Florida State University Librariesdiginole.lib.fsu.edu/islandora/object/fsu:168556/... · florida...
Transcript of Florida State University Librariesdiginole.lib.fsu.edu/islandora/object/fsu:168556/... · florida...
Florida State University Libraries
Electronic Theses, Treatises and Dissertations The Graduate School
2009
Evaluation of Engineering Properties of HotMix Asphalt Concrete for the Mechanistic-Empirical Pavement DesignYuan Xiao
Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected]
FLORIDA STATE UNIVERSITY
COLLEGE OF ENGINEERING
EVALUATION OF ENGINEERING PROPERTIES OF HOT MIX
ASPHALT CONCRETE FOR THE MECHANISTIC-EMPIRICAL
PAVEMENT DESIGN
By
YUAN XIAO
A Dissertation submitted to the Department of Civil & Environmental Engineering
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Degree Awarded: Spring Semester, 2009
ii
The members of the Committee approve the Dissertation of Yuan Xiao defended on January 30, 2009.
__________________________ Wei-Chou V. Ping Professor Directing Dissertation __________________________ Xufeng Niu Outside Committee Member __________________________ Tarek Abichou Committee Member __________________________ John Sobanjo Committee Member
Approved: _____________________________________________________________ Kamal Tawfiq, Chair, Department of Civil & Environmental Engineering
_____________________________________________________________ Ching-Jen Chen, Dean, College of Engineering
The Graduate School has verified and approved the above named committee members.
iii
ACKNOWLEDGEMENTS
I would like to express my sincere gratitude to my advisor, Dr. W. Virgil Ping, for his
guidance, support and patience during the whole study of my Ph.D. program. My
special thanks go to Ms. Ginger Ling for her generous help in every stage of my work. I
also wish to thank all other members of my committee: Dr. Xufeng Niu, Dr. Tarek
Abichou, and Dr. John Sobanjo for their friendly encouragement and valuable advice.
I extend many thanks to my colleague, Mr. Ed Mallory, who provided unconditional
support during the whole experimental program for this research. I would also like to
thank the engineers and staff from Florida Department of Transportation and other
contracting companies who offered financial and contractual support.
iv
TABLE OF CONTENTS
LIST OF TABLES................................................................................................................ vii
LIST OF FIGURES ............................................................................................................... ix
ABSTRACT ...................................................................................................................... xv
CHAPTER 1 INTRODUCTION............................................................................................ 1
1.1 Background................................................................................................................. 1
1.2 Objectives ................................................................................................................... 3
1.3 Dissertation Outline.................................................................................................... 4
CHAPTER 2 LITERATURE REVIEW................................................................................. 6
2.1 Introduction ................................................................................................................ 6
2.2 Asphalt Cement Properties ......................................................................................... 6
2.3 Hot Mix Asphalt (HMA) Mixture Design................................................................ 21
2.4 Mechanical Tests for Characterization of Asphalt Mixtures.................................... 35
2.5 HMA Fracture Mechanics Concepts ........................................................................ 47
CHAPTER 3 MATERIALS AND EXPERIMENTAL PROGRAM ................................... 54
3.1 General...................................................................................................................... 54
3.2 Mix Designs and Materials....................................................................................... 55
3.3 SBS Polymer-modified Asphalt Binder.................................................................... 56
3.4 Aggregates Gradation Modification ......................................................................... 60
3.5 Specimen Preparation and Volumetric Properties.................................................... 60
3.6 Test Procedures......................................................................................................... 64
3.7 Testing Program ....................................................................................................... 70
CHAPTER 4 FRACTURE MECHANICS PROPERTIES FROM IDT .............................. 73
4.1 Resilient Modulus Testing Procedures and Results ................................................. 73
v
4.2 Creep Compliance Testing Procedures and Results ................................................. 76
4.3 Tensile Strength Testing Procedures and Results..................................................... 77
CHAPTER 5 EVALUATION OF FRACTURE MECHANICS PROPERTIES ................. 81
5.1 Evaluation of Gradation Effects ............................................................................... 81
5.2 Evaluation of SBS Polymer-modified Binder Effects .............................................. 92
5.3 Effect of Aggregate Type........................................................................................ 108
5.4 Summary of Analysis and Findings from Fracture Mechanics Tests ..................... 113
CHAPTER 6 COMPLEX MODULUS AND RESILIENT MODULUS TEST RESULTS.................................................................................................................... 116
6.1 Test Procedures....................................................................................................... 116
6.2 Presentation of DMT and IDT Testing Results ...................................................... 118
CHAPTER 7 CORRELATION OF INDIRECT TENSION RESILIENT MODULUS AND COMPLEX MODULUS TEST RESULTS................................................ 121
7.1 General.................................................................................................................... 121
7.2 HMA Master Curve Development ......................................................................... 121
7.3 Master Curve Construction..................................................................................... 123
7.4 Verification of Dynamic Complex Modulus Experimental Results....................... 126
7.5 Comparison between Resilient Modulus and Dynamic Modulus .......................... 128
CHAPTER 8 SUMMARY AND CONCLUSIONS........................................................... 134
8.1 Summary................................................................................................................. 134
8.2 Findings and Conclusions....................................................................................... 135
8.3 Recommendations .................................................................................................. 138
APPENDIX A MATERIALS AND MIX DESIGNS......................................................... 139
APPENDIX B CREEP COMPLIANCE TEST RESULTS................................................ 155
APPENDIX C TEST RESULTS FOR IDT AND DMT .................................................... 163
vi
REFERENCES................................................................................................................... 174
BIOGRAPHICAL SKETCH .............................................................................................. 183
vii
LIST OF TABLES
Table 2-1: Generic classification of asphalt additives and modifiers (Roberts et al. 1996) ........... 9
Table 2-2: Chosen unit weight ranges by mix type....................................................................... 30
Table 2-3: Recommended aggregate ratios................................................................................... 31
Table 3-1: Number of specimens prepared for fracture mechanics tests ...................................... 62
Table 3-2: Specimens tested for fracture mechanics properties.................................................... 62
Table 3-3: Specific gravities and air voids of the mixtures .......................................................... 63
Table 4-1: Resilient modulus test results at -10˚C........................................................................ 75
Table 4-2: Resilient modulus test results at 5˚C ........................................................................... 75
Table 4-3: Resilient modulus test results at 25˚C ......................................................................... 76
Table 4-4: Resilient modulus test results at 40˚C ......................................................................... 76
Table 4-5: Tensile strength test results for F2 series mixtures ..................................................... 79
Table 4-6: Tensile strength test results for F4 series mixtures ..................................................... 80
Table 5-1: Power law regression coefficients for modified gradation mixes ............................... 83
Table 5-2: Power model regression coefficients for modified gradation tests.............................. 86
Table 5-3: Power model regression coefficients for PMA mixture tests ...................................... 95
Table 6-1: Cycles for DTM test sequence................................................................................... 118
Table A-1: Superpave mix designs sorted by test series............................................................. 139
Table A-2: Performance grade binder grading report ................................................................. 140
Table A-3: Summary of mix designs and aggregates ................................................................. 141
Table A-4: Aggregate gradations for series 1 - 5 ........................................................................ 142
Table A-5: Aggregate gradations for series 6 - 10 ...................................................................... 142
Table A-6: Aggregate gradations for series 11 - 15 .................................................................... 143
Table A-7: Aggregate gradations for series 16 - 20 .................................................................... 143
Table A-8: Lab analysis report for 0.0% polymer base asphalt (Graded as PG67-22) ............... 148
viii
Table A-9: Lab analysis report for 3.0% polymer asphalt (Graded as PG76-22) ....................... 149
Table A-10: Lab analysis report for 4.5% polymer asphalt (Graded as PG82-22) ..................... 150
Table A-11: Lab analysis report for 6.0% polymer asphalt (Graded as PG82-28) ..................... 151
Table A-12: Gradations for F2C and its adjustments ................................................................. 152
Table A-13: Gradations for F4C and its adjustments ................................................................. 152
Table A-14: Volumetric properties of mixture design series 1 - 5 ............................................. 153
Table A-15: Volumetric properties of mixture design series 6 - 10 ........................................... 153
Table A-16: Volumetric properties of mixture design series 11 - 15 ......................................... 154
Table A-17: Volumetric properties of mixture design series 16 - 20 ......................................... 154
Table B-1: Creep compliance test results at -10˚C (1/GPa)........................................................ 155
Table B-2: Creep compliance test results at 5˚C (1/GPa)........................................................... 156
Table B-3: Creep compliance test results at 25˚C (1/GPa)......................................................... 156
Table B-4: Creep compliance test results at 40˚C (1/GPa)......................................................... 157
Table C-1: Summary of resilient modulus and Poisson’s Ratio test results ............................... 163
Table C-2: Summary of dynamic modulus testing results .......................................................... 170
Table C-3: Summary of phase angle testing results.................................................................... 172
ix
LIST OF FIGURES
Figure 2-1: Polymer classifications based on link structure ......................................................... 11
Figure 2-2: SBS polymer modifier structure................................................................................. 11
Figure 2-3: Complex modulus of SBS-modified asphalt at 60°C (Chen et al. 2002)................... 13
Figure 2-4: Relationship between the observed critical cracking temperature (Tcr) and SBS polymer concentration (Collins et al. 1991).................................................................................. 15
Figure 2-5: Comparison of the rut depths measured on sections with PMA and the companion sections without PMA mixtures (Quintus et al. 2007).................................................................. 17
Figure 2-6: Components of complex modulus G* ........................................................................ 18
Figure 2-7: Viscous and elastic behavior of asphalt binders......................................................... 18
Figure 2-8: Dynamic shear rheometer........................................................................................... 19
Figure 2-9: Stress-strain response of viscoelastic material ........................................................... 19
Figure 2-10: Typical aggregate gradations.................................................................................... 24
Figure 2-11: The four principles of Bailey method for coarse-graded mix .................................. 29
Figure 2-12: Combined blend evaluation for coarse-graded mixes. ............................................. 30
Figure 2-13: Combined blend evaluation for fine-graded mixes. ................................................. 30
Figure 2-14: Illustration of gradation requirements for 12.5 mm (1/2 in.) nominal size .............. 34
Figure 2-15: Conceptual schematic of dynamic complex modulus test ....................................... 40
Figure 2-16: Typical test set-up for dynamic complex modulus .................................................. 41
Figure 2-17: The schematic components of dynamic complex modulus test ............................... 41
Figure 2-18: Kelvin model under sinusoidal loading.................................................................... 42
Figure 2-19: Indirect diametral test during loading and at failure ................................................ 43
Figure 2-20: Theoretical stress distribution on horizontal diametral plane for indirect tensile test (After Yoder et al. 1975)............................................................................................................... 45
Figure 2-21: Theoretical stress distribution on vertical diametral plane for indirect tensile test (After Yoder et al. 1975)............................................................................................................... 45
Figure 2-22: Illustration of potential loading condition (Roque et al. 2002) ................................ 50
x
Figure 2-23: Determination of fracture energy and dissipated creep strain energy ...................... 50
Figure 3-1: Gradation curves for F2 and its trial adjustments ...................................................... 58
Figure 3-2: Gradation curves for F4 and its trial adjustments ...................................................... 58
Figure 3-3: Change of percent retained on top 3 sieves for F2 series........................................... 59
Figure 3-4: Change of percent retained on top 3 sieves for F4 series........................................... 59
Figure 3-5: Cutting of raw specimen ............................................................................................ 59
Figure 3-6: Coring of the Superpave specimen............................................................................. 63
Figure 3-7: Cutting of the dynamic modulus specimen ................................................................ 63
Figure 3-8: Indirect Diametral Resilient Modulus Test Setup...................................................... 65
Figure 3-9: Load & deformations in a typical resilient modulus test............................................ 66
Figure 3-10: Load and deformation curves of creep compliance test ........................................... 67
Figure 3-11: Specimen fails after tensile strength test .................................................................. 68
Figure 3-12: Dynamic complex modulus test setup...................................................................... 69
Figure 3-13: Flowchart of the experimental program for measuring fracture mechanics properties of HMA mixtures.......................................................................................................................... 71
Figure 3-14: Flowchart of the testing program for MR vs. E* ....................................................... 72
Figure 4-1: Instantaneous and total resilient deformations ........................................................... 74
Figure 4-2: Determination of fracture energy and dissipated creep strain energy ........................ 78
Figure 5-1: Power law regression for F2 gradation series ............................................................ 82
Figure 5-2: Power law regression for F4 gradation series ............................................................ 83
Figure 5-3: Resilient modulus for mixtures with modified gradations......................................... 84
Figure 5-4: Comparison of resilient modulus between control and modified gradations............. 85
Figure 5-5: Power model parameter D1 for modified gradations.................................................. 86
Figure 5-6: Power model parameter m for modified gradations................................................... 86
Figure 5-7: Comparison of creep compliance for granite gradation series ................................... 87
Figure 5-8: Comparison of creep compliance for limestone gradation series............................... 87
xi
Figure 5-9: Tensile strength for control and modified gradation mixes ....................................... 89
Figure 5-10: Comparison of TS between control and modified gradation mixes......................... 89
Figure 5-11: Fracture Energy for modified gradation mixes ........................................................ 90
Figure 5-12: DCSE for modified gradation mixes........................................................................ 90
Figure 5-13: Comparison of Fracture Energy for modified gradation mixtures ........................... 91
Figure 5-14: Comparison of DCSE for modified gradation mixtures .......................................... 91
Figure 5-15: Comparison of resilient modulus for F2 SBS PMA mixes ...................................... 93
Figure 5-16: Comparison of resilient modulus for F4 SBS PMA mixes ...................................... 93
Figure 5-17: Comparison of MR between control and PMA mixtures ......................................... 94
Figure 5-18: Power model parameter D1 for mixes with SBS PMA ............................................ 96
Figure 5-19: Power model parameter m for mixes with SBS PMA ............................................. 96
Figure 5-20: Creep compliance master curves for granite PMA mixtures ................................... 97
Figure 5-21: Creep compliance master curves for limestone PMA mixtures............................... 97
Figure 5-22: Comparison of creep compliance at -10˚C for F2 series.......................................... 98
Figure 5-23: Comparison of creep compliance at 5˚C for F2 series ............................................. 98
Figure 5-24: Comparison of creep compliance at 25˚C for F2 series ........................................... 99
Figure 5-25: Comparison of creep compliance at 40˚C for F2 series ........................................... 99
Figure 5-26: Comparison of creep compliance at -10˚C for F4 series........................................ 100
Figure 5-27: Comparison of creep compliance at 5˚C for F4 series ........................................... 100
Figure 5-28: Comparison of creep compliance at 25˚C for F4 series ......................................... 101
Figure 5-29: Comparison of creep compliance at 40˚C for F4 series ......................................... 101
Figure 5-30: Tensile strength for granite PMA mixes ................................................................ 103
Figure 5-31: Tensile strength for limestone PMA mixes............................................................ 103
Figure 5-32: Comparison of tensile strength between control and PMA mixes......................... 104
Figure 5-33: Fracture Energy for granite PMA mixes ................................................................ 104
xii
Figure 5-34: Fracture Energy for limestone PMA mixes............................................................ 105
Figure 5-35: DCSE for granite PMA mixes................................................................................ 105
Figure 5-36: DCSE for limestone PMA mixes ........................................................................... 106
Figure 5-37: Comparison of Fracture Energy between control and PMA mixes........................ 106
Figure 5-38: Comparison of Fracture Energy between control and PMA mixes........................ 107
Figure 5-39: Relationship between the observed Failure Strain and SBS polymer content ....... 107
Figure 5-40: Gradation curves for control mixes and modified gradation mixes ....................... 109
Figure 5-41: Comparison of resilient modulus for granite and limestone mixtures ................... 110
Figure 5-42: Comparison of CP between granite and limestone mixes at -10˚C ....................... 110
Figure 5-43: Comparison of CP between granite and limestone mixes at 5˚C........................... 111
Figure 5-44: Comparison of CP between granite and limestone mixes at 25˚C......................... 111
Figure 5-45: Comparison of CP between granite and limestone mixes at 40˚C......................... 112
Figure 5-46: Comparison of Tensile Strength between granite and limestone mixes ................ 112
Figure 5-47: Comparison of Fracture Energy between granite and limestone mixes ................. 113
Figure 6-1: Specimen and LVDTs setup for DMT test............................................................... 117
Figure 6-2: Resilient modulus at different testing temperatures................................................. 119
Figure 6-3: Poisson’s Ratio from IDT test for all mixture series................................................ 119
Figure 6-4: Average phase angles for different type of materials ............................................... 120
Figure 7-1: Parameters used in sigmoidal fitting function of master curve................................ 123
Figure 7-2: Master curves for granite materials .......................................................................... 124
Figure 7-3: Master curves for limestone materials ..................................................................... 125
Figure 7-4: Master curves for all mixtures.................................................................................. 125
Figure 7-5: Measured vs. predicted dynamic modulus values for all mixtures .......................... 128
Figure 7-6: Comparison of average air void content between IDT and DMT specimens .......... 130
Figure 7-7: Resilient modulus versus dynamic complex modulus at 10 Hz............................... 131
xiii
Figure 7-8: Resilient modulus versus dynamic complex modulus at 5 Hz................................. 132
Figure 7-9: Resilient modulus versus dynamic complex modulus at 1 Hz................................. 132
Figure 7-10: Resilient modulus versus dynamic modulus at various loading frequencies ......... 133
Figure 7-11: Relationship of linear multiplication factors with DMT loading frequency.......... 133
Figure A-1: Gradation chart for S1 to S3.................................................................................... 144
Figure A-2: Gradation chart for S4 and S5 ................................................................................. 144
Figure A-3: Gradation chart for S7 to S9.................................................................................... 145
Figure A-4: Gradation chart for S10 to S12................................................................................ 145
Figure A-5: Gradation chart for S14, S15, and S18.................................................................... 146
Figure A-6: Gradation chart for S16 and S17 ............................................................................. 146
Figure A-7: Gradation chart for S6 and S13 ............................................................................... 147
Figure A-8: Gradation chart for S19 and S20 ............................................................................. 147
Figure B-1: Creep compliance of F2 control and all polymer-modified levels at -10˚C. ........... 157
Figure B-2: Creep compliance of F4 control and all polymer-modified levels at -10˚C. ........... 158
Figure B-3: Creep compliance of F2 control and all polymer-modified levels at 5˚C. .............. 158
Figure B-4: Creep compliance of F4 control and all polymer-modified levels at 5˚C. .............. 158
Figure B-5: Creep compliance of F2 control and all polymer-modified levels at 25˚C. ............ 159
Figure B-6: Creep compliance of F4 control and all polymer-modified levels at 25˚C. ............ 159
Figure B-7: Creep compliance of F2 control and all polymer-modified levels at 40˚C. ............ 159
Figure B-8: Creep compliance of F4 control and all polymer-modified levels at 40˚C. ............ 160
Figure B-9: Creep compliance of F2 control and modified gradation levels at -10˚C................ 160
Figure B-10: Creep compliance of F4 control and modified gradation levels at -10˚C.............. 160
Figure B-11: Creep compliance of F2 control and modified gradation levels at 5˚C................. 161
Figure B-12: Creep compliance of F4 control and modified gradation levels at 5˚C................. 161
Figure B-13: Creep compliance of F2 control and modified gradation levels at 25˚C............... 161
xiv
Figure B-14: Creep compliance of F4 control and modified gradation levels at 25˚C............... 162
Figure B-15: Creep compliance of F2 control and modified gradation levels at 40˚C............... 162
Figure B-16: Creep compliance of F4 control and modified gradation levels at 40˚C............... 162
xv
ABSTRACT
Hot Mix Asphalt (HMA) is a viscoelastic material and has been broadly used in pavement
structures. It is important to understand the mechanism of complex behaviors of HMA mixtures
in field for improving pavement mechanical performance. Aggregate gradation and asphalt
binder are two key factors that influence the engineering properties of HMA. The asphalt binder
plays a significant role in elastic properties of HMA and it is the essential component that
determines HMA’s viscous behavior. Many research works suggest that the Styrene-Butadiene-
Styrene (SBS) polymer is a promising modifier to improve the asphalt binder, and hence benefit
the HMA viscoelastic properties. The specific beneficial characteristics and appropriate polymer
concentration need to be identified. In addition, aggregate gradation requirements have been well
defined in Superpave mix design criteria. However, a potentially sound coarse mixture with the
gradation curve passing below the coarse size limit may be disqualified from being used. There is
a need to evaluate the Superpave coarse aggregate gradation limits by studying mixtures
purposely designed exceeding the control limits. Moreover, the mechanical parameters adopted
by AASHTO to characterize HMA properties are shifting from indirect diametral tensile (IDT)
test to dynamic modulus test (DMT), because the DMT has the ability to simulate real traffic
conditions and to record more viscoelastic information of HMA. Thus, the DMT and the IDT test
for implementing the AASHTO Mechanistic-Empirical Design Guide (M-E PDG) are needed to
be discussed.
The primary objective of this research study was to evaluate the fracture mechanics
properties of HMA concrete and to study the correlation between the DMT and the IDT test for
Superpave mixtures. An experimental program was performed to evaluate the engineering
properties of the asphalt mixtures with various types of materials. The laboratory testing program
was developed by applying a viscoelastic fracture mechanics-based framework that appeared to
be capable of describing the comprehensive mechanical properties of HMA mixtures according
to past research studies. The goals for these experiments are to evaluate the effect of aggregate
type, the effect of coarse aggregate gradation adjustment to mix designs, and the effect of SBS
polymer modifier on fracture mechanics properties of HMA concrete.
To achieve the objectives and goals, a complete dynamic testing system was acquired to
perform the temperature controlled dynamic tests to determine the engineering properties for all
xvi
selected asphalt concrete mixtures. The laboratory experimental program for fracture mechanics
properties involved two standard asphalt mix designs as control levels: one granite mixture and
one limestone mixture. Each control mix design was modified to two different gradation levels
with the control asphalt binder (PG 67-22) and three SBS polymer content levels (3.0%, 4.5%,
and 6.0%) with the original aggregate gradation. The experimental program for dynamic complex
modulus test involved 20 Superpave asphalt concrete mixtures commonly used in Florida with a
range of aggregates and mix designs including the following types of aggregates: 14 Georgia
granite materials, one Nova Scotia granite, one North Florida limestone, two Central Florida
limestone materials, one South Florida oolite, and one Alabama limestone. One type of base
asphalt binder, PG 67-22, was used for all mixtures tested for dynamic modulus. The volumetric
properties of all the mixtures were verified to ensure that the specimens’ air voids are as close to
the optimum (4.0%) as possible. The DMT specimens were cored from the 150 mm diameter
Superpave samples. The 20 Superpave asphalt concrete mixtures were tested for both dynamic
complex modulus and indirect tensile resilient modulus.
The SHRP IDT test procedure was generally followed to perform the indirect diametral
tensile test. The measurement and analysis system developed for SHRP IDT was also applied.
Three types of IDT test, the resilient modulus test, the creep compliance test, and the tensile
strength test were performed to determine the fracture mechanics properties of asphalt concrete at
four temperature levels: -10, 5, 25, and 40°C (14, 41, 77, and 104°F). Data evaluation of the test
results indicated the following characteristics: 1) the increase of nominal maximum size
aggregate amount to the standard mix designs in this study had negligible or adverse effect on
HMA fracture mechanics properties. 2) The SBS polymer-modified asphalt binder improved the
fracture mechanics behavior of asphalt mixtures comprehensively. The resilient modulus values
of polymer-modified asphalt (PMA) mixtures decreases with an increase of SBS polymer content
throughout the concentration range tested at low temperatures. At the high testing temperature of
40˚C, an optimum SBS content appeared to exist around 4.5% which would make the HMA
stiffest, which suggested that limiting the concentration within an optimal range is especially
important at high service temperatures. The SBS polymer also helps the HMA obtain an
upgraded creep performance. The mixtures with SBS polymer modifiers are more compliant at
the low temperature level (-10˚C) and become less compliant at the high testing temperature
(40˚C), which should lead to improved resistance to rutting and thermal cracking of HMA
xvii
mixtures. At a specific temperature level, a higher SBS polymer concentration generally results
in higher creep compliance values. Furthermore, the SBS polymer modifier improves the asphalt
mixture fracture properties by increasing the fracture energy (FE) limit or dissipated creep strain
energy (DCSE) limit which were indicators of mixtures’ resistance to fatigue cracking. The
failure strain of PMA mixtures tends to increase with an increase of SBS polymer content at low
temperatures (-10˚C and 5˚C). 3) It was found that the limestone mixtures were more compliant
than granite mixtures at low temperatures and turned to be less compliant than granite at high
temperature (40˚C). Therefore limestone materials hold advantages over granite materials in
improving the performance of thermal cracking at low service temperature and the rutting
resistance at high service temperatures.
The dynamic complex modulus tests were conducted at three temperature levels: 5, 25, and
40°C (41, 77, and 104°F). For all temperatures tested, the following frequencies were used: 25,
10, 5, 1, and 0.5 Hz. The master curves for all 20 mixtures were developed and constructed using
the time-temperature superposition principle. The Witczak prediction model was adopted to
verify the relationship between predicted and measured dynamic modulus for all mixture series.
The comparison indicated that the Witczak prediction model worked well for the asphalt
concrete mixtures tested in this study. A comparative study was also made between the dynamic
modulus and resilient modulus test results. The linear regression analysis indicated that the total
resilient modulus increased with an increase in dynamic modulus at a specific loading frequency.
The resilient modulus values were comparable with the dynamic modulus values at the loading
frequency of 4 Hz.
1
CHAPTER 1
INTRODUCTION
1.1 Background
The Superpave asphalt mix design method has been increasingly accepted, and the
system has been implemented in Florida. The current Superpave mix design approach is
based on meeting certain asphalt binder, aggregate and volumetric properties such as the
asphalt binder performance grading (PG) specification, aggregate gradation control limits,
gradation restricted zone, asphalt mix air voids, voids in mineral aggregate (VMA), voids
filled with asphalt (VFA), etc. The Superpave mix design system has continuously been
under evaluation to search for further improvements.
Gradation is perhaps the most important property of an aggregate. It affects almost all
the important properties of hot mix asphalt (HMA) mixtures including stiffness, stability,
durability, permeability, workability, fatigue resistance, frictional resistance, and
resistance to moisture damage. Therefore, gradation is a primary consideration in asphalt
mix design, and the Superpave specifications place limits on the aggregate gradations that
can be used in HMA mixtures. The gradation of the aggregate is important to ensure that
1) the maximum aggregate size is not too large or too small, 2) VMA requirements are
met, and 3) a satisfactory aggregate skeleton is obtained. According to the Superpave, the
aggregate gradation must be within the control limits to meet the Superpave requirements.
For example, if a 19-mm (3/4-inch) maximum aggregate size is specified, then 100
percent of the aggregate must pass the 25-mm (1-inch) sieve size. At least 90-100 percent
of the aggregate must be finer than the nominal maximum size (19 mm). Less than 90%
of the aggregate must pass the 12.5-mm sieve. In order to meet the Superpave
requirements, a coarse graded aggregate will have to be “gap-graded” to be within the
nominal size control limits. However, a “smoother” coarse gradation passing below the
lower control limit of the nominal maximum sieve size may exist to provide as similar
results to the “gap-graded” curve. Research is needed to evaluate the nominal size control
limits for Superpave mix design and to study its effect on mechanical properties of HMA
mixes.
2
In addition, modified binders have been used in Superpave mixtures of many state
agencies in an effort to improve the mixtures. SBS (styrene-butadiene-styrene) polymer
has been used to modify the asphalt binders. Some laboratory and full-scale field tests
have been performed to evaluate the beneficial effects of adding the SBS polymer to
asphalt binders and modified asphalt mixtures. SBS polymer modifiers appear to provide
greater benefit to open graded mixtures than to dense graded mixtures. It has been
recommended that asphalt binder modified with 3% SBS polymer is an effective way of
treating the Superpave mixtures. However, modifiers with higher percentage of SBS
polymer have been successfully used in Europe. It appears that research is needed to
evaluate the beneficial effect of using higher dosages of SBS polymer. The fracture
mechanics concept/approach has been studied extensively and proposed by Roque et al.
(2002). The fracture energy-based approach was developed and verified by an analysis of
22 field test sections throughout the state of Florida and appeared to be capable of
describing the HMA structural characteristics related to pavement cracking performance
(Roque et al. 2004). The proposed experimental framework may be adopted to evaluate
the fracture mechanics properties of the HMA mixtures.
Recently, the new AASHTO Mechanistic-Empirical (M-E) Design Guide research
team advocated the use of the dynamic complex modulus (E*) as the primary test protocol
to characterize the modulus response of HMA mixtures. The research team supported the
role, selection, and utilization of the dynamic complex modulus test for asphalt concrete
mixtures over the indirect tensile resilient modulus (MR) in the National Cooperative
Highway Research Program (NCHRP) 1-37A Project concerning the AASHTO M-E
Design Guide for Pavement Structures, which is currently aiming to introduce more
rigorous measures of performance into HMA mixtures and pavement design procedures.
The use of the indirect tensile test was also encouraged as a means of determining the
relative moduli response of field cores taken for rehabilitation designs. However, the use
of the test to characterize modulus at high temperatures was not recommended.
The difference between a resilient modulus test and a dynamic complex modulus test
for HMA mixtures is that the former uses loading of any waveform with a given rest
period, while the latter applies a sinusoidal or haversine loading with no rest period. The
dynamic complex modulus is one of the many methods for describing the stress-strain
3
relationship of viscoelastic materials. The modulus is a complex quantity, of which the
real part represents the elastic stiffness and the imaginary part characterize the internal
damping of the materials. The absolute value of the complex modulus is commonly
referred to as the dynamic modulus. A detailed comparison of key differences between
the dynamic complex modulus test and the indirect diametral resilient modulus test for
asphalt concrete mixtures was summarized in a position paper by the NCHRP 1-37A
2002 Project research team (Witczak 1999). The transition from the resilient modulus test
to the use of the dynamic complex modulus test for design of flexible pavement structures
has hardly been smooth. The potential impact of adopting the dynamic complex modulus
for implementation of the new AASHTO M-E Design Guide is tremendous for state
transportation agencies, such as the Florida Department of Transportation (FDOT). The
IDT has traditionally been used to characterize the HMA mixtures for flexible pavement
design in Florida, and the test method has been shown to be both an expedient and a
reliable way of obtaining mixture properties from field cores. In response to the need, an
effort is desired to characterize the empirical relationship between dynamic complex
modulus and indirect resilient modulus of HMA mixtures.
1.2 Objectives
The primary objectives of this study were to evaluate the effects of coarse aggregate
gradation limits, aggregate type, and SBS polymer-modified asphalt binder on the
engineering properties of HMA mixtures. Specifically, the research goals were to evaluate
the gradation limits on nominal maximum sieve size specified in the Superpave mix
design guide, to evaluate the benefits of using SBS polymer modifier in HMA concrete
pavements, and to evaluate the effect of aggregate type on fracture mechanics properties
of asphalt mixtures.
Since the mechanical parameters adopted by AASHTO to characterize HMA
properties are shifting to dynamic complex modulus, the relationship between the
dynamic complex modulus test (DMT) and the indirect diametral tensile test (IDT) for
implementing the AASHTO Mechanistic-Empirical Design Guide for Pavement
Structures is also needed to be discussed.
4
To achieve these objectives, a series hot mix asphalt mix designs were selected for
testing. An experimental program was developed to measure the fracture mechanics
properties of all HMA specimens fabricated in the laboratory. The dynamic modulus test
was also performed on corresponding mixture specimens in order to develop a rational
comparison with the resilient modulus.
1.3 Dissertation Outline
This dissertation summarizes the study to evaluate the HMA gradation effect, to
study the SBS polymer-modified binder effect, and to develop a rational relationship
between dynamic complex modulus test (DMT) and indirect diametral test (IDT) for
implementing the AASHTO Mechanistic-Empirical Design Guide for Flexible Pavement
Structures. The dissertation is organized as the following structures:
Chapter 1 introduces the background, objective, and scope of the study.
Chapter 2 gives a comprehensive literature review of the aggregate gradation effect
and SBS polymer modifier effect on HMA. The fracture mechanics model developed by
Roque et al. (2004) is introduced to evaluate the engineering properties of asphalt
concrete. The complex modulus and resilient modulus for characterizing the hot mix
asphalt concrete mixtures are also summarized.
Chapter 3 introduces the materials and develops a whole experimental program.
Detailed testing methods and procedures are specified.
Chapter 4 presents the results from IDT sweep set of tests for the two control mixes,
four modified gradation mixes, and six mixes with SBS polymer-modified asphalt at
different concentrations (3%, 4.5%, and 6%).
Chapter 5 analyzes the IDT test results in detail to account for the gradation effect
and SBS polymer modifier effect on HMA mixtures.
Chapter 6 shows the results of dynamic modulus from DMT and resilient modulus
from IDT for 20 selected standard mix designs.
Chapter 7 discusses a comparison between DMT and IDT. The dynamic modulus
master curves for the selected mixes are constructed. The correlation between dynamic
modulus and resilient modulus results is developed.
5
Chapter 8 summarizes the whole dissertation. Concluding remarks are presented.
6
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
The purpose of this study is to evaluate the engineering properties of asphalt concrete
mixtures obtained from laboratory tests. Asphalt concrete pavement performance is
influenced by a great number of factors. HMA mixtures are essentially made up of
various kinds of aggregates in appropriate size combinations and different types of
asphalt binders. Our research work is focused on the material effects of aggregate
gradation and Styrene-butadiene-styrene (SBS) polymer-modified asphalt binders. There
are many lab test methods developed for measuring the mechanical properties of asphalt
concrete over the past twenty years. The most common ones are the indirect diametral
tensile (IDT) test and the dynamic modulus test (DMT). They were introduced in the
AASHTO flexible pavement design guide in 1993 and 2004 respectively. In this chapter
the following are to be discussed:
• To conduct a comprehensive literature review on publications related to aggregate
gradation effect and SBS polymer-modified asphalt binder effect on flexible
asphalt mixture characteristics.
• To introduce test methods and procedures, corresponding comparisons of testing
methodologies, and prediction models that have been used to evaluate mechanical
responses of asphalt concrete mixtures.
• To review the fracture mechanics and energy models that were developed and
used to evaluate cracking performance of HMA mixtures.
The following sections provide an explanation of the basic material mechanisms and
approaches used to evaluate the performance of asphalt pavement.
2.2 Asphalt Cement Properties
Asphalt cement is bituminous material that is either naturally occurring or produced
by distillation process from crude petroleum using different refining techniques. It is
7
widely used throughout the world in roadway paving applications. Asphalt cement is a
black, sticky and highly viscous material at ambient temperatures. It is also resistant to
the action of most acids, alkalis and salts. The largest use of asphalt cement is in the
production of Hot Mix Asphalt for construction of flexible pavements. By applying heat
to the asphalt cement, it can be liquefied for mixing with mineral aggregates; it adheres to
aggregate particles and binds them to form HMA. After cooling to ambient temperature,
with asphalt cement’s excellent adhesive and waterproofing characteristics, HMA
become a very strong and durable paving material which can sustain heavy traffic loads.
Three methods, based on penetration, viscosity and performance are used to classify
asphalt cements into different standard grades. The penetration grading of asphalt cement
is specified in ASTM D946 and is primarily controlled by the penetration test. The
viscosity grading is specified in ASTM D3381. It is based on either the viscosity of the
original asphalt cement or on the viscosity of the asphalt cement after aging in the rolling
thin film oven (RTFO) test. The performance-based method of classifying asphalt binders
was developed in the Strategic Highway Research Program (SHRP).
2.2.1 Chemical Properties of Additives and Polymer Modifiers
Asphalt modifiers have been used over 60 years. They are more commonly used in
Europe compared to the United States in the 20th century. A greatly increased effort has
been dedicated to the research and application of asphalt modifiers over the past 20 years
in the United States. The Superpave asphalt binder specifications based on SHRP require
the asphalt binders to meet stiffness criteria at both high and low pavement service
temperatures. However, most regular asphalt binders are not qualified for the
requirements in areas with extreme climate conditions. In the meantime, traffic volume
and loads have increased significantly in recent years. This has caused lots of premature
rutting and cracking of HMA pavement constructed with neat asphalt binders.
Modifications of asphalt binders become of considerable interest in the improvement of
pavement performance and service life. Although low initial cost discourages the use of
modifiers, some state highway agencies started to specify modified asphalt binders and to
be willing to pay a higher initial cost for pavements with a longer service life and reduced
8
risk of premature distress, and therefore, lower life cycle costs. Additionally, the disposal
of waste materials and industrial byproducts, such as tires, glass, sulfur, etc., used as
additives in HMA is economical and benefits the environment.
Some specific technical reasons for using additives and modifiers in HMA are listed
as follows (Roberts et al. 1996):
1. Obtaining stiffer mixtures at high service temperatures to minimize rutting.
2. Obtain softer mixtures at low service temperatures to minimize thermal cracking.
3. Improve fatigue resistance of HMA mixtures.
4. Improve asphalt-aggregate bonding to reduce stripping or moisture susceptibility.
5. Improve resistance to aging or oxidation; rejuvenate aged asphalt binders.
6. Permit thicker asphalt films on aggregate for increased mix durability.
7. Improve abrasion resistance of mixture to reduce raveling.
8. Reduce flushing or bleeding; reduce structural thickness of pavement layers.
9. Reduce life cycle costs and improve overall performance of HMA pavements.
Additives and modifiers can be classified in different ways. A generic classification
system was first suggested by Terrel and Walter (1986). A modified version of the system
(Table 2-1) and a discussion of each additive or modifier were given by Roberts et al.
(1996).
It can be seen in Table 2-1 that polymers are comprised of rubber, plastic and their
combination materials. Elastomers (rubber) and plastomers (plastic) are the two basic
categories. Elastomers resist deformation from applied stress with their high extensibility
and contractibility and rapidly recover upon removal of the load. The initial modulus is
usually low but they stiffen when stretched. Plastomers resist deformation by their tough
and rigid three-dimensional network. Earlier research showed that elastomers (rubbers)
increase asphalt binders’ tensile strength with elongations whereas little additional
strength is obtained from the rubbers by asphalt binders until they are stretched. On the
other hand, plastomers exhibit quick early strength on loading but may fracture under
strain (Hines 1993). Therefore, when elastomers are used for modifying asphalt cements,
HMA pavements generally get more flexible and resilient. In contrast, asphalt binders
modified with plastomers usually improve the stiffness moduli of HMA pavement.
9
Table 2-1: Generic classification of asphalt additives and modifiers (Roberts et al. 1996)
Type Generic Examples
1. Filler
� Mineral Filler: crusher fines lime Portland cement fly ash
� Carbon black
2.Extender � Sulfur � Lignin
Rubber: a. Natural latex b. Synthetic latex c. Block copolymer d. Reclaimed rubber
� Natural rubber � Styrene-butadiene or SBR � Polychloroprene latex � Styrene-butadiene-styrene (SBS),
Styrene-isoprene-styrene (SIS) � Crumb rubber modifier
Plastic
� Polyethylene/Polypropylene � Ethylene acrylate copolymer � Ethyl-vinyl-acetate (EVA) � Polyvinyl chloride (PVC) � Ethylene propylene or EPDM � Polyolefins
3. Polymers
Combination Blends of polymers above
4. Fiber
� Natural: asbestos rock wool
� Man-made: polypropylene polyester fiberglass mineral cellulose
5. Oxidant Manganese salts
6.Antioxidant � Lead compounds � Carbon � Calcium salts
7. Hydrocarbon � Recycling and rejuvenating oils � Hard and natural asphalts
8. Anti-stripping Agent
� Amines � Lime
9. Waste Materials � Roofing shingles � Recycled tires � Glass
10. Miscellaneous � Silicones � Deicing calcium chloride granules
A polymer molecule is produced by the reaction of many monomers, which are
smaller molecules, with one another in long chains or clusters. The term “poly” means
10
many as opposed to “mono”. Homopolymers are made up of only one kind of repeating
monomer in the polymer molecular chain. Copolymers are composed of the combination
of two or more different monomers in a random or block arrangement. The types of
polymers are listed below.
• Homopolymer: only one monomer is used along the chain.
• Random Copolymer: the repeating units are in random sequence.
• Alternating Copolymer: the two units repeat in an ordered manner.
• Block Copolymer: the chain consists of long sequence (blocks) of repeating units.
• Grafted Copolymer: branched copolymer in which the side chains are structurally
distinct from the main chain.
• Periodic copolymers: with A and B units arranged in a repeating sequence, e.g.
(− ABABBAAAABBB−)n
The physical properties of polymers vary remarkably depending on the sequence,
structure, and chemical process for the composing monomers (Usmani 1997). Polymers
can also be categorized into linear polymers, branched polymers, and cross-linked
polymers as shown in Figure 2-1 according to their structures.
Polymers may also be classified as thermosets and thermoplastics. Thermosets are
usually rigid and tightly cross-linked. When mixed with asphalt at high temperatures, the
thermoset’s particles may swell to more than twice the original volume as a result of
chemical interaction which leads to a remarkable increase in mixture viscosity.
Thermoplastic elastomers are commonly applied in the modification of asphalt binders.
They are usually linear or branched in types of block copolymer (SB)nX, where ‘S’
11
denotes the styrene block, ‘B’ denotes the butadiene block, and ‘X’ denotes the coupling
agent, as shown in Figure 2-2. It was found that a separation takes place between
butadiene (soft block) and the styrene (hard block) because they are mutually
incompatible; as a result, the styrene remains dispersed in a continuous elastomeric
matrix (Diani et al. 1997).
Linear
Branched
Cross - Linked
Figure 2-1: Polymer classifications based on link structure
Figure 2-2: SBS polymer modifier structure
As listed in Table 2-1, Elastomers or rubbers used as asphalt modifiers include
natural rubber, styrene-butadiene latexes (SBR), polychloroprene latex, styrene-
butadiene-styrene block polymers (SBS), styrene-isoprene-styrene block polymers (SIS),
and crumb rubber modifiers (ground tires). SBS block copolymers are usually in the solid
forms of pellets, crumbs, or ground material in bags or bulk. The common concentration
level is about 3 to 5 percent by weight of asphalt cement in the HMA industry. High shear
12
mixing equipment is used for blending the SBS modifier with hot asphalt cement
maintained at 350-380˚F (177-193˚C). Since the asphalt binder must be within specified
viscosity ranges for workability purposes during mixing and compaction, it is probably
necessary to increase the mixing and compaction temperatures while conducting
laboratory work and testing.
Polymer modifiers have complex characteristics and their effects on asphalt binders
depend on quite a few factors such as polymer concentration, molecular weight, chemical
composition, and molecular structure. Other important things include the source of
original asphalt binder, production process, binder grade, reaction between binder and
modifier, etc. Special properties can be obtained through various combinations of
elastomers and plastomers to meet desired requirements. However, it should be noted that
it is very difficult to predict whether a particular combination will be able to provide
improvements in the desired property. Sometimes polymer properties may get diluted or
even changed when blended with asphalt binders. The structures of the pure polymer-
modified binder generally are different from those of the PMB in the asphalt mixture.
Therefore, it is necessary to test the polymer-modified asphalt binder; or in more practical
situations, it would be more advisable to evaluate the performance of actual HMA
produced with modified asphalt binder (Wegon and Brule 1999). It is hoped that polymer
modifiers can be used in Superpave mix design and evaluation procedures to obtain a
stiffer HMA at high service temperatures to minimize rutting, a more elastic HMA to
resist fatigue cracking at intermediate temperatures, and a softer HMA at low service
temperatures to resist thermal cracking.
Chen et al. (2002, 2003) investigated the morphology of the SBS modified binders
described by the concentration and the presence of the microstructure of the copolymer.
As the polymer content increases, the dispersed polymer particles gradually swell to form
local SBS networks which highly enhance the mechanical properties of the asphalt binder
(viscosity, softening point, toughness and complex modulus, etc). A continuous polymer
structure was observed to begin at an SBS content between about 5% and 6%, yet the
minimum percentage depends more on the base asphalt and the polymer itself. The
optimum SBS content is based on the formation of the critical network between asphalt
and polymer, which appeared to be slightly higher than the phase inversion content that
13
occurs when the SBS entered the continuous matrix phase. However, once the critical
networks begin to form, increases in polymer content have less significant effect on PMA
property improvement (Figure 2-3), or may even lead to the separation of polymer and
bitumen. Recent work by Chen and Huang (2007) showed that the SBS-asphalt blended
with sulfur resulted in improved rheological characteristics.
Brule et al. (1988) studied the relationship between the composition, structure, and
properties of asphalt binders incorporated with SBS block copolymers. It was found that
increasing the agitation time made the microstructure finer, which led to a greater
deformability. They also found that the amount of polymer required for matrix inversion
and for obtaining highly modified practical properties depended significantly on the
asphalt itself. However, the value of this inversion threshold was not predictable. The
extent of swelling in asphalt-SBS blends was not highly dependent on content for high
polymer concentrations, but increased substantially as the amount of polymer decreased;
it also appeared to be independent of temperature in the high-level range (80-160˚C). In
addition, the SBS polymer was no longer swollen in the binder but dissolved beyond a
colloidal instability index value.
Frequency (Hz) Figure 2-3: Complex modulus of SBS-modified asphalt at 60°C (Chen et al. 2002)
Lu et al. (1998) reported that SBS polymer modification improves the low-
temperature properties of bitumens. The polymer modification reduces the creep stiffness
14
and limiting stiffness temperature of bitumens. The changes generally increase with SBS
content and are influenced slightly by SBS structure.
Many studies (Huffman 1980; Lalwani et al. 1982; Scofield 1989) have reported that
the polymer-modified asphalt can lower temperature susceptibility, which is the primary
drawback of regular asphalt, reduce binder penetration, increase the viscosity and
softening point, and improve resistance to aging and oxidation. These effects should lead
to increased resistance to deformation (rutting) and thermal cracking in practice. King et
al. (1986) documented a correlation between styrene-butadiene elastomer-modified
asphalt and pavement durability. The addition of polymer improves stiffness, rutting
resistance, fatigue life, adhesion and stripping resistance to the bituminous mix.
Carpenter et al. (1987) conducted a series of lab tests on asphalt mixtures including
the diametral resilient modulus test, indirect tensile test at temperatures ranging from
72˚F to -20˚F (22.2˚C to -28.9˚C), and permanent deformation testing at 72˚F (22.2˚C)
and 100˚F (37.8˚C). The testing indicated that the polymer additives reduced stiffness at
low temperatures yet maintained adequate stiffness at elevated temperatures. The low-
temperature performance was greatly improved over that of untreated asphalt cements of
all grades, whereas the permanent deformation characteristics were greatly improved at
elevated temperatures. Carpenter et al. (2006) conducted further tests and showed that the
healing/recovery rate of the polymer-modified binders is significantly greater than the
neat binder. Button (1992) drew a summary of asphalt additive performance which
indicated positive influences by polymer addition in bitumens.
Collins et al. (1991) studied the performance of paving asphalt modified by SB
polymers. The modification resulted in a substantial improvement of fatigue life by
reducing flexural fatigue cracking and a dramatic increase of strength and resistance to
creep at high temperatures. They also found that the actual critical cracking temperature
was significantly lower than that of the base asphalt and decreased with increasing
polymer content (Figure 2-4). The poly-butadiene chains in polymer contribute to the
flexibility of the binder and the elastomeric lattice between asphalt molecules and SBS
polymer improves the elastic characteristics of the binder without increasing the stiffness
binders at low service temperatures. Similar findings were reported by many other
researchers. Verhaeghe et al. (1994) conducted studies on asphalt binder modified with
15
Ethylene Vinyl Acetate (EVA) which improves the compressive strength and rutting
resistance of asphalt mixes. Pradhan (1993) reported that the addition of commercial SBR
modifiers improves the physical properties related to rutting problems on Montana
asphalt pavements. Testing programs conducted by Kennedy et al. (1992) showed that
SBS and SBR polymers generally increase the mixture’s tensile strength at high
temperatures and tensile strain at failure at low temperatures. The permanent deformation
resistance was also improved, indicated by indirect creep testing.
King et al. (1993) studied a type of standard mixture containing four control
bitumens with styrene-butadiene polymer of three different contents (x%, 1.5x% and
2x%). They found that the softer the base asphalt, the lower the cracking temperature; and
that increasing the polymer content generally lowered the cracking temperature. Shih
(1996) conducted testing studies to compare the effects of different additives on typical
Florida asphalt mixtures. Test results showed that the addition of modifiers generally
benefits the rutting resistance of pavement and the SBR-modified asphalt mixtures have
lower resilient moduli at low temperatures; thus, the addition of modifiers would be
beneficial to the resistance of thermal cracking.
Figure 2-4: Relationship between the observed critical cracking temperature (Tcr) and SBS polymer concentration (Collins et al. 1991)
16
Aglan (1997) analyzed the fatigue tests and electron microscopic scans on polymer
modification for asphalt mixtures. The binder-rich areas were observed to contain ridges
produced by the micro-stretching of the SBS modified binder on the fracture surface, and
the mixture test results showed a superior resistance to fracture.
Jones et al. (1998) performed Superpave IDT creep and strength tests on five
different modified mixtures. Higher tensile strength was observed at intermediate
temperatures, yet it appeared that there was no noticeable difference at low temperatures
(around and below 0˚C). Khattak and Baladi (1998, 2001) evaluated the effects of SBS
polymer-modified binder on mechanical properties of mixtures. The measurement results
showed increased fatigue life and tensile strength at intermediate temperature whereas
low temperature elastic properties were almost the same. They also found that the fatigue
life and permanent deformation were strongly related to the rheological properties of
polymer-modified binders. Kim et al. (2003) investigated the use of SBS modifier in
asphalt pavement mixtures through lab testing for cracking resistance and healing
characteristics. Although the SBS does not show an influence on healing of the asphalt
mixture, it appears to reduce the rate of micro-damage accumulation which justified the
benefits of SBS modification on creep and failure properties of the mixtures. More
recently, Quintus et al. (2007) conducted an investigation of a large amount of real-world
pavement sections to quantify the benefits of using PMA mixtures. It was found that the
PMA significantly enhanced the rutting performance of asphalt pavement (Figure 2-5)
and its fatigue and fracture performance.
2.2.2 Mechanical Properties of Asphalt Cement
The Strategic Highway Research Program (SHRP) was launched in 1987. The
program made a research effort to develop performance-based tests and specifications for
asphalt binders and HMA mixtures. The Superpave (Superior Performing Asphalt
Pavements) binder tests and specifications have a few prominent features (McGennis et
al. 1994; Warren et al. 1994; Asphalt Institute 1994) compared with the old physical
testing system for asphalt cement.
17
Figure 2-5: Comparison of the rut depths measured on sections with PMA and the companion sections without PMA mixtures (Quintus et al. 2007)
The rolling thin film oven (RTFO) test is specified in AASHTO T240 and ASTM
D2872. The RTFO simulates the asphalt binder aging during the manufacture and
construction of HMA pavements. It continually exposes fresh binder to heat and air flow
during rolling. This test mode does not allow any asphalt surface skin to be formed, this
inhibits aging. And modifiers, if used in asphalt cement, usually remain dispersed due to
rolling action, which makes the modified binder age more sufficiently. The RTFO test
determines the mass of volatiles lost from the binder, which indicates the amount of aging
that occurs during HMA production and construction. However, some asphalt binders
gain weight during the RTFO aging due to the oxidative products formed during the test.
The Dynamic Shear Rheometer (DSR) Test is used to characterize the viscous and elastic
behavior of asphalt binders at high and intermediate service temperatures. The DSR
measures the complex shear modulus G* and phase angle � of asphalt binders at the
desired temperature and loading frequency. Complex modulus G* can be considered as
the total resistance of the binder to deformation at repeated shear load. Complex modulus
G* consists of two components as shown in Figure 2-6: (a) elastic modulus G’, also
known as the storage or recoverable part; (b) loss modulus G”, also known as the viscous
or non-recoverable part (McGennis et al. 1994).
The values of G* and � for asphalt binders are affected by both service temperature
and loading frequency. Most asphalt binders are viscoelastic at usual pavement service
18
temperatures. They behave like elastic solids as well as viscous fluids simultaneously.
The magnitude G* and phase angle � define a complete picture of the behavior of asphalt
binders in certain conditions, as shown in Figure 2-7. The elastic component or storage
modulus is related to the amount of energy stored in the sample during each testing cycle.
The viscous component or loss modulus is related to the energy lost during each testing
cycle through permanent flow or deformation (ASTM 1994).
������������ �
����������� �
�
Figure 2-6: Components of complex modulus G
*
Vis
co
us B
eh
avio
r
Elastic Behavior
1δ2δ
*
1G
*
2G
Figure 2-7: Viscous and elastic behavior of asphalt binders
The DSR test procedure is given in AASHTO TP5. The asphalt cement sample is
sandwiched between a fixed plate and an oscillating plate or spindle as shown in Figure
2-8. Two types of oscillatory shear rheometer are usually used: constant stress and
constant strain. Constant stress rheometers use a fixed torque to oscillate the top spindle
and the strain will vary. Constant strain rheometers move the spindle with a fixed
distance (e.g., from point A to B) and measure the torque resulting from this movement.
All Superpave DSR tests are conducted in constant stress mode which uses a fixed torque
to oscillate the top plate at a frequency of 10 radians per second (about 1.59 Hz). When
torque is applied to the oscillating plate, it starts at point A and moves to point B, and
then the spindle moves back and goes to point C passing point A. From point C it returns
back to point A. This movement comprises one cycle of oscillation. When the spindle is
oscillated back and forth with constant stress, the resulting strain is monitored.
19
Oscillating Plate
Asphalt
Fixed Plate
Applied Stress or Strain
B CA
Figure 2-8: Dynamic shear rheometer
δmaxγ
maxτ
Time
Time
Applied
Shear
Stress
Resulting
Shear
Strain
LagTime
G
=
=
δ
γ
τ
max
max*
Figure 2-9: Stress-strain response of viscoelastic material
The relationship between the applied stress and the resulting strain is used to
compute complex modulus G* and phase angle �, which is the time lag between the
applied stress and resulting strain. Theoretically, the phase angle � is zero for a perfect
elastic material because the strain response is instant. For an ideal viscous fluid, the time
lag is 90 degrees. In reality, asphalt binders behave like viscoelastic materials with a
20
stress-strain response between the two extreme conditions at certain service temperatures
as shown in Figure 2-9, in which the resulting phase angle is between 0 and 90 degrees.
G* is the ratio of maximum shear stress (�max) to maximum shear strain (�max), which
are calculated by the following formulas:
max
max*
γ
τ=G (2-1)
3max
2
r
T
πτ = (2-2)
h
r⋅=
θγ max (2-3)
Where,
T = maximum applied torque,
r = radius of binder specimen/plate (either 12.5 or 4 mm),
� = deflection (rotation) angle,
h = specimen height (either 1 or 2 mm).
The SHRP researchers considered rutting to be a stress-controlled, cyclic loading
phenomenon. Work is being done to deform the HMA pavement surface with each traffic
loading cycle. A part of this work is recoverable in elastic rebound mode while some is
dissipated in the form of permanent deformation and heat energy. The amount of
dissipated work must be minimized in order to minimize rutting. The work dissipated per
loading cycle at a constant stress can be expressed as follows (Bahia and Anderson 1995):
��
���
�×=
δσπ
sin/
1*
2
0G
Wc (2-4)
Where,
Wc = work dissipated per load cycle,
�0 = stress applied during the load cycle,
G* = complex modulus,
� = phase angle.
The work dissipated per loading cycle is inversely proportional to G*/�, as indicated
from the equation. A high complex modulus G* value and low phase angle � are both
desirable for rutting resistance. This relationship appears logical because the asphalt
21
binder will be stiffer with higher G* value; the lower the � value, the more elastic the
asphalt binder will be, and thus the more resistant to rutting and permanent deformation.
Therefore, the G*/� parameter was chosen as a Superpave asphalt binder specification.
Fatigue cracking is typically considered a strain-controlled phenomenon in thin HMA
pavement layers and a stress-controlled phenomenon in thick ones. The SHRP
researchers assumed that fatigue cracking should be considered mainly a strain-controlled
phenomenon since it is known to be more prevalent in thin pavement layers (Bahia and
Anderson 1995). The work dissipated per loading cycle at a constant strain can be
expressed as follows:
[ ]δεπ sin*2
0 ××= GWc (2-5)
Where �0 is the strain and other variables are as described previously. The equation
indicates that the dissipated work will increase as G* and/or � are increased. As G*
decreases, the asphalt binder becomes less stiff and thus able to deform without building
up large stresses which might cause cracking. In addition, low � values indicate more
elastic asphalt binders which can regain their original condition without dissipating work.
Therefore, G*�� was chosen in Superpave specifications to limit the total amount of
energy dissipated for minimizing fatigue cracking.
The Superpave asphalt binder specification is given in AASHTO MP1-93. It is meant
to be performance-based and thus addresses three primary performance parameters of
HMA pavements: permanent deformation (rutting), fatigue cracking, and low temperature
(thermal) cracking. Other common specification criteria include safety, pumping and
handling, excessive aging, etc.
2.3 Hot Mix Asphalt (HMA) Mixture Design
2.3.1 Physical Properties of Aggregates
Aggregates for HMA are usually classified by size as coarse aggregates, fine
aggregates, and mineral fillers. ASTM defines coarse aggregate as particles retained on a
No. 4 (4.75 mm) sieve, fine aggregate as that passing a No. 4 sieve (4.75 mm), and
mineral filler as material with at least 70 percent passing the No. 200 (75 µm) sieve.
22
Some agencies use another sieve size as the dividing line between coarse and fine
aggregates. For example, the Asphalt Institute uses the No. 8 (2.36 mm) sieve as the
dividing line.
Specifications for coarse aggregates, fine aggregates, and mineral fillers are given in
ASTM D692, D1073 and D242, respectively. Aggregates for HMA are generally required
to be strong, sound, and properly graded; to have a clean surface without deleterious
materials; to consist of angular particles with low porosity and appropriate absorption for
asphalt cement.
The specific gravity of an aggregate is a basic parameter for HMA mix design. It is
used to make weight-volume conversions and to calculate the void content in a
compacted HMA. The specific gravity is defined as the ratio of the weight of a unit of
volume of the material to the weight of an equal volume of water at approximately 23˚C
(73.4˚F). Two different aggregate specific gravities are often used for HMA based on the
method used to define the volume of the aggregate particles: (a) bulk specific gravity; and
(b) effective specific gravity.
When the sample aggregates consist of separate aggregate fractions of coarse
aggregate, fine aggregate and mineral filler, the bulk specific gravity of total aggregate
can be calculated from the following equation:
n
n
n
sb
G
P
G
P
G
P
PPPG
+++
+++=
�
�
2
2
1
1
21 (2-6)
Where,
Gsb = bulk specific gravity for the total aggregates,
Pi = individual percentages by mass of aggregate, i = 1, 2, …, n;
Gi = individual bulk specific gravity of aggregate, i = 1, 2, …, n.
The effective specific gravity of aggregate, Gse includes all void spaces in the aggregate
particles excluding voids permeable to asphalt. It is determined by the following
equation:
b
b
mm
mm
bmm
se
G
P
G
P
PPG
−
−= (2-7)
23
Where,
Gse = effective specific gravity of the aggregate,
Gmm = maximum specific gravity of the mixture,
Pmm = percent by mass of total loose mixture = 100,
Pb = asphalt content,
Gb = specific gravity of asphalt cement.
2.3.2 Aggregate Gradation
Aggregate gradation is the distribution of particle sizes expressed as a percent of the
total weight. It is one of the most important properties of an aggregate. The gradation of
an aggregate is normally expressed as total percent passing various sieve sizes. It affects
the HMA performance in many respects including stiffness, durability, stability,
permeability, workability, resistance to rutting and fatigue cracking, and frictional
resistance. Therefore, gradation is a critical consideration in asphalt mix design.
Aggregate gradations are described as dense (well-graded), open (uniformly-graded), and
gap-graded, as shown in Figure 2-10. Most states place limits on the aggregate gradations
for HMA. Fuller and Thompson (1907) proposed one of the best known gradations for
maximum density. The equation for Fuller’s maximum density curve is:
nDdP )/(100 ⋅= (2-8)
Where d is the diameter of the sieve size in question, P is the total percent passing or finer
than the sieve, and D is the maximum size of the aggregate. Studies by Fuller and
Thompson showed that a maximum density can be obtained for an aggregate when n=0.5.
In the early 1960s, the Federal Highway Administration introduced an aggregate grading
chart which is based on the Fuller gradation but uses a 0.45 exponent in the equation. The
maximum density lines can be conveniently obtained by drawing a straight line from the
origin at the lower left corner of the chart to the actual percentage point of the nominal
maximum size, which was defined in the specification as the largest sieve size retaining
any material. The maximum aggregate size is normally limited to about one-half of the
lift thickness in construction. The use of large stone mixes has been increased in recent
years in order to minimize rutting. However, large maximum aggregate size (e.g. greater
24
than 1 inch, or 25.4 mm) usually results in segregation during placement of HMA.
Special attention is required when these large stone mixes are used.
#
20
0
#1
00
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Sieve Sieze (No.)
Tota
l P
erc
ent
Passin
g Gap -graded
Well-graded or
Dense-graded
Uniformly or
Open-graded
Figure 2-10: Typical aggregate gradations.
Some guidance for developing gradation limits and potential problem areas were
proposed by Hveem in 1940. Theoretically, the gradation that gives the densest packing
provides enhanced stability and reduces void space in the mineral aggregate through
increased interlocking between mixture particles. However, gradations of maximum
density may not provide enough voids in the aggregate. There must be sufficient air void
space in HMA to permit enough asphalt cement to be incorporated to provide adequate
film thickness for maximum durability. In addition, appropriate HMA air voids content
must be ensured in the mixture to avoid bleeding or rutting. Therefore, deviations from
the maximum density curves are necessary in order to increase the total voids in the
mineral aggregate (VMA). VMA is an important parameter and minimum values of VMA
are required and suggested by most pavement agencies depending on the maximum
25
nominal aggregate size of the mixture design. It is preferred that the gradation curve be
approximately parallel to the maximum density line with a few percentage points offset,
either above or below the line. Most specifications for HMA define aggregate gradation
band and tolerance for each nominal maximum size mixture according to accumulated
field experiences. In particular, the Superpave mix design developed by the Strategic
Highway Research Program (SHRP) requires a selected number of control points on the
gradation chart. The Superpave mix design system uses the following aggregate size
definitions:
• Maximum size: one sieve size larger than the nominal maximum size.
• Nominal maximum size: one sieve size larger than the first sieve to retain more
than 10 percent.
The maximum density line is obtained in Superpave by connecting the origin at the
lower left of the 0.45 power gradation chart to the maximum aggregate size at the upper
right of the chart (FHWA 1995).
Birgisson and Ruth (2001) developed a power law model to evaluate and classify
gradation curves according to mixture performance. Ruth et al. (2002) expanded the
parametric study and provided an experience-based methodology which introduced
aggregate gradation factors based on regression analysis of power law constants (aca and
afa) and exponents (nca and nfa). These gradation factors were used to evaluate
relationships with tensile strength, fracture energy, and failure strain of the mixtures. The
findings appeared to imply that the gradation characterization factors relate well to
mixture properties. Birgisson et al. (2004) and Ekingen (2004) established a correlation
between dynamic modulus and aggregate gradation factors at high temperature (40˚C)
based on the power law model. The relationship between a low nfa and a high dynamic
modulus at 40˚C has been identified, which indicates that power law parameters can be
used to optimize mixture gradations for the dynamic modulus and the rate of change in
the gradation on the fine side affects the stiffness and rutting resistance of the mixture. In
addition, it was also found that a high nca results in a low dynamic modulus when
controlling for nfa. More recently, Roque et al. (2006) developed a conceptual and
theoretical approach to evaluate the relationship between coarse aggregate structure based
on gradation and the pavement rutting performance. They found that the relative
26
proportion of particles from two contiguous size ranges can be no greater than 70/30 and
the porosity must be no more than 50% in order to form an interactive network.
2.3.3 HMA Mix Design
Most HMA produced before 1990s in the United States was designed using the
Marshall or Hveem method. The Superpave mix design procedures were developed by
the Strategic Highway Research Program (SHRP) and were adopted by a few states for
some pavement projects starting in 1995. The key points for all three design methods are
the same: to determine an appropriate asphalt content with which to begin field
construction.
The concept of the aggregate maximum density line for the densest packing of HMA
was first validated by Nijboer (1948). Goode and Lufsey (1962) then proposed that
aggregates should be graded using a mathematical concept of packing the void space
between aggregates of large diameter with aggregates of smaller diameter. They noted
that if the gradation corresponding to the exponent of 0.5 is used as proposed by Fuller in
1907, then the VMA may be too low to ensure both sufficient air void content and enough
asphalt cement for durability and stability. Therefore, the FHWA included the suggested
use of the 0.45 power curve as well as the maximum density line to evaluate and adjust
aggregate gradations. Huber and Shuler (1992) presented the relationship of VMA to
aggregate gradation and particle characteristics for a controlled experiment. They
investigated different methods of drawing maximum density lines that produces the
densest packing.
Hveem first noticed that there was a relationship between the gradation of the
mineral aggregate and the amount of oil required to maintain a consistent color and
appearance of the mixture. Then he realized that having the proper oil content did not
guarantee good performance relative to rutting. This led to the development of Hveem
stabilometer test to evaluate the ability of HMA mixtures to resist the shear forces applied
by traffic loads. The basis for selecting the optimum asphalt content in the Hveem method
is to use a well-graded aggregate with high friction and appropriate amount of fines and
add as much asphalt cement as the mixture will tolerate without losing stability. A
27
detailed account of the evolution of the Hveem mixture design method was given by
Vallerga and Lovering (1985). Details of specimen preparation and testing by Hveem
apparatus are given in ASTM D1561 and D1560, respectively.
A detailed introduction of the Marshall mix design method is given by a few
researchers (Foster 1982; White 1985). The acceptance tests on the aggregates and
asphalt cement are conducted at the beginning. If the materials pass these tests, the test
procedure for the Marshall method can be performed (The Asphalt Institute 1993). The
test protocol calls for fabricating 18 test specimens for the volumetric analysis. Three
loose mixture specimens are made near the optimum asphalt content to measure Rice
specific gravity or theoretical maximum density (TMD). Three compacted specimens
each are prepared at five different asphalt contents with 0.5 percent increments with at
least two above the estimated optimum asphalt content and two below the estimated
optimum. The approximate optimum asphalt content can be based on experience or
specific guide.
The amount of compaction is selected based on traffic level. The test specimens are
compacted using a Marshall hammer with 35, 50 or 75 blows per side for light, medium
or heavy traffic, respectively. The bulk specific gravity is then measured for each
specimen after proper handling. The Rice Specific Gravity (Gmm) is calculated for each of
the asphalt content mixes using the equation (relationship between Gmm, Gse and Pb).
Other volumetric parameters, including air voids (VTM), VMA, and VFA, are also
calculated using the related equations presented earlier. The compaction procedure will
produce specimens with decreasing air voids as a function of increasing asphalt binder
content. The compacted specimens are usually 4 inches (100 mm) in diameter and 2.5
inches (63.5 mm) in height.
These specimens are then used for conducting the Marshall stability and flow test.
The test is performed at 140˚F (60˚C), which is considered a critical temperature for
permanent deformation. A load at 2 inches/minute (50.8 mm/min) is applied to the
specimen until the maximum load is reached. The stability is the maximum load in
pounds (Newtons) and the flow is the deformation in 0.01 inch (0.25 mm). The stability
generally increases with increasing asphalt content, reaches a peak, and then decreases.
The asphalt content at the peak stability value is a good indicator of optimum binder
28
content based on the idea that constant compaction effort across varying asphalt content
produces a maximum stability value near the optimum asphalt content. In addition, the
percent VMA will decrease with increasing asphalt content, reach a minimum, and then
increase. Since the mixture strength increases as the VMA decreases, the mix with
minimum VMA should have the maximum strength or stability at the optimum asphalt
content. Finally the optimum asphalt content is determined by averaging the three asphalt
contents at maximum stability, maximum density and midpoint of the specified air voids
range (typically 4 percent). All parameters are checked at this optimum binder content for
acceptability according to the Marshall mix design criteria.
2.3.3.1 Bailey Method
The Marshall mix design method was broadly used in the United States before the
1980s. It provides some guidance on the use of coarse and fine mixes. However,
numerous trial and error process still have to be conducted to obtain a proper aggregate
blend. The Bailey method gives a good starting point for mix design when adjustments
are required to improve the volumetric properties of the mix (Vavrik et al. 2001; Asphalt
Institute and the Heritage Group 2005). The detailed methodology is summarized herein.
The Bailey mix design method was originally developed by Robert. D. Bailey in the
early 1980s. The primary purpose of this methodology is to control the mix properties
during construction including volumetric properties, segregation, workability, and
compatibility by focusing on aggregate packing. There are four key principles in the
Bailey method:
1. Determine the coarse and fine aggregate. The coarse fraction creates voids and the
fine fraction fills in the voids.
2. Analysis of coarse fraction which influences the packing of fine fraction.
3. Analysis of coarse part of the fine fraction, which relates to the packing of the
overall fine fraction in the blend.
4. Analysis of fine part of the fine fraction, which relates to the packing of the fine
portion of the gradation in the blend.
29
Figure 2-11 shows the four principles on a typical gradation curve for a coarse
gradation mix. The Bailey method defines the break between coarse and fine fractions as
the Primary Control Sieve (PCS). The PCS is the closest sieve to the result of
0.22×NMPS, where the NMPS denotes the Nominal Maximum Particle Size, which is
equivalent to the Nominal Maximum Aggregate Size (NMAS) used in the Superpave
system. The Bailey method uses AASHTO T19 to determine the Loose Unit Weight
(LUW) and the Rodded Unit Weight (RUW) of each individual aggregate used in the
mix. The suggested Chosen Unit Weight ranges for each mix type are shown in Table 2-
2. It should be noted that Stone Mastic Asphalt (SMA) references the RUW condition of
coarse aggregate, while coarse-graded and fine-graded mixtures reference the LUW
condition. The combined blend evaluation for coarse-graded and fine-graded mixes is
shown in Figure 2-12 and Figure 2-13, respectively. SCS and TCS denote Secondary
Control Sieve and Tertiary Control Sieve, respectively.
100
90
80
70
60
50
40
30
20
10
0
Sieve Size (mm) Raised to 0.45 Power
% P
assin
g
Fine Fraction Coarse Fraction
K J I H G F E D C B A
Coarse-Graded
4
3
1
2
Figure 2-11: The four principles of Bailey method for coarse-graded mix
30
Table 2-2: Chosen unit weight ranges by mix type
Fine-Graded Coarse-Graded SMA
90% or less LUW 95% − 105%
LUW 110% − 125%
RUW
Figure 2-12: Combined blend evaluation for coarse-graded mixes.
�������������
�������
�������
�������
����
�����
������
����
�����
������
� ��� �!"#
$ ����� �%���
& ����� ��%���
' ����� ��%���
(�)�������������
(�)�������
(�)����
�����
������
Figure 2-13: Combined blend evaluation for fine-graded mixes.
The Coarse Aggregate ratio (CA), Coarse part of Fine Aggregate ratio (FAc), and
Fine part of Fine Aggregate ratio (FAf) can be calculated by the following equations:
31
sievehalfpassing
PCSpassingsievehalfpassingRatioCA
%100
%%
−
−= (2-9)
PCSpassing
SCSpassingRatioFAc
%
%= (2-10)
SCSpassing
TCSpassingRatioFA f
%
%= (2-11)
Table 2-3 shows the recommended values of the different ratios for coarse and fine
mixes. The Bailey mix design method provides a useful and practical approach for
evaluating volumetric properties and compactability and thus helps in better
understanding the aggregate structure in asphalt mixtures as well as quality control at the
plant or in the field.
Table 2-3: Recommended aggregate ratios
NMPS (mm) 37.5 25.0 19.0 12.5 9.5 4.75
Coarse 0.80−0.95 0.70−0.85 0.60−0.75 0.50−0.65 0.40−0.55 0.30−0.45 CA Ratio
Fine 0.6 − 1.0
Coarse 0.35 − 0.50 FAc Ratio
Fine 0.35 − 0.50
Coarse 0.35 − 0.50 FAf Ratio
Fine 0.35 − 0.50
2.3.3.2 Superpave Mix Design
Since the early 1980s, traffic volume and axle loads have been increasing remarkably
in the United States. There emerged a need to develop an improved mix design method
that could be used in various traffic conditions and environments. With this as a primary
goal, the Strategic Highway Research Program (SHRP) was initiated in 1988 and
completed in 1993 resulting in the Superior Performing Asphalt Pavement (Superpave)
System. The Superpave system consists of the following components: new grading
system for asphalt binder (performance graded (PG) grading system), consensus
32
properties of aggregate, new mix design procedure, and mixture analysis procedures
(FHWA 1995; Asphalt Institute 1995; TRB 1994).
The aggregate properties that are specified by the SHRP are the coarse and fine
aggregate angularity, flat and elongated particles, and sand equivalent results. The
angularity of aggregate is related to the shear strength of the HMA mix and thus
influences the rutting performance of HMA pavement. The coarse aggregate angularity is
determined by measuring the percentage of coarse aggregate particles with fractured
faces, whereas the angularity for fine aggregate is measured by determining the amount of
voids by the National Aggregate Association (NAA) flow test in accordance with
AASHTO TP33 Method A. Flat or elongated particles tend to lie flat or even break down
during compaction, which may affect the workability of the mixtures. In addition, flat or
broken aggregates will make the mixture VMA lower than designed or expected. The test
procedure for flat or elongated particles is specified in ASTM D4791, “Flat or Elongated
Particles in Coarse Aggregate”. The clay content is related to the stripping problem of the
mixture. Excessive amounts of clay may result in a poor bond between the asphalt binder
and aggregate. The clay content is measured by the sand equivalent test conducted in
accordance with AASHTO T176 or ASTM D2419.
Aggregate blend is one of the most important factors to consider in HMA mix design
to ensure that a satisfactory gradation skeleton is obtained and volumetric requirements
are met. According to the definition given by Superpave, at least 90 − 100 percent of the
aggregate must be finer than the nominal maximum aggregate size. Control points are
also set on the 2.36 mm (No. 8) and the 0.075 mm (No.200) sieve sizes. Superpave
requires the aggregate gradation curve to be within the control limits. Another part of the
Superpave specification for gradation curve is the restricted zone. The restricted zone
provides a guide to help avoid too much natural rounded sand being used in the mixture
and to help ensure minimum VMA requirements are met. However, Kandhal et al. (2001)
showed that potential good mixes may get rejected because their gradations pass through
the restricted zone. Chowdhury et al. (2001) found that there is no relationship between
the restricted zone and permanent deformation when crushed aggregates are used in the
mixture design. In practice, there are aggregate blends that pass through the restricted
zone while not using an excessive amount of rounded aggregates that meet the minimum
33
VMA requirements. A typical gradation curve along with the corresponding gradation
limits is shown in Figure 2-14.
The Superpave Gyratory Compactor (SGC) is a key component of the Superpave mix
design. The compaction equipment is designed to compact HMA samples to conditions
similar to those obtained in the field under traffic loads. The compaction effort is
controlled by three parameters: vertical pressure, angle of gyration, and number of
gyrations. In the Superpave mix design procedure the vertical pressure is set at 600 kPa
(87 psi), the angle of gyration is set at 1.25˚, the rate of gyration is 30 revolutions per
minute, and the number of applied gyrations depends on the design traffic level and
average high air temperature. Ni, Nd, and Nm are three numbers of gyrations specified for
the Superpave Gyratory Compactor. Ni is N-initial which measures the mixture
compactibility to ensure that the mix will not compact too quickly. Nd is N-design and
represents the number of gyrations required to produce a density in the mixture similar to
that ultimately obtained in the field when subjected to traffic. Nm is the N-maximum and
is the number of gyrations that provides a compacted density which should not be
exceeded in the field, for a too-densified mix will result in low VMA which may cause a
rutting problem. Generally Nd is determined based on lab and field test data through
comparison of in-place density and laboratory density at various numbers of gyrations. Ni
and Nm are then given by the following equations:
45.0)( di NN = (2-12)
10.1)( dm NN = (2-13)
Superpave defines the optimum asphalt content as the one that produces 4 percent air
voids at Nd. An estimate of the optimum asphalt content is selected from aggregate blend
trials. Three samples each are prepared at 0.5% below estimated optimum, at estimated
optimum, at 0.5% above, and at 1.0% above estimated optimum. All samples are put into
an oven to be aged at 135˚C (275˚F) allowing absorption of the asphalt cement into
aggregate pores before compaction. Each sample is compacted up to Nm. The estimated
bulk specific gravity at each number of gyrations is calculated by using the specimen
weight, diameter (6 inches, or 150 mm) and height which is measured and recorded
during the compaction process. This estimated density is slightly lower than the actual
34
density because usually the raw compacted specimens have many surface voids on the
top, bottom and cylindrical sides. The actual bulk specific density at Nm is measured by
weighing the samples in air and water. The correction factor is calculated at Nm by the
following equation:
)(
)(
estimatedG
actualGCF
mb
mb= (2-14)
The actual bulk specific gravity at Ni and Nd can be back-calculated using the
correction factor and the estimated density at the corresponding number of gyrations. The
theoretical maximum density (TMD) is measured from the rice test on loose mixtures.
The air voids can then be determined by knowing the Gmm and the actual Gmb at various
compaction levels. The air voids of three samples at each asphalt content are averaged
and plotted on a graph. The actual optimum asphalt content that provides 4 percent air
voids at Nd can be determined by interpolation. It is required that the air voids be greater
than 11% at Ni and greater than 2% at Nm. Other requirements that must be satisfied
include VMA and VFA according to Superpave specifications.
#
20
0
#1
00
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
Gradation Curve
Maximum Density Line
Figure 2-14: Illustration of gradation requirements for 12.5 mm (1/2 in.) nominal size
35
2.4 Mechanical Tests for Characterization of Asphalt Mixtures
2.4.1 Introduction
Flexible pavements constructed with asphalt mixtures are subjected to a wide range
of traffic loads and environmental conditions. Characterization of HMA mixtures is the
measurement and analysis of their response to these conditions. The performance of any
HMA mixture is dependent upon the entire pavement structure, and the structural
capacity of the pavement layers is dependent on the quality of materials and their
compositions in the mixture. An understanding of fundamental engineering properties of
HMA mixtures is required for satisfactory performance of pavement structures in service.
There have been many testing protocols developed in the laboratory for measuring
mixture properties related to thermal cracking, fatigue cracking, and permanent
deformation over the past few decades. These test procedures are used to evaluate the
distress of HMA mixtures under various types of load at different loading rates and
temperature levels similar to those encountered in the field.
The test methods can be categorized into in-place and laboratory tests. The testing
program developed for this study is conducted on lab-prepared HMA specimens with a
complete set of equipment. Laboratory mechanical tests can be further grouped based on
the test mode, such as direct tension, indirect tension, compression, flexural, shear, and
torsion. Pavement design using elastic layer theory needs two elastic parameters for each
material layer used: Young’s modulus (stiffness) and Poisson’s ratio. In the NCHRP 9-19
report (2001), many tests had been proposed as a Simple Performance Test, including the
dynamic modulus test, the indirect tensile creep compliance test, resilient modulus test,
tensile strength test and other test methods. Details of the test methods utilized in this
study are discussed in the following sections.
2.4.2 Dynamic Complex Modulus Test (DMT)
Papazian (1962) was one of the first to present the concepts and definitions
concerning the dynamic complex modulus (E*) for characterizing the viscoelastic
behaviors of HMA mixtures. In the early 1970s, the Asphalt Institute selected the
36
dynamic complex modulus as one of the choices for the modulus test. The dynamic
modulus test was then specified in ASTM D3497-79 “Standard Test Method for the
Dynamic Modulus of Asphalt Mixtures”. The complex modulus test accounts not only for
the instantaneous elastic response without delayed elastic effects, but also the
accumulation of cyclic creep and delayed elastic effects with the number of cycles. The
test is usually conducted on cylindrical specimens subjected to a compressive sinusoidal
vertical load at a given temperature and loading frequency. After the specimens are well
prepared, they are placed in a controlled temperature cabinet and brought to the specified
test temperature. The conditioned specimen is then placed into the loading apparatus and
the strain gauge wires are connected to the measurement system. A hardened steel disk is
put on both top and bottom of the specimen and centered under the loading apparatus.
The electronic measuring system is adjusted and balanced as necessary. The sinusoidal
loading is applied to the specimen without impact and with loads varying between 0 and
35 psi (0 to 241.5 kPa) for each load application for a minimum of 30 seconds and a
maximum of 45 seconds at temperatures of 41, 77, and 104°F (5, 25, and 40°C) and at
loading frequencies of 1, 4, and 16 Hz for each temperature. Both the loading stress and
axial strain are monitored during the test. One piece of test equipment that is required is a
loading device capable of transmitting haversine waveforms at a frequency range of 0.1 to
20 Hz and a maximum stress of about 100 psi (690 kPa). Because of the importance of
testing asphalt mixes at various temperatures, some form of temperature control system is
required. This device can be either external or internal to the loading device depending
upon how the test is conducted. Strains are usually obtained by use of bonded wire strain
gages or calculated from vertical deformations measured with Linear Variable
Displacement Transducers (LVDTs). Test specimens usually have a minimum 4-inch
diameter and a height-to-diameter ratio of two. A minimum of three specimens is
required for testing. A conceptual schematic of the dynamic complex modulus test is
shown in Figure 2-15. A typical illustration of set-up for dynamic modulus is shown in
Figure 2-16.
It is critical to maintain an appropriate stress level during the dynamic modulus test
to obtain a proper strain response, since the concept of dynamic complex modulus is
based on the assumption of linear viscoelasticity of HMA mixtures. The concept of
37
material linearity is based on the following two conditions of superposition and
homogeneity according to Schapery (1972):
)()()( 2121 σεσεσσε +=+ (2-15)
)()( σβεβσε = (2-16)
Where,
�, �1, �2 = stress histories,
� = strain histories,
� = constant
The original ASTM 3497 protocol for the dynamic modulus test only defines the
haversine load level (0 to 35 psi or 241 kPa) as long as the deformation is not excessive
(2500 micro-strains). Since the 1990s, the dynamic modulus (|E*|) has been one of the
most widely-used structural parameters for asphalt cement mixtures used in Mechanistic-
Empirical (M-E) structural pavement design procedures. It becomes a primary material
input to compute stress, strain, rutting and cracking damage in flexible pavement systems.
Witczak et al. (1996) reviewed the approach for conducting the dynamic modulus test and
suggested strain amplitudes of 75 to 200 micro-strains to maintain the HMA material in
linear viscoelastic range. The dynamic modulus test protocol was then modified and
developed in the National Cooperative Highway Research Program (NCHRP) Project 9-
19 and 1-37A and has been standardized in the new Mechanistic-Empirical Pavement
Design Guide (PDG) as AASHTO Provisional Standard TP62-03: “Standard Method of
Test for Determining Dynamic Modulus of Hot-Mix Asphalt Concrete Mixtures.” This
test protocol calls for a minimum of two replicate specimens to be tested at temperatures
of 14, 40, 70, 100, and 130˚F (-10, 4.4, 21.1, 37.8, and 54.4˚C) and loading frequencies of
25, 10, 5, 1.0, 0.5, and 0.1 Hz. The dynamic modulus tests carried out in NCHRP 9-19
project (Witczak et al. 2002) selected the stress levels for a given test temperature to
produce resilient strains of less than 100 micro-strains
As introduced earlier, |E*| is a fundamental property defining the response of HMA
mixtures and strongly influences the performance of asphalt pavement. Huang (1993)
illustrated the theory of complex modulus by the use of the Kelvin model, shown in
38
Figure 2-18, subject to a sinusoidal loading, which can be represented by a complex
number:
tietit ωσωσωσσ 000 )sin()cos( =+= (2-17)
in which �0 is the stress amplitude and is angular velocity, which is related to the
frequency by
fπω 2= (2-18)
The governing differential equation can be written as
tieE
t
ωσεε
λ 011 =+∂
∂ (2-19)
The solution of Equation (2-19) can be expressed as
)(
0
ϕωεε −= tie (2-20)
in which � is the strain amplitude and is the phase angle by which the strain lags behind
the stress, as shown in Figure 2-17. Substituting Equation (2-20) to Equation (2-19) gives
tititi eeEei ωϕωϕω σεωελ 0
)(
01
)(
01 =+ −− (2-21)
After canceling eit on both sides of Equation (2-21), and equating the real term to �0 and
imaginary terms to zero, the following two equations are obtained to solve �0 and
00101 cossin σϕεϕωελ =+ E (2-22)
0sincos 0101 =− ϕεϕωελ E (2-23)
The solutions of Equation 2-22 and Equation 2-23 are
2
1
2
1
00
)( ωλ
σε
+=
E (2-24)
1
1tanE
ωλϕ = (5b) (2-25)
39
It can be seen from Equation (2-25) that for elastic materials �1 = 0, so = 0; while for
viscous materials E1 = 0, so = �/2. The complex modulus E* is defined as
)(
0
0*
ϕω
ω
ε
σ
ε
σ−
==ti
ti
e
eE (2-26)
Or ϕε
σϕ
ε
σ
ε
σsincos
0
0
0
0* iE +== (2-27)
The dynamic modulus is the absolute value of the complex modulus. Experimentally, the
dynamic modulus is determined as the ratio of the applied stress amplitude to the strain
response amplitude under the sinusoidal loading (Figure 2-17), as calculated by the
following equation (Yoder and Witczak 1975):
0
02
0
02
0
0* )sin()cos(||ε
σϕ
ε
σϕ
ε
σ=+=E (2-28)
The phase lag is simply the angle at which the �0 lags �0:
)360()360( �� ⋅⋅=⋅= ftt
tlag
p
lagϕ (2-29)
Where
tlag = time lag between a cycle of sinusoidal stress and a cycle of strain
tp = time period of a stress cycle (seconds)
f = frequency of the dynamic load (in Hz)
By definition, the complex modulus E* can be expressed as the following:
EiEE ′′+′=∗ (2-30)
Where E’ refers to the real part of the complex modulus; it is the storage modulus that
describes the elastic component of the complex modulus:
40
)cos()cos(||0
0* δε
σδ ⋅=⋅=′ EE (2-31)
and E” refers to the imaginary part of the complex modulus; it is the loss modulus which
describes the viscous component of the complex modulus:
)sin()sin(||0
0* δε
σδ ⋅=⋅=′′ EE (2-32)
The storage and loss moduli can be determined by measuring the lag in the response
between the applied stress and the measured strains. This lag, referred to as the phase
angle , shown in Figure 2-17, can also be determined by Equation 2-33:
)(tan 1
E
E
′
′′= −ϕ (2-33)
The complex modulus can also be written as:
ϕ⋅⋅= ieEE || ** (2-34)
Results
Deformation
Sin. Load
1,4,16 Hz
Temperature
5,25,40°C
Schematic of Dynamic Complex Modulus
Test (ASTM D3497-79)
Figure 2-15: Conceptual schematic of dynamic complex modulus test
41
Loading Head
Load Cell
LVDT
Specimen
Environment
Chamber
Figure 2-16: Typical test set-up for dynamic complex modulus
tωσ sin0
)sin(0 ϕωε −t
Figure 2-17: The schematic components of dynamic complex modulus test
42
�� ��
tie
⋅ωσ 0
Figure 2-18: Kelvin model under sinusoidal loading.
2.4.3 Indirect Diametral Tests in Tension
The indirect diametral test is used extensively by state highway and other agencies
for routine tests. The 1986 AASHTO Pavement Design Guide, which recommended the
use of resilient modulus to characterize pavement materials, has led to accelerated use of
this type of test. This test is usually conducted on cylindrical specimens subjected to a
compressive load along two opposite generators resulting in a relatively uniform tensile
stress acting perpendicular to and along the diametral plane of the applied load. A
splitting failure generally occurs as a result along the diametral plane (Figure 2-19). If a
repetitive pulsating load is applied diametrically to the sample, the dynamic load results
in dynamic deformations across the horizontal diametral plane. The transducers mounted
on each side of the horizontal specimen axis record these deformations. The resilient
modulus (MR) of HMA mixtures can be determined by the dynamic load and
deformation. The indirect diametral test is originally specified by ASTM D4123-82
Standard Test Method for Indirect Tension Test for Resilient Modulus of Bituminous
Mixtures, which was withdrawn in 2003. The resilient modulus (MR) has been used in the
AASHTO Design Guide (AASHTO 1993) since 1993. The resilient modulus laboratory
test procedure is described in AASHTO TP 31. The test is defined as a repetitive 0.1
second haversine load followed by a 0.9 second rest period, continued at 1 Hz intervals.
Many empirical relationships have been developed throughout the years relating MR to
other tests like the California Bearing Ratio (CBR) and the Marshall stability test
43
(AASHTO 1993), since it has long been considered the defining characteristic for HMA
layers.
Figure 2-19: Indirect diametral test during loading and at failure
After the specimens were well prepared, they were placed in a controlled temperature
cabinet and brought to the specified test temperature. The specimen was placed into the
loading apparatus and the loading strips were positioned to be parallel and centered on the
vertical diametral plane. The specimen was preconditioned by applying a repeated
haversine or other suitable waveform load without impact for a minimum period
sufficient to obtain uniform deformation readout. Depending upon the loading frequency
and temperatures, a minimum of 50 to 200 load repetitions is typical; however, the
minimum for a given situation must be determined so that the resilient deformations are
stable. Resilient modulus evaluation will usually include tests at three temperatures, for
example, 41, 77, and 104°F (5, 25, and 40°C), at one or more loading frequencies. The
horizontal and vertical deformations were continuously monitored during the test.
The required test equipment is a loading device capable of applying a load pulse over
a range of frequencies, load durations and load levels. Some form of temperature control
system is required. The temperature-control system should be capable of control over a
temperature range from 41 to 104°F (5 to 40°C). The measurement and recording system
44
should include sensors for measuring and recording horizontal and vertical deformations.
The values of vertical and horizontal deformation can be measured by linear variable
differential transducers (LVDTs) or other suitable devices. LVDTs should be at mid-
height opposite each other on the specimen’s horizontal diameter. A metal loading strip
with a concave surface having a radius of curvature equal to the normal radius of the test
specimen is required to apply load to the specimen. The specimens should have a height
of at least two inches and a minimum diameter of four inches for aggregate up to one inch
maximum size, and a height of at least three inches and a minimum diameter of six inches
for aggregate up to 1.5 inches maximum size.
Hondros (1959) derived the stress equations to model the actual test conditions as
well as to determine Young’s modulus and Poisson’s ratio of the material. The theoretical
distribution of stresses for a concentrated load is shown in Figure 2-20 and Figure 2-21.
Roque and Buttlar (1992) developed a measurement and analysis system to
determine asphalt concrete properties, primarily thermal cracking, using the indirect
tensile testing mode, which was incorporated in AASHTO TP9-96, Standard Test Method
for Determining the Creep Compliance and Strength of Hot Mix Asphalt (HMA) Using
the Indirect Tensile Test Device. They proposed the Gauge-Point-Mounted device to
measure horizontal and vertical deformations across a gauge length of 25.4mm (1 inch).
Poisson’s ratio was also obtained from the horizontal and vertical deformations instead of
using assumed values. Correction factors from 3-D finite element analysis were used to
account for: (1) the effect of specimen bulging on deformation measurement, and (2)
approximation of 2-D plane stress assumption. Roque et al. (1997) made further
modifications and improvements on the SHRP IDT system for characterizing relevant
asphalt mixture properties. The test procedures and data reduction methodologies were
also summarized in Long-Term Pavement Performance (LTPP) Protocol P07 (2001): Test
Method for Determining the Creep Compliance, Resilient Modulus and Strength of
Asphalt Materials Using the Indirect Tensile Test Device.
45
� �����������
�����������������
2
22
22
4
42��
���
�
+
−
⋅⋅ xd
xd
dt
P
π
��
���
�−
+⋅⋅
−1
4
4222
2
xd
d
dt
P
π
Figure 2-20: Theoretical stress distribution on horizontal diametral plane for indirect tensile test (After Yoder et al. 1975)
dt
P
⋅⋅π
2
��
���
�−
++
−⋅
−
dydydt
P 1
2
2
2
22
π
Figure 2-21: Theoretical stress distribution on vertical diametral plane for indirect tensile test (After Yoder et al. 1975)
46
2.4.4 Relationship between Resilient and Dynamic Moduli
The resilient modulus (MR) of the mixture is calculated by using the measured
horizontal and vertical deformations (Hondros 1959; Kennedy 1977), and is defined as
the ratio between applied stress and recoverable strain. Several different test protocols
and data reduction methods were developed for determining the indirect diametral
resilient modulus of asphalt concrete mixtures (ASTM 4123-82 1982; Barksdale et al.
1997; AASHTO TP-31 1996; Roque and Buttlar 1992; Buttlar and Roque 1994; SHRP-
LTPP P07 2001; Witczak 2004). The dynamic complex modulus is a viscoelastic
response of an asphalt concrete mixture under sinusoidal loading conditions at different
test temperatures and loading frequencies, which accounts for both elastic and viscous
effects of the material. Several different test protocols and methods were also developed
for determining the dynamic complex modulus of asphalt concrete mixtures (ASTM
D3497-79 1979; Witczak et al. 2000; Witczak et al. 2002; Pellinen and Witczak 2002). In
addition, the dynamic complex modulus measurements were also conducted in Europe in
recent research studies to evaluate the mechanical properties of bituminous materials (Di
Benedetto et al. 2001, 2004).
A detailed comparison of key differences between the dynamic complex modulus test
and the indirect diametral resilient modulus test for asphalt concrete mixtures was
summarized in a position paper by the NCHRP 1-37A 2002 Project research team
(Witczak 1999). Basically, the primary difference between the resilient modulus test and
dynamic complex modulus test for asphalt concrete mixtures is that the former uses
loading of any waveform with a given rest period, while the latter applies a sinusoidal or
haversine loading with no rest period; hence, no delayed elastic rebound would occur
during the test. The transition from the resilient modulus test to the use of the dynamic
complex modulus test for design of flexible pavement structures has hardly been smooth.
The potential impact of adopting the dynamic complex modulus for implementing the
new AASHTO M-E Design Guide is tremendous for state transportation agencies such as
the Florida Department of Transportation (FDOT). The IDT has traditionally been used to
characterize the HMA mixtures for flexible pavement design in Florida, and the test
method has been shown to be both an expedient and a reliable way of obtaining mixture
47
properties from field cores. In response to the need, a major effort was undertaken by the
FDOT to characterize Florida HMA mixtures using the dynamic complex modulus
(Birgisson et al. 2004). Despite the fundamental differences between the resilient
modulus and dynamic complex modulus (Witczak 1999; Drescher et al. 1997; Zhang et
al. 1997; Kim et al. 2004), a number of research studies were attempted in order to
establish a direct correlation between the resilient modulus and dynamic complex
modulus of asphalt concrete mixtures (Kim et al. 2004; Birgisson et al. 2004; Loulizi et
al. 2006). Birgisson et al. (2004) developed testing and analysis procedures to accurately
determine the tensile dynamic complex modulus from the SHRP IDT tests. The dynamic
complex modulus was found to be correlated with resilient modulus and testing frequency
for the range of testing temperatures and frequencies. Loulizi et al. (2006) conducted a
comparison study on the dynamic complex modulus and resilient modulus tests, and they
found a strong relationship between the dynamic complex modulus performed at 5 Hz
and the resilient modulus performed at a loading time of 0.03 seconds.
Other alternative approaches have also been attempted in order to determine the
dynamic complex modulus from the IDT test with modified loading conditions using the
theory of viscoelasticity (Drescher et al. 1997; Zhang et al. 1997; Kim et al. 2004).
Recently, an analytical method of calculating resilient modulus from the dynamic
complex modulus was also proposed (Lacroix et al. 2007). The proposed theoretical
prediction involved the application of multiaxial linear viscoelastic theory to linear elastic
solutions for the IDT test. The proposed approach could provide reasonable estimates of
the resilient modulus from the dynamic complex modulus of the asphalt concrete
mixtures. The accuracy of the prediction was not affected by assuming a constant
Poisson’s ratio.
2.5 HMA Fracture Mechanics Concepts
2.5.1 Background
It is commonly considered necessary to study real cracking growth mechanisms in
order to essentially understand the crack damage in HMA. Research conducted by Roque
et al. (2002) on top-down cracking of asphalt pavement indicated that the tearing-apart
48
effect from vehicle tires can cause a certain level of tensile stress leading to cracking of
the pavement surface and crack propagation. The conventional linear elastic fracture
mechanics presume that there are intrinsic flaws in a material. A crack initiates from the
flaws and is propagated continuously under a critical loading condition. The crack growth
rate of linear elastic materials is assumed to follow Paris’s law:
nKA
dN
da)(∆= (2-35)
where a is crack length, N is number of load repetitions, K is stress intensity factor, and A
and n are constants.
However, Jacobs (1996) investigated the fracture mechanics for HMA mixtures and
pointed out that the non-homogeneity of asphalt concrete could cause the discontinuity of
crack propagation in the mixture. It was shown that a crack in asphalt concrete grows
discontinuously. Zhang (2000) and Zhang et al. (2001) found that the continuous crack
growth assumption can not characterize the cracking performance of asphalt concrete
mixtures observed in the field, which occurs in a stepwise manner rather than a
continuous one. They indicated that there is a specific threshold below which the damage
is considered to be on a micro scale and healable with a rest period or temperature
increase, whereas the damage would be permanent on a macro scale when the threshold is
reached or exceeded.
Shen et al. (2005) introduced the Plateau Value (PV) concept using the Ratio of
Dissipated Energy Change (RDEC) to show its relationship with damage and failure at
normal or low strain levels (70 – 500 micro-strains). Carpenter et al. (2006) applied this
RDEC approach to analyze healing and HMA fatigue behavior at normal and low strain
levels using the standard four-point bending beam fatigue test procedure specified in
AASHTO standards (21): constant strain at 500 micro-strains, 20±0.5˚C temperature,
10Hz frequency with haversine load waveform, etc. Healing was observed at low strain
conditions or long rest period, and hence may increase the fatigue life of HMA material.
2.5.2 HMA Fracture Mechanics Model from IDT
49
An HMA fracture model for predicting pavement cracking was developed by Zhang
et al. (2001) and Roque et al. (2002, 2004). Crack growth laws were identified for asphalt
mixtures using IDT. The linear elastic finite element method was used to simulate the
IDT specimens at different cracking lengths. They established a relationship between the
theoretical crack length and the deformation measured between the vertical gage points.
Besides the three types of regular IDT tests (resilient modulus, creep compliance, and
tensile strength), another type of fracture test was performed. The specimens for the
fracture test have 150 mm diameter and 25 mm thickness with an 8 mm hole in the
center. The fracture test was conducted under the same load mode as MR test but at higher
deformation levels in order to determine the crack growth characteristics of the specimen.
The test was performed at 10˚C. The repeated load was applied until the specimen failed.
The crack growth rate parameters for Paris law ( nKAdNda )(/ = were determined by the
following steps:
• Establish the relationships of cracking length (a) versus horizontal deformation
(�H) and stress intensity factor (K) using theoretical finite element analysis.
• Establish a relationship between horizontal deformation (�H) and loading
repetitions (N) from fracture test.
• Incorporate the theoretical calculation into the test results to develop a
relationship between cracking length growth rate (da/dN) and stress intensity
factor (K).
• Obtain the fracture parameters, A and n, by regression analysis.
The regression models were used to evaluate the mixture cracking resistance.
Discrepancies between laboratory tests and field performance were observed. Regression
analyses were conducted to determine the relationship between the mixture properties
(tensile strength, m-value, fracture energy and resilient modulus) and measured crack
growth rates. It was determined that dissipated creep strain energy to failure is not
dependent on mode of loading and could be used as a threshold to explain the
inconsistency of lab and field observations, as shown in Figure 2-22 (Roque et al. 2002).
There are two possible reasons for fracture to occur: 1) a number of continuous repeated
loads can cause damage accumulation due to creep strain energy, and fracture can develop
if the DCSE threshold is reached, even when the loading stress is below the tensile
50
strength. It also should be noted that the mixture may never crack if the healing effect
makes the induced dissipated energy below the threshold regardless of the load
repetitions; 2) fracture may occur if any large single load exceeds the fracture energy (FE)
threshold. Case 3 in Figure 2-22 shows that cracking would not occur during a single load
application unless the upper FE threshold is exceeded, even when the dissipated energy
(DE) is exceeded.
Figure 2-22: Illustration of potential loading condition (Roque et al. 2002)
e0 ef Strain
St
MR
EE
Dissipated Creep Strain Energy
(DCSE)
MR
Str
ess
Figure 2-23: Determination of fracture energy and dissipated creep strain energy
51
The concepts of fracture energy (FE) and dissipate creep strain energy (DCSE) were
introduced in the model to account for the pavement structure crack performance. The
two energy values are determined using the tensile strength test along with the resilient
modulus test. The schematics used to calculate these limits are shown in Figure 2-23. The
values are calculated by the following equations:
� ⋅=tS
dFE0
εσ (2-36)
R
t
fEEM
S=−= 0εεε (2-37)
R
t
tEEM
SSEE
⋅=⋅⋅=
2
)(
2
12
ε (2-38)
EEFEDCSE −= (2-39)
Where
FE = Fracture Energy, total energy applied to the specimen till fracture
EE = Elastic Energy, recoverable energy
DCSE = Dissipated Creep Strain Energy absorbed by the specimen prior to fracture
St = tensile strength of the mixture
�f = failure strain
MR = resilient modulus of the mixture
It was shown that the dissipated creep strain energy at failure (DCSEf) is the
threshold that controls crack propagation, which can be described as a step function
consisting of crack initiation (DCSE below the threshold) and crack propagation (DCSE
over the threshold). It was also found that micro-damage in HMA can be healed while
macro-damage cannot be healed at rest period or temperature increase conditions. DCSE
per cycle and number of load repetitions can be further estimated using the following
equations:
))100((20
1/ 1
1
−= m
AVEAVEnCreepStrai mDcycleDE σσ (2-40)
)//( cycleDCSEDCSEN ff = (2-41)
52
where �ave is the average stress near the crack tip, m and D1 are power law parameters
obtained from the creep compliance test, and Nf is the number of cycles to failure.
Villiers (2004) used the HMA Cracking Model along with the IDT sweep of tests to
evaluate the sensitivity of Superpave mixtures with regards to cracking performance. The
mixtures were tested at 10˚C to determine the cracking performance when subjected to
the Acceptable Variances. A statistical evaluation was conducted to examine the variation
in the IDT testing parameters. Significant variation was observed for all the IDT
parameters which were consistent with research conducted by Roque et al. (2004). It was
found that the average values used from the IDT test parameter could be used to
distinguish between pavements that exhibited top-down cracking and those that did not.
Roque et al. (2004) showed that cracking performance of HMA is complex and
controlled by multiple mixture properties. The Energy Ratio concept was derived as a
fundamental material property using the HMA Fracture Mechanics Model. It is defined as
the ratio of dissipated creep strain energy threshold of the mixture to the minimum
dissipated creep strain energy required, which can be determined from Superpave IDT
including resilient modulus, creep compliance, and tensile strength tests. Nf of 6000 was
set as the critical value that distinguish mixture performance. The equations to calculate
the Energy Ratio are presented below:
1
98.2 Dm
DCSEaER
f
⋅
×= (2-42)
81.3 1046.2)36.6(0299.0 −− ×+−⋅⋅= tSa σ (2-43)
Where
� = Tensile stress of the asphalt layer in psi (pavement structure)
St = Tensile strength in MPa (IDT tensile strength test)
The other parameters are the same as those defined earlier. The HMA fracture mechanics
were implemented to examine all test sections, based on which performance criteria of
ER greater than 1 and DCSE greater than 0.75 were defined to evaluate cracking
performance. They showed that no single property can be an accurate performance
indicator since fracture properties are interrelated as a system. The Energy Ratio appeared
53
to be a suitable parameter for evaluating top-down cracking situations of sections within a
pavement system at low in-service temperatures.
Kim (2005) developed an HMA thermal fracture model based on the same principle
and failure criteria used in the HMA fracture model introduced above. The Superpave
IDT tests were designed at three temperatures (0, 10, and 20˚C) which are typical low in-
service temperatures in Florida. The performance evaluation of the model showed
potential to reliably evaluate the performance of asphalt mixtures subjected to thermally
induced damage.
54
CHAPTER 3
MATERIALS AND EXPERIMENTAL PROGRAM
3.1 General
The two methods of measuring dynamic properties of HMA in this research study
were the triaxial dynamic modulus test (DMT) and indirect diametral tension test (IDT).
Both of these tests were reviewed in more detail in Chapter 2. Originally, the dynamic
modulus test was specified by ASTM D3497-79 Standard Test Method for Dynamic
Modulus of Asphalt Mixtures, while the IDT resilient modulus test was specified by
AASHTO TP31-94 and ASTM D4123-82. This study adopts the SHRP IDT Testing and
Analysis System (Roque et al. 1997) to measure the resilient modulus, creep compliance,
and tensile strength, and the NCHRP 9-29 Equipment Specifications for the Simple
Performance Test System to measure the complex dynamic modulus for all asphalt
concrete specimens. A complete dynamic testing system was acquired to perform the
temperature-controlled dynamic tests to determine the engineering properties of Florida
HMA mixtures. In this study, a Servopac Gyratory Compactor and an Interlaken Asphalt
Test System were used to compact the asphalt mixture and measure the dynamic response
of asphalt concrete, respectively. The laboratory experimental program consists of two
parts.
The first part of the experimental program involved two standard mix designs as
control mixes: Two modified gradations were designed for each control mixture while
using the same base asphalt binder (PG 67-22). In addition, each of the standard
Superpave mixture was modified using three levels of SBS polymer asphalt binder
instead of the original asphalt to evaluate SBS polymer effects on fracture mechanics
properties of asphalt concrete mixes. Therefore, the overall experimental program in this
part of the study involved twelve HMA concrete mixtures. The second part of the
experimental program was developed to study the relationship between the dynamic
complex modulus test and the indirect resilient modulus test. Twenty (20) standard
Florida HMA Superpave mixtures approved by FDOT were selected for determination of
55
the resilient modulus and the dynamic modulus of the HMA mixtures. All specimens
were prepared at targeted optimum air voids of 4%.
The physical properties of the materials used, including their aggregate properties,
aggregate gradation, asphalt binder characteristics, and mixture design series, will be
presented in detail according to the purpose of the studies.
3.2 Mix Designs and Materials
One Georgia granite mix (SP 04-3034A, TL-D, Ga553), referred to as “F2C”, and
one South Florida Limestone mix (LD 02-2529A, TL-D, SFL), called F4C, were selected
as the control mixes for the fracture mechanics tests. The two Superpave mix designs are
commonly used in Florida and approved by Florida Department of Transportation
(FDOT). They are both coarse mixes with the gradation curves passing below the
Superpave restricted zone, which were selected with the intention of making adjustments
to their coarse aggregate proportions to study the effect of gradation on mixtures’ fracture
mechanics properties.
The 20 mix designs for the dynamic modulus test and the resilient modulus test are
summarized in Appendix Table A-1 for information. These mix designs were contributed
by companies involved in the production and use of HMA in Florida. The summary table
consists of 20 mix designs in the format of 19 columns each. The data presented in the
table were sorted by mix design series number. Below is a description of each column:
1. The test series number related to the whole project
2. The mix design number used by the FLDOT for reference
3. The nominal maximum aggregate sizes, which are 19.0, 12.5, or 9.5 millimeters.
4. The type of the design mix, coarse or fine, determined by which side of the
forbidden zone the mix passes when plotted on the 0.45 Power chart.
5. The load level the design represents; Superpave has 5 levels, A-E.
6. The design applied either to a structural or friction course; different qualities are
desired for the two types of courses.
7 – 18. Columns 7-18 list the materials used in the design. The numbers are FDOT’s
reference numbers for sources.
56
19. The type of asphalt used in the design.
The grade of asphalt cement used in mixtures, as introduced in preceding chapters, is
one important factor that can affect the strength of asphalt concrete and amount of rutting
which occurs in the mix. In this part of the study, only one type of unmodified asphalt
cement, PG67-22 (AC-30), which is commonly used in Florida, was used for all mixtures
tested. The asphalt binder PG67-22 grading report is summarized in Table A-2. The
nominal maximum aggregate sizes for the mixtures tested are 19.0 mm, 12.5 mm, and 9.5
mm, respectively. The Superpave mixture designs were selected as they are commonly
used FDOT gradations and are known to perform well in the field. The types of
aggregates used are as follows:
• Granite Georgia-553 Georgia-206 Nova Scotia Granite (NS)
• Reclaimed Asphalt Pavement (RAP)
• Limestone North Florida Limestone (NFL) Mid Florida Limestone (MFL) South Florida Oolite/Limestone (SFL) Alabama Limestone (AL)
A summary of the 20 mix designs and the aggregate types is presented in Appendix
Table A-3. The 20 mixtures were tested for both complex modulus and indirect tensile
resilient modulus. The gradations of all mix designs, sorted by the mix design series
number and used in this study, are summarized in Table A-4 through Table A-7. The
corresponding gradation charts for all mix design series (sieve size raised to 0.45 power
mm) are presented in Figure A-1 through Figure A-8 for illustration.
3.3 SBS Polymer-modified Asphalt Binder
57
The grade of asphalt cement used in mixtures is one important factor that can affect
the strength of asphalt concrete and amount of rutting which occurs in the mix. The
unmodified asphalt PG 67-22 (AC-30), which is commonly used in Florida, was selected
as the base asphalt for both fracture mechanics tests and the dynamic modulus test. The
asphalt binder PG67-22 grading report is summarized in Appendix Table A-8. Three
levels of SBS polymer-modified asphalt are produced and used in the SBS effects study.
The SBS modified asphalt binder grading reports are summarized in Table A-9 through
Table A-11. The base asphalt and the other three levels of polymer-modified asphalt
(PMA) are listed as follows:
1. Control level
Base asphalt A (PG 67-22) + Aggregates = Control Mix
Mixtures are referred to as F2C and F4C.
The twenty Superpave mixes for DMT also used this base asphalt.
2. Mix plus 3% SBS polymer A
[Base asphalt A + 3% SBS polymer A] = PMA PG 76-22
PMA PG 76-22 + Aggregates = Mix with 3% PMA
Mixtures are referred to as F2P1 and F4P1.
3. Mix plus 4.5% SBS polymer A
[Base asphalt A + 4.5% SBS polymer A] = PMA PG 82-22
PMA PG 82-22 + Aggregates = Mix with 4.5% PMA
Mixtures are referred to as F2P2 and F4P2.
4. Mix plus 6% SBS polymer A
[Base asphalt B (softer) + 6% SBS polymer A] = PMA PG 82-22
PMA PG 82-22 + Aggregates = Mix with 6% PMA
Mixtures are referred to as F2P3 and F4P3.
58
0.45 Power Gradation Chart
#2
00
#
10
0
#5
0
#3
0
#1
6
#8
#4
3/8
"
1/2
"
3/4
"
1"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
F2 Control
F2G1
F2G2
Figure 3-1: Gradation curves for F2 and its trial adjustments
0.45 Power Gradation Chart
#
20
0
#1
00
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
F4 Control
F4G1
F4G2
Figure 3-2: Gradation curves for F4 and its trial adjustments
59
0
5
10
15
20
25
30
35
1/2 3/8 4
Sieve Size
Perc
ent
Reta
ined
F2 Control
F2G1
F2G2
Figure 3-3: Change of percent retained on top 3 sieves for F2 series
0
5
10
15
20
25
30
35
1/2 3/8 4
Sieve Size
Perc
ent
Reta
ined
F4 Control
F4G1
F4G2
Figure 3-4: Change of percent retained on top 3 sieves for F4 series
Figure 3-5: Cutting of raw specimen
60
3.4 Aggregates Gradation Modification
As mentioned in section 3.1, the two mix designs used as control mixtures are F2C
Georgia Granite (Ga553) and F4C South Florida Limestone. The nominal maximum
aggregate size for both F2C and F4C is 12.5 mm. They are commonly used FDOT
gradations and are known to perform well in the field. Both gradation curves of the two
control level mixtures go below the restricted zone and then zig upward across the
maximum density line at No. 4 sieve size, and continue a certain amount higher than the
maximum density line through the coarse sizes. The main purpose of this shape is to
assure sufficient air voids content of the asphalt mixture. In order to facilitate study of the
coarse aggregate effect on asphalt concrete mixtures, the coarse part (No. 4 sieve size and
larger) of each mix design was modified to two different compositions with the fine parts
of the mixes kept unchanged. The Job Mix Formulas of the original standard mix designs
and associated gradation modifications are summarized in Table A-12 and Table A-13.
The corresponding gradation charts for all mix design series (sieve size raised to 0.45
power mm) are presented in Figure 3-1 and Figure 3-2 for illustration. As shown in the
charts, the first set of modified gradations, named F2G1 and F4G1, have gradation curves
slightly lower than the original mix design in the coarse part, but still above the maximum
density line. The second set, denoted as F2G2 and F4G2, have gradation curves further
lower than the first modified one, and go below the maximum density line in the coarse
part. Figure 3-3 and Figure 3-4 show comparisons of percent retained on top three sieves
between control level and modified gradations. The asphalt content levels for mixtures
with modified gradation were kept the same as for the original control mixes.
3.5 Specimen Preparation and Volumetric Properties
Raw specimens with dimensions of 150 mm (5.9 in.) in diameter by 165 mm (6.5 in.)
in height were first prepared on the required air void content (4%) using a Servopac
Gyratory Compactor for the selected HMA mixtures. The sample preparation for the IDT
test was based on the findings from the NCHRP Project 1-28A, “Harmonized Test
Methods for Laboratory Determination of Resilient Modulus for Flexible Pavement
61
Design”. At least 6 mm was sawed off both sides of each test specimen to provide
smooth, parallel surfaces for mounting the measurement gauges. The testing specimen
was then sawed to the required thickness (two specimens out of each compacted pill,
Figure 3-5, referred to as A and B). This sample preparation procedure was done to make
four pills, which were then sawed to make eight samples for each HMA mixture. The Gmb
values were measured for the prepared test specimens to assure that the air voids were
within targeted range. Resilient modulus test, creep compliance test, and tensile strength
test were performed on these 150 mm (6 in.) in diameter by 63 mm (2.5 in.) thick test
specimens. Table 3-1 through Table 3-3 show a summary of the specimens prepared for
each mix and the corresponding volumetric properties measured in the lab.
In the second part of the experimental program, the mixture design process was
verified for the mixture volumetric properties before the production of test specimens.
The original Superpave design procedure was used for all twenty mixture designs. The
Servopac Superpave gyratory compactor was used in the process. The Servopac
compaction parameters used for the design were a 150 mm diameter mold, a 1.25°
gyratory angle, a 600-kPa ram pressure, and 30 gyrations per minute. To verify the
volumetric properties of the mixtures, the maximum theoretical specific gravity was
measured using Rice maximum theoretical specific gravity method specified in AASHTO
T 209/ASTM D 2041 standards. In this case, the mixtures were allowed to cool down in
the loose state. Table A-14 to Table A-17 show the volumetric properties of all the
mixtures used in this part of the research project.
The sample preparation for dynamic modulus test was based on the conclusions of an
extensive study on sample geometry and aggregate size conducted during NCHRP Project
9-19. Results show that (1) a minimum height-to-diameter ratio of 1.5 was required in
order to ensure that the response of a sample evaluated in either the dynamic modulus or
permanent deformation test and repeated load tests represents a fundamental engineering
property; (2) a minimum sample diameter of 4 in. (100 mm) was satisfactory for all HMA
mixtures up to a maximum aggregate size of 1.5 in. (37.5 mm); and (3) smooth, parallel
specimen ends were needed to eliminate end friction and violation of the theoretical
boundary effects of the specimen during the test. Similar to the IDT specimen
preparation, raw specimens with dimensions of 150 mm (6 in.) in diameter by 165 mm
62
(6.5 in.) in height were first prepared on the required air void content (4%) using a
Servopac Gyratory Compactor for targeted Florida HMA mixtures. The nominal 100 mm
(4in.) diameter test specimens were cored from the center of the gyratory specimens
(Figure 3-6) and were subsequently cut to 150 mm (6 in.) in height (Figure 3-7). Dynamic
modulus testing was performed on the test specimens measuring 100 mm (4 in.) in
diameter by 150 mm (6 in.) in height.
Table 3-1: Number of specimens prepared for fracture mechanics tests
Mixes F2 Control (F2C) F4 Control (F4C)
# of Specimens 8 8
Gradation Modifications F2G1 F2G2 -- F4G1 F4G2 --
# of Specimens 8 8 -- 8 8 --
SBS Polymer Modification F2P1 F2P2 F2P3 F4P1 F4P2 F4P3
# of Specimens 8 8 8 8 8 8
Table 3-2: Specimens tested for fracture mechanics properties
Gradation Study
Mix F2C F2G1 F2G2 F4C F4G1 F4G2
Specimen Number
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
SBS Modifier Study
Mix F2P1 F2P2 F2P3 F4P1 F4P2 F4P3
Specimen Number
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
1A, 1B 2A, 2B 3A, 3B 4A, 4B
63
Table 3-3: Specific gravities and air voids of the mixtures
Gradation Study
Mix F2C F2G1 F2G2 F4C F4G1 F4G2
Gmm 2.589 2.585 2.585 2.253 2.260 2.260
Gmb 2.479 2.487 2.490 2.173 2.179 2.179
VTM (Va) 4.3 3.8 3.7 3.5 3.6 3.6
SBS Modifier Study
Mix F2P1 F2P2 F2P3 F4P1 F4P2 F4P3
Gmm 2.573 2.573 2.573 2.253 2.253 2.253
Gmb 2.472 2.463 2.479 2.179 2.130 2.187
VTM (Va) 3.9 4.3 3.7 3.3 5.4 3.0
Figure 3-6: Coring of the Superpave specimen
Figure 3-7: Cutting of the dynamic modulus specimen
64
3.6 Test Procedures
3.6.1 Resilient Modulus Test
After the specimens were well-prepared, they were placed in a controlled-
temperature cabinet and brought to the specified test temperature. The specimens were
placed into the loading apparatus; and the loading strips were positioned in a parallel
format and centered on the vertical diametral plane (Figure 3-8). Tests were performed at
temperatures of -10, 5, 25, and 40ºC at 1.0 Hz frequency. Testing began with the lowest
temperature and proceeded to the highest temperature. Typical load and deformation
outputs that form a resilient modulus test are shown in Figure 3-9.
On the night before testing, extensometers were placed on the test specimen using
glue. The specimen was then placed in a controlled temperature cabinet overnight at
-10ºC to ensure temperature equilibrium. On the morning of testing, the specimen was
placed in the environmental chamber at -10ºC and allowed to equilibrate for two hours.
To begin testing, the extensometers were zeroed, and a minimal contact load was
applied to the specimen. Each stress cycle was made up of a 0.1 second haversine pulse
followed by a 0.9 second hold cycle to simulate moving wheel loads. The data acquisition
system was set up to record the last six cycles at each frequency with about 400 points per
cycle. The raw force and displacement data were manipulated to obtain the resilient
modulus for each specimen. After the entire cycle of testing was complete at -10ºC, the
environmental chamber was set to the next temperature. After two hours of conditioning,
the above steps were repeated until the entire sequence of temperatures was completed.
The test was conducted based on the SHRP IDT testing procedures. The resilient
modulus is the ratio of the applied stress to the recoverable strain as shown in Equation
3.1. During the test, the load was carefully measured so that the horizontal strain was
within 100 and 300 micro-strains. These limits were established based on research
conducted by Roque et al. (1997) to accurately evaluate the resilient modulus and
Poisson’s Ratio of bituminous materials. The upper limit was set to make sure that the
horizontal strains were within the linear viscoelastic range and the lower limit was set to
obtain sufficient amplitude of strain against system noises.
65
rrRM εσ /= (3-1)
The Resilient Modulus and the Poisson’s Ratio were calculated using the equations
developed by Roque et al. (1997) based on a three-dimensional finite element analysis
(Equation 3-2 through Equation 3-4).
CMPL
RCDtH
GLPM
×××∆
×= (3-2)
332.0)(6354.0 1 −×= −
YXCCMPL (3-3)
222 )()(778.0)(480.11.0
YX
Dt
YX ××−×+−=ν (3-4)
Where
MR = Resilient Modulus P = Maximum Load GL = Gage Length
�H = Horizontal Deformation t = Thickness D = Diameter
CCMPL = Non-dimensional Factor � = Poisson’s Ratio
(X/Y) = Ratio of Horizontal to Vertical Deformation
Figure 3-8: Indirect Diametral Resilient Modulus Test Setup
66
Figure 3-9: Load & deformations in a typical resilient modulus test
3.6.2 Creep Compliance Test
Creep Compliance is a function of time-dependent strain (�t) divided by constraint
stress (�) (Equation 3-5). Once the Resilient Modulus Test was completed, the Creep Test
was conducted by applying a static load on the specimen for 100 seconds. Similar to the
MR Test, the horizontal strain was limited from 150 to 300 micro-strains at 100 seconds
to avoid excessive permanent deformation of the specimen. The equation used to
calculate the Creep Compliance is presented in Equation 3-6.
σ
ε )()(
ttD = (3-5)
GLP
CDtHtD CMPL
×
×××∆=)( (3-6)
where D(t) is the creep compliance at time t with a unit of 1/GPa, other parameters are the
same as defined in resilient modulus equations. The specimen set-up and transducers
attachment are the same as for the resilient modulus test. Figure 3-10 displays typical load
and deformation curves of the creep compliance test.
67
Load
De
form
atio
n
Time (sec.)
0
Loa
d
0
Vertical Deformation
Horizontal Deformation
Figure 3-10: Load and deformation curves of creep compliance test
3.6.3 Tensile Strength Test
The strength test is a destructive test. The strength test, along with the MR test, was
used to determine asphalt mixture fracture mechanics properties which included the
Tensile Strength (St), Fracture Energy (FE), Dissipate Creep Strain Energy (DCSE), and
Failure Strain. The procedures used to calculate these limits are presented in the
following equations (Roque et al. 1997):
Dt
DtDtP
Dt
CPS SX
t⋅⋅
⋅⋅+⋅−⋅−⋅=
⋅⋅=
π
νν
π
))/(436.12693.0)/(01114.0948.0(2)(2 (3-7)
� ⋅=tS
dFE0
εσ (2-36)
R
t
tEEM
SSEE
⋅=⋅⋅=
2
)(
2
12
ε (2-38)
EEFEDCSE −= (2-39)
68
Where CSX is the stress correction factor, t is specimen thickness, D is specimen diameter,
� is Poisson’s ratio, and other variables are the same as defined in section 2.5.
The specimen set-up and transducers attachment are the same as for the resilient
modulus test. However, the tensile strength test was conducted in a displacement control
mode by applying a constant rate of displacement of 12.5 mm/min at -10˚C, 25 mm/min
at 5˚C, and 50 mm/min at 25 and 40˚C. Figure 3-11 displays a specimen broken along the
diametral direction after the strength test.
Figure 3-11: Specimen fails after tensile strength test
3.6.4 DMT Test Procedures
The dynamic moduli and phase angle were measured by applying compressive
sinusoidal (haversine) loading. The deformations were measured through three LVDTs
(Linear Variable Differential Transducers). These LVDTs were placed vertically on
diametrically symmetric specimen sides (Figure 3-12). On the night before testing,
parallel studs were glued 100 mm (4”) apart, located approximately 25 mm (1”) from the
top and bottom of the specimen. They were used to secure the LVDTs in place. The
69
diameter of the specimens was 100 mm (4”) and the height was 150 mm (6”). They were
cut and cored from the raw gyratory compacted pills with diameters of 150 mm (6”) and
heights of 165 mm. The specimens were then placed in a controlled temperature cabinet
overnight at 5ºC to ensure temperature equilibrium. On the morning of testing, the
specimen was placed in the environmental chamber at 5ºC and allowed to equilibrate for
two hours. All testing was conducted using this temperature-controlled chamber capable
of accommodating temperatures from -16 to 60°C (3.2 to 140°F). Tests were performed
at temperatures of 5, 25, and 40 ºC and frequencies of 25, 10, 5, 1, and 0.5Hz. Testing
began with the lowest temperature and proceeded to the highest temperature. At a given
temperature level, the testing began with the highest frequency of loading and proceeded
to the lowest frequency. This temperature-frequency sequence was carried out to cause
minimum damage to the specimens before the next sequential test (Pellinen 2001).
Figure 3-12: Dynamic complex modulus test setup
To begin testing, the extensometers were zeroed, and a minimal contact load was
applied to the specimens. A sinusoidal axial compressive load was applied to the
70
specimens without impact in a cyclical manner. The load was adjusted in each case to
attempt to keep the axial strains between 50 and 150 micro-strains. The first step was to
apply a preconditioning load to the specimens with 200 cycles at 25 Hz. Testing
continued with different numbers of cycles for each frequency. The data acquisition
system was set up to record the last six cycles at each frequency with about 200 points per
cycle. The raw force and displacement data were manipulated to obtain the dynamic
modulus and phase angle for each specimen. After the entire cycle of testing was
complete at 5ºC, the environmental chamber was set to the next temperature. After two
hours of conditioning, the above steps were repeated until the entire sequence of
temperatures and frequencies was completed.
3.7 Testing Program
One coarse mix of Georgia granite and one coarse mix of limestone were selected
from Florida HMA Superpave mixtures as control mixes to study aggregate gradation and
SBS polymer-modified binder effects using the SHRP IDT testing and data processing
method. Each mix was modified to two gradation levels and three SBS polymer content
levels. In addition, a total of 20 standard mix designs were selected for a study of
correlations between resilient modulus and dynamic modulus. These mix designs were
contributed by companies involved in the production and use of HMA in Florida. The
nominal maximum aggregate sizes for the mixtures tested are 19.0 mm, 12.5 mm, and 9.5
mm, respectively. The Superpave mixture designs were selected because they are
commonly used FDOT gradations and are known to perform well in the field.
The HMA mixtures were compacted in the laboratory and the specimens were
prepared for the IDT and DMT. A flowchart is shown in Figure 3-13 to illustrate the
experimental program for measuring fracture mechanics properties of HMA mixtures.
The standard granite (Ga553, 04-3034A) and South Florida Limestone (SFL, 02-2529A)
mixtures at control level are named F2C and F4C, respectively. The testing program for
the study of the relationship between resilient modulus and complex dynamic modulus is
shown in Figure 3-14.
71
�� �!"#$$%��&'(%&%')*
+
�' �!,�-��&�(�$�.)*
"�#/#�����,�0/�
��1�/�2��/�"�#/#�����
2���3#�4�1�
��"5����"�6��'"5���'"�
71)�322����,�0/�
%�,8,�7�9����� �������
%:&;��':$;��#�/�<;
2���3#�4�1�
��75����7�����7%6��'75���'7����'7%
"��#����� ���#�����
,�������� 0����=
>���9�����1�/090��?���
(5&��$���$��'&@
&:5��A�#/
&:.��>���
����� ���9�#����?���
(5&��$���$��'&@
,�#����A�#/
5&&������/�
?����9��,����=�4�?���
(5&��$���$��'&@
����#���B���9#������
��99���#��0��
B#�#�>�/0�����
,�#������#9�)�#9����
Figure 3-13: Flowchart of the experimental program for measuring fracture mechanics properties of HMA mixtures
72
Figure 3-14: Flowchart of the testing program for MR vs. E*
TWENTY (20) FDOT HMA MIXTURES
SUPERPAVE GYRATORY
COMPACTION
SPECIMEN CORING & CUTTING
SUPERPAVE LEVEL I
MIX DESIGN
DYNAMIC COMPLEX MODULUS TESTING
(NCHRP 9-29)
INDIRECT RESILIENT MODULUS TESTING
(SHRP IDT Tester)
Temperature
40ºC 25ºC 5ºC
Frequency 25Hz 10Hz 5Hz 1Hz
0.5Hz
Temperature
40ºC 25ºC 5ºC
Frequency
0.1s Load 0.9s Rest
DYNAMIC MODULUS, E* &
PHASE ANGLE, φ
RESILIENT MODULUS, MR &
POISSON’S RATIO, C
DATA ANALYSIS
73
CHAPTER 4
FRACTURE MECHANICS PROPERTIES FROM IDT
The laboratory testing program conducted in this study included resilient modulus
testing, creep compliance testing, and tensile strength testing. All types of testing were
conducted in unconfined conditions. The Interlaken dynamic test system was used for all
of the sliced specimens to get the fracture mechanics properties including resilient
modulus (MR), creep (Dt), Fracture Energy (FE), and Dissipate Creep Strain Energy
(DCSE). The data reductions were conducted according to the procedures presented by
Roque et al. (1997).
4.1 Resilient Modulus Testing Procedures and Results
4.1.1 Test Procedures
After the specimens were well prepared, they were placed in a controlled temperature
cabinet and brought to the specified test temperature. The specimens were placed into the
loading apparatus; the loading strips were positioned in a parallel format and centered on
the vertical diametral plane. Tests were performed at temperatures of -10, 5, 25, and 40ºC
and at 1.0 Hz frequency. Testing began with the lowest temperature and proceeded to the
highest. On the night before testing, extensometers were placed on the test specimen
using glue. The specimen was then placed in a controlled temperature cabinet overnight
at -10ºC to ensure temperature equilibrium. On the morning of testing, the specimen was
placed in the environmental chamber at -10ºC and allowed to equilibrate for two hours.
To begin testing, the extensometers were zeroed, and a minimal contact load was
applied to the specimen. Each stress cycle was made up of a 0.1-second haversine pulse
followed by a 0.9-second hold cycle to simulate moving wheel loads. The data acquisition
system was set up to record the last six cycles at each frequency with about 400 points per
cycle. The raw force and displacement data were manipulated to obtain the resilient
modulus for each specimen as described in section 3.2. The load was selected to keep the
horizontal strain in the linear viscoelastic range which is typically 150 to 350 micro-
74
strains. After the entire cycle of testing was complete at -10ºC, the environmental
chamber was set to the next temperature. After two hours of conditioning, the above steps
were repeated until completion of the entire sequence of temperatures. Upon completion
of the resilient modulus tests, all samples were placed in the environmental chamber for
overnight conditioning before creep compliance testing and tensile strength testing.
4.1.2 Resilient Modulus Data Analysis and Results
For the measurement and analysis system used, two vertical and horizontal
measurements were obtained for each specimen. Data from five load cycles were
recorded after 100 cycles of equilibrium. The maximum load and the maximum
deformation were determined for each cycle from the load and deformation curves. Linear
regression was performed on the unloading and recovery portion of each deformation
wave to determine the instantaneous and total recoverable deformations (Figure 4-1). The
trimmed mean deformations and the average load were obtained from the replicate
specimens tested. The average total resilient modulus for each mixture was calculated
using Equation 3-2 through Equation 3-4. Table 4-1 through Table 4-4 show the resilient
modulus test results for all mixtures.
Figure 4-1: Instantaneous and total resilient deformations
75
Table 4-1: Resilient modulus test results at -10˚C
Mixtures with Modified Gradations
Control G1 G2
F2 F4 F2 F4 F2 F4
PrI 0.35 0.35 0.25 0.32 0.29 0.31
PrT 0.35 0.35 0.25 0.32 0.28 0.31
MrI (GPa) 28.91 20.98 30.08 21.79 29.19 20.55
MrT (GPa) 28.15 20.61 29.62 21.41 28.50 20.26
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
F2 F4 F2 F4 F2 F4
PrI 0.33 0.28 0.35 0.26 0.31 0.33
PrT 0.33 0.27 0.35 0.26 0.31 0.33
MrI (GPa) 27.10 19.24 24.93 16.42 23.06 14.49
MrT (GPa) 26.61 18.91 24.49 16.14 22.45 14.01
Note: PrI: Poisson’s Ratio, instantaneous; PrT: Poisson’s Ratio, total; MrI: Resilient Modulus, instantaneous; MrT: Resilient Modulus, total. 1 GPa = 145 ksi
Table 4-2: Resilient modulus test results at 5˚C
Mixtures with Modified Gradations
Control G1 G2
F2 F4 F2 F4 F2 F4
PrI 0.36 0.31 0.32 0.33 0.36 0.32
PrT 0.36 0.32 0.32 0.33 0.36 0.33
MrI (GPa) 19.22 13.40 18.57 13.31 19.50 11.90
MrT (GPa) 18.25 12.90 17.59 12.81 18.52 11.36
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
F2 F4 F2 F4 F2 F4
PrI 0.35 0.36 0.35 0.27 0.29 0.39
PrT 0.34 0.36 0.35 0.27 0.29 0.39
MrI (GPa) 19.71 12.49 17.24 10.97 14.80 7.93
MrT (GPa) 18.86 11.98 16.39 10.57 13.57 7.26
76
Table 4-3: Resilient modulus test results at 25˚C
Mixtures with Modified Gradations
Control G1 G2
F2 F4 F2 F4 F2 F4
PrI 0.43 0.43 0.28 0.34 0.32 0.46
PrT 0.44 0.45 0.31 0.33 0.34 0.44
MrI (GPa) 6.21 4.80 5.58 4.96 5.29 4.90
MrT (GPa) 5.53 4.22 4.92 4.51 4.56 4.36
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
F2 F4 F2 F4 F2 F4
PrI 0.37 0.37 0.30 0.39 0.28 0.29
PrT 0.37 0.35 0.31 0.39 0.29 0.30
MrI (GPa) 6.15 4.32 4.90 4.53 3.54 2.21
MrT (GPa) 5.24 3.78 4.29 4.04 3.06 1.99
Table 4-4: Resilient modulus test results at 40˚C
Mixtures with Modified Gradations
Control G1 G2
F2 F4 F2 F4 F2 F4
PrI 0.35 0.39 0.32 0.32 0.38 0.36
PrT 0.38 0.36 0.32 0.31 0.33 0.39
MrI (GPa) 1.39 1.48 1.40 1.66 1.65 1.42
MrT (GPa) 1.19 1.03 1.23 1.47 1.45 1.28
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
F2 F4 F2 F4 F2 F4
PrI 0.41 0.41 0.45 0.36 0.44 0.47
PrT 0.40 0.43 0.46 0.38 0.47 0.41
MrI (GPa) 1.96 1.28 2.09 1.40 1.22 1.25
MrT (GPa) 1.67 1.13 1.93 1.25 1.08 1.10
4.2 Creep Compliance Testing Procedures and Results
4.2.1 Test Procedures
The mounting of LVDTs and the preloading for the creep compliance test were the
same as those for the resilient modulus test. A static load was applied on specimen for
100 seconds. The horizontal strains at the 30th second were controlled to be between 100
and 200 micro-strains to ensure the specimen was tested in viscoelastic range. If the range
77
limit was exceeded, the load was immediately removed from the specimen and a
minimum of three minutes rest period was allowed for the specimen to recover before
reloading at another appropriate level. The data acquisition program records the loads and
specimen deformations at a rate of 10 Hz. Matlab scripts were generated to analyze the
load and deformation data and to calculate the creep compliance values at points of
specified time. All specimens were placed in the environmental chamber for at least one
overnight recovery prior to the tensile strength test.
4.2.1 Creep Compliance Data Analysis and Results
For each creep compliance data file collected, the creep test start point and the initial
extensometer reading were determined first. Then the deformations for each creep time
point were calculated by determining the corresponding extensometer readings. The
deformations and axial load were averaged for the replicate specimens tested. The creep
compliance for each time point was calculated using Equation 3-6. The creep compliance
test results are summarized in Appendix B.
4.3 Tensile Strength Testing Procedures and Results
4.3.1 Test Procedures
The tensile strength test was conducted in a displacement control mode by applying a
constant rate of displacement until the specimen failed. It was observed that the
specimens crashed too quickly to obtain sufficient data points if the rate of displacement
was relatively high at a certain level of temperature. In order to make data records and
reduction more accurate, the displacement rate was set as 12.5 mm/min (0.5 in/min) at
-10ºC, 25 mm/min (1.0 in/min) at 5ºC, 50 mm/min (2.0 in/min) at 25ºC and 40ºC. The
horizontal and vertical deformation and the applied load were recorded at a rate of 20 Hz
during the test. The dissipated creep strain energy (DCSE) and fracture energy (FE) can
be determined from the tensile strength and resilient modulus of the specimen. The
schematics used to calculate these limits are described in section 2.5 and are displayed in
Figure 4-2 for reference.
78
4.3.2 Tensile Strength Data Analysis and Results
Similar to the data reduction procedures for resilient modulus and creep compliance,
the load and deformations at each time point were determined first for each tensile
strength data file. Specifically, the instant of failure is identified as the point in time at
which the difference between the vertical and horizontal deformations reaches a peak ((Y-
X) peak). The tensile strength was then calculated using Equation 3-7. The strength of the
mixture was obtained by taking the average value of the replicated specimens tested.
Stress and strain at each time point were calculated from the start of the load cycle to the
instant of specimen failure using the following equations (Roque et al. 1997):
))/(436.12693.0)/(01114.0948.0(2
)( ννπ
σ ⋅⋅+⋅−⋅−⋅⋅⋅
⋅= DtDt
Dt
Loadt (4-1)
))/(089.0081.0)/(189.003.1()(
072.1)( 2DtDt
GL
tnDeformatiot ⋅+⋅−⋅−⋅⋅= νε (4-2)
Where �(t) is stress and � is strain. Other variables are the same as defined in section 3.6.
The fracture energy is obtained by integrating the area under the stress-strain curve until
failure as shown in Figure 4-2 for convenience. All fracture mechanics parameters
obtained from the tensile strength test were calculated using Equations 2-36 through 2-39
and are presented in Table 4-5 and Table 4-6.
e0 ef Strain
St
MR
EE
Dissipated Creep Strain Energy
(DCSE)
MR
Str
ess
Figure 4-2: Determination of fracture energy and dissipated creep strain energy
79
Table 4-5: Tensile strength test results for F2 series mixtures
F2 Control Temperature (˚C) -10 5 25 40
EE (KPa) 0.45 0.27 0.07 0.04
DCSE (KPa) 2.37 5.16 3.37 1.84
FE (KPa) 2.82 5.43 3.44 1.88
TS (MPa) 5.04 3.11 0.90 0.32
FS (103 micro) 0.86 4.35 5.56 10.78
F2G1 Temperature (˚C) -10 5 25 40
EE (KPa) 0.28 0.30 0.10 0.05
DCSE (KPa) 1.07 4.82 2.79 1.26
FE (KPa) 1.35 5.12 2.89 1.31
TS (MPa) 4.07 3.23 1.00 0.35
FS (103 micro) 0.54 2.04 3.75 5.13
F2G2 Temperature (˚C) -10 5 25 40
EE (KPa) 0.39 0.28 0.15 0.06
DCSE (KPa) 2.34 4.69 2.46 2.41
FE (KPa) 2.73 4.97 2.61 2.47
TS (MPa) 4.72 3.22 1.15 0.43
FS (103 micro) 0.87 2.00 2.95 7.63
F2P1 (3.0%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.30 0.50 0.15 0.06
DCSE (KPa) 1.99 6.93 3.14 1.06
FE (KPa) 2.29 7.43 3.29 1.12
TS (MPa) 3.88 4.33 1.23 0.43
FS (103 micro) 0.77 2.28 3.73 3.55
F2P2 (4.5%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.41 0.20 0.46 0.03
DCSE (KPa) 2.58 5.06 4.73 2.37
FE (KPa) 3.00 5.25 5.19 2.40
TS (MPa) 4.44 2.52 1.99 0.34
FS (103 micro) 3.04 2.64 3.30 9.00
F2P3 (6.0%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.42 0.19 0.15 0.08
DCSE (KPa) 3.12 7.27 3.76 2.65
FE (KPa) 3.53 7.46 3.91 2.72
TS (MPa) 4.33 2.25 0.96 0.41
FS (103 micro) 1.66 5.90 5.29 9.19
Note EE: elastic energy DCSE: dissipated creep strain energy FE: total fracture energy TS: tensile strength FS: failure strain 1 MPa = 145 psi 1 KPa = 0.145 psi 1 micro-strain = 10
-6 mm/mm (in./in.)
80
Table 4-6: Tensile strength test results for F4 series mixtures
F4 Control Temperature (˚C) -10 5 25 40
EE (KPa) 0.41 0.49 0.11 0.07
DCSE (KPa) 1.69 4.57 3.86 1.91
FE (KPa) 2.11 5.06 3.98 1.98
TS (MPa) 4.10 3.57 0.98 0.37
FS (103 micro) 0.86 1.95 9.85 6.36
F4G1 Temperature (˚C) -10 5 25 40
EE (KPa) 0.25 0.40 0.09 0.04
DCSE (KPa) 2.25 4.21 2.01 1.52
FE (KPa) 2.51 4.61 2.10 1.56
TS (MPa) 3.30 3.21 0.88 0.36
FS (103 micro) 1.12 1.84 3.13 5.56
F4G2 Temperature (˚C) -10 5 25 40
EE (KPa) 0.36 0.45 0.07 0.04
DCSE (KPa) 1.92 4.72 2.80 1.90
FE (KPa) 2.28 5.18 2.87 1.94
TS (MPa) 3.81 3.21 0.77 0.30
FS (103 micro) 0.93 2.16 4.66 8.52
F4P1 (3.0%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.52 0.48 0.10 0.10
DCSE (KPa) 2.44 5.76 4.18 3.68
FE (KPa) 2.96 6.24 4.31 3.79
TS (MPa) 4.42 3.38 0.85 0.48
FS (103 micro) 1.09 2.50 6.25 10.24
F4P2 (4.5%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.67 0.47 0.15 0.06
DCSE (KPa) 1.85 5.83 3.20 3.00
FE (KPa) 2.52 6.29 3.35 3.06
TS (MPa) 4.62 3.14 1.09 0.38
FS (103 micro) 0.89 2.62 4.08 11.00
F4P3 (6.0%) Temperature (˚C) -10 5 25 40
EE (KPa) 0.68 0.31 0.13 0.13
DCSE (KPa) 8.09 6.38 4.80 5.62
FE (KPa) 8.77 6.69 4.92 5.75
TS (MPa) 4.36 2.12 0.71 0.54
FS (103 micro) 3.10 4.24 11.10 13.70
81
CHAPTER 5
EVALUATION OF FRACTURE MECHANICS PROPERTIES
In this chapter, two effects – the effect of coarse aggregate gradation modification
and the effect of SBS polymer-modified asphalt binder – will be further evaluated with
respect to the fracture mechanics properties presented in Chapter 4. The evaluation of
each impact will be focused on the following parameters obtained from the sweep of IDT
tests: resilient modulus, creep compliance, tensile strength, fracture energy, and dissipated
creep strain energy. A brief summary of discussion will be given at the end of this
chapter.
5.1 Evaluation of Gradation Effects
5.1.1 Evaluation of Gradation Curves for Modified Gradation Mixtures
The power law model developed by Ruth et al. (2002) was used to fit the gradation
curve for each mixture. As described in section 4.3, coarse portions of the control
mixtures (percent passing 1/2 in. and 3/8 in. sieves) were modified, and the fine portions
were maintained. Therefore, the power law constant and exponent (aca, nca) for only
coarse aggregate were calculated by regression for the mixtures. The format of the power
law equation is
can
caca daP )(⋅= (5-1)
where Pca is the percent of material by weight passing a given sieve having an opening of
width d. The break sieve size to distinguish coarse and fine aggregate is defined by the
primary control sieve (PCS) based on the Bailey method:
22.0×= NMPSPCS (5-2)
where PCS is the primary control sieve for the overall blend which defines the break
between coarse and fine aggregate, and NMPS is the nominal maximum particle size for
the overall blend as defined in Superpave mix design, which is one sieve larger than the
82
first sieve that retains more than 10%. The NMPS for F2 control and F4 control are both
12.5 mm in this study. The “break” sieve size should be 12.5 × 0.22 = 2.75 mm, which
corresponds to the No. 8 sieve. However, since the percent passing No. 4 and smaller
sieves are all the same for control and modified gradations, the parameters used for the
power law regression are No. 4 and higher sieve sizes. The regression coefficients are
also shown in Figure 5-1 and Figure 5-2. For each series, whether granite or limestone,
the regression parameters (aca and nca) decreased by a small amount as the coarse size
aggregate increased. As expected, the fitted curves are in the same order as their real
gradation curves. Table 5-1 summarizes the regression coefficients for the control mixes
and those with modified gradations. According to Birgisson et al. (2004), a high nca
implies a low dynamic modulus at a high temperature of 40˚C when controlling for nfa.
The fine portions of the control gradations were maintained in this study, and the
mixtures with modified gradations obtained a lower nca. This implies that the modified
gradations would have higher dynamic moduli than control mixes at high temperature
levels which are favorable characteristics for HMA performance.
Power Law Regression F2 Series
y = 25.693x0.491
R2 = 0.9178
y = 25.228x0.4851
R2 = 0.9708
y = 24.917x0.4676
R2 = 0.9979
40
60
80
100
120
5 10 15 20 25
Sieve Size (mm)
Perc
ent
Passin
g (
%) F2
F2G1
F2G2
Pow er (F2)
Pow er (F2G1)
Pow er (F2G2)
Figure 5-1: Power law regression for F2 gradation series
83
Power Law Regression F4 Series
y = 30.548x0.4295
R2 = 0.8941
y = 29.605x0.4256
R2 = 0.9792
y = 28.95x0.4095
R2 = 0.9818
40
60
80
100
120
5 10 15 20 25
Sieve Size (mm)
Perc
ent
Passin
g (
%) F4
F4G1
F4G2
Pow er (F4)
Pow er (F4G1)
Pow er (F4G2)
Figure 5-2: Power law regression for F4 gradation series
Table 5-1: Power law regression coefficients for modified gradation mixes
Mixture F2 F2G1 F2G2 F4 F4G1 F4G2
aca 25.693 25.228 24.917 30.548 29.605 28.95
nca 0.491 0.4851 0.4676 0.4295 0.4256 0.4095
R2 0.9178 0.9708 0.9979 0.8941 0.9792 0.9818
5.1.2 Evaluation of Resilient Modulus for Modified Gradation Mixtures
Figure 5-3 shows the resilient modulus comparisons at various temperature levels.
Using the resilient modulus values at control level in abscissa and the values at modified
gradation in ordinate, a trend line can be plotted to present the relationship of resilient
modulus between control and modified gradation mixes, as shown in Figure 5-4. The
linear regression coefficient ranges from 0.96 to 1.03. The correlation coefficients (R2)
are all higher than 0.99, which indicates the linear relationship is very strong. From these
results, it can be concluded that no significant difference in resilient modulus was present
between the control mix design and the mixes with modified gradations. Increasing the
84
amount of 1/2 in. coarse aggregate within the range of 5% to 15% appeared to have
minimal influence on mixtures’ resilient modulus.
0
10
20
30
-10 5 25 40
Temperature (˚C)
Resili
ent
Modulu
s (
GP
a)
F2 Control
F2G1
F2G2
F4 Control
F4G1
F4G2
Figure 5-3: Resilient modulus for mixtures with modified gradations (1 GPa = 145 ksi)
y = 1.0224x
R2 = 0.9952
0
10
20
30
40
0 20 40
MR for F2 Control (GPa)
MR f
or
F2G
1 (
Gpa)
(a)
y = 1.0087x
R2 = 0.9976
0
10
20
30
40
0 20 40
MR for F2 Control (GPa)
MR f
or
F2G
2 (
Gpa)
(b)
85
y = 1.028x
R2 = 0.9982
0
10
20
30
0 10 20 30
MR for F4 Control (GPa)
MR f
or
F4G
1 (
Gpa)
(c)
y = 0.957x
R2 = 0.9932
0
10
20
30
0 10 20 30
MR for F4 Control (GPa)
MR f
or
F4G
2 (
Gpa)
(d)
Figure 5-4: Comparison of resilient modulus between control and modified gradations
5.1.3 Evaluation of Creep Compliance for Modified Gradation Mixtures
The creep compliance test results were analyzed using the power law relationship
presented by Roque et al. (1997):
m
tDDtD 10)( += (5-3)
It was showed that the parameters obtained from this model are fairly accurate
indicators for the viscous response and rutting performance of HMA mixtures. Kim et al.
(2005) recommended a fixed D0 value (0.0483 1/GPa, or 3.33×10-7 psi) to obtain more
consistent D1 and m values for the tests conducted at 0, 10 and 20˚C. Master curve
construction for creep compliance curves in this study included D0 in the parametric
analysis since the lowest testing temperature of -10˚C was used as the reference
temperature. Table 5-2, Figure 5-5 and Figure 5-6 show the regression coefficients D1 and
m for all mixes at control level and modified gradations. The comparisons of creep
compliance master curves for F2 and F4 gradation series are shown in Figure 5-7 and
Figure 5-8, respectively. The creep compliance curves within each aggregate type are very
similar to each other, with the exception of the control level granite mixture (F2 Control),
which is a little less compliant at high temperature (40˚C) than the other two mixes that
were blended with higher portions of coarse aggregate. This implies that the increase of
86
5% to 15% of 1/2" aggregate did not make a significant difference in the creep
compliance properties for the HMA tested.
Table 5-2: Power model regression coefficients for modified gradation tests
F2C F2G1 F2G2 F4C F4G1 F4G2
D1 (1/GPa) 0.012 0.007 0.009 0.013 0.007 0.010
m 0.360 0.410 0.398 0.369 0.404 0.370
0.0E+00
3.0E-03
6.0E-03
9.0E-03
1.2E-02
1.5E-02
F2 Mixes F4 Mixes
D1 (
1/G
pa)
Control
G1
G2
Figure 5-5: Power model parameter D1 for modified gradations (1/GPa = 6.89×10-6/psi)
0.0
0.1
0.2
0.3
0.4
0.5
F2 Mixes F4 Mixes
m
Control
G1
G2
Figure 5-6: Power model parameter m for modified gradations
87
-2
-1
0
1
2
3
0 2 4 6 8 10
Log (T) (Sec.)
Log (
D)
(1/G
pa)
F2 Control
F2G1
F2G2
Figure 5-7: Comparison of creep compliance for granite gradation series (1/GPa = 6.89×10-6/psi)
-2
-1
0
1
2
3
0 2 4 6 8 10
Log (T) (Sec.)
Log (
D)
(1/G
pa)
F4 Control
F4G1
F4G2
Figure 5-8: Comparison of creep compliance for limestone gradation series (1/GPa = 6.89×10-6/psi)
88
5.1.4 Evaluation of Tensile Strength and Fracture Energy for Modified Gradation
Mixtures
The indirect tensile strength (TS) of control mixtures and the modified gradation
mixtures is presented in Figure 5-9. The tensile strength clearly shows the expected trends
that the strength value decreases as the temperature increases. A comparison of strength
values between control and modified gradation mixtures is shown in Figure 5-10. At mid
to high service temperatures (25 and 40˚C), the tensile strength values for all mixtures are
similar and the differences appear to be negligible. However, at low service temperatures
(-10 and 5˚C), the tensile strength of mixtures with modified gradations are clearly lower
than that of control level mixtures. This tends to indicate that increasing the coarse
aggregate in the standard control mixture has an adverse effect on the tensile strength
property of the HMA at low temperatures.
Figure 5-11 and Figure 5-12 show the test results of Fracture Energy (FE) and
Dissipated Creep Strain Energy (DCSE), respectively, for all mixtures of modified
gradation. It is observed that fracture energy values are lower at both low (-10˚C) and
high (40˚C) temperatures than at mid-range temperatures (5 and 25˚C). The reason for
this trend is that the fracture energy is calculated as the area under the stress-strain curve
of the tensile strength test. At low testing temperatures, the tensile strength of HMA is
large but the failure strain is very small. At high temperatures, in contrast, the tensile
strength of HMA is the lowest but the failure strain is the highest due to the ductile effect
of asphalt binder. At some mid-range temperatures, the integration of the stress and strain
curve attains a peak value. The trend of dissipated energy (DCSE) is essentially the same
as for the fracture energy. Comparisons of fracture energy and DCSE between control
mixtures and modified gradations are presented in Figure 5-13 and Figure 5-14,
respectively. The distributions of the points in the two figures are very similar. A majority
of the point falls close to or under the equality line which means that the fracture energy
values (or DCSE values) of modified gradations are less than those of the control mixes.
89
0
2
4
6
-10 5 25 40
Temperature (˚C)
Tensile
Str
ength
(M
Pa)
F2 Control
F2G1
F2G2
F4 Control
F4G1
F4G2
Figure 5-9: Tensile strength for control and modified gradation mixes (1 MPa = 145 psi)
0
2
4
6
0 2 4 6
Tensile Strength of Control Level (MPa)
Tensile
Str
ength
of
Modifie
d G
radations (
MP
a)
F2G1 vs F2C
F2G2 vs F2C
Equality Line
F4G1 vs F4C
F4G2 vs F4C
Low Service
Temperatures:
-10 and 5˚C
Figure 5-10: Comparison of TS between control and modified gradation mixes
90
0
2
4
6
-10 5 25 40
Temperature (˚C)
Fra
ctu
re E
nerg
y (
KP
a)
F2 Control
F2G1
F2G2
F4 Control
F4G1
F4G2
Figure 5-11: Fracture Energy for modified gradation mixes (1 KPa = 0.145 psi)
0
2
4
6
-10 5 25 40
Temperature (˚C)
DC
SE
(K
Pa)
F2 Control
F2G1
F2G2
F4 Control
F4G1
F4G2
Figure 5-12: DCSE for modified gradation mixes (1 KPa = 0.145 psi)
91
0
2
4
6
0 2 4 6
Fracture Energy
of Control Mixes (KPa)
Fra
ctu
re E
nerg
y o
f
Modifie
d G
radations (
KP
a)
F2G1 vs F2C
F2G2 vs F2C
Equality Line
F4G1 vs F4C
F4G2 vs F4C
Figure 5-13: Comparison of Fracture Energy for modified gradation mixtures
0
2
4
6
0 2 4 6
DCSE of Control Mixes (KPa)
DC
SE
of
Modifie
d G
radations (
KP
a)
F2G1 vs F2C
F2G2 vs F2C
Equality Line
F4G1 vs F4C
F4G2 vs F4C
Figure 5-14: Comparison of DCSE for modified gradation mixtures
92
5.2 Evaluation of SBS Polymer-modified Binder Effects
5.2.1 Evaluation of Resilient Modulus for SBS Polymer-modified Mixtures
Similar to the comparison of resilient modulus between control level and modified
gradation mixtures presented in the last section, Figure 5-15 and Figure 5-16 show
comparisons of resilient modulus values at various temperature levels for the SBS
polymer-modified asphalt mixes of F2 and F4 series, respectively. It can be clearly seen
that at low to mid-range temperature levels (-10, 5, and 25˚C), the resilient modulus
values of PMA mixtures are less than those of control mixtures, and that an increment of
SBS polymer content lowers the resilient modulus magnitude. The only exception is that
the MR of F2P1 at 5˚C is a little higher than that of the F2 control. At a high testing
temperature (40˚C), however, the resilient moduli of PMA mixtures are higher than those
of the control mixtures and increase with the increase of SBS polymer content, except
that the MR values of 6.0% PMA mixtures drop back down. This phenomenon indicates
that the effect of SBS concentration for PMA mixtures appears to be consistent with the
findings presented by Chen et al. (2002, 2003) for SBS modified asphalt binders. It
implies that an optimum SBS content exists within the 3% to 6% range which would
make the PMA mixture stiffness the highest at the high service temperature, but increases
in polymer content after the optimum had an adverse effect on PMA resilient modulus
property, which was probably due to the mixing and distribution problems of SBS
polymer, base bitumen, and aggregate.
Trend lines are developed in Figure 5-17 for each PMA mixture versus the control
mixture. As shown in the figures, the linear regression coefficient decreases as the content
of SBS polymer modifier increases, and all correlation coefficient (R2) values are greater
than 0.97. The linear regression indicates an obvious trend that increasing SBS polymer
content makes the resilient modulus of HMA lower at low and mid-range temperatures.
Based on the above investigations, the MR results show that SBS modifiers make HMA
softer at mid-range to low service temperatures and stiffer at the highest testing
temperature (40˚C), which are both favorable attributes for the improvement of the HMA
performance issues of low temperature thermal cracking and high temperature rutting.
93
0
10
20
30
-10 5 25 40
Temperature (˚C)
Resili
ent
Modulu
s (
GP
a)
F2 Control
F2P1
F2P2
F2P3
Figure 5-15: Comparison of resilient modulus for F2 SBS PMA mixes (1 GPa = 145 psi)
0
5
10
15
20
25
-10 5 25 40
Temperature (˚C)
Resili
ent
Modulu
s (
GP
a)
F4 Control
F4P1
F4P2
F4P3
Figure 5-16: Comparison of resilient modulus for F4 SBS PMA mixes (1 GPa = 145 psi)
94
y = 0.9715x
R2 = 0.9949
0
10
20
30
0 10 20 30
MR for F2 Control (GPa)
MR f
or
F2P
1 (
Gpa)
(a)
y = 0.9204x
R2 = 0.9997
0
5
10
15
20
25
0 5 10 15 20 25
MR for F4 Control (GPa)
MR f
or
F4P
1 (
Gpa)
(b)
y = 0.8767x
R2 = 0.9961
0
10
20
30
0 10 20 30
MR for F2 Control (GPa)
MR f
or
F2P
2 (
Gpa)
(c)
y = 0.7987x
R2 = 0.994
0
5
10
15
20
25
0 5 10 15 20 25
MR for F4 Control (GPa)
MR f
or
F4P
2 (
Gpa)
(d)
y = 0.7756x
R2 = 0.9923
0
10
20
30
0 10 20 30
MR for F2 Control (GPa)
MR f
or
F2P
3 (
Gpa)
(e)
y = 0.6425x
R2 = 0.9777
0
5
10
15
20
25
0 5 10 15 20 25
MR for F4 Control (GPa)
MR f
or
F4P
3 (
Gpa)
(f)
Figure 5-17: Comparison of MR between control and PMA mixtures
95
5.2.2 Evaluation of Creep Compliance for SBS Polymer-modified Mixtures
Table 5-3, Figure 5-18 and Figure 5-19 show the regression coefficients D1 and m
from creep compliance test results for all mixes with SBS polymer-modified binders. The
creep compliance master curves are developed and shown in Figure 5-20 and Figure 5-21
for the F2 and F4 series, respectively. As demonstrated in the figures, at low reduced time
of about 0 to 104.4 seconds, the PMA mixtures are all more compliant than the control
mixes. This means the polymer modifier makes the HMA more ductile at low
temperatures which would be beneficial to the reduction of thermal cracking. At higher
reduced time, the master curves come across each other and the PMA mixture master
curves tend to go under the control ones, indicating that the PMA mixes are stiffer and
more resistant to rutting at high temperatures. The temperature effect for creep
compliance of PMA mixtures can be also observed clearly from the direct testing results
plotted in Figure 5-22 through Figure 5-29. At -10˚C, all the CP values of PMA mixtures
are higher than those of control mixes. On the other hand, at 40˚C, the points all drop
below the equality line except that F4P2 is a little higher than the F4 control. These
observations further verify the SBS polymer effect discussed in the resilient modulus
results. At mid-range temperatures (5˚C and 25˚C), the creep compliance of specimens
did not differ significantly. In addition, the linear regression indicates that at a specific
temperature level for each mix series, an increment of SBS polymer content usually
results in higher creep compliance values. The two exceptions are that the creep
compliance of F2P1 is a little higher than that of F2P2 at 5˚C, and the creep compliance
of F4P2 is higher than that of F4P3 at 40˚C.
Table 5-3: Power model regression coefficients for PMA mixture tests
F2P1 F2P2 F2P3 F4P1 F4P2 F4P3
D1 (1/GPa) 0.011 0.014 0.034 0.016 0.015 0.017
m 0.413 0.365 0.279 0.353 0.365 0.318
96
0.0E+00
1.0E-02
2.0E-02
3.0E-02
4.0E-02
F2 Mixes F4 Mixes
D1 (
1/G
pa) Control
P1
P2
P3
Figure 5-18: Power model parameter D1 for mixes with SBS PMA (1/GPa = 6.89×10-6/psi)
0.0
0.1
0.2
0.3
0.4
0.5
F2 Mixes F4 Mixes
m
Control
P1
P2
P3
Figure 5-19: Power model parameter m for mixes with SBS PMA
97
-2
-1
0
1
2
3
0 2 4 6 8 10
Log (T) (Sec.)
Log (
D)
(1/G
pa)
F2 Control
F2P1
F2P2
F2P3
Low Reduced
Time
High Reduced
Time
Figure 5-20: Creep compliance master curves for granite PMA mixtures (1/GPa = 6.89×10-6/psi)
-2
-1
0
1
2
3
0 2 4 6 8 10
Log (T) (Sec.)
Log (
D)
(1/G
pa)
F4 Control
F4P1
F4P2
F4P3
Low Reduced
Time
High Reduced
Time
Figure 5-21: Creep compliance master curves for limestone PMA mixtures (1/GPa = 6.89×10-6/psi)
98
0.03
0.06
0.09
0.12
0.15
0.03 0.06 0.09 0.12 0.15
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa)
F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F2P1 Fit, y=1.096x
F2P2 Fit, y=1.122x
F2P3 Fit, y=1.424x
Figure 5-22: Comparison of creep compliance at -10˚C for F2 series
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa) F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F2P1 Fit, y=0.782x
F2P2 Fit, y=0.693x
F2P3 Fit, y=0.966x
Figure 5-23: Comparison of creep compliance at 5˚C for F2 series
99
0
2
4
6
0 2 4 6
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa) F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F2P1 Fit, y=0.760x
F2P2 Fit, y=1.031x
F2P3 Fit, y=1.126x
Figure 5-24: Comparison of creep compliance at 25˚C for F2 series
0
3
6
9
12
15
0 3 6 9 12 15
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa) F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F2P1 Fit, y=0.519x
F2P2 Fit, y=0.540x
F2P3 Fit, y=0.742x
Figure 5-25: Comparison of creep compliance at 40˚C for F2 series
100
0.03
0.06
0.09
0.12
0.15
0.03 0.06 0.09 0.12 0.15
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa)
Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
F4P1 Fit, y=1.145x
F4P2 Fit, y=1.269x
F4P1 Fit, y=1.418x
Figure 5-26: Comparison of creep compliance at -10˚C for F4 series
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa) Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
F4P1 Fit, y=0.831x
F4P2 Fit, y=0.924x
F4P3 Fit, y=1.208x
Figure 5-27: Comparison of creep compliance at 5˚C for F4 series
101
0
2
4
6
0 2 4 6
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa) Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
F4P1 Fit, y=0.872x
F4P2 Fit, y=1.063x
F4P3 Fit, y=1.084x
Figure 5-28: Comparison of creep compliance at 25˚C for F4 series
0
3
6
9
12
15
0 3 6 9 12 15
D(t) of Control Mixtures (1/GPa)
D(t
) of
PM
A M
ixtu
res (
1/G
Pa)
Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
F4P1 Fit, y=0.578x
F4P2 Fit, y=1.055x
F4P3 Fit, y=0.692x
Figure 5-29: Comparison of creep compliance at 40˚C for F4 series
102
5.2.3 Evaluation of Tensile Strength and Fracture Energy for SBS Polymer-
modified Mixtures
Figure 5-30 and Figure 5-31 show the test results of Tensile Strength (TS) for control
mixtures and polymer-modified asphalt mixtures. The comparison of tensile strength
between control and PMA mixtures is presented in Figure 5-32. The SBS polymer did not
seem to critically affect the HMA tensile strength.
Figure 5-33 through Figure 5-36 show the test results of fracture energy (FE) and
Dissipated Creep Strain Energy (DCSE) for control mixtures and PMA mixtures. As
discussed in modified gradation mixes, the trend of dissipated energy (DCSE) is the same
as for the fracture energy, since the elastic energy part of HMA mixture is essentially
determined by tensile strength and resilient modulus, which did not differ noticeably; for
a specific HMA specimen, the magnitude of the elastic energy observed (0-0.7 KPa) is
usually much lower than the magnitude of the total fracture energy (1-9 KPa).
Comparisons of fracture energy and DCSE values are displayed in Figure 5-37 and Figure
5-38. Most of the points fall above the equality line indicating that SBS polymer tends to
increase the fracture energy or DCSE and hence improve the fatigue cracking
performance of HMA mixtures. However, no specific relationship was observed between
the fracture energy parameters and the SBS polymer content.
At mid-range to high temperatures (25˚C and 40˚C), the PMA mixtures exhibit
complicated behavior on failure strains which did not show any clear trend, probably due
to the enhanced viscous effect of the polymer-modified binder, which makes the mixture
properties more dependent on the overall particle distributions of the SBS polymer, the
asphalt, and the aggregate. At low testing temperatures (-10˚C and 5˚C), it is found that
the failure strain of PMA mixtures tends to increase with an increase of SBS polymer
content, as shown in Figure 5-39. This phenomenon is in agreement with the findings
reported by Kennedy et al. (1992).
103
0
2
4
6
-10 5 25 40
Temperature (˚C)
Tensile
Str
ength
(M
Pa)
F2 Control
F2P1
F2P2
F2P3
Figure 5-30: Tensile strength for granite PMA mixes (1 MPa = 145 psi)
0
2
4
6
-10 5 25 40
Temperature (˚C)
Tensile
Str
ength
(M
Pa)
F4 Control
F4P1
F4P2
F4P3
Figure 5-31: Tensile strength for limestone PMA mixes (1 MPa = 145 psi)
104
0
2
4
6
0 2 4 6
Tensile Strength of Control Level (MPa)
Tensile
Str
ength
of
PM
A M
ixes (
MP
a)
F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
Figure 5-32: Comparison of tensile strength between control and PMA mixes
0
2
4
6
8
10
-10 5 25 40
Temperature (˚C)
Fra
ctu
re E
nerg
y (
KP
a)
F2 Control
F2P1
F2P2
F2P3
Figure 5-33: Fracture Energy for granite PMA mixes (1 KPa = 0.145 psi)
105
0
2
4
6
8
10
-10 5 25 40
Temperature (˚C)
Fra
ctu
re E
nerg
y (
KP
a)
F4 Control
F4P1
F4P2
F4P3
Figure 5-34: Fracture Energy for limestone PMA mixes (1 KPa = 0.145 psi)
0
2
4
6
8
10
-10 5 25 40
Temperature (˚C)
DC
SE
(K
Pa)
F2 Control
F2P1
F2P2
F2P3
Figure 5-35: DCSE for granite PMA mixes (1 KPa = 0.145 psi)
106
0
2
4
6
8
10
-10 5 25 40
Temperature (˚C)
DC
SE
(K
Pa)
F4 Control
F4P1
F4P2
F4P3
Figure 5-36: DCSE for limestone PMA mixes (1 KPa = 0.145 psi)
0
2
4
6
8
10
0 2 4 6 8 10
Fracture Energy of Control Mixes (KPa)
Fra
ctu
re E
nerg
y o
f P
MA
Mix
ture
s (
KP
a)
F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
Figure 5-37: Comparison of Fracture Energy between control and PMA mixes
107
0
2
4
6
8
10
0 2 4 6 8 10
DCSE of Control Mixes (KPa)
DC
SE
of
PM
A M
ixtu
res (
KP
a)
F2P1 vs F2C
F2P2 vs F2C
F2P3 vs F2C
Equality Line
F4P1 vs F4C
F4P2 vs F4C
F4P3 vs F4C
Figure 5-38: Comparison of Fracture Energy between control and PMA mixes
0.0E+00
1.0E+03
2.0E+03
3.0E+03
4.0E+03
5.0E+03
0.0 2.0 4.0 6.0 8.0
SBS Polymer Content, % weight
Failu
re S
train
(m
icro
-str
ain
)
F2 Series -10˚C
F2 Series 5˚C
F4 Series -10˚C
F4 Series 5˚C
Figure 5-39: Relationship between the observed Failure Strain and SBS polymer content
108
5.3 Effect of Aggregate Type
Another important factor that influences HMA engineering properties is aggregate
type. As introduced in the preceding chapter, two major aggregate types, granite and
limestone, were used in this HMA fracture mechanics study. For control mixes (F2C and
F4C), the mix designs with two different types of aggregate have the same nominal
maximum aggregate size (1/2 in. or 12.5 mm), the same control limit points, the same
restricted zone, and hence very similar gradation curves (Figure 5-40). For the modified
gradation mixes, the adjustment of aggregate amount for each gradation level (G1 or G2)
at each sieve size is also close, and this makes the adjusted shapes of gradation curves for
the two different types of aggregate appear to be alike. For the mixtures with SBS
polymer-modified asphalt binder, the polymer content is identical at each level (3.0%,
4.5%, and 6.0%). All these analogues provide a basis to evaluate the differences of
fracture mechanics properties between the two types of aggregate. It is commonly known
that limestone aggregate is usually softer than granite aggregate. Figure 5-41 shows a
comparison of resilient modulus between granite and limestone mixtures. All points fall
below the line of equality, which confirms that the granite mixtures are stiffer than
limestone mixtures. In particular, the difference of resilient modulus values between the
two types of aggregate at mid-range to low temperature levels (-10˚C and 5˚C) is much
more remarkable than that at elevated testing temperatures (25˚C and 40˚C). From this
point of view, the limestone mixtures would appear more ductile under low service
temperature conditions, and as a result would be capable of improving the performance of
thermal cracking of pavement structures.
The comparisons of creep compliance between granite and limestone mixtures at
each testing temperature are shown in Figure 5-42 through Figure 5-45. The result is not
as simple as the limestone always being more compliant than granite. At low
temperatures (-10˚C and 5˚C), the creep compliance values of limestone mixtures are all
higher than those of granite mixtures. At 25˚C, the data points are distributed closely
along the line of equality. When the temperature goes up to 40˚C, most of the data points
go under the line of equality, which means that the limestone specimens become less
compliant than the granite specimens. These characteristics exhibited by limestone
109
mixtures are advantageous to pavement structures in improving performance of thermal
cracking at low service temperatures and increasing rutting resistance at high service
temperatures. This effect, in creep aspect of view, is analogous to that of SBS polymer-
modified asphalt binders.
It was expected that the granite mixtures would have higher tensile strength values
than limestone mixtures because generally granite material exhibits a higher hardness
nature than limestone material. Figure 5-46 shows the comparison of tensile strength
between granite and limestone mixtures. The strength values of the two aggregate types
are generally in the same magnitude. The zero interception linear trendline indicates that
the limestone mixtures have less tensile strength than the granite mixtures by a small
margin. The comparison of fracture energy between granite and limestone mixtures is
displayed in Figure 5-47. The plot shows a poor correlation of fracture energy between
granite and limestone mixtures. The reason for this is that the fracture energy result
depends on a few other basic variables including tensile strength, failure strain, and the
dynamic stress-strain behavior of each specific specimen.
0.45 Power Gradation Chart
#
20
0
#1
00
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
F2 Control
F2G1
F2G2
F4 Control
F4G1
F4G2
Figure 5-40: Gradation curves for control mixes and modified gradation mixes
110
y = 0.8293x
R2 = 0.921
y = 0.6856x
R2 = 0.9376
0
10
20
30
40
0 10 20 30 40
Resilient Modulus of Granite Mixtures (GPa)
Resili
ent
Modulu
s o
f Lim
esto
ne M
ixtu
res (
GP
a)
Testing Temperature of
25˚C and 40˚C
Testing Temperature of
-10˚C and 5˚C
Figure 5-41: Comparison of resilient modulus for granite and limestone mixtures
y = 1.2519x
R2 = 0.7599
0.03
0.06
0.09
0.12
0.15
0.03 0.06 0.09 0.12 0.15
D(t) of Granite Mixtures (1/GPa )
D(t
) of
Lim
esto
ne M
ixtu
res (
1/G
Pa)
Figure 5-42: Comparison of CP between granite and limestone mixes at -10˚C
111
y = 1.1903x
R2 = 0.8368
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8
D(t) of Granite Mixtures (1/GPa)
D(t
) of
Lim
esto
ne M
ixtu
res (
1/G
Pa)
Figure 5-43: Comparison of CP between granite and limestone mixes at 5˚C
y = 1.0069x
R2 = 0.9195
0
2
4
6
0 2 4 6
D(t) of GraniteMixtures (1/GPa)
D(t
) of
Lim
esto
ne M
ixtu
res (
1/G
Pa)
Figure 5-44: Comparison of CP between granite and limestone mixes at 25˚C
112
y = 0.7773x
R2 = 0.6655
0
3
6
9
12
15
0 3 6 9 12 15
D(t) of Granite Mixtures (1/GPa)
D(t
) of
Lim
esto
ne M
ixtu
res (
1/G
Pa)
Figure 5-45: Comparison of CP between granite and limestone mixes at 40˚C
y = 0.9279x
R2 = 0.9252
0
2
4
6
0 2 4 6
Tensile Strength of Granite Mixtures (MPa)
Tensile
Str
ength
of
Lim
esto
ne M
ixtu
res (
MP
a)
Figure 5-46: Comparison of Tensile Strength between granite and limestone mixes
113
0
2
4
6
8
10
0 2 4 6 8 10
Fracture Energy of Granite Mixtures (KPa)
Fra
ctu
re E
nerg
y o
f
Lim
esto
ne M
ixtu
res (
KP
a)
Figure 5-47: Comparison of Fracture Energy between granite and limestone mixes
5.4 Summary of Analysis and Findings from Fracture Mechanics Tests
In gradation modifications, the amount of aggregate retained in nominal maximum
sieve was increased. However, the power law regression curves for the mix design series
turned out to be close to each other. The effect of adjusting coarse aggregate gradation by
increasing the 12.5 mm (1/2 in.) sieve aggregate with a range of 5% to 15% appeared to
be negligible on mixtures’ resilient modulus. Similarly, the proposed coarse gradation
modifications did not make a significant difference in the creep compliance properties for
the HMA tested in this study. The tensile strength test results indicated that the increase
of nominal maximum size aggregate downgraded the HMA tensile strength property at
low temperature levels. The reason for this is probably that at low temperatures, the
condition of particle composition is critical to the strength of HMA. The addition of
coarse aggregate decreased the bonding condition between aggregate and asphalt binder.
On the other hand, at high testing temperatures, the asphalt binder becomes softer and
plays a more important role in the mixture. Since both the control mixes and the mixtures
with modified gradations used the same type of asphalt binder (PG 67-22), all mixes
114
showed similar strength properties at elevated temperatures. It is found that most of the
fracture energy values of modified gradations were lower than those of control mixtures,
which means that the mixtures with modified coarse gradations would probably have less
resistance to fatigue cracking than the control mixtures according to the fracture
mechanics model proposed by Roque et al. (2004). Overall, increasing the amount of
nominal maximum size aggregate in this study had negligible or adverse effect on HMA
fracture mechanics properties.
The SBS polymer-modified asphalt binder was found to be beneficial to HMA
fracture mechanics properties in a few ways. Firstly, the SBS polymer improves the
stiffness behavior of asphalt mixtures. The SBS polymer-modified binder makes the
PMA mixtures less stiff than control mixes with unmodified asphalt at low to mid-range
temperature levels (-10˚C and 5˚C). The resilient modulus values of PMA mixtures
decrease with an increase of SBS polymer content throughout the concentration range
tested except at high temperatures. At the testing temperature of 40˚C, the stiffness of
PMA mixtures appeared to be maintained or even increased. The increment of SBS
polymer dosage results in an enhanced resilient modulus superior to that of the control
mixtures. Moreover, it is interesting to note that when the SBS polymer concentration
increases from 4.5% to 6.0%, the resilient modulus values of PMA mixtures drop back
slightly. This implies that an optimum SBS content exists around 4.5% which would
make the HMA stiffest at the high testing temperature. This finding is analogous with the
results reported by many other research studies (Collins et al. 1991; Shih 1996; Chen et
al. 2002, 2003) on polymer-modified binders and PMA mixtures. The analysis performed
on resilient modulus results based on this study indicates that when SBS polymer is used
in the HMA, although increasing the SBS polymer content will always improve low
temperature performance of pavement, limiting the concentration within an optimal range
is especially important at high service temperatures. Secondly, the SBS polymer helps the
HMA obtain an upgraded creep performance. The PMA mixtures are more compliant
than the control mixes with unmodified asphalt binders at the low temperature level (-
10˚C). On the other hand, the PMA mixtures become less compliant than the control
mixes at the high testing temperature (40˚C). At a specific temperature level, a higher
SBS polymer concentration generally results in higher creep compliance values. At some
115
mid-range temperature levels, the PMA mixtures show similar creep property to the
control mixes. These effects should lead to improved resistance to rutting and thermal
cracking of HMA mixtures, and provide a good verification to conclusions drawn by
previous research studies (Lalwani et al. 1982; Carpenter et al. 1987; Pradhan 1993).
Thirdly, the SBS polymer modifier improves the asphalt mixture fracture properties. The
indirect tensile strength test showed that the SBS polymer did not significantly affect
HMA tensile strength as Jones et al. (1998) observed at low temperature levels (around
0˚C). However, the SBS polymer generally increases the fracture energy or creep strain
energy which were indicators of mixtures’ resistance to fatigue cracking. At low
temperatures (-10˚C and 5˚C), the failure strain of PMA mixtures tends to increase with
an increase of SBS polymer content. These performances further justify the benefits of
polymer-modified binder documented by a few other researchers (Kennedy et al. 1992;
King et al. 1993).
Comparison between the two different types of aggregate showed that the limestone
mixtures are less stiff than the granite mixtures for all specimens tested in this study. The
difference in resilient modulus values between the two types of aggregate is more
remarkable at low temperature levels (-10˚C and 5˚C). In addition, the limestone mixtures
are more compliant than granite mixtures at low temperatures. As the temperature
increases, the limestone mixtures shows a creep property similar to that of the granite
mixtures at 25˚C, and then the limestone mixtures become less compliant than granite at
the high temperature (40˚C). These characteristics exhibited by limestone mixtures are
advantageous to pavement structures in improving performance of thermal cracking at
low service temperatures and increasing rutting resistance at high service temperatures.
116
CHAPTER 6
COMPLEX MODULUS AND RESILIENT MODULUS
TEST RESULTS
6.1 Test Procedures
The laboratory testing program conducted in this project included dynamic modulus
testing and indirect tension testing. Both types of testing were conducted in unconfined
conditions. The Interlaken dynamic test system was used to load the specimens. The IDT
resilient modulus test procedures are the same as described in Chapter 4, while in this part
of the study, the specimens were tested at only three temperatures: 5, 25, and 40˚C. The
dynamic modulus tests were conducted at the same three temperature levels and the
procedures are as follows:
The dynamic moduli and phase angle were measured by applying compressive
sinusoidal (haversine) loading. The deformations were measured through three LVDTs
(Linear Variable Differential Transducers). These LVDTs were placed vertically on
diametrically symmetric specimen sides (Figure 6-1).
On the night before testing, parallel studs, used to secure the LVDTs in place, were
glued 100 mm (4 in.) apart and located approximately 25 mm (1 in.) from the top and
bottom of the specimen. The diameter of the specimens was 100 mm (4 in.) and the
height was 150 mm (6 in.), cut and cored from the raw gyratory compacted pills with a
diameter of 150 mm (6 in.) and a height of 165 mm. The specimens were then placed in a
controlled temperature cabinet overnight at 5ºC to ensure temperature equilibrium. On the
morning of testing, the specimens were placed in the environmental chamber at 5ºC and
allowed to equilibrate for two hours. All testing was conducted using this temperature
controlled chamber capable of holding temperatures from -16 to 60°C (3.2 to 140°F).
Tests were performed at temperatures of 5, 25, and 40 ºC and frequencies of 25, 10, 5, 1,
and 0.5Hz. Testing began with the lowest temperature and proceeded to the highest
temperature. At a given temperature level, the testing began with the highest frequency of
loading and proceeded to the lowest. This temperature-frequency sequence was carried
117
out to cause minimum damage to the specimens before the next sequential test (Pellinen
2001).
To begin testing, the extensometers were zeroed, and a minimal contact load was
applied to the specimens. A sinusoidal axial compressive load was applied to the
specimens without impact in a cyclic manner. The load was adjusted in each case to
attempt to keep the axial strains between 50 and 150 micro-strains. The first step was to
apply a preconditioning load to the specimens with 200 cycles at 25 Hz. Testing
continued with different numbers of cycles for each frequency as shown in Table 6-1. The
data acquisition system was set up to record the last six cycles at each frequency with
about 200 points per cycle. The raw force and displacement data were manipulated to
obtain the dynamic modulus and phase angle for each specimen. After the entire cycle of
testing was complete at 5ºC, the environmental chamber was set to the next temperature.
After two hours of conditioning, the above steps were repeated until completion of the
entire sequence of temperatures and frequencies.
Figure 6-1: Specimen and LVDTs setup for DMT test
118
Table 6-1: Cycles for DTM test sequence
Frequency, Hz Number of Cycles
Preconditioning (25) 200
25 50
10 50
5 50
1 25
0.5 6
6.2 Presentation of DMT and IDT Testing Results
The measurement and analysis system developed for the SHRP indirect tensile test
(IDT) (Roque et al. 1992, 1997) was applied as described in Chapter 4. The SHRP IDT
could obtain an accurate determination of tensile properties of asphalt mixtures at low
temperatures, with an accurate measuring of Poisson’s ratio. The deformation
measurement mounting system was modified in the SHRP IDT in order to account for
specimen bulging. The resilient modulus and Poisson’s ratio test results are summarized
in Table C-1 for the 20 Superpave HMA mixtures. The average values of resilient
modulus and Poisson’s ratio for each mixture are also shown in Figure 6-2 and Figure 6-3
for the 20 HMA mixes.
Similarly, the dynamic complex moduli for all 20 mixtures were tested at the selected
temperature levels: 5, 25, and 40°C (41, 77, and 104˚F). For all temperatures tested, the
frequencies listed in Table 6-1 were used: 25, 10, 5, 1, and 0.5 Hz. The tests were
conducted from the lowest temperature to the highest temperature and from the highest
frequency to the lowest frequency, as explained earlier. The data processing procedures
described in the National Cooperative Highway Research Program (NCHRP) Project 9-
29, “Simple Performance Tester for Superpave Mix Design”, are used in this study to
process the raw data and compute the dynamic moduli and phase angles. All of the
dynamic modulus and phase angle test results are summarized in Table C-2 and Table C-
3 respectively. The phase angles were grouped together and are shown in Figure 6-4 for
granite, limestone and RAP materials at each testing temperature.
119
0
5
10
15
20
25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Mix Design Series Number
Resili
ent
Modulu
s (
GP
a)
40 degree C 25 degree C 5 degree C
Figure 6-2: Resilient modulus at different testing temperatures (1GPa = 145 psi)
0.1
0.2
0.3
0.4
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Mix Design Series Number
Pois
son's
Ratio
40 degree C 25 degree C 5 degree C
Figure 6-3: Poisson’s Ratio from IDT test for all mixture series
120
0
10
20
30
40
0 5 10 15 20 25 30
Frequency (Hz)
Phase A
ngle
Granite RAP Limestone
40°
25°
5°C
Figure 6-4: Average phase angles for different type of materials
As displayed in the tables and plots, the results clearly show the expected trends of
the dynamic modulus and phase angle for asphalt mixtures. In summary, the two
variables showed the following trends:
1. The dynamic modulus increased as the test frequency increased under a certain
testing temperature.
2. The dynamic modulus increased with a decrease in test temperature under a
certain loading frequency.
3. The phase angle increased as the test temperature increased.
4. The phase angle has a decreasing trend with increasing load frequency under a
certain temperature. A more complex behavior of the phase angle as a function of
the loading frequency was observed at elevated temperatures.
These trends are in agreement with the research results reported by others.
121
CHAPTER 7
CORRELATION OF INDIRECT TENSION RESILIENT MODULUS
AND COMPLEX MODULUS TEST RESULTS
7.1 General
A comparative study between the indirect tension resilient modulus and dynamic
complex modulus experimental results is performed in this chapter. A detailed
comparison of key differences between the dynamic complex modulus test and the
indirect diametral resilient modulus test for asphalt concrete mixtures was summarized in
a position paper by the NCHRP 1-37A 2002 Project research team (Witczak 1999).
Basically, the primary difference between the resilient modulus test and dynamic complex
modulus test for asphalt concrete mixtures is that the former uses loading of any
waveform with a given rest period, while the latter applies a sinusoidal or haversine
loading with no rest period; hence, no delayed elastic rebound would occur during the
test. The transition from the resilient modulus test to the use of the dynamic complex
modulus test for design of flexible pavement structures has hardly been smooth. The
potential impact of adopting the dynamic complex modulus for implementing the new
AASHTO M-E Design Guide is tremendous for state transportation agencies such as the
Florida Department of Transportation (FDOT). The IDT has traditionally been used to
characterize the HMA mixtures for flexible pavement design in Florida, and the test
method has been shown to be both an expedient and a reliable way of obtaining mixture
properties from field cores. In response to the need, a major effort was undertaken by the
FDOT to characterize the Florida HMA mixtures using the dynamic complex modulus
(Birgisson et al. 2004). The Witczak prediction model was adopted to verify the dynamic
modulus results obtained in the laboratory for all mixture series.
7.2 HMA Master Curve Development
In the new AASHTO Mechanistic-Empirical (M-E) Pavement Design Guide, the
dynamic modulus of HMA, at all levels of temperature and loading frequency, was
122
determined from a master curve constructed at a reference temperature. The dynamic
modulus and phase angle of HMA were shifted with respect to the frequency axis until
the curves merged to form a single smooth characteristic curve, which is referred to as the
time-temperature superposition principle. The master curve of the HMA stiffness
described the time dependency of the material. The shift factor, a(T), as a function of
temperature, defined the required shift at a given temperature to obtain the reduced
frequency. It is shown in the following form:
)log()log()](log[)( r
r
ffTaorf
fTa −== (7-1)
where:
f = testing frequency at desired temperature
fr = reduced frequency
T = temperature of interest
Obviously, a(T) = 1 at the reference temperature.
Pellinen and Witczak (2002) developed a mathematical model by a sigmoidal fitting
function for master curve construction. The shift can be calculated by solving the shift
factors simultaneously with the coefficients of the sigmoidal function:
rfeE
log1|)log(|
γβ
αδ
+
∗
++= (7-2)
where
log(|E*|) = log of dynamic modulus
fr = reduced frequency
� = minimum modulus value
D = span of modulus value
�, � = shape parameters
As indicated in the sigmoidal function, the upper limit of the log of dynamic modulus
was D+�, and the minimum value is �. � and � are shape factors that determine the shape
of the master curve. The parameters used in sigmoidal fitting function are demonstrated
123
in Figure 7-1. The characteristics of the sigmoidal function are described as follows. At
the reference temperature, the shift factor a(T) = 1 . The parameter � influenced the
steepness of the function (rate of change between minimum and maximum) and �
influenced the horizontal position of the turning point. The upper part of the sigmoidal
function asymptotically approached the maximum stiffness of the mix, which was
dependent on limiting binder stiffness at cold temperatures. At high temperatures, the
compressive loading caused aggregate influence to be more dominant than the viscous
binder influence. The modulus started to approach a limiting equilibrium value, which is
dependent on the aggregate gradation. Thus, the sigmoidal function captured the physical
behavior of the asphalt mixture observed in the mechanical testing using compressive
cyclic loading through the entire range of temperatures that are typically of interest.
Figure 7-1: Parameters used in sigmoidal fitting function of master curve
7.3 Master Curve Construction
The procedure developed by Pellinen and Witczak (2002) was used for obtaining
predicted master curves for all mixtures in this study. In all master curve constructions,
124
the reference temperature was taken as 25°C (77°F). The shifting factors were obtained
simultaneously with the coefficients of the sigmoidal function through nonlinear
regression, without assuming any functional form of a(T) with respect to temperature.
The nonlinear regression was performed by using the Solver Function of a Microsoft
Excel spreadsheet. For instance, a set of testing values of dynamic modulus for a specific
specimen was obtained, at test temperatures of 5, 25 and 40°C and loading frequencies of
25, 10, 5, 1 and 0.5 Hz. Then the parameters of the sigmoidal function D, �, �, and � were
assumed as well as the shift factor a(T) at each corresponding temperature. Equation 7-1
was used to calculate the sigmoidal fitting values of log(|E*|). Nonlinear regression was
then performed to achieve an optimum fitting between the testing values and the
sigmoidal model calculation, which resulted in obtaining the optimized parameters of the
sigmoidal function and the shift factors.
100
1000
10000
100000
-4 -3 -2 -1 0 1 2 3 4 5 6
Reduced Frequency Log(fr) (Hz)
|E*| (
MP
a)
Figure 7-2: Master curves for granite materials
125
100
1000
10000
100000
-4 -3 -2 -1 0 1 2 3 4 5 6
Reduced Frequency Log(fr) (Hz)
|E*| (
MP
a)
Figure 7-3: Master curves for limestone materials
100
1000
10000
100000
-4 -3 -2 -1 0 1 2 3 4 5 6
Reduced Frequency Log(fr) (Hz)
|E*
| (M
Pa)
)log(5680.05069.0
*
1
3543.20214.2|)log(|
rfe
E⋅−−+
+=
Figure 7-4: Master curves for all mixtures
126
The master curves for granite and limestone materials are grouped together and
shown in Figure 7-2 and Figure 7-3, respectively. The master curves were all similar in
shape and close to each other for the same type of materials. This could be due to the fact
that one type of asphalt binder (PG 67-22) was used for all the dynamic modulus test of
HMA mixtures. All of the master curves were then grouped together and are shown in
Figure 7-4. One regression equation was derived for all the mixtures, presented as
follows:
)log(5680.05069.0
*
1
3543.20214.2|)log(|
rfeE
⋅−−++= (7-3)
7.4 Verification of Dynamic Complex Modulus Experimental Results
Efforts were made by asphalt pavement researchers to develop regression equations
to estimate the dynamic complex modulus for a specific HMA mixture. One of the most
comprehensive mixture dynamic complex modulus models is the Witczak prediction
model (Witczak and Fonseca 1996; Fonseca and Witczak 1996). This prediction model
was based on the volumetric properties of a given mixture and was adopted in the
AASHTO M-E design guide. Past studies (Birgisson et al. 2004; Schwartz 2005; King et
al. 2005; Loulizi et al. 2006) showed that the Witczak predictive model provided
sufficiently accurate and reasonably robust estimates of dynamic complex modulus for
use in the M-E performance prediction and design, although it had more limited ability to
make fine distinctions between the performance of different mixtures at the same
temperature and other design conditions. In this study, the predicted dynamic complex
modulus values from the Witczak prediction equation were compared with the measured
dynamic complex modulus values to verify the test results. The Witczak prediction
equation is presented as follows:
127
[ ])log393532.0log313351.0603313.0(
34
2
38384
4
2
200200
*
1
00547.0)(000017.0003958.00021.0871977.3
)(802208.0058097.0002841.0
)(001767.0029232.0249937.1log
η−−−+
+−+−+
+−−−
−+−=
f
abeff
beff
a
e
PPPP
VV
VVP
PPE
(7-4)
where
|| *E = dynamic (complex) modulus, in 105 psi
η = bituminous viscosity, in 106 poise (at any temperature, degree of aging)
f = load frequency, in Hz
aV = percent air voids content, by volume
beffV = percent effective bitumen content, by volume
34P = percent retained on 19-mm sieve, by total aggregate weight (cumulative)
38P = percent retained on 9.5-mm sieve, by total aggregate weight (cumulative)
4P = percent retained on 4.75-mm sieve, by total aggregate weight (cumulative)
200P = percent passing 0.75-mm sieve, by total aggregate weight (cumulative)
In this model, the parameter E (bitumen viscosity) for each dynamic complex
modulus test temperature is determined by:
TVTSA log)log(log ⋅+=η (7-5)
where A is the regression intercept, T is Rankine temperature and VTS is the slope of log-
log viscosity versus temperature relationship. The A and VTS parameters are functions of
binder type and material characteristics, and they are determined by regression using
experimental data of binder viscosity versus temperature T. For this study, the input
binder viscosity was obtained from Brookfield rotational viscometer results on short-term
Rotational Thin Film Oven (RTFO) aged PG 67-22 specimens (Birgisson et al. 2004).
The binder viscosity values thus obtained are: A = 10.407 and VTS = -3.4655.
The comparison between the measured and predicted dynamic complex moduli for
all mixture series is presented in Figure 7-5. A linear regression with zero intercept was
performed for the comparative analysis. The R2 indicated the goodness of fit, whereas the
linear coefficient (slope) was a measure of the quality of fit between the prediction and
128
test measurement. Since the comparison was made by using measured dynamic complex
modulus as the horizontal x-values, the points below the line of equality indicated a
prediction that is conservative, in which the predicted dynamic complex modulus was
lower than the measured one. However, the regression analysis shown in the figure
indicated a reasonable estimate of the predicted dynamic complex modulus for the
mixtures tested in this study. Since the Witczak prediction model was shown to provide
sufficiently accurate and reasonably robust estimates of dynamic complex modulus for
use in the M-E pavement performance prediction and design, the reasonable correlation
between the measured and predicted dynamic complex moduli for all the mixtures in this
study could be interpreted as the measured dynamic complex modulus values also being
reasonably accurate.
y = 0.927x
R2 = 0.88
100
1000
10000
100000
100 1000 10000 100000
Measured |E*| (MPa)
Pre
dic
ted
|E*|
(M
Pa)
Figure 7-5: Measured vs. predicted dynamic modulus values for all mixtures
7.5 Comparison between Resilient Modulus and Dynamic Modulus
Despite the fundamental differences between the resilient modulus and dynamic
complex modulus (Witczak 1999; Drescher et al. 1997; Zhang et al. 1997; Kim et al.
2004), a number of research studies had been attempted in order to establish direct
129
correlation between the resilient modulus and dynamic complex modulus of asphalt
concrete mixtures (Kim et al. 2004; Birgisson et al. 2004; Loulizi et al. 2006). Birgisson
et al. (2004) developed testing and analysis procedures to accurately determine the tensile
dynamic complex modulus from the SHRP IDT tests. The dynamic complex modulus
was found to be correlated with resilient modulus and testing frequency for the range of
testing temperatures and frequencies. Loulizi et al. (2006) conducted a comparison study
on the dynamic complex modulus and resilient modulus tests, and they found a strong
relationship between the dynamic complex modulus performed at 5 Hz and the resilient
modulus performed at loading time of 0.03 seconds.
Other alternative approaches had also been attempted in order to determine the
dynamic complex modulus from the IDT test with modified loading conditions using the
theory of viscoelasticity (Drescher et al. 1997; Zhang et al. 1997; Kim et al. 2004).
Recently, an analytical method of calculating resilient modulus from the dynamic
complex modulus was also proposed (Lacroix et al. 2007). The proposed theoretical
prediction involved the application of multiaxial linear viscoelastic theory to linear elastic
solutions for the IDT test. The proposed approach could provide reasonable estimates of
the resilient modulus from the dynamic complex modulus of the asphalt concrete
mixtures. The accuracy of the prediction was not affected by assuming a constant
Poisson’s ratio.
In this study, a strong effort was devoted to preparing the cored dynamic complex
modulus specimens with a targeted air void content (approximately 4.0 percent) as close
as possible to the design air void content (4.0 percent) of the resilient modulus specimens.
As shown in Figure 7-6, the air void content between the dynamic complex modulus and
resilient modulus test specimens are reasonably comparable. Thus, the influence of air
void content on the modulus properties of HMA mixtures could be neutralized.
Therefore, temporarily setting aside the fundamental differences in loading mode
(compression versus tension), loading condition (no rest period versus with rest period),
and basic property (elastic versus viscoelastic) between the two test methods, the resilient
modulus and dynamic complex modulus test results for this study may be compared
empirically on the basis of loading frequency.
130
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Va from IDT Specimens (%)
Va
fro
m D
MT
Sp
ecim
ens
(%)
Figure 7-6: Comparison of average air void content between IDT and DMT specimens
The dynamic complex modulus varies with the loading frequency. A proper
frequency that most closely simulates the actual traffic loading should be selected for the
test, so the dynamic complex modulus thus determined would be equivalent to the
resilient modulus for pavement design purposes. The comparisons between the (total)
resilient modulus and dynamic complex modulus test results are shown in Figure 7-7
through Figure 7-9, for the loading frequencies of 10 Hz, 5 Hz, and 1 Hz, respectively.
The linear regression analysis clearly showed a trend of the dynamic complex modulus
increasing with an increase in total resilient modulus at a specific loading frequency. In
particular, the resilient modulus values were very close to the dynamic complex modulus
values at the loading frequency of 5 Hz (Figure 7-8). The comparisons for the loading
frequencies of 25 Hz, 10 Hz, 5 Hz, 1 Hz, and 0.5 Hz are presented in Figure 7-10, and the
linear regression equations at each frequency are also shown in the figure for illustration.
The closest interpreted loading frequency for obtaining an equal value of the resilient
modulus and dynamic complex modulus would be approximately 4 Hz (Figure 7-11).
Hence, based on this empirical correlation, the dynamic complex modulus values
measured at a loading frequency of 4 Hz may be comparable with the resilient modulus
values obtained from the indirect diametral test at the same temperature level (Ping et al.
2008).
131
As mentioned in the introduction, the transition from the use of resilient modulus to
the use of dynamic complex modulus in implementing the AASHTO M-E Pavement
Design Guide was hardly smooth. A significant amount of resilient modulus data
collected in the past are currently in state highway agencies’ inventory; these data are on
the verge of becoming obsolete. The rational comparative approach presented in this
experimental study could be utilized to establish an empirical correlation between
resilient modulus and dynamic complex modulus test results so that the developed
regression correlation could be used to predict the dynamic complex modulus from
measured resilient modulus values of the HMA mixtures already in the inventory or in the
system.
y = 1.1233x
R2 = 0.8919
0
5000
10000
15000
20000
25000
0 5000 10000 15000 20000 25000
Resilient Modulus (MR), MPa
Dy
nam
ic M
od
ulu
s (|
E*
|),
MP
a
Figure 7-7: Resilient modulus versus dynamic complex modulus at 10 Hz
132
y = 1.0265x
R2 = 0.9065
0
5000
10000
15000
20000
25000
0 5000 10000 15000 20000 25000
Resilient Modulus (MR), MPa
Dy
nam
ic M
od
ulu
s (|
E*
|),
MP
a
Figure 7-8: Resilient modulus versus dynamic complex modulus at 5 Hz
y = 0.8087x
R2 = 0.9043
0
5000
10000
15000
20000
25000
0 5000 10000 15000 20000 25000
Resilient Modulus (MR), MPa
Dy
nam
ic M
od
ulu
s (|
E*
|),
MP
a
Figure 7-9: Resilient modulus versus dynamic complex modulus at 1 Hz
133
0
5000
10000
15000
20000
25000
0 5000 10000 15000 20000 25000
Resilient Modulus (MR), MPa
Dy
nam
ic M
od
ulu
s (|
E*
|),
MP
a
line of equality
0.5 Hz, y=0.7162x
1 Hz, y=0.8087x
5 Hz, y=1.0265x
10 Hz, y=1.1233x
25 Hz, y=1.2416x
Figure 7-10: Resilient modulus versus dynamic modulus at various loading frequencies
y = 0.1349Ln(x) + 0.8096
R2 = 0.9999
0.6
0.8
1
1.2
1.4
0.1 1 10 100
Frequency
Lin
ear
Mu
ltip
lica
tio
n F
acto
rs
Figure 7-11: Relationship of linear multiplication factors with DMT loading frequency
134
CHAPTER 8
SUMMARY AND CONCLUSIONS
8.1 Summary
The primary objective of this study was to evaluate the engineering properties of Hot
Mix Asphalt concrete for implementing the AASHTO Mechanistic-Empirical Design
Guide for Pavement Structures. The specific goals of the study were to evaluate the
coarse aggregate gradation limits by adjusting the coarse aggregate amount of the mix
designs and to evaluate the styrene-butadiene-styrene (SBS) polymer-modified asphalt
binder effect utilizing the fracture mechanics tests in Indirect Diametral Tension testing
mode. Another goal of the study was to develop a rational comparison between the
dynamic complex modulus and indirect resilient modulus obtained in laboratory. To
achieve the objectives and goals, a complete dynamic testing system was established to
perform the temperature controlled dynamic tests. A laboratory experimental program
was developed to evaluate two standard control mixes, four mixes with modified coarse
gradation, six mixes with SBS polymer modifier, and twenty selected Superpave asphalt
concrete mixes with a range of aggregates and mix designs.
The two control mix designs included one granite material (F2C) and one limestone
material (F4C). Their coarse part gradations were adjusted to evaluate the effect of
gradation limits specified in the AASHTO mix design guide. The two control mixtures
were further modified by using SBS polymer-modified binder at different concentrations
instead of the base asphalt to study the polymer binder effect on fracture mechanics
properties of the mixtures. The 20 mixtures tested for dynamic complex modulus
included the following types of aggregates: 14 Georgia granite materials, one Nova Scotia
granite, one North Florida limestone, two Central Florida limestone materials, one South
Florida oolite, and one Alabama limestone. Only one base asphalt binder, PG 67-22 (AC-
30), was used for the dynamic modulus test.
The sweep of IDT tests included the resilient modulus test, the creep compliance test,
and the tensile strength test to characterize the fracture mechanics properties of the
asphalt mixtures. The tests were conducted at four temperature levels (-10, 5, 25, and
135
40°C) covering a typical range of asphalt concrete pavement service temperatures. The
creep compliance master curves and the fracture mechanics parameters were analyzed
using the fracture energy model developed by Roque et al. (2004). The analysis indicated
that the modification of coarse part gradation of the mix designs did not influence the
fracture properties noticeably. However, the use of SBS polymer modifier improved the
creep parameters as well as the stiffness characteristics at both low and high
temperatures.
The dynamic complex modulus tests were conducted at three temperature levels: 5,
25, and 40°C. For all temperatures, the following frequencies were used for the dynamic
modulus test: 25, 10, 5, 1, and 0.5 Hz. The dynamic modulus master curves for all 20
mixtures were constructed using the time-temperature superposition principle. The
Witczak prediction model was adopted to verify the laboratory test results. A comparative
study was then made between the dynamic modulus and resilient modulus test results.
The linear regression analysis indicated that the total resilient modulus increased with an
increase in dynamic modulus at a specific loading frequency. The resilient modulus
values were comparable to the dynamic modulus values at the loading frequency of 4 Hz.
8.2 Findings and Conclusions
8.2.1 Coarse Aggregate Gradation Effect
Based on the test results and analyses of this study, the findings may be drawn as
follows for the specific mix designs with 1/2" nominal maximum size:
1. The increase in 1/2" aggregate from 5% to 15% has minimum influence on
mixtures’ resilient modulus and does not make a significant difference in the
creep compliance properties for the coarse mixes tested.
2. At low testing temperatures (-10˚C and 5˚C), the tensile strength values of
mixtures with modified coarse aggregate gradations are less than those of control
level mixtures, which indicated that increasing the amounts of nominal
maximum size aggregate in the standard control mixtures has an adverse effect
on the tensile strength property of the HMA at low temperatures. At mid to high
testing temperatures, the differences are negligible.
136
3. The fracture energy and dissipated creep strain energy values of modified
gradations are slightly less than those of the control mixes, which means that the
mixtures with modified gradations would probably have less resistance to
fatigue cracking than the control mixtures.
Based on the above findings, the aggregate percentage of maximum nominal size
should not exceed 10% to 20% range, because mixtures with modified gradations in this
study show similar or even downgraded fracture mechanics characteristics.
8.2.2 SBS Polymer Modifier Effect
The SBS polymer modifier improved the HMA properties in the following aspects:
1. SBS polymer modifiers make HMA softer at mid to low service temperatures
and stiffer at high temperatures, which are both favorable attributes for the
improvement of HMA performance in terms of low temperature thermal
cracking and high temperature rutting.
2. The polymer modifier makes the HMA more ductile at low temperatures which
would be beneficial for the reduction of thermal cracking. At high temperatures,
the PMA mixes are less compliant and thus more resistant to rutting.
3. The SBS polymer did not critically affect the HMA tensile strength. However, it
tends to increase the fracture energy limit, and hence, improve the fatigue
cracking performance of HMA mixtures.
4. The failure strain of PMA mixtures tends to increase with an increase of SBS
polymer content at low testing temperatures (-10˚C and 5˚C), which is a desired
attribute to improve the low temperature cracking of HMA pavement.
These findings are in agreement with the theoretical suppositions and with other
practical studies. Furthermore, the effect of SBS concentration for PMA mixtures appears
to be consistent with the findings for SBS modified asphalt binders documented by other
researchers. An optimum SBS content exists around 4.5% which would make the PMA
mixture stiffness the highest at high service temperatures. Excessive increase of polymer
content above the optimum level may improve the stiffness, creep, and failure strain
behavior at low temperatures, however, it had an adverse effect on the PMA resilient
modulus property at high temperatures, which was probably due to a combined
137
performance of mixing and distribution of SBS polymer modifier, base bitumen, and
aggregate.
8.2.3 Effect of Aggregate Type
The limestone mixtures are less stiff than granite mixtures. This difference appeared
to be more significant at low temperatures. In addition, limestone materials show
favorable behaviors in creep performance compared with granite materials at all testing
temperatures. These properties implies that limestone materials are much more ductile
than granite under low service temperature conditions and hence presents an advantage
for the thermal cracking performance of HMA, and at high temperatures, limestone
materials have the potential to increase rutting resistance while maintaining the stiffness
of pavement structures. It should be noted that these findings were observed in the
laboratory for the two specific types of aggregate commonly used in Florida.
8.2.4 Correlation between Dynamic Modulus and Resilient Modulus
The dynamic complex modulus test results were comparable with those from other
research studies. The linear regression analysis between the resilient modulus and
dynamic complex modulus test results clearly showed a trend of the dynamic complex
modulus increasing with an increase in resilient modulus at a specific loading frequency.
The resilient modulus values were very close to the dynamic complex modulus values at
the loading frequency of 5 Hz. Based on the empirical correlation, the dynamic complex
modulus values measured at the loading frequency of 4 Hz may be comparable with the
resilient modulus values obtained from the indirect tension diametral test at the same
temperature level. The empirical correlation was established based on the experimental
results of Florida asphalt concrete mixtures. This empirical correlation may not be
applicable to those HMA mixtures with different types of asphalt binders. However, the
rational comparative approach presented could be adopted to establish an empirical
correlation between resilient modulus and dynamic complex modulus test results for
different types of HMA mixtures so that the developed regression correlation could be
138
used to predict the dynamic complex modulus from measured resilient modulus values of
the HMA mixtures.
8.3 Recommendations
Based on the conclusions and limitations of this research study, the primary
recommendations are as follows:
The effect of coarse aggregate gradation should be further evaluated by investigating
more Superpave mixture designs commonly used in field. The dynamic complex modulus
may also be adopted as an indicator. The gradation adjustment for fine aggregate could be
taken into consideration with a desire to reduce the mix design cost while maintaining the
HMA mechanical quality. A broad range of mixtures should be tested to evaluate the
effect of SBS polymer modifier and to obtain a more confident range of optimum
concentration. Similarly, more research work should be conducted to statistically verify
the effect of aggregate type on fracture mechanics properties of HMA. The HMA fracture
energy approach should be further verified for evaluating HMA pavement structures at
some high service temperatures by incorporating the practical fracture and rutting data in
field. The relationship between fracture energy and temperature should be evaluated by
studying the parameters at more temperature levels.
139
APPENDIX A MATERIALS AND MIX DESIGNS
Table A-1: Superpave mix designs sorted by test series
Series Mix Design Size Type Load Application Rap # 7 # 67 S-1-A S-1-B # 78 #89 5/16" W-10 M-10 Stone Sand Asphalt
Level Stone Stone Stone Stone Stone Stone Stone Screenings Screenings Screenings
S-1 SP 03-2460A 12.5 C Structural A0704 43 51 20 CWR AC-30
A0704-3 GA-553 GA-553 GA-553 Quincy
S-2 LD 00-2502A 12.5 Fine D Structural 41 51 20 Starvation Hill PG 67-22
S-3 LD 02-2529A 12.5 Coarse D Structural 41 51 20 PG 67-22
S-4 SP 02-2180A 9.5 Fine B Structural 51 20 21 A/C PG 67-22
GA-553 GA-553 GA-553 Mayo
S-5 SP 03-2921A 9.5 Coarse D Structural 51 20 21 PG 67-22
GA-553 GA-553 GA-553
S-6 SP 03-2922A 19.0 Coarse D Structural 42 43 51 21 PG 67-22
GA-553 GA-553 GA-553 GA-553
S-7 SP 04-3034A 12.5 Coarse D Structural 42 43 51 20 21 PG 67-22
GA-553 GA-553 GA-553 GA-553 GA-553
S-8 SP 03-2610A 12.5 Fine C Structural 52 51 20 PG 67-22
AL-149 AL-149 AL-149
S-9 SP 03-2627A 12.5 Fine C Structural 42 51 20 PG 67-22
87-145 87-145 87-145
S-10 SP 04-3225A 12.5 Fine C Structural 43 51 21 TM-35 PG 67-22
GA-553 GA-553 GA-206 Hill
S-11 SP 02-2194A 12.5 Coarse D Structural 44 54 22 PG 67-22
NS-315 NS-315 NS-315
S-12 SP 03-2452A 12.5 Coarse D Structural 54 21 22 PG 67-22
GA-383 GA-383 GA-383
S-13 SP 03-2941A 19.0 Fine C Structural 42 43 51 20 JCH PG 67-22
GA-553 GA-553 GA-553 GA-553 Ruben
S-14 SP 03-2351A 12.5 Fine B Structural 43 51 20 A/C PG 67-22
GA-553 GA-553 GA-553 Sunny Hills
S-15 SP 05-4015A 12.5 Fine C Structural 43 51 20 A/C PG 67-22
GA-553 GA-553 GA-553 Grandin
S-16 SPM 05-4044A 9.5 Fine B Structural 51 20 21 A/C PG 67-22
GA-553 GA-553 GA-554 Compass Lake
S-17 SPM 05-4051A 9.5 Fine C Structural 51 20 A/C PG 67-22
GA-553 GA-553 Compass Lake
S-18 SP 05-4061A 12.5 Fine C Structural A0638 41 53 20 A/C PG 67-22
38-286 38-268 GA-553 Quincy
S-19 SP 05-4100A 12.5 Fine C Structural 43 51 20 A/C PG 67-22
GA-553 GA-553 GA-553 Quincy
S-20 SP 02-2052A 12.5 FC C Friction 43 51 20 A/C PG 67-22
GA-553 GA-553 GA-553 Quincy
Table A-2: Performance grade binder grading report
Project FSU Dynamic Modulus Testing Date Received 02-17-05
Submitted By Dr. Ping (from C. W. Roberts) Date Tested 02-22 thru 02-24-05
Tested By Hill & Stickles Date Reported 02-24-05
Test Test Temp.
Test Result
P / F
Florida Specification
Spot Test n/a Negative P Negative
Solubility, % n/a 99.71 P Minimum 99.0%
Smoke Point COC, °C n/a 174 P Minimum 125°C
Flash Point COC, °C n/a 316 P Minimum 230°C
Absolute Viscosity, poises 60°C 3329 P 2400 – 3600 poises
Rotational Viscosity, Pa•s 135°C 0.58 P Maximum 3.0 Pa•s
67°C 1.32 P Original Dynamic Shear, G*/sin�, kPa
70°C 0.87 F
Minimum 1.0 kPa
The initial High Temperature Grade is PG 67
RTF Mass Loss, % 163°C 0.293 P Maximum 0.500%
67°C 2.82 P RTF Dynamic Shear, G*/sin�, kPa
70°C 1.88 F
Minimum 2.20 kPa
The final High Temperature Grade is PG 67
25°C 3260 P PAV Dynamic Shear, G*sin�, kPa
22°C 5361 F
Maximum 5000 kPa
The initial Low Temperature Grade is -22
Creep Stiffness S, MPa -12 154 P
Creep Stiffness, M-value -12 0.346 P
Creep Stiffness S, MPa -18 369 F
Creep Stiffness, M-value -18 0.275 F
S Maximum 300 MPa M-value Minimum 0.300
This sample graded out to a final grade of PG 67-22
Note: When heated and stirred there was a granular texture to the asphalt. The Solubility test revealed what appears to be very fine ground tire rubber but the percentage is within acceptable parameters specified by the State of Florida Department of Transportation.
141
Table A-3: Summary of mix designs and aggregates
Test SP RAP Ga-553 Ga-206 Nova Sco. North Fl Central Fl South Fl Alabama
Series Spec # Granite Granite Granite Limestone Limestone Oolite Limestone
S-1 2460
S-2 2502
S-3 2529
S-4 2180
S-5 2921
S-6 2922
S-7 3034
S-8 2610
S-9 2627
S-10 3225
S-11 2194
S-12 2452
S-13 2941
S-14 2351
S-15 4015
S-16 4044
S-17 4051
S-18 4061
S-19 4100
S-20 2052
142
Table A-4: Aggregate gradations for series 1 - 5
Sieve Size (mm) S-1 S-2 S-3 S-4 S-5
19 100 100 100 100 100
12.5 100 93 94 100 100
9.5 89 89 89 100 100
4.75 67 71 56 74 71
2.36 50 53 30 48 42
1.18 37 42 20 39 28
0.6 27 35 15 28 18
0.3 14 22 10 16 13
0.15 8 9 6 7 9
0.075 5.6 4.5 4.3 4.5 6.9
Table A-5: Aggregate gradations for series 6 - 10
Sieve Size (mm) S-6 S-7 S-8 S-9 S-10
19 100 100 100 100 100
12.5 90 95 96 96 98
9.5 77 84 90 88 90
4.75 51 52 72 69 57
2.36 32 32 52 54 40
1.18 22 21 34 38 34
0.6 16 15 24 27 28
0.3 12 9 11 19 16
0.15 9 6 6 12 4
0.075 6.4 5.2 4.5 4.5 4.5
143
Table A-6: Aggregate gradations for series 11 - 15
Sieve Size (mm) S-11 S-12 S-13 S-14 S-15
19 100 100 100 100 100
12.5 98 99 90 100 99
9.5 89 75 79 90 90
4.75 58 44 61 55 61
2.36 38 29 44 40 42
1.18 24 19 35 34 33
0.6 16 13 26 28 26
0.3 10 9 18 16 18
0.15 5 6 8 4 7
0.075 4.5 4.5 4.4 2.9 2.8
Table A-7: Aggregate gradations for series 16 - 20
Sieve Size (mm) S-16 S-17 S-18 S-19 S-20
19 100 100 100 100 100
12.5 100 100 91 100 98
9.5 100 100 83 90 90
4.75 74 75 68 60 59
2.36 50 48 51 43 40
1.18 39 40 39 34 34
0.6 31 30 30 27 26
0.3 23 16 16 19 11
0.15 9 6 8 7 4
0.075 5.6 3.0 4.8 3.0 3.5
144
0.45 Power Gradation Chart
#
20
0
#
10
0
#5
0
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g Restricted Zone
upper control
lower control
S1
S2
S3
Figure A-1: Gradation chart for S1 to S3
0.45 Power Gradation Chart
#
20
0
#1
00
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
S4
S5
Figure A-2: Gradation chart for S4 and S5
145
0.45 Power Gradation Chart
#2
00
#
10
0
#5
0
#3
0
#1
6
#8
#4
3/8
"
1/2
"
3/4
"
1"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g Restricted Zone
upper control
lower control
S7
S8
S9
Figure A-3: Gradation chart for S7 to S9
0.45 Power Gradation Chart
#
20
0
#
10
0
#5
0
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g Restricted Zone
upper control
lower control
S10
S11
S12
Figure A-4: Gradation chart for S10 to S12
146
0.45 Power Gradation Chart
#
20
0
#
10
0
#5
0
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g Restricted Zone
upper control
lower control
S14
S15
S18
Figure A-5: Gradation chart for S14, S15, and S18
0.45 Power Gradation Chart
#2
00
#
10
0
#5
0
#3
0
#1
6
#8
#4
3/8
"
1/2
"
3/4
"
1"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
S16
S17
Figure A-6: Gradation chart for S16 and S17
147
0.45 Power Gradation Chart
#
20
0
#
50
#
30
#
16
#
8
#
4
3
/8"
1
/2"
3
/4"
1
"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
S6
S13
#1
00
Figure A-7: Gradation chart for S6 and S13
0.45 Power Gradation Chart
#2
00
#
10
0
#5
0
#3
0
#1
6
#8
#4
3/8
"
1/2
"
3/4
"
1"
0
10
20
30
40
50
60
70
80
90
100
Pe
rce
nt P
assin
g
Restricted Zone
upper control
lower control
S19
S20
Figure A-8: Gradation chart for S19 and S20
148
Table A-8: Lab analysis report for 0.0% polymer base asphalt (Graded as PG67-22)
Report Date: 10/1/08 Bituminous Tech. Lab No:
374208
Terminal: Mariani Asphalt Co.
Address: 500 North 19th St.
Tampa, FL 33605
Sample: 0.0% Polymer Date Tested: 9/10/08 – 9/11/08
Test Test Method Specification Test Results Original Binder
Absolute Viscosity, Poise
T202 3290 Poise
Solubility, % soluble T44 99.0% minimum 99.98%
Spot Test Negative Negative
Flash Point, ˚C T48 230˚C minimum 316˚C+
Smoke Point, ˚C 260˚C
Softening Point, ˚F 125˚F
Rotational Viscosity, Pa.s, @135˚C
T316 3.0 maximum 0.5 Pa.s
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 1.0 minimum at 64˚C 1.0 minimum at 67˚C 1.0 minimum at 76˚C
1.97 kPa 1.32 kPa
0.9617 kPa
RTFOT Residue Mass Change, % T240 1.0 maximum +0.028%
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 2.2 minimum at 64˚C 2.2 minimum at 67˚C 2.2 minimum at 76˚C
3.42 kPa 2.64 kPa 1.659 kPa
R28, PAV @100˚C Residue Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 5000 maximum at 25˚C 5000 maximum at 22˚C
4070 kPa 5840 kPa
Creep Stiffness, S, @60 sec.
T314 300 maximum at -12˚C 300 maximum at -18˚C
213 479
Creep Stiffness, m-value, @60 sec.
T314 0.3 minimum at -12˚C 0.3 minimum at -18˚C
0.321 0.248
149
Table A-9: Lab analysis report for 3.0% polymer asphalt (Graded as PG76-22)
Report Date: 10/1/08 Bituminous Tech. Lab No:
374108
Terminal: Mariani Asphalt Co.
Address: 500 North 19th St.
Tampa, FL 33605
Sample: 3.0% Polymer Date Tested: 9/10/08 – 9/11/08
Test Test Method Specification Test Results Original Binder
Absolute Viscosity, Poise
T202 19583 Poise
Solubility, % soluble T44 99.0% minimum 99.83%
Spot Test Negative Negative
Flash Point, ˚C T48 230˚C minimum 316˚C+
Smoke Point, ˚C 260˚C
Softening Point, ˚F 145˚F
Rotational Viscosity, Pa.s, @135˚C
T316 3.0 maximum 1.57 Pa.s
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 1.0 minimum at 76˚C 1.0 minimum at 82˚C
1.359 kPa 0.9178 kPa
RTFOT Residue Mass Change, % T240 1.0 maximum -0.026%
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 2.2 minimum at 76˚C 2.2 minimum at 82˚C
2.37 kPa 1.38 kPa
R28, PAV @100˚C Residue Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 5000 maximum at 31˚C 5000 maximum at 28˚C 5000 maximum at 25˚C 5000 maximum at 22˚C 5000 maximum at 19˚C
1880 kPa 2330 kPa 3450 kPa 4870 kPa 6100 kPa
Creep Stiffness, S, @60 sec.
T314 300 maximum at -12˚C 300 maximum at -18˚C
176 371
Creep Stiffness, m-value, @60 sec.
T314 0.3 minimum at -12˚C 0.3 minimum at -18˚C
0.329 0.266
150
Table A-10: Lab analysis report for 4.5% polymer asphalt (Graded as PG82-22)
Report Date: 10/1/08 Bituminous Tech. Lab No:
374008
Terminal: Mariani Asphalt Co.
Address: 500 North 19th St.
Tampa, FL 33605
Sample: 4.5% Polymer Date Tested: 9/16/08 – 9/19/08
Test Test Method Specification Test Results Original Binder
Absolute Viscosity, Poise
T202 Too viscous
Solubility, % soluble T44 99.0% minimum 99.49%
Spot Test Negative Negative
Flash Point, ˚C T48 230˚C minimum 316˚C+
Smoke Point, ˚C 288˚C
Softening Point, ˚F 197.5˚F
Rotational Viscosity, Pa.s, @135˚C
T316 3.0 maximum 7.40 Pa.s
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 1.0 minimum at 76˚C 1.0 minimum at 82˚C 1.0 minimum at 88˚C 1.0 minimum at 94˚C 1.0 minimum at 100˚C
3.47 kPa 2.40 kPa 1.70 kPa 1.26 kPa
0.3799 kPa
RTFOT Residue Mass Change, % T240 1.0 maximum -0.030%
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 2.2 minimum at 76˚C 2.2 minimum at 82˚C 2.2 minimum at 88˚C
4.45 kPa 2.82 kPa 1.78 kPa
R28, PAV @100˚C Residue Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 5000 maximum at 31˚C 5000 maximum at 25˚C 5000 maximum at 19˚C 5000 maximum at 16˚C 5000 maximum at 13˚C
969 kPa 1860 kPa 3530 kPa 4780 kPa
NA
Creep Stiffness, S, @60 sec.
T314 300 maximum at -12˚C 300 maximum at -18˚C
87.7 368
Creep Stiffness, m-value, @60 sec.
T314 0.3 minimum at -12˚C 0.3 minimum at -18˚C
0.371 0.262
151
Table A-11: Lