indicates that four
mixtures were in Group I (DASR porosity < 50%), two mixtures were in Group II
(DASR porosity > 50%), and two mixtures were in Group III (marginal interaction).
Note that two DASR porosity values are presented for Projects 1 and 2, which had the
mixtures determined to have marginal interaction within the DASR range. This is
evident in Figure 4-1, which shows that the relative proportion of the 4.75/2.36 mm and
the 2.36/1.18 mm aggregate sizes was right at 70/30. As indicated in Figure 4-11, the
Large/Small Particle Proportion
Project 1
Project 2
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-1. Interaction Diagram for Field Mixtures of Project 1 and 2
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
38
JMF
100/0
Group 1
Group 2
Group 3
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-2. Interaction Diagram for Field Mixtures of Project 3 Layer A
JMF
100/0
Group 1
Group 2
Group 3
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-3. Interaction Diagram for Field Mixtures of Project 3 Layer B
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
39
JMF
Large/Small Particle Proportion
100/0
Group 1
Group 2
Group 3
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-4. Interaction Diagram for Field Mixtures of Project 4 Layer A
Group 3
Group 2
Group 1
JMF
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous sizes, mm
Figure 4-5. Interaction Diagram for Field Mixtures of Project 4 Layer B
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
40
JMF
Group 1
Group 2
Group 3
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-6. Interaction Diagram for Field Mixtures of Project 5 Layer A
JMF
Group 1
Group 2
Group 3
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous sizes, mm
Figure 4-7. Interaction Diagram for Field Mixtures of Project 5 Layer B
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
41
JMF
Group 1
Group 2
Group 3
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Figure 4-8. Interaction Diagram for Field Mixtures of Project 6
JMF
Group 1
Group 2
Group 3
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-9. Interaction Diagram for Field Mixtures of Project 7
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
42
8-1A
8-2A
8-3A
8-1B
8-2B
8-3B
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
0/100
9.5-4.75
10/90
12.5-9.5
Large/Small Particle Proportion
100/0
Contiguous Sizes, mm
Figure 4-10. Interaction Diagram for Field Mixtures of Project 8
80
without interaction
70
Porosity, %
60
50
40
30
20
1
2
3
4
Project
Figure 4-11. Porosity Result for Field Mixtures
5
6
7
8
43
DASR porosity of both mixtures is less than 50% if these sizes are considered interactive,
but significantly greater than 50% if these sizes are not interactive.
Field rut depths obtained from transverse profilograph measurements on each
project are presented in Figure 4-12. The results are presented in terms of rut
depth/ESAL’s (mm/ESAL*106) in order to normalize the effect of traffic between the
different sections. Two sets of rut depth measurements are presented (Round I and
Round II), which refer to measurements obtained at two different times after construction.
Round I measurements were obtained approximately 1~2 years after construction, while
round II measurements were obtained about one year later.
A cursory evaluation of Figure 4-12 indicates that Projects 3, 4, 5, and 7 exhibited
the best rutting performance, while projects 1, 2, 6, and 8 had relatively higher rutting.
As shown in Figure 4-11, Projects 3, 4, 5, and 7 were the four projects in Group I with
DASR porosity less than 50%, while Projects 6 and 8 were in Group II (DASR porosity >
50%) and Projects 1 and 2 were in Group III (marginal interaction).
Figure 4-13 and 4-14 presents the average rut depth/ESAL for the three groups of
mixtures, for Round I and II, respectively. These figures clearly indicate that mixtures
with DASR porosity < 50% exhibited significantly lower field rutting performance than
mixtures with DASR porosity > 50% or mixtures with marginally interactive aggregates.
The minimum and maximum rut depth/ESAL for each group is also shown in Figures 413 and 4-14, which show that all mixtures within each group exhibited similar
performance.
The results of these evaluations indicate the following:
44
7.0
Round I
Round II
4
5
Rut Depth / ESALs (mm/Million)
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1
2
3
6
7
8
Projects
Figure 4-12. Rut Depth/ESALs from Field Measurement
Rut Depth / ESALs, mm/Million
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
3
4
5
η DASR < 50%
7
6
8
η DASR > 50%
1
2
Marginal
Interaction
Project
Figure 4-13. Average Rut Depth/ESALs for Different Porosity Groups (Round I)
45
Rut Depth / ESALs, mm/Million
5.0
4.0
3.0
2.0
1.0
0.0
3
4
5
η DASR < 50%
7
6
8
η DASR > 50%
1
2
Marginal
Interaction
Project
Figure 4-14. Average Rut Depth/ESALs for Different Porosity Groups (Round II)
•
The DASR porosity criterion of 50% based on the gradation evaluation system
developed as part of this research effort appears to accurately distinguish between
the relative rutting performance of Superpave mixtures in the field. Mixtures
meeting the porosity criterion exhibited less rutting than mixture that did not.
•
The interaction criterion of 70/30 for the relative proportion of contiguous
aggregate sizes within the DASR range appears to distinguish well between coarse
aggregate structures that interact properly and those that do not. Marginal
interaction as determined according to this criterion resulting mixtures with higher
rutting than mixtures with gradations that were not marginal.
46
4.3 Laboratory Performance: Superpave Monitoring Projects 8 to 12
These projects have been monitored from the time of construction to the present.
Consequently, and in contrast to projects 1 to 7 that had already been constructed at the
time the Superpave monitoring project began, it was possible to obtain samples of plant
mixture for laboratory testing. These samples were used to perform rut tests with the
asphalt pavement analyzer and the Servopac Gyratory compactor. Unfortunately, these
test sections were recently constructed and have not been subjected to enough traffic in
the field, so reliable measurements of field rutting were not yet available for evaluation.
Interaction diagrams for mixture gradations associated with these projects are
presented in Figures 4-15 to 4-19. It is emphasized that these gradations were determined
from the same plant mix samples that were used to perform the laboratory tests reported
in this section. It should be noted that the gradation of plant mixtures from project 8 was
different than the field gradation because of breakdown that occurred during compaction
in the field. Resulting DASR porosity of each project mixture is presented in Figure 4-20,
which indicates that two mixtures were in Group I (DASR porosity < 50%), two mixtures
were in Group II (DASR porosity > 50%), and one mixtures was in Group III (marginal
interaction). Once again, two DASR porosity values are presented for Project 10, which
had the mixture determined to have marginal interaction within the DASR range.
This is evident in Figure 4-17, which shows that the relative proportion of the
4.75/2.36 mm and the 2.36/1.18 mm aggregate sizes was right at 70/30 (actually slightly
above for the 2.36/1.18 mm sizes). As indicated in Figure 4-20, the DASR porosity of
this mixture is less than 50% if these sizes are considered interactive, but significantly
greater than 50% if the these sizes are not interactive.
47
Large/Small Particle Proportion
8-5
8-4
8-3
8-2
8-1
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-15. Interaction Diagram for Plant Mixtures of Project 8
9-1A
9-2A
9-3A
9-1B
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-16. Interaction Diagram for Plant Mixtures of Project 9
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
48
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-17. Interaction Diagram for Plant Mixtures of Project 10
11-2A
11-2B
11-3B
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-18. Interaction Diagram for Plant Mixtures of Project 11
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
49
12-1B
12-1A
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
9.5-4.75
0/100
4.75-2.36
10/90
12.5-9.5
Large/Small Particle Proportion
100/0
Contiguous Sizes, mm
Figure 4-19. Interaction Diagram for Plant Mixtures of Project 12
100
90
without interaction
Porosity, %
80
70
60
50
40
30
20
8
9
10
Project
Figure 4-20. Porosity Result for Plant Mixtures
11
12
50
Rut depths obtained from the modified APA system, which was developed by the
University of Florida as part of a FDOT research project (Drakos, 2003; Drakos et al.,
2001, 2005), are presented for each of the mixtures in Figure 4-21. The modified system
involved the use of a simulated tire rib for loading, instead of the hose used in the
conventional system. Research showed that the rib induces stresses that are more
representative of an actual radial truck tire. The new system also involved measurements
of rut profiles in addition to the absolute rut depth (ARD) measurement obtained in the
conventional APA system (Figure 4-22). The rut profiles allow for the determination of
differential rut depth (DRD) and change in cross-sectional area of profile, which can be
used to identify the presence of mixture instability (Figure 4-23). Positive area changes
indicate dilation associated with instability, while zero or negative area change indicates
contraction or no volume change, which indicates that no instability has occurred.
Absolute Rut Depth
Percent Area Change
Rut Depth (mm), Area Chage (%)
5.0
4.0
3.0
2.0
1.0
0.0
-1.0
Project 8
Project 9
Project 10
Project 11
Figure 4-21. APA Test Result (Rib) for Plant Mix Gradations
Project 12
51
Figure 4-22. Concepts for DRD and ARD
Figure 4-23. Area Change Interpretation
Results of area change calculations for the mixtures are also presented in Figure 4-21.
A cursory evaluation of Figure 4-21 indicates that Projects 8 and 11 exhibited the
best rutting performance (lowest rut depth and negative area change, indicating no
instability), while Projects 9, 10, and 12 exhibited higher APA rut depths and positive
area change, indicating the presence of instability). As shown in Figure 4-20, Projects 8
52
and 11 were the two projects in Group I with DASR porosity less than 50%, while
Projects 9 and 12 were in Group II (DASR porosity > 50%) and Projects 10 was in Group
III (marginal interaction).
Figure 4-24 and 4-25 present the average APA rut depth and percent area change,
respectively, for the three groups of mixtures. Figure 4-24 clearly indicates that mixtures
with DASR porosity < 50% exhibited significantly lower APA rut depths than mixtures
with DASR porosity > 50% or mixtures with marginally interactive aggregates. Figure
4-25 clearly shows that mixtures with DASR porosity < 50% exhibited negative area
change (no instability), while mixtures with DASR porosity > 50% and mixtures with
marginally interactive aggregates exhibited positive area changes (instability). The
minimum and maximum rut depth and area change values for each group are also shown
in Figures 4-24 and 4-25, which show that all mixtures within each group exhibited
similar performance.
Results of Servopac rutting analysis procedures, which were developed by
Birgisson et al.(2003, 2004), are presented for each of the mixtures in Figure 4-26. The
two parameters obtained from the procedure, which is based on shear stress
measurements obtained during compaction with the Servopac unit at compaction angles
of 1.25 and 2.5 degrees, are:
•
Gyratory shear slope, which is the rate of change in shear resistance during the
densification portion of compaction at 1.25 degrees; and
•
Vertical failure strain, which is the amount of vertical strain developed in the
mixture between the time instability is induced by increasing the compaction angle
to 2.5 degrees and the time the mixture begins to regain strength after instability.
Gyratory shear slope presented the results from the regression analyses of the
gyratory shear resistance versus the number of cycles in the densification zone.
53
Absolute Rut Depth, mm
5.0
4.0
3.0
2.0
1.0
0.0
8
11
η DASR < 50%
9
12
η DASR > 50%
10
Marginal
Interaction
Project
Figure 4-24. Absolute Rut Depth for Different Porosity Groups (APA)
1.0
0.8
Area Change, %
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
8
11
η DASR < 50%
9
12
η DASR > 50%
Project
Figure 4-25. Area Change for Different Porosity Groups (APA)
10
Marginal
Interaction
54
Project 8
Project 9
Project 10
Project 11
Project 12
40
Brittle Mixtures
Gyratory Shear Slope, kPa
35
Optimal Mixtures
Plastic MIxtures
30
DASR Porosity = 50 %
25
20
15
10
5
Low Shear Resistance
0
1
1.2
1.4
1.6
1.8
2
2.2
2.4
Vertical Failure Strain, %
Figure 4-26. Servopac Result for Plant Mix Gradations
The best-fit regression curves they found provided a relationship in the form:
G S = k1 log( N ) + k 2
where,
GS=
N=
k1=
k2=
gyratory shear resistance
number of gyrations
slope of regression line
intercept of regression line
To obtain data to calculate k1 by regression, compaction procedure applied at air
void levels ranging from: a) 7 percent to 4 percent, if the maximum gyratory shear
strength was not reached at 4 percent air voids, or b) from 7 percent to the air voids at
maximum gyratory shear strength.
In lieu of the gyratory shear strength, the vertical “failure” strain, measured from
the onset of the compaction with the 2.5 degree gyratory angle, to the local minimum on
55
the gyratory shear curve is an indicator of how “brittle” or how “plastic” a mixture will
respond during the rearrangement of the aggregate structure. A “low” failure strain
indicates a brittle mixture and a “high” value indicates a plastic mixture. To obtain
failure strains, a set of replicate samples for each mixture were compacted to a target air
void level of 7 (±0.5) percent. Once the target air voids level was reached, the compacted
for another 100 gyrations, and the failure strain was calculated.
Based on the criteria developed in the research, mixtures are considered to exhibit
optimal behavior when the percent vertical failure strain is between 1.4 and 2, and the
gyratory shear slope is greater than 15 kPa.
The results presented in Figure 4-26 indicate that only mixtures from projects 8
and 11 consistently have vertical failure strains in the optimal range. Except for two
specimens tested, vertical strains for projects 9, 10, and 12 were outside the optimal range
(in the brittle range). As shown in Figure 4-20, Projects 8 and 11 were the two projects in
Group I with DASR porosity less than 50%, while Projects 9 and 12 were in Group II
(DASR porosity > 50%) and Projects 10 was in Group III (marginal interaction). It is
interesting to note that the two specimens from the Group II and III mixture that were in
the optimal range were: 1) a plant mix specimen obtained from a location along Project 9
where the DASR porosity was 50%; and 2) one Project 10 mixture, which was
considered marginal, indicating that small changes could potentially make the mixture
good or bad (i.e., sensitivity). The results are presented by grouping according to the
gradation analysis in Figure 4-27.
In summary, the evaluation based on laboratory rut depths also appear to verify the
validity of the criteria established based on the gradation evaluation system developed as
56
Porosity < 50%, Projects 8,11
Porosity > 50%, Projects 9,12
Marginal Interaction, Project 10
40
Brittle Mixtures
Gyratory Shear Slope, kPa
35
Optimal Mixtures
Plastic MIxtures
30
DASR Porosity = 50 %
25
20
15
10
5
Low Shear Resistance
0
1
1.2
1.4
1.6
1.8
2
2.2
2.4
Vertical Failure Strain, %
Figure 4-27. Servopac Result for Different Porosity Groups
part of this research effort. These are promising outcomes based on 12 Superpave
mixtures of varying gradation and aggregate type that are currently used throughout the
state of Florida.
57
4.4 WesTrack Test Sections
4.4.1 General Description
WesTrack is the Federal Highway Administration's (FHWA) road test facility
located in Nevada (Epps et al., 1997, 1999, 2002). The project, entitled "Accelerated
Field Test of Performance-Related Specifications for Hot-Mix Asphalt Construction",
had two primary objectives:
Development of performance-related specifications (PRS) for HMA construction.
Early field verification of the SHRP SUPERPAVE(TM) Level III mix design.
The track was designed and constructed during the period between October 1994
and October 1995. The 2.9-km oval track consists of two tangent and the superelevated
curves connecting them. Each tangent contains 13 test sections, each of which is 70
meters (m) long (Figure 4-28). There are no test sections along the curves.
Figure 4-28. WesTrack - Layout of Test Track (not to scale)
58
As it neared the end of its planned loading in June 1998, WesTrack had been
trafficked for more than 2 years, during that time, more than 4.5 million 80-kN (18,000lb) equivalent single-axle loads (ESALs) were applied to the track.
4.4.2 Experiment Design and Performance History
The experiment design was based on seven experimental factors and target levels
shown in Table 4-1.
Table 4-1. Original Experimental Factors
Factor
Target levels
Coarse Aggregate Type
One level: local Dayton, Nevada pit
Aggregate Gradation
Three levels: Coarse, fine, and fine plus
Aggregate Shape/Texture
One level: high percent fractured faces
Asphalt Cement Type
One level: PG 64-22
Asphalt Content
Three levels each: 4.7, 5.4 and 6.1 percent for the fine
mixes; 5.0, 5.7 and 6.4 percent for the coarse mixes
Air Void Content
Three levels: 4, 8 and 12 percent
Hot-Mix Asphalt Thickness
One level: 150 mm or 6 inch
These factors and associated levels were selected to obtain the most information
relative to the effects of materials and construction variability on pavement performance.
A complete factorial was not feasible because of economic constraints, therefore
three factors were ultimately chosen, based on the potential on performance and/or
experience from previous investigations.
The factorial experiment is shown in Table 4-2; note that six cells out of the matrix
were eliminated because of construction impracticality, leaving 21 potential mixes. To
this, 5 replicates were added, resulting in 26 total sections. The numbers within each cell
59
represent the randomized paving sequence of each section. In June 1997 an additional
eight sections were built to replicate the coarse aggregate experiment with a different
aggregate source.
Table 4-2. Experiment Design
1997
Rehabilitation
Original 1995 construction
Design
air void
content
%
Aggregate gradation design
Fine plus
Coarse
Design asphalt contents (%)
Fine
4.7
4
5.4
6.1
4
18
14
8
2
1/15
12
3/16
17
4.7
5.4
6.1
12
21/9
22
19/11
13
10
20
5.0
Coarse
5.7
6.4
23
25
8
5/24
7
26
6
5.1
5.8
6.5
39
55
38
35/54
37
56
36
The description of the materials used in this project is presented in Table 4-3.
Table 4-3. Materials
Original Test Sections
Replacement Test Sections
Binder grade
and source
PG 64-22 West coast
PG 64-22 Idaho
Aggregate
source and
gradations
Quarry near Dayton, Nevada
(partially crushed fluvial deposit)
Sand from Wadsworth, Nevada
coarse, fine and fine-plus
Quarry near Lockwood, Nevada
(crushed andesite)
Sand from Wadsworth, Nevada
coarse
All the mixes in this project are 19mm NMPS; by spring 1997, the application of
more than 2.7 million ESALs resulted in rutting in almost every test section and fatigue
cracking in many of the test sections. Several sections had rutted more than 25 mm and
severe fatigue cracking had occurred in others. As a result, 10 sections (Sections 5-9, 13,
21, and 24-26) had to be removed and replaced during May and June 1997.
60
A new mix design was developed for eight of the replacement sections. This mix
design duplicated the coarse-graded mix experiment in the original construction, but
changed to a more angular aggregate. A quarried andesite replaced the crushed gravel
used in the original construction. The change in aggregate resulted in changes in the
volumetric properties from those obtained with the original coarse-graded mixes. The
other two replacement sections (Sections 43 and 51) utilized conventional Nevada
Department of Transportation (DOT) mixtures containing polymer-modified binders.
The replacement sections were placed in June 1997 and loading began in mid-July.
Most of the new sections exhibited significant deformation in the first 5 days of
trafficking. As a result of this early rutting and a concern that Superpave mixture design
or construction procedures might be missing a critical step or steps, FHWA assembled a
team of academicians, asphalt industry representatives, and State highway agency
engineers to investigate the performance at WesTrack.
The main conclusions from different reports about WesTrack are:
•
The main cause of rutting at WesTrack was a relatively high design binder content.
Over-asphalting during construction compounded the problem.
•
Much of the rutting appeared to be related to high binder contents due to high
VMA values, in conjunction with relatively low mastic stiffnesses.
•
For fatigue cracking, both field performance and laboratory test results have shown
the effects of compaction and asphalt content. With low air void content or
medium to high asphalt content the mixes showed much better fatigue resistance.
Also, aggregate gradation was significant, particularly for the coarse gradation.
The most important mix parameter, however, is compaction. As the degree of
compaction is increased, fatigue life is significantly improved.
•
For permanent deformation (rutting), field performance and laboratory RSST-CH
results have demonstrated the effects of asphalt content, compaction, pavement
temperature and, to some extent, the effects of aggregate gradation.
61
4.4.3 Interaction Diagrams
Figure 4-29 shows the JMF’s for each of the four mixture types used at WesTrack.
Interaction diagrams for the coarse-graded mixtures are presented in Figure 4-30, while
those for the fine-graded mixtures are presented in Figure 4-31. Interaction diagrams
indicate that both the coarse- and fine-graded mixtures used at Westrack exhibited
marginal interaction and potential sensitivity to variations in gradation. Figures 4-30 and
4-31 indicate that although the original coarse-graded mixture did not exhibit marginal
interaction for any particle size combination, the relatively minor change in gradation
implemented with the replacement mixture resulted in marginal interaction between the
9.5/4.75 mm sizes. This appears to indicate that the mixture was potentially sensitive to
variations in gradation. Unfortunately, the actual gradation of the original coarse-graded
MDL
JMF Coarse replacement
JMF Fine plus
JMF Coarse original
JMF Fine
100
90
80
% passing
70
60
50
40
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
⅜"
Sieve size, ^0.45
Figure 4-29. JMF Mixtures Gradations
½"
¾"
62
Large/Small Particle Proportion
JMF Coarse original
JMF Coarse replacement
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
19-12.5
0/100
Contiguous Sizes, mm
Figure 4-30. Interaction Diagram for JMF Coarse and JMF Coarse Replacement
JMF Fine
JMF Fine plus
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
Contiguous Sizes, mm
Figure 4-31. Interaction Diagram for JMF Fine and JMF Fine plus
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
0/100
12.5-9.5
10/90
19-12.5
Large/Small Particle Proportion
100/0
63
mixture placed at the track could not be found in the available reports, so a direct analysis
of DASR porosity of these mixtures for comparison to observed performance was not
possible.
However, the in-place gradations of the replacement sections were available and
are presented in Figure 4-32 for one set of coarse-replacement sections. The interaction
diagrams for these mixtures are presented in Figures 4-33 and 4-34, which indicate that
the in-place mixtures exhibited marginal interaction between two or more sets of particle
size combinations. Sections 36 and 37 (Figure 4-33) exhibited marginal interaction
between the 9.5/4.75 mm sizes and between the 4.75/2.36 mm sizes, while sections 55
and 56 (Figure 4-34) exhibited marginal interaction between the 2.36/1.18 mm sizes in
addition to the other two combinations.
100
90
80
% passing
70
MDL
60
JMF
50
Section 36
Section 37
40
Section 55
Section 56
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 4-32. Gradation of Coarse Replacement Sections (36, 37, 55, 56)
64
Large/Small Particle Proportion
Section 36
Section 37
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.3-0.15
0.15-0.075
0.075-0
0.3-0.15
0.15-0.075
0.075-0
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
19-12.5
0/100
Contiguous Sizes, mm
Figure 4-33. Interaction Diagram for Sections 36 and 37
Section 55
Section 56
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
19-12.5
0/100
Contiguous Sizes, mm
Figure 4-34. Interaction Diagram for Sections 55 and 56
65
Figure 4-35 shows DASR porosity values calculated for each of the coarse
replacement sections. As shown in this figure, the DASR porosity varies tremendously
depending on whether or not the marginally interactive size combinations are considered
to be interactive. The DASR porosity is well below 50% if full interaction is considered
and well over 50% if interaction is not considered.
Effect of Interaction @ 12.5/1.18 & 12.5/9.5
80
70
Porosity, %
60
50
40
30
20
10
0
Section 36
Section 37
Section 55
Section 56
Interact @ 12.5/1.18
31.3
31.6
30.9
30.4
Interact @ only 12.5/9.5
75.2
73.6
74.5
74.8
Figure 4-35. DASR Porosity (ηDASR) of Coarse Replacement Sections (36, 37, 55, 56)
Rut depth measurements for the original and replacement coarse-graded sections
are presented in Tables 4-4 and 4-5, respectively. These results clearly indicate that both
the original and replacement sections exhibited significant rutting, and the replacement
sections actually rutted more severely than the original sections. Rutting in the
replacement mixtures ranged from 20.5 to 34.6 mm after only 582,000 ESAL’s. It should
be noted that this more severe rutting occurred even though a more angular aggregate was
used in the replacement mixtures. This seems to indicate that better aggregate cannot
compensate for poor gradation.
66
Table 4-4. Rut depth for Original Coarse Mixtures
Section
Rut depth (mm) - peak to valley
ESALs ×106
5
22
2.8
6
30
1.5
7
36
2.8
8
23
1.5
23
12
2.8
24
26
2.8
25
27
1.5
26
19
2.8
Table 4-5. Rut Depth for Coarse Replacement Sections (36, 37, 55, 56)
Section
Field rut depth (Peak to valley), mm – After 582,000 ESAL’s
Section 36
34.6
Section 37
24.3
Section 56
20.5
Section 57
25.2
A second set of coarse replacement sections was placed and similar results were
obtained. Gradation and interaction diagrams for this second set of sections are presented
in Figures 4-36, 4-37, and 4-38. DASR porosity results are shown in Figure 4-39.
Clearly, the results are very similar to the previous set of sections and these sections also
rutted severely (although not as severely as the first set), exhibiting rut depths of between
11.4 and 15.8 mm after only 582,000 ESAL’s (see Table 4-6).
The sensitivity of the fine-graded mixtures resulting from the marginal interaction
between different coarse particle sizes (see Figure 4-31) was revealed in the observed
rutting performance of these mixtures. Measured rut depths for the fine and fine- plus
67
100
90
80
% passing
70
60
MDL
JMF
50
Section 35
Section 38
40
Section 39
Section 54
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 4-36. Gradation of Coarse Replacement Sections (35, 38, 39, 54)
Large/Small Particle Proportion
Section 35
Section 38
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-37. Interaction Diagram for Sections 35 and 38
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
19-12.5
0/100
68
Section 39
Section 54
Large/Small Particle Proportion
100/0
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
19-12.5
0/100
Contiguous Sizes, mm
Figure 4-38. Interaction Diagram for Sections 39 and 54
Effect of Interaction @ 12.5/1.18 & 12.5/9.5
80
70
Porosity, %
60
50
40
30
20
10
0
Section 35
Section 38
Section 39
Section 54
Interact @ 12.5/1.18
30.8
30.2
32.2
30.0
Interact @ only 12.5/9.5
72.8
74.8
75.5
73.3
Figure 4-39. DASR Porosity (ηDASR) of Coarse Replacement Sections (35, 38, 39, 54)
69
Table 4-6. Field Rut Depth for Coarse Replacement Sections (35, 38, 39, 54)
Section
Field rut depth (Peak to valley), mm – After 582.000 ESAL’s
Section 35
15.8
Section 38
11.6
Section 39
11.4
Section 54
12.3
mixtures are presented in Tables 4-7 and 4-8, respectively. The results are also presented
in Figure 4-40, which shows that significantly different rutting performance was observed
for the fine mixture than for the fine-plus mixture, even though the gradation differences
between them were relatively minor (see Figure 4-31). Unfortunately, the in-place
gradations of these mixtures were not available for these mixtures, so DASR porosity
calculations could not be performed for the fine mixtures. It is anticipated that DASR
porosity would be less than 50% if interaction were considered and greater than 50% if it
were not. The ultimate performance would be dictated by the in-place gradation, which
was apparently more favorable for the fine than for the fine-plus mixture.
Table 4-7. Rut Depth for Fine Mixtures
Section
Rut depth (mm) - peak to valley
ESALs ×106
1
9
2.8
2
6
2.8
3
10
2.8
4
9
2.8
14
10
2.8
15
10
2.8
16
9
2.8
17
10
2.8
18
7
2.8
70
Maximum rut depth (peak to valley), mm
Table 4-8. Rut Depth for Fine plus Mixtures
Section
Rut depth (mm) - peak to valley
ESALs ×106
9
30
1.5
10
12
2.8
11
11
2.8
12
10
2.8
13
20
1.5
19
10
2.8
20
11
2.8
21
35
1.5
22
10
2.8
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
1
2
3
4 14 15 16 17 18
Fine Mixtures
Section
9 10 11 12 13 19 20 21 22
Fine plus Mixtures
Section
Figure 4-40. Maximum Rut Depth for Fine and Fine plus Mixtures
71
4.4.4 Summary
All mixtures placed at WesTrack were identified as having gradations exhibiting
marginal interaction as determined by the gradation analysis system developed in this
research. All coarse-graded mixtures rutted, even after a more angular aggregate was
introduced. It was noted that the gradation used with the more angular aggregate was
even more marginal and potentially sensitive than the original coarse gradation. The
modified gradation with the more angular aggregate resulted in even more severe rutting
than the original mixture. The fine-graded (fine and fine plus) mixture exhibited highly
variable rutting performance, as expected based on the marginally interactive gradation.
72
4.5 NCAT Test Sections
NCAT Pavement Test Track is a 1.7 mile oval divided in 200 ft test sections
(Brown et al., 2002, 2004); the primary purpose of the work at the NCAT test track is to
use the performance at the track to verify or help develop performance tests (Figure 441). Secondary objectives of the project are to look at fine-graded vs. coarse-graded
mixes, to evaluate the effect of grade bumping (modified AC vs. non-modified AC),
compare performance of various mix types, and to evaluate the effect of aggregate type
(limestone, slag, gravel, granite, etc.).
Figure 4-41. NCAT - Layout of Test Track (not to scale)
The track was designed to be sufficiently strong so that fatigue cracking would not
occur resulting in rutting as the expected form of distress. The average rutting at the
track was approximately 0.12 inches (3 mm) after approximately 9 million ESALs.
Rutting is typically not considered to be a problem until the magnitude reaches
approximately 0.5 inches (12.5 mm), so the rutting observed at the track was minimal.
All the cases presented in this report are 12.5mm NMPS mixes. They were divided into
73
four groups based on their gradations; coarse, fine, dense-coarse and SMA. Table 4-9
shows the reference figures and tables applied by the DASR porosity approach.
Table 4-9. Reference Figures and Tables for NCAT
Gradation Type
Gradations
Rut Depth
Coarse
Figure 4-42
Table 4-10
Fine
Figure 4-45
Table 4-11
Dense-Coarse
Figure 4-48
Table 4-12
SMA
Figure 4-51
Table 4-13
Interaction
Figure 4-43
Figure 4-46
Figure 4-49
Figure 4-52
Porosity
Figure 4-44
Figure 4-47
Figure 4-50
Figure 4-53
All the sections meet the interaction and DASR porosity requirements, and as expected,
they performed well in terms of rutting even for different aggregate types. Even though
marginal interactions were considered, DASR porosities were below 50%. The specifics
of the interaction diagrams for each set of mixtures are discussed in the sections below.
4.5.1 Interaction Diagrams: Coarse Mixtures
Gradations for the three coarse mixtures placed at the NCAT test track are
presented in Figure 4-42. The resulting interaction diagrams, which are presented in
Figure 4-43, indicate that for all three mixtures, there was marginal interaction between
the 4.75/2.36 mm sizes and the 2.36/1.18 mm sizes. However, the DASR porosity
calculations presented in Figure 4-44 show that the DASR porosity was less than 50%
whether or not these interactions were considered. In other words, it appears that these
mixtures have very good gradations.
The rutting results presented in Table 4-10 indicate that all the rut depth was less
between 2.8 and 6.2 mm after 9 million ESALs. These results support the findings from
the gradation analysis based on the approach developed in this study.
74
100
90
80
% passing
70
60
MDL
50
E2
40
E3
E4
30
20
10
0
#100 #30 #16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 4-42. Gradation of Sections E2, E3, and E4
E2
E3
E4
100/0
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-43. Interaction Diagram for Sections E2, E3, and E4
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
75
100
90
80
Porosity, %
70
60
50
40
30
20
10
0
E2
E3
E4
Interact @ 9.5/4.75
30.0
30.8
30.6
Interact @ 9.5/1.18
48.0
48.3
48.9
Figure 4-44. DASR Porosity of Sections E2, E3, and E4
Table 4-10. Field Rut Depth for Sections E2, E3, and E4
Section
Aggregate type
ESAL’s
Field rut depth , mm
Section E2
Limestone
4,172,787
6.2
Section E3
Limestone
4,172,787
3.1
Section E4
Granite
4,172,787
2.8
4.5.2 Interaction Diagrams: Fine Mixtures
Gradations for the three fine mixtures placed at the NCAT test track are presented
in Figure 4-45. The resulting interaction diagrams, which are presented in Figure 4-46,
indicate that for all three mixtures, there was excellent interaction from the 4.75 mm to
the 1.18 mm sizes. Although the interaction between the 9.5/4.75 mm sizes is within the
70/30 criterion identified for marginal interaction, it was treated as marginally interactive
to evaluate the effect on DASR porosity.
76
100
90
80
% passing
70
60
MDL
50
E8
40
E9
30
E10
20
10
0
#100 #30 #16 #8
1.18 2.36
#4
4.7
5
⅜"
½"
¾"
Sieve size, ^0.45
E8
E9
E10
2.36-1.18
1.18-0.6
0.6-0.3
Figure 4-45. Gradation of Sections E8, E9, and E10
100/0
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
Contiguous Sizes, mm
Figure 4-46. Interaction Diagram for Sections E8, E9, and E10
0.075-0
0.15-0.075
0.3-0.15
4.75-2.36
9.5-4.75
12.5-9.5
0/100
77
DASR porosity calculations presented in Figure 4-47 show that the DASR porosity
was right at 50% when interaction was not considered and well below 50% when it was
considered. The rutting results presented in Table 4-11 indicate that all the rut depths for
sections with these mixtures were less than 3.3 mm after 9 million ESALs. These results
also support the findings from the gradation analysis based on the approach developed in
this study, which indicate that these fine mixtures have good aggregate structure.
100
90
80
Porosity, %
70
60
50
40
30
20
10
0
E8
E9
E10
Interact @ 9.5/1.18
45.2
43.8
46.4
Interact @ 4.75/1.18
50.8
49.2
51.1
Figure 4-47. DASR Porosity of Sections E8, E9, and E10
Table 4-11. Field Rut Depth for Sections E8, E9, and E10
Section
Aggregate type
ESAL’s
Field rut depth , mm
Section E8
Granite
4,172,787
3.3
Section E9
Granite
4,172,787
1.9
Section E10
Granite
8,972,237
N/A
78
4.5.3 Interaction Diagrams: Dense-Coarse Mixtures
Gradations for the four dense-coarse mixtures placed at the NCAT test track are
presented in Figure 4-48. The resulting interaction diagrams, which are presented in
Figure 4-49, indicate that for all four mixtures exhibited marginal interaction between the
9.5/4.75 mm sizes, and one or two exhibited marginal interaction between the 4.75/2.36
mm sizes. However, the DASR porosity calculations presented in Figure 4-50 show that
the DASR porosity was well under 50% for all four mixtures, whether or not these
interactions were considered. In other words, it appears that these mixtures have very
good gradations.
100
90
80
% passing
70
MDL
60
N5
50
N6
40
N7
30
N8
20
10
0
#100 #30 #16 #8
1.18 2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 4-48. Gradation of Sections N5, N6, N7, and N8
The rutting results presented in Table 4-14 indicate that all the rut depth was less
between 3.0 and 5.6 mm after 9 million ESALs. Once again, these results support the
findings from the gradation analysis based on the approach developed in this study.
79
N8
N7
N6
N5
100/0
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-49. Interaction Diagram for Sections N5, N6, N7, and N8
100
90
80
Porosity, %
70
60
50
40
30
20
10
0
N5
N6
N7
N8
Interact @ 9.5/1.18
38.6
36.9
36.3
37.4
Interact @ 4.75/1.18
44.2
41.9
41.6
42.3
Figure 4-50. DASR Porosity of Sections N5, N6, N7, and N8
80
Table 4-12. Field Rut Depth for Sections N5, N6, N7 and N8
Section
Aggregate type
ESAL’s
Field rut depth , mm
Section N5
Grn/Lms/Snd
4,172,787
3.0
Section N6
Grn/Lms/Snd
4,172,787
4.8
Section N7
Granite
4,172,787
4.3
Section N8
Granite
4,172,787
5.6
4.5.4 Interaction Diagrams: SMA Mixtures
Gradations for the four SMA mixtures placed at the NCAT test track are
presented in Figure 4-51. The resulting interaction diagrams, which are presented in
Figure 4-52, indicate that only the 9.5/4.75 mm sizes were interactive for these mixtures.
This is expected for SMA mixtures, which are designed to have one or two dominant
sizes.
100
90
80
% passing
70
60
MDL
50
W3 lower
W3upper
40
W4 lower
30
W4upper
20
10
0
#100 #30 #16 #8
1.18 2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 4-51. Gradation of Sections W3 lower, W3 upper, W4 lower, and W4 upper
81
W3 lower
W3upper
W4 lower
W4upper
100/0
Large/Small Particle Proportion
90/10
80/20
70/30
60/40
50/50
40/60
30/70
20/80
10/90
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0/100
Contiguous Sizes, mm
Figure 4-52. Interaction Diagram for Sections W3 lower, W3 upper, W4 lower, and W4
upper
As shown in Figure 4-53, the DASR porosity of all SMA mixtures was less than 50%.
The SMA mixture closest to the maximum density line had a DASR porosity close to
50%, while the DASR porosity of the others was well below 50%.
The rutting results presented in Table 4-13 indicate that rut depths for all four SMA
mixtures were less than 5 mm after 9 million ESALs. As with all other mixtures
evaluated, these results support the findings from the gradation analysis based on the
approach developed in this study.
4.5.4 Summary
All mixtures placed at the NCAT test track were identified as having good
gradation characteristics by gradation analysis system developed in this research.
82
100
90
80
Porosity, %
70
60
50
40
30
20
10
0
W3 lower
W3 upper
W4 lower
W4 upper
Figure 4-53. DASR Porosity of Sections W3 lower, W3 upper, W4 lower, and W4 upper
Table 4-13. Field Rut Depth Sections W3 lower, W3 upper, W4 lower, and W4 upper
Section
Aggregate type
ESAL’s
Field rut depth , mm
Section W3 lower
Limestone
4,172,787
4.6
Section W3 upper
Limestone
4,172,787
4.6
Section W4 lower
Granite
4,172,787
4.1
Section W4 upper
Granite
4,172,787
4.1
The DASR porosity of all mixtures was less than 50%, even when marginally interactive
aggregate sizes were treated as non-interactive in the DASR calculations. All mixtures
exhibited good rutting performance, where the maximum rut depth for any mixture was
6.2 mm after 9 million ESALs.
These results indicate that the gradation analysis system developed in this study
accurately identified the rutting performance of a broad range of mixtures under realistic
traffic conditions. These mixtures encompassed a broad range of gradations, from fine-
83
graded to SMA, and aggregate types. This appears to indicate that the criteria established
may be fundamental enough in nature to be independent of mixture or aggregate type.
4.6 Additional Observations
Results of evaluations presented in the previous sections of this chapter clearly
indicate that the following criteria, which were based on the gradation analysis system
developed in this study, resulted in reasonable agreement with observed laboratory and
field performance of asphalt mixture:
•
DASR porosity of asphalt mixture should be less than 50% to ensure coarse
aggregate interlock.
•
The relative proportion of contiguous size particles within the DASR must be no
greater than 70/30 to ensure proper interaction among the different size particles in
the DASR.
It was also observed that mixtures may exhibit marginal performance if the
gradation exhibits either of the following two characteristics:
•
DASR porosity is very close to 50% and small changes in gradation would result in
significantly higher DASR porosity.
•
The relative proportion of one or more sets of contiguous size particles in the
DASR is very close to 70/30 and the interaction of this set of particle sizes is
critical to achieve a DASR porosity lower than 50%.
The implication is that for mixtures having these gradation characteristics, small
changes in field gradation may result in DASR porosity greater than 50% and
unacceptable performance. This effect was evident in several cases evaluated in this
chapter, including mixtures used in the WesTrack studies and mixture involved in the
FDOT Superpave monitoring projects. These cases illustrated how these types of
mixtures, which were called marginal mixtures, resulted in variable and even catastrophic
performance, particularly when marginal interaction was observed between the 4.75/2.36
mm or the 2.36/1.18 mm sizes.
84
Based on these observations, the following recommendations are presented to
reduce the potential of selecting gradations that are likely to result in marginal
performance:
•
In addition to having a DASR porosity less than 50%, gradations should be
evaluated to ensure that acceptable gradation variances do not result in DASR
porosity greater than 50%.
•
The relative proportions between contiguous size aggregates in the DASR range
should be well below 70/30 (e.g., 65/35) when the interaction of these sizes is
critical to maintain the DASR porosity below 50%.
4.6.1 Excessively Low DASR Porosity
Although the available data did not allow for direct evaluation of a lower DASR
porosity limit, existing knowledge of mixture behavior indicates that excessively low
porosity may result in the following problems:
•
Mixtures may be difficult to compact and have generally poor workability.
•
Mixtures may exhibit brittle behavior.
Therefore, an investigation of the use of a minimum allowable DASR porosity is
highly recommended for future work.
As a start, a series of finite element (FEM) analyses was conducted as part of this
study to investigate the potential effects of low DASR porosity on stress concentrations
within the asphalt aggregate structure. FEM analyses were conducted for three levels of
DASR porosity, corresponding to three levels of interstitial volume (IV). Note that IV is
directly related to DASR porosity, since IV is the volume occupying the pores
represented by the DASR porosity.
The system modeled in the FEM analysis is represented in Figure 4-54. As shown
in the figure, the mixture was modeled as a two-part system composed of aggregate,
representing the DASR, and the asphalt, aggregate, and air void system within the IV,
85
which is referred to as the interstitial component (IC). The same level of tensile stress
was applied to the mixtures with different IV’s to evaluate the effect on the resulting
tensile stress within the IC.
(a) More IV
(b) Less IV
Figure 4-54. Finite Element Model of Aggregate and Interstitial Volume
The results plotted in Figure 4-55 clearly indicate that the tensile stress within the
IC increases as the IV decreases, even though the applied tensile stress was the same in
all cases. These higher internal stresses imply that mixtures with lower IV will fail at
lower strain levels, because lower applied tensile stress would be required to reach the
failure strength of the material.
These results also imply that IV may be a good indicator of brittle mixtures.
Currently, there is no commonly accepted mixture parameter that reliably predicts brittle
behavior. However, additional study is required to investigate this further and establish
rational criteria for this purpose. These preliminary results indicate that IV is promising,
86
but additional characteristics of the interstitial component (IC) and mixture type also
likely play a significant role.
20
stress zz
15
10
5
0
0
10
20
30
40
Interstitial Volume, %
Figure 4-55. Interstitial Spacing (Volume) vs Local Stress
50
60
CHAPTER 5
FURTHER TESTING
5.1 Introduction
Based on the evaluation with an extensive range of database presented in Chapter 4,
additional laboratory tests were performed to validate different types of gradations in
terms of the interaction diagram and the DASR porosity. Mixtures were designed with
Georgia granite and Rinker South Florida limestone aggregates to evaluate the effect of
aggregate type. All mixtures were 12.5 mm nominal maximum aggregate size
gradations. Asphalt type PG 67-22 was used to prepare all the mixtures.
5.2 Materials
Two common aggregates in Florida were selected for the research, which are
Georgia granite and Rinker South Florida limestone. The selected sources of aggregates
were shown in Table 5-1. To reduce asphalt binder effect, the same type of binder (PG
67-22), which is commonly used in Florida, was selected for all mixtures.
Table 5-1. Aggregate Sources
Source
GA Granite
Rinker
South FL
Limestone
Local Sand
Type
FDOT code
Pit No.
Producer
# 78 Stone
43
GA-553
Junction City Mining
# 89 Stone
51
GA-553
Junction City Mining
W-10 Screenings
20
GA-553
Junction City Mining
# 67 Stone
42
87-090
Rinker Materials Corp.
S-1-B
55
87-090
Rinker Materials Corp.
Med. Screenings
21
87-090
Rinker Materials Corp.
Local Sand
Starvation Hill
87
V. E. Whitehurst & Sons
88
5.3 Gradations
Each type of aggregate has four different gradations in order to have different
characteristics of the interaction diagram and the DASR porosity on purpose. Total eight
gradations were designed and have IDs shown in Table 5-2. Two of them (2-Bad and 3Bad) for each types of aggregates were designed to have bad performance and
characteristics of gradation according to the DASR porosity. On the other hand, one of
them (Good) was designed for good and the last one was designed for the marginal
porosity (≈ 48).
Table 5-2. Gradation IDs for Testing
Aggregate
Gradations
GA Granite
GA-2-Bad
GA-3-Bad
GA-Good
GA-48
Rinker FL Limestone
FL-2-Bad
FL-3-Bad
FL-Good
FL-48
5.3.1 Georgia Granite
Figure 5-1 shows four gradations for GA granite. Gradation GA-48 is a little
coarser than GA-Good. GA-2-Bad and GA-3-Bad are much coarser than GA-Good in
coarse aggregate fractions. However, the difference between GA-48 and GA-3-Bad is
not big as much as GA-2-Bad. All three gradations are finer than GA-Good in fine
aggregate fractions.
Interaction diagram is presented in Figure 5-2. GA-2-Bad shows that the relative
proportion of the 4.75/2.36 mm sizes, and 2.36/1.18 mm aggregate sizes was clearly out
of range 70/30. On the other hand, GA-Good and GA-48 are fully interacted between
them. GA-3-Bad shows marginal interactions (i.e. close to 70/30).
89
GA Granite
100
90
80
% passing
70
MDL
GA-2-Bad
60
GA-3-Bad
GA-Good
50
40
GA-48
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
½"
⅜"
¾"
Sieve size, ^0.45
Figure 5-1. Gradations for GA Granite
GA-2-Bad
GA-3-Bad
GA-Good
GA-48
100
Big particle % retained
90
80
70
60
50
40
30
20
10
Contiguous Sizes, mm
Figure 5-2. Interaction Diagram for GA Granite Gradations
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0
90
Based on interaction diagrams, DASR porosities of four gradations were calculated
and shown in Figure 5-3. While the DASR porosity for GA-2-Bad shows clearly over
50%, GA-Good shows clearly below 50%. The DASR porosity of GA-3-Bad is around
50% if marginal interaction sieve sizes, shown in Figure 5-2, are considered interactive,
but significantly greater than 50% if these sizes are not interactive. The DASR porosity
of GA-48 was targeted around 48%.
80
without interaction
Porosity, %
70
60
50
40
30
20
GA-Bad-2
GA-Bad-3
GA-Good
GA-48
Gradation
Figure 5-3. DASR Porosity for GA Granite Gradations
The characteristics of the DASR porosity and interaction diagram for GA granite
gradations are summarized in Table 5-3.
5.3.2 Rinker South Florida Limestone
Gradations for Rinker South FL limestone are shown in Figure 5-4. Even though
2.36~1.18mm sizes for gradation FL-48 is a little finer than FL-Good to make a
difference in DASR porosity, coarse aggregate fractions are pretty similar each other.
91
Table 5-3. Summary of the DASR Porosity and Interaction Diagram for GA Granite
Gradations
Gradation
DASR Porosity, %
DASR, mm
Interaction
GA-2-Bad
67
4.75
> 70 (4.75 ~ 1.18)
GA-3-Bad
51 / 71
4.75 ~1.18 / 4.75
Marginal
GA-Good
42
9.5 ~1.18
37 ~ 55
GA-48
48
4.75 ~1.18
61
Rinker FL Limestone
100
90
80
70
MDL
FL-2-Bad
FL-3-Bad
% passing
60
50
FL-Good
FL-48
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve size, ^0.45
Figure 5-4. Gradations for Rinker South FL Limestone
FL-2-Bad has coarser aggregates in coarse fractions, but FL-3-Bad is finer than others
overall.
Interaction diagram is presented in Figure 5-5. All gradations exhibited that
interaction between the 9.5/4.75 mm sizes were clearly out of range, but between the
2.36/1.18 mm sizes are within range. While the relative proportion of the 4.75/2.36 mm
sizes only for FL-2-Bad was clearly over the range, others were still safe within the range
70/30.
92
FL-2-Bad
FL-3-Bad
FL-Good
FL-48
100
Big particle % retained
90
80
70
60
50
40
30
20
10
0.075-0
0.15-0.075
0.3-0.15
0.6-0.3
1.18-0.6
2.36-1.18
4.75-2.36
9.5-4.75
12.5-9.5
0
Contiguous Sizes, mm
Figure 5-5. Interaction Diagram for Rinker South FL Limestone Gradations
The DASR porosity of four gradations was calculated based on interaction
diagrams and shown in Figure 5-6. While the DASR porosity for FL-2-Bad and 3-Bad
show over 50%, FL-Good shows below 50%. Although the DASR porosity of FL-Good
was initially tried to be within 40~45%, it exhibited around 46% because of limitations of
gradation types of materials. The DASR porosity of FL-48 was around 48%.
The characteristics of the DASR porosity and interaction diagram for Rinker South
Florida limestone gradations are summarized in Table 5-4.
Therefore, to evaluate different DASR porosities (i.e. over 50%, below 50%,
around 50%), all gradations were used for comparisons. GA-3-Bad was designed to
check the effect of the marginal interaction. FL-2-Bad and 3-Bad can be used to find out
whether or not there is effect between different DASR ranges. Table 5-5 summarized all
test matrix described above.
93
80
Porosity, %
70
60
50
40
30
20
FL-Bad-2
FL-Bad-3
FL-Good
FL-48
Gradation
Figure 5-6. DASR Porosity for Rinker South FL Limestone Gradations
Table 5-4. Summary of the DASR Porosity and Interaction Diagram for Rinker South FL
Limestone Gradations
Gradation
DASR Porosity, %
DASR, mm
Interaction
FL-2-Bad
55
4.75 ~ 2.36
62
FL-3-Bad
56
4.75 ~1.18
55 ~ 56
FL-Good
46
4.75 ~1.18
57 ~ 64
FL-48
48.5
4.75 ~1.18
57 ~ 64
Table 5-5. Summary for Test Matrix
To test
Matrix
GA Granite
FL Limestone
Using All
DASR porosity
>50% / ≈ 50% / <50%
Using All
Interaction
marginal
GA-3-Bad
DASR range with
equal porosity
4.75~2.36 or 4.75~1.18mm
FL-2-Bad
FL-3-Bad
94
5.4 Mix Design
According to FDOT request, they were designed for traffic level C, which is more
than 3 million and less than 10 million. Compaction levels corresponding to traffic level
C are 115 gyrations for Nmax and 75 gyrations for Ndesign.
As mentioned earlier, PG 67-22 was used for all mixtures. Table 5-6 presents the
design information for the selected gradations. Mixtures with GA granite aggregate meet
the Superpave criteria such as VMA and VFA. However, although all mixtures with
Rinker South FL limestone aggregate passed VFA criteria, only one of mixtures (FL-3Bad) meets VMA criteria.
Table 5-6. Designed volumetric information
Gradation
AC (%)
Gmm
Gsb
VMA (%)
VFA (%)
FL-2-Bad
5.2
2.336
2.413
11.9
66.5
FL-3-Bad
7.2
2.306
2.425
15.3
73.9
FL-48
6.5
2.324
2.408
13.4
70.1
FL-Good
6.6
2.311
2.400
13.6
70.8
GA-2-Bad
4.7
2.553
2.745
14.9
73.3
GA-3-Bad
4.7
2.561
2.746
14.7
72.8
GA-48
4.6
2.578
2.758
14.4
72.2
GA-Good
4.8
2.579
2.770
14.9
73.1
5.5 APA Test
The APA was used to test the rutting susceptibility or rutting resistance of
mixtures. It has been observed that the APA results are sensitive to aggregate gradation
and also correlated highly with actual in-place rut depths (Asiamah, 2002). The final
profiles were measured by the system itself and also by the new measurement system
(contour gauge) developed by Drakos (2003). Even though the original measurement
95
system measures a pin point, the new measurement system was implemented to record
the entire surface profile of the specimen.
Figure 5-7 shows rut depth results by the APA measurement system. FL-Good
exhibited the best performance result among all mixtures. Other FL limestone mixtures
presented similar rutting. GA-2-Bad and GA-3-Bad mixtures were the most rutted. The
difference between GA Bad mixtures and Good and 48 mixtures was conspicuously
significant. However, it is notable difference between FL limestone and GA granite
mixtures. The mixtures with bad gradations for FL limestone showed better rut
resistance compared to GA granite bad mixtures. They performed so well as much as
GA-48 and GA-Good mixtures. This might be induced by the characteristics of
aggregate such as texture, angularity, and so on.
10.0
9.0
Rut Depth, mm
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
FL-2B FL-3B FL-48 FL-G GA-2B GA-3B GA-48 GA-G
Figure 5-7. APA Results by System Measurement
The results from the new measurement system presented the similar trend to the
original measurement (Figure 5-8). GA-2-Bad and GA-3-Bad were much more rutted
than others. FL-Good exhibited less rutted than other FL-Bad mixtures and FL-48
96
showed also better rut resistance than FL-3B. GA-Good and GA-48 also performed better
than GA-Bad mixtures. This result was confirmed by Student T-test as shown in
Appendix E.
GA-3-Bad, which showed bad rutting performance, has a marginally interactive
gradation (Figure 5-2). In other words, the relative proportion of the 4.75/2.36 mm and
the 2.36/1.18 mm aggregate sizes was close to 70/30. The DASR porosity is 51 % if
these sizes are considered interactive, but significantly greater than 50% if these sizes are
not interactive.
20
DRD
18
ARD
Rut Depth, mm
16
14
12
10
8
6
4
2
0
FL-2B FL-3B FL-48
FL-G
GA-2B GA-3B GA-48
GA-G
Figure 5-8. Differential Rut Depth and Absolute Rut Depth Results from APA
Consolidation rutting induces (-) volume change, therefore, the area change
calculated from the surface profile presents (-) value. On the other hand, instability
rutting is presented by (+) area change. Generally speaking, FL limestone mixtures
showed (-) area changes except for FL-2-Bad, but GA granite mixtures exhibited (+) area
changes except for GA-48. However, as shown in Figure 5-9, it is difficult to say which
one shows (+) or (-) area change, because the result of the area changes is extensively
97
varied. The result of the area changes from the surface profile by the contour gauge is so
sensitive to get stable data.
4.0
3.0
1.0
GA-G
GA-48
GA-3B
GA-2B
FL-G
FL-48
-1.0
FL-3B
0.0
FL-2B
Area Change, %
2.0
-2.0
-3.0
-4.0
Figure 5-9. Area Change Results from APA
Since the result from the area change calculations was not clear, the maximum hill
height was computed instead of the area change. The maximum hill height, which is the
difference between DRD and ARD, is directly related to DRD and ARD as shown in
Figure 5-10. Figure 5-11 shows the relationship between the maximum hill height
(DRD-ARD) and the area change. Even though they show some scatter data around 0 %
area change, lower hill height clearly shows (-) area change and higher hill height shows
(+) area change. In other words, the mixture induced higher rutting by APA may have
(+) area change, and lower rutting mixtures have (-) area change. This is understandable
because most of less rutting mixtures will induce the consolidation type of rutting, but
more rutting mixtures will exhibit the instability type of rutting due to the boundary effect
of APA system. Therefore, the trend is clearer with the DRD-ARD results as shown in
98
Figure 5-12. Since GA granite mixtures show higher hill height, they are more possible
to have instability type of rutting.
8
DRD
7
ARD
DRD-ARD, mm
6
y = 0.7494x - 1.0922
R2 = 0.9051
5
4
y = 0.4352x - 0.6796
R2 = 0.9665
3
2
1
0
0
5
10
15
20
Rut Depth, mm
Figure 5-10. Relationship between Hill Height (DRD-ARD) and, DRD or ARD
8
7
DRD-ARD, mm
6
5
y = 2.3119x + 3.9443
R2 = 0.6182
4
3
2
1
0
-1.00 -0.75 -0.50 -0.25
0.00
0.25
0.50
0.75
1.00
Area Change, %
Figure 5-11. Relationship between Hill Height (DRD-ARD) and Area Change
99
8
7
DRD-ARD, mm
6
5
4
3
2
1
G-G
G-48
G-3B
G-2B
F-G
F-48
F-3B
F-2B
0
Figure 5-12. Results of The Maximum Hill Height (DRD-ARD)
5.6 ServoPac Test
Figure 5-13 presented the ServoPac test results. FL-Good and FL-48 performed
better in APA test are within the area of optimal mixtures. Although GA-Good is in
brittle mixtures, it is on the line between the area of optimal and brittle mixtures.
Therefore, good and even marginal mixtures were within or closer to the optimal area
than bad mixtures. Even though all mixtures are compared, the result is quite reasonable.
Figure 5-14 presents the relationship between the failure strain from ServoPac test and
the rut depth from APA test.
However, all worse performance mixtures, which have more rutted in APA test,
positioned into the area for brittle mixtures. According to inspection of each APA
sample, bad performed mixtures by APA exhibited more severe damage at the bottom
side of sample in general. FL-Good, FL-48, GA-Good, and GA-48 have some hair
cracks and/or smaller amount of materials falling apart than bad mixtures (Figure 5-15
and 5-16).
100
FL-2B
FL-3B
FL-48
FL-G
GA-2B
GA-3B
GA-48
GA-G
40
Brittle
Mixtures
Gyratory Shear Slope, kPa
35
Plastic
MIxtures
Optimal
Mixtures
30
25
20
15
failure strain 0.97%
10
Low Shear
Resistance
5
0
1
1.2
1.4
1.6
1.8
2
2.2
Vertical Failure Strain, %
Figure 5-13. ServoPac Test Results
18
DRD
16
y = -16.897x + 31.399
R2 = 0.7545
Rut Depth, mm
14
12
ARD
10
8
6
4
y = -9.4499x + 18.334
R2 = 0.7473
2
0
0.7
0.9
1.1
1.3
1.5
1. 7
Failure Strain, %
Figure 5-14. Relationship between the Failure Strain and the Rut Depth
2.4
101
Figure 5-15 Pictures for Bad Performance Samples after APA Test
Figure 5-16 Pictures for Good Performance Samples after APA Test
5.7 Summary
To evaluate gradation effects, four gradations were designed for each source of
aggregates in terms of the characteristics of the interaction diagram and the DASR
porosity. Two gradations expected good performance for each were showing better
performance in APA and ServoPac tests. The bad gradations relatively exhibited poor
resistance to rutting. However, FL limestone mixtures performed better in general
because aggregates may have much rougher textures than GA granite aggregates.
According to results, the mixtures with the marginal DASR porosity (≈ 48 %) also had
102
good resistance to rutting. Therefore, if gradation is strictly controlled based on JMF and
interactions especially at 4.75 ~ 1.18 mm are strong, it will perform as much as a good
gradation does. On the other hand, as shown in GA-3-Bad which has marginal
interactions between the 4.75/2.36 mm and the 2.36/1.18 mm, the marginal interaction
should be avoided especially when it affects the significant changes in the DASR
porosity.
CHAPTER 6
CLOSURE
6.1 Summary of Findings
The importance of aggregate structure on asphalt mixture performance has been
well established on the basis of experience and is well documented in the literature.
Furthermore, coarse aggregate structure is most important for resistance to rutting, and
recent work has shown that it can also play a significant role in resistance to damage and
fracture. Therefore, large enough aggregates should engage dominantly in the structure
for good mixture performance. This study focused on the development of a conceptual
and theoretical approach to evaluate coarse aggregate structure based on gradation.
It is a well-known fact in soil mechanics that the porosity of granular materials in
the loose state is approximately constant between 45% and 50%, regardless of particle
size or distribution. This implies that the porosity of an assemblage of granular particles
(e.g., the aggregate within an asphalt mixture) must be no greater than 50% for the
particles to be in contact with each other. This also implies that one can use porosity as a
criterion to assure contact between large enough particles within the mixture to provide
suitable resistance to deformation and fracture. Calculations performed for gradations
associated with typical dense graded mixtures indicated that the porosity of particles
retained on any single sieve was significantly greater than 50%, even for gradations
associated with the maximum density line. Since many dense-graded mixtures are
known to provide suitable resistance to deformation and fracture, then there must be a
103
104
range of contiguous coarse aggregate particle sizes that form a network of interactive
particles with a porosity of less than 50%.
A theoretical analysis procedure was developed to calculate the center-to-center
spacing between specific size particles within a compacted assemblage of particles of
known gradation. Calculations performed with this procedure indicated that the relative
proportion of two contiguous size particles, as defined by the standard arrangement of
Superpave sieves, can be no greater than 70/30 in order to form an interactive network.
Thus, the 70/30 proportion can be used to determine whether particles on contiguous
Superpave sieves can form an interactive network of particles in continuous contact with
each other. The range of particle sizes determined to be interactive was referred to as the
dominant aggregate size range (DASR) and its porosity must be no more than 50% for
the particles to be in contact with each other.
Analysis of an SMA mixture indicated that the DASR was composed of only one
size aggregate, and as expected, its porosity was less than 50%. Analysis of densegraded mixtures (coarse-graded and fine-graded) of known performance indicated that
DASR porosity of aggregate particles coarser than the 1.18 mm sieve was less than 50%
for the good performers and greater than 50% for those exhibiting relatively poor
performance. Although the approach makes it evident that coarser particle DASR
porosities of less than 50% are easier to achieve with coarser gradations, they are also
achieved with properly proportioned fine-graded mixtures. In addition, DASR porosities
less than 50% are not assured with coarse-graded mixtures; they must also be properly
proportioned.
105
According to the analysis of existing database such as Superpave monitoring
project, WesTrack, and NCAT, the approach of the DASR porosity concept exhibits the
reasonable result. The mixture gradations with DASR porosity less than 50% showed
more rut resistance. The marginal mixture gradations exhibited varied results depending
on field gradation and DASR porosity.
From the lab produced mixtures, gradations with good interaction and lower DASR
porosity exhibited good resistance to rutting potentials. On the other hand, gradation
with higher DASR porosity and/or marginal interaction performed poorly.
6.2 Conclusions
Several key conclusions were drawn based on the findings of this study. These
conclusions, which are summarized below, appear to apply to the broad range of mixtures
typically used for roads from fine-graded to SMA:
•
DASR porosity of asphalt mixture, determined using the gradation analysis system
developed in this study, should be less than 50% to ensure coarse aggregate
interlock, which is required for good mixture performance.
•
The relative proportion of contiguous size particles within the DASR must be no
greater than 70/30 to ensure proper interaction (interlock) among the different size
particles in the DASR.
•
Gradation evaluation for asphalt mixture should include a sensitivity analysis to
evaluate the effects of potential changes in gradation on DASR porosity.
Adjustments should be made to JMF’s when accepted gradation variances result in
DASR porosity greater than 50%.
•
Relative proportions between contiguous size aggregates in the DASR should be
significantly lower than 70/30 (e.g., 65/35) when the interaction of these sizes is
critical to maintain the DASR porosity below 50%.
•
Mixtures with excessively low DASR porosity (low IV) should be avoided, as they
may be brittle. However, additional study is necessary to identify specific criteria,
which are likely to depend on other variables like mixture type and characteristics
of the interstitial components.
106
6.3 Recommendations
Research should continue to further develop and refine this very promising
approach to establishing gradation guidelines for mixture performance. Specifically, the
following areas need further development:
•
Effects of aggregate characteristics and properties including shape, angularity and
texture on the criteria identified.
•
Establishment of criteria for minimum interstitial volume (IV) or minimum DASR
porosity for different types of mixture.
•
Develop further understanding of the effects of the interstitial component (IC)
characteristics and properties, which most likely has the greatest effect on fracture
resistance of mixture. This should lead to the identification of criteria and
guidelines for IC characteristics to optimize mixture performance.
APPENDIX A
GRADATIONS FOR SUPERPAVE MONITORING PROJECT
Project 1
Project 2
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-1. Gradations for Project 1 and 2
Group 3
Group 2
Group 1
JMF
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-2. Gradations for Project 3 Layer A
108
⅜"
½"
¾"
109
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#4
4.75
#8
2.36
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-3. Gradations for Project 3 Layer B
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-4. Gradations for Project 4 Layer A
⅜"
½"
¾"
110
Group 3
Group 2
Group 1
JMF
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-5. Gradations for Project 4 Layer B
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-6. Gradations for Project 5 Layer A
⅜"
½"
¾"
111
JMF
Group 1
Group 2
Group 3
100
90
80
70
% Passing
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-7. Gradations for Project 5 Layer B
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-8. Gradations for Project 6
⅜"
½"
¾"
112
Group 3
Group 2
Group 1
JMF
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-9. Gradations for Project 7 Layer A
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-10. Gradations for Project 8 Layer A
⅜"
½"
¾"
113
JMF
Group 1
Group 2
Group 3
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-11. Gradations for Project 8 Layer B
8-1
8-2
8-3
8-4
8-5
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-12. Gradations for Project 8 Plant Mixture
⅜"
½"
¾"
114
9-1A
9-2A
9-3A
9-1B
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
½"
¾"
Sieve Size^0.45 mm
Figure A-13. Gradations for Project 9
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-14. Gradations for Project 10
⅜"
115
11-2A
11-2B
11-3B
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30 #16
1.18
#8
2.36
#4
4.75
⅜"
½"
¾"
Sieve Size^0.45 mm
Figure A-15. Gradations for Project 11
12-1B
12-1A
100
90
80
% Passing
70
60
50
40
30
20
10
0
#100
#30
#16
1.18
#8
2.36
#4
4.75
Sieve Size^0.45 mm
Figure A-16. Gradations for Project 12
⅜"
½"
¾"
APPENDIX B
POROSITY RESULTS FOR SUPERPAVE PROJECTS
100
90
80
Porosity, %
70
60
50
40
30
117
20
10
11-3B
11-2B
11-2A
8-5
8-4
8-3
8-2
8-1
7-3A
7-2A
7-1A
5-3B
5-2B
5-1B
5-3A
5-2A
5-1A
4-3B
4-2B
4-1B
4-3A
4-2A
4-1A
3-3B
3-2B
3-1B
3-3A
3-2A
3-1A
0
Project-Group-Layer
Figure B-1. Porosity Results for Group 1 (Field Gradations for Projects 3, 4, 5, 7, and 8, and Plant-Mix Gradations for Project
11)
100
90
80
Porosity, %
70
60
50
40
30
20
118
10
12-1B
12-1A
9-1B
9-3A
9-2A
9-1A
8-3B
8-2B
8-1B
8-3A
8-2A
8-1A
6-3A
6-2A
6-1A
0
Project-Group-Layer
Figure B-2. Porosity Results for Group 2 (Field Gradation for Projects 6, and Plant-Mix Gradations for Projects 8, and 12)
APPENDIX C
TRAFFIC AND RUT DEPTH DATA FOR SUPERPAVE MONITORING PROJECT
0.35
120
Average Rut Depth, inch
0.30
0.25
Round-I
0.20
Round-II
0.15
Round-III
0.10
0.05
0.00
1
2
3
4
5
6
7
Project
Figure C-1. Cumulative Average Rut Depth for Each Round
8
9
10
11
12
7.0E+06
6.0E+06
ESALs
5.0E+06
4.0E+06
Round-I
3.0E+06
Round-II
Round-III
2.0E+06
121
1.0E+06
0.0E+00
1
2
3
4
5
6
7
Project
Figure C-2. Cumulative ESALs for Each Round
8
9
10
11
12
0.30
6.0E+06
0.25
5.0E+06
0.20
4.0E+06
0.15
3.0E+06
0.10
2.0E+06
0.05
1.0E+06
0.00
0.0E+00
1
2
3
4
5
Project
Figure C-3. Total Rut Depth and ESALs
6
7
8
11
ESALs
7.0E+06
Round-III
ESALs-III
122
Average Rut Depth, inch
0.35
APPENDIX D
THEORETICAL CALCULATION FOR SURFACE AREA
D.1 The Number of Spheres in the Representative Volume
The number of particles for a certain size (n) is based on volume and weight.
n=
total vol. of Agg. at a certain size
representative vol. of a sphere (instead of a real Agg. vol.)
∴ n=
weight of Agg.
W
=
vol. of a sphere × specific gravity of Agg. 4 3
πr × SG
3
(D-1)
where,
W
SG
R
= weight of aggregates
= specific gravity of aggregates
= radius of sphere
Figure D-1 shows the situation that the arbitrary plane cuts through the mixture
which has the representative volume. The number of spheres on this arbitrary plane (n’),
is,
n′ = total no. of shpheres ×
2 × radius of the sphere
2nr
=
height of the representative volume
h
where,
h
= height of the mixture
123
(D-2)
124
R
h
Figure D-1. Mixture Cut Through by an Arbitrary Interstitial Plane
D.2 The Number of Spheres (n) for Each Level (m) of Protrusion
It is considered that half the spheres are protruded less than a hemisphere and other
half are embedded less than a hemisphere on a 2D plane, because an interstitial plane
may follow the shortest way on the surface of spheres. Figure D-2 shows particles on an
interstitial plane.
Figure D-2. Particles on an Interstitial Plane
The hemisphere, which has the maximum protrusion area, can cut by m times on an
arbitrary plane (Figure D-3). This means that there are m types of hemispheres cut by a
plane (Figure D-4). This is also applied to embedded area.
125
Figure D-3. Maximum Protrusion Area (Hemisphere)
m cuts
Figure D-4. m Times Cuts for a Hemisphere
It is assumed that there are the same numbers of particles (c) for each protrusion or
embedment level (m) for the same size of sphere on the same arbitrary plane (interstitial
plane), because the particles are assumed to be distributed uniformly (Figure D-5). In
other words, there is the same number of spheres for the protruded and embedded spheres
with the same protruded or embedded area. Therefore, the surface area on an arbitrary
plane will be the same result to the case that all spheres are only protruded (Figure D-6).
Same area
Same area
Figure D-5. The Case with Protruded and Embedded Spheres on the Plane
126
Same area
Same area
Figure D-6. The Case with Only Protruded Spheres on the Plane
Therefore, the number of particles for each cut (c) is,
c=
n′ 2nr h 2nr
=
=
m
m
h⋅m
(D-3)
Equation D-4 is derived from Equation D-3 to another form,
m
n ′ = ∑ ci
(D-4)
i =1
D.3 The Protruded Surface Area of the Spherical Cap (SAp) over an Interstitial
Plane
The surface area of the spherical cap is given by the Equation D-5 (Figure D-7).
To calculate the surface area of a hemisphere, r is substituted to a. Then the surface area
of the perfect hemisphere (SAhsphere) is equal to 2πr2.
SAcap = 2πra
a
r
Figure D-7. Surface Area of the Spherical Cap
(D-5)
127
Figure D-8 shows the trends of surface area changes for a sphere by m cut. It
shows linear relationship for the hemisphere.
In other words, if there are m types of spherical caps, the average protruded surface
area (SAp) over a plane is same to the area of a half of a hemisphere (i.e., a=r/2).
Therefore, the protruded spherical surface area with hemispheres on an arbitrary 2D
plane of the representative volume is shown in Equation D-6.
SA p = n′ × Acap =
r 2nπr 3
2nr
× 2πr × =
2
h
h
(d-6)
Hemisphere (r=3)
60
Surface Area
50
40
30
20
10
0
0
20
40
60
80
100
120
No. of cuts, m
Figure D-8. Surface Area for m Cut with r = 3
D.4 Key Protrusion Depth for Particles to Act as Anchors
If there is a key particle size that serves as “anchors” that provide sufficient
interlock to prevent excessive shear deformation, this size may not be represented by a
radius or diameter of a particle, but protruded or embedded depth. If the protrusion depth
of a particle is less than the key protrusion depth (d) even though a particle has larger
radius, then it can not work as “anchor” in the shear plane. In Figure D-9, one particle
128
has more protruded than the key protrusion depth, but another does not. To work as
“anchor”, particles should protrude greater than the key protrusion depth.
The number of particles on an arbitrary plane, n’, is calculated in Equation D-2. n’
particles include m types of a hemisphere cut by a plane. Therefore, the probability (Pd),
that protruded or embedded depth for particles is larger than the key protrusion depth is,
Pd =
r−d
r
(D-7)
d
d
d = key protrusion depth
Figure D-9. Example of the Protruded or Embedded Depth of Particles
Then the number of particles (nd) that are protruded or embedded over the key
particle size is,
nd = n' × Pd =
2nr ⎛ r − d ⎞ 2n(r − d )
×⎜
⎟=
h
h
⎝ r ⎠
(D-8)
D.5 The Number of Prolate Spheroids of Various Level-Cuts (m) for Each Size of
Particle
It is considered that an aggregate is a prolate spheroid. Prolate spheroid has same
length between b on y-axis and c on z-axis (b=c), or a on x-axis and c on z-axis (a=c).
There are two types of prolate spheroids. The one has longer a than b (a>b) and the other
has longer b than a (a<b) (Figure D-10). The latter is just considered because they have
same results for the surface area except for the number of particles on the representative
129
plane at any cut. However, the difference between the numbers of particles for the case
of a>b and a<b is in the ratio of a to b. Other assumptions are same to the case of sphere.
b
c
z
x
a
a > b=c
a=c < b
Figure D-10. Different Types of Prolate Spheroid
At first, the number of spheroids on an arbitrary plane, n’, is,
n′ = total no. of shperoids ×
2 × height of a shperoid
2nb
=
height of the representative volume
h
(D-9)
Therefore, the number of the half of a spheroid for each cut (cs) is,
cs =
n ′ 2nb h 2nb
=
=
m
m
h⋅m
(D-10)
Equation D-10 is the same expression to,
m
n′ = ∑ c s i
(D-11)
i =1
The surface area of a spheroid is not simple as the case of sphere. The surface area
for the case of a<b (longer length on y-axis), SA’, is as follows:
⎡
⎛ b2 − a2
ab 2
SA' = 2π ⎢a 2 +
sin −1 ⎜
⎜
b
⎢⎣
b2 − a2
⎝
⎞⎤
⎟⎥
⎟⎥
⎠⎦
(D-12)
Therefore, to calculate the surface area of the cap for a spheroid (SAs), the integral
equation for the surface area of a spheroid should be modified by a different boundary.
0~π/2 is used for the entire spheroid, but h~π/2 should be used for the cap of the
spheroid.
130
⎡
⎛ b2 − a2
ab 2
SAs = π ⎢a 2 +
sin −1 ⎜
⎜
b
⎢⎣
b2 − a2
⎝
2
⎡
⎞⎤
⎛ h b2 − a2
⎟⎥ − π ⎢a 2 h + ab
sin −1 ⎜
⎟⎥
⎜
b
⎢⎣
b2 − a2
⎠⎦
⎝
⎞⎤
⎟⎥
⎟⎥
⎠⎦
(D-13)
The average surface area of m types of caps of a spheroid is not exactly same to the
half of half-spheroid (the quarter of a spheroid), but almost same. The below figures
show the trends of surface area changes for a spheroid by m cut. Figure D-11 shows
almost linear change for the spheroid.
The average spheroidal surface area (SAps) can be derived from Equation D-13.
When there are m types of spheroidal caps, the average protrusion (and embedded) area
on a plane is almost same to the area of a half of a spheroid. Therefore, the average
surface area on an arbitrary 2D plane of the representative volume can be derived.
SA ps ≈
⎡
⎛ b2 − a2
n′
ab 2
× 2π ⎢a 2 +
sin −1 ⎜
⎜
4
b
⎢⎣
b2 − a2
⎝
⎞⎤ n′
⎟⎥ = SA
⎟⎥ 4 s
⎠⎦
Prolate Spheroid (a=1, b=2)
12
Surface Area
10
8
6
4
2
0
0
20
40
60
No. of cuts, m
Figure D-11. Surface Area for m Cut with a = 1, b = 2
80
100
120
(D-14)
APPENDIX E
LABORATORY MIXTURES INFOMATION
Table E-1. Blending Percent for GA Granite Gradations
Gradation
Percent Passing for Blend
Type
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
GA-Bad-2
35.0
23.0
12.0
30.0
GA-Bad-3
26.5
25.4
20.8
27.3
GA-48
26.7
19.8
35.2
18.3
GA-Good
33.0
7.0
50.0
10.0
Table E-2. Blending Percent for Rinker South FL Limestone
Gradation
Percent Passing for Blend
Type
# 67 Stone
S-1-B
Med. Screenings
Local Sand
FL-Bad-2
18.0
45.0
17.0
20.0
FL-Bad-3
13.3
29.5
47.8
9.4
FL-48
10.6
43.1
38.8
7.5
FL-Good
10.0
45.0
42.0
3.0
131
132
Table E-3. JMF for GA Granite Gradation
Sieve, mm
Sieve Size
GA-Bad-2
GA-Bad-3
GA-Good
GA-48
19.0
3/4”
100.00
100.00
100.00
100.00
12.5
1/2”
98.95
99.21
99.01
99.20
9.5
3/8”
85.58
89.06
86.45
89.00
4.75
#4
52.05
58.11
65.07
61.84
2.36
#8
40.72
43.94
46.60
44.80
1.18
# 16
36.20
37.07
31.80
34.01
0.600
# 30
32.13
31.65
22.63
26.73
0.300
# 50
18.40
18.32
13.70
15.80
0.150
# 100
5.08
5.60
6.50
6.00
0.075
# 200
2.32
2.79
4.20
3.48
0
Pan
0.00
0.0
0.00
0.00
Table E-4. JMF for Rinker South FL Limestone
Sieve, mm
Sieve Size
FL-Bad-2
FL-Bad-3
FL-Good
FL-48
19.0
3/4”
100.00
100.00
100.00
100.00
12.5
1/2”
94.24
95.74
96.80
96.61
9.5
3/8”
84.79
89.10
89.75
89.55
4.75
#4
56.53
70.09
64.05
64.61
2.36
#8
39.45
54.83
44.38
45.90
1.18
# 16
33.73
43.01
33.24
35.54
0.600
# 30
29.02
33.54
25.05
27.67
0.300
# 50
18.44
23.00
17.94
19.17
0.150
# 100
6.47
9.01
7.86
7.87
0.075
# 200
2.82
3.17
3.06
3.04
0
Pan
0.00
0.0
0.00
0.00
133
Table E-5. Batch Weight for Granite Gradations
GA-2-Bad
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
1,575
2,610
3,150
12.5
1/2"
47
1,575
2,610
3,150
9.5
3/8"
646
1,578
2,610
3,150
4.75
#4
1,433
2,300
2,610
3,150
2.36
#8
1,512
2,569
2,772
3,150
1.18
# 16
1,544
2,589
2,923
3,150
0.600
# 30
1,544
2,600
3,015
3,231
0.300
# 50
1,559
2,600
3,064
3,785
0.150
# 100
1,559
2,600
3,096
4,352
0.075
# 200
1,559
2,600
3,112
4,460
0
Pan
1,575
2,610
3,150
4,500
2.809
2.799
2.770
2.626
1,575
1,035
540
1,350
Gsb
GA-3-Bad
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
1,193
2,336
3,272
12.5
1/2"
36
1,193
2,336
3,272
9.5
3/8"
489
1,196
2,336
3,272
4.75
#4
1,085
1,993
2,336
3,272
2.36
#8
1,145
2,290
2,616
3,272
1.18
# 16
1,169
2,313
2,878
3,272
0.600
# 30
1,169
2,324
3,038
3,345
0.300
# 50
1,181
2,324
3,122
3,849
0.150
# 100
1,181
2,324
3,178
4,365
0.075
# 200
1,181
2,324
3,206
4,463
0
Pan
1,193
2,336
3,272
4,500
2.809
2.799
2.770
2.626
1,193
1,143
936
1,229
Gsb
134
Table E-4. Continued
GA-48
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
1,202
2,093
3,677
12.5
1/2"
36
1,202
2,093
3,677
9.5
3/8"
493
1,204
2,093
3,677
4.75
#4
1,093
1,825
2,093
3,677
2.36
#8
1,153
2,057
2,568
3,677
1.18
# 16
1,177
2,075
3,011
3,677
0.600
# 30
1,177
2,084
3,281
3,726
0.300
# 50
1,189
2,084
3,423
4,064
0.150
# 100
1,189
2,084
3,518
4,409
0.075
# 200
1,189
2,084
3,566
4,475
0
Pan
1,202
2,093
3,677
4,500
2.809
2.799
2.770
2.626
1,202
891
1,584
824
Gsb
GA-Good
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
1,485
1,800
4,050
12.5
1/2"
45
1,485
1,800
4,050
9.5
3/8"
609
1,486
1,800
4,050
4.75
#4
1,351
1,706
1,800
4,050
2.36
#8
1,426
1,787
2,475
4,050
1.18
# 16
1,455
1,794
3,105
4,050
0.600
# 30
1,455
1,797
3,488
4,077
0.300
# 50
1,470
1,797
3,690
4,262
0.150
# 100
1,470
1,797
3,825
4,451
0.075
# 200
1,470
1,797
3,893
4,487
0
Pan
1,485
1,800
4,050
4,500
2.809
2.799
2.770
2.626
1,485
315
2,250
450
Gsb
135
Table E-6. Batch Weight for Limestone Gradations
FL-2-Bad
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
259
810
2,835
3,600
12.5
1/2"
259
810
2,835
3,600
9.5
3/8"
502
992
2,835
3,600
4.75
#4
761
2,005
2,835
3,600
2.36
#8
778
2,673
2,919
3,600
1.18
# 16
786
2,754
3,087
3,600
0.600
# 30
786
2,774
3,225
3,654
0.300
# 50
786
2,774
3,332
4,023
0.150
# 100
786
2,774
3,493
4,401
0.075
# 200
790
2,784
3,571
4,473
0
Pan
810
2,835
3,600
4,500
2.335
2.339
2.471
2.626
810
2,025
765
900
Gsb
FL-3-Bad
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
599
1,926
4,077
12.5
1/2"
192
599
1,926
4,077
9.5
3/8"
371
718
1,926
4,077
4.75
#4
563
1,382
1,926
4,077
2.36
#8
575
1,820
2,163
4,077
1.18
# 16
581
1,873
2,636
4,077
0.600
# 30
581
1,886
3,023
4,102
0.300
# 50
581
1,886
3,324
4,276
0.150
# 100
581
1,886
3,776
4,453
0.075
# 200
584
1,893
3,995
4,487
0
Pan
599
1,926
4,077
4,500
2.335
2.339
2.471
2.626
599
1,328
2,151
423
Gsb
136
Table E-6. Continued
FL-48
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
450
2,475
4,365
12.5
1/2"
144
450
2,475
4,365
9.5
3/8"
279
632
2,475
4,365
4.75
#4
423
1,645
2,475
4,365
2.36
#8
432
2,313
2,683
4,365
1.18
# 16
437
2,394
3,099
4,365
0.600
# 30
437
2,414
3,439
4,373
0.300
# 50
437
2,414
3,704
4,428
0.150
# 100
437
2,414
4,100
4,485
0.075
# 200
439
2,424
4,293
4,496
0
Pan
450
2,475
4,365
4,500
2.335
2.339
2.471
2.626
450
2,025
1,890
135
Gsb
FL-Good
Retained Weight, g
Sieve, mm
Sieve Size
# 78 Stone
# 89 Stone
W-10 Screenings
Local Sand
19.0
3/4"
0
477
2,417
4,163
12.5
1/2"
153
477
2,417
4,163
9.5
3/8"
296
652
2,417
4,163
4.75
#4
448
1,621
2,417
4,163
2.36
#8
458
2,261
2,609
4,163
1.18
# 16
463
2,339
2,993
4,163
0.600
# 30
463
2,358
3,307
4,183
0.300
# 50
463
2,358
3,551
4,321
0.150
# 100
463
2,358
3,918
4,463
0.075
# 200
465
2,368
4,096
4,490
0
Pan
477
2,417
4,163
4,500
2.335
2.339
2.471
2.626
477
1,940
1,746
338
Gsb
137
Table E-7. DRD, ARD, and Area Change Results from APA
Sample
ID
FL-2B
FL-3B
FL-48
FL-G
GA-2B
GA-3B
GA-48
GA-G
Sample ID
DRD
DRD
ARD
Average
F-2B-1
F-2B-2
F-2B-3
F-3B-1
F-3B-2
F-3B-3
F-48-7%
F-48-1
F-48-2
F-G-7%
F-G-1
F-G-2
G-2B-7%
G-2B-1
G-2B-2
G-3B-1
G-3B-7%
G-3B-2
G-48-7%
G-48-Try
G-48-1
G-48-2
G-G-1
G-G-7%
G-G-2
G-G-3
(mm)
9.2
8.5
7.6
8.9
10.5
8.6
8.1
6.5
8.1
5.2
6.9
6.7
17.7
14.9
18.7
14.3
16.1
15.8
10.5
7.8
11.2
11.7
8.4
9.9
6.8
9.4
(mm)
8.42
9.33
7.56
6.30
17.09
15.42
10.29
8.63
ARD
Average
(mm)
6.4
6.1
4.9
6.4
6.7
6.2
6.0
4.1
4.8
3.9
4.4
4.7
11.2
9.6
9.9
9.4
10.0
9.3
5.2
5.7
6.1
7.3
5.6
5.4
4.4
5.8
(mm)
5.76
6.43
4.94
4.32
10.24
9.54
6.11
5.29
Area
Change
(%)
-1.36
1.50
0.05
-1.38
-0.09
-1.00
-1.87
1.50
0.00
-1.55
-0.14
-1.02
-0.62
-1.33
3.51
0.36
1.70
0.09
-1.76
-1.10
2.56
0.16
-0.28
0.57
0.32
0.26
Area
Change
Average
(%)
0.06
-0.82
-0.12
-0.91
0.52
0.72
-0.03
0.22
Old
System
(mm)
5.68
5.67
4.05
5.79
5.77
5.15
4.69
4.28
6.92
4.00
4.36
3.75
9.35
8.40
8.71
8.25
8.81
8.54
4.46
5.24
6.23
5.57
4.86
7.08
4.64
5.21
138
Table E-8. Student T-test Results for DRD
DRD
F-2B
F-3B
F-48
F-G
G-2B
G-3B
G-48
DRD
F-2B
F-3B
F-48
F-G
G-2B
G-3B
G-48
t value
F-3B
-1.210
p value
F-3B
0.290
F-48
1.200
2.200
F-G
2.990
3.800
1.680
G-2B
-6.980
-5.980
-7.540
-8.520
G-3B
-9.440
-7.340
-10.000
-11.700
1.300
G-48
-1.700
-0.836
-2.430
-3.560
4.820
4.510
G-G
-0.236
0.736
-1.150
-2.510
6.700
7.170
1.500
F-48
0.300
0.092
F-G
0.040
0.019
0.170
G-2B
0.0022
0.0039
0.0017
0.0010
G-3B
0.0007
0.0018
0.0006
0.0003
0.260
G-48
0.150
0.440
0.590
0.016
0.0048
0.0063
G-G
0.820
0.500
0.300
0.054
0.0011
0.0008
0.190
Table E-9. Student T-test Results for ARD
ARD
F-2B
F-3B
F-48
F-G
G-2B
G-3B
G-48
ARD
F-2B
F-3B
F-48
F-G
G-2B
G-3B
G-48
t value
F-3B
-1.410
p value
F-3B
0.230
F-48
1.160
2.630
F-G
2.870
7.960
1.050
G-2B
-6.670
-7.370
-7.160
-10.800
G-3B
-7.440
-11.400
-7.730
-16.200
1.280
G-48
-0.528
0.585
-1.660
-3.160
6.100
6.030
G-G
0.913
3.080
-0.606
-2.450
9.100
10.600
1.510
F-48
0.310
0.058
F-G
0.046
0.0014
0.350
G-2B
0.0026
0.0018
0.0020
0.0004
G-3B
0.0017
0.0003
0.0015
0.0001
0.270
G-48
0.620
0.580
0.160
0.025
0.0017
0.0018
G-G
0.400
0.028
0.570
0.058
0.0003
0.0001
0.180
LIST OF REFERENCES
Asphalt Institute, “Superpave Mix Design,” Superpave Series No. 2 (SP-02). Asphalt
Institute, Lexington, KY, 2001.
Asphalt Institute and the Heritage Group, “The Bailey Method: Achieving Volumetrics
and HMA Compactability,” Asphalt Institute Educational Courses and
Seminars, Lexington, KY, 2005.
Birgisson, B., Darku, D., Roque, R., and Page, G., “The Need for Inducing Shear
Instability to Obtain Relevant Parameters for HMA Rut-Resistance,” Journal of
Association of Asphalt Paving Technologists, Baton Rouge, LA, Vol. 73, 2004, pp.
23-52.
Birgisson, B., Roque, R., and Ruth, B.E. “SuperpaveTM Gyratory Compactor with Shear
Measurements as an Index Test for Instability Rutting Potential of Mixtures,”
Proceedings, Canadian Technical Asphalt Association, 2003
Birgisson, B., and Ruth, B.E., “Development of Tentative Guidelines for the Selection of
Aggregate Gradations for Hot-Mix Asphalt,” American Society for Testing and
Materials, STP 1412, 2001.
Brown, E.R., Cooley, L.A. Jr., Hanson, D., Lynn, C., Powell, B., Powell, B., and Watson,
D., “NCAT Test Track Design, Construction, and Performance,” NCAT Report
2002-12, National Ceter for Asphalt Technology, 2002.
Brown, E.R., Prowell, B., Cooley, A., Zhang, J., and Powell R.B., “Evaluation of Rutting
Performance on The 2000 NCAT Test Track,” Journal of Association of Asphalt
Paving Technologists, Baton Rouge, LA, Vol. 73, 2004.
Chowdhury, A., Grau, J.D. C., Button, J.W., and Little, D.N., “Effect of Aggregate
Gradation on Permanent Deformation of Superpave HMA,” the 80th Annual
Meeting of Transportation Research Board, Washington, D.C., 2001.
Coree, B.J., and Hislop, W.P., “A Laboratory Investigation into the Effective of
Aggregate-Related Factors of Critical VMA in Asphalt Paving Mixtures,” Iowa
DOT project TR-415, Ames, IA, 2000.
Drakos, C., “Identification of a Physical Model to Evaluate Rutting Performance of
Asphalt Mixtures”, Ph D Dissertation, University of Florida, 2003.
139
140
Drakos, C., Roque, R., and Birgisson, B., “Effects of Measured Tire Contact Stresses on
Near Surface Rutting,” Transportation Research Record No. 1764, Transportation
Research Board, Washington, DC, 2001, pp. 59-69.
Drakos, C., Roque, R., Birgisson, B., and Novac, M., “Identification of a Physical Model
to Evaluate Rutting Performance of Asphalt Mixtures,” Journal of ASTM
International, American Society for Testing and Materials, Volume 2, Issue 3,
March 2005
Epps, J.A., Leahy, R.B., Mitchell, T., Ashmore, C., Seeds, S., Alavi, S., and Monismith,
C.L., “WESTRACK: The Road to Performance-Related Specifications,”
International Conference on Accelerated Pavement Testing, Reno, NV, 1999.
Epps, J.A., Monismith, C.L., Seeds, S.B., Ashmore, S.C., and, Mitchell, T.M.
“WESTRACK Full-Scale Test Track: Interim Findings,”
http://www.westrack.com/isap.pdf, International Symposium on Asphalt Pavement,
Seattle, WA, 1997.
Epps, J.A., Hand, A., Seeds,S., Schulz, T., Alavi, S., Ashmore, C., Monismith, C.L.,
Deacon, J.A., Harvey, J.T., and Leahy, R., “Recommended Performance-Related
Specification for Hot-Mix Asphalt Construction: Results of the WesTrack Project,”
NCHRP Report 455, National Cooperative Highway Research Program, National
Academies Press, Washington, DC, 2002.
Freeze, A.R., and Cherry, J.A., “Groundwater,” Prentice Hall, 1979.
Gardiner, M.S., and Brown, E.R., “Segregation in Hot-Mix Asphalt Pavements,” NCHRP
Report 441, Transportation Research Board, Washington, D.C., 2000.
Kandhal, P.S., and Cooley, L.A., Jr., “Coarse versus Fine-Graded Superpave Mixtures:
Comparative Evaluation of Resistance to Rutting,” NCAT Report 2002-02, National
Center for Asphalt Technology, 2002.
Kandhal, P.S., Foo, K.Y., and Malick, R.B., “Critical Review of VMA requirements in
Superpave,” NCAT Report 98-1, National Ceter for Asphalt Technology, 1998.
Kandhal, P.S., and Malick, R.B., “Effect of Mix Gradation on Rutting Potential of
Graded Asphalt Mixtures,” the 80th Annual Meeting of Transportation Research
Board, Washington, D.C., 2001.
Lambe, T.W., and Whitman R.V., “Soil Mechanics,” John Wiley & Sons, New York,
1969.
Nukunya, B., Roque, R., Tia, M., and Birgisson, B., “Evaluation of VMA and Other
Volumetric Properties as Criteria for the Design and Acceptance of Superpave
Mixture,” Journal of Association of Asphalt Paving Technologists, Clearwater, FL,
Vol. 70, 2001.
141
Roque, R., Huang, S.C., and Ruth, B.E., “Maximizing Shear Resistance of Asphalt
Mixtures by Proper Selection of Aggregate Gradation,” International Society for
Asphalt Pavements, Seattle, WA, 1997, pp. 249-268.
Ruth, B.E., Roque, R., and Nukunya, B., “Aggregate Gradation Characterization Factors
and Their Relationships to Fracture Energy and Failure Strain of Asphalt
Mixtures,” Journal of Association of Asphalt Paving Technologists, Colorado
springs, CO, Vol. 71, 2002.
Vavrik, W.R., Huber, G., Pine, W.J., Carpenter, S.H., and Bailey, R., “Bailey Method for
Gradation Selection in HMA Mixture Design,” Transportation Research Circular,
E-C044, Washington, DC, 2002.
Vavrik, William R., Pine, William J., Huber, Gerald, and Carpenter, Samuel H., “The
Bailey Method of Gradation Evaluation: The Influence of Aggregate Gradation and
Packing Characteristics on Voids in the Mineral Aggregate,” Journal of Association
of Asphalt Paving Technologists, Clearwater, FL, Vol. 70, 2001, pp. 132-175.
BIOGRAPHICAL SKETCH
Sungho Kim was born in Daegu, Republic of Korea, on January 27, 1973, to
Byungtae Kim and Sangyeon Lee. He received a Bachelor of Science degree in
transportation engineering from Hanyang University in 1995, and then joined the Korean
Air Forces where he served as an intelligence and security officer for three years. After
finishing his M.S. in pavement research of Hanyang University in, he decided to come
across the Pacific to have more advanced experience in United State.
Sungho started working as a research assistant for Dr. Reynaldo Roque in 2001.
While enrolled as a student at the University of Florida, he also met his fiancée in 2002
and married in 2005. After completing his Ph.D., he plans to work in academia,
government agencies, or private companies in Civil Engineering to continue his service
to the community.
142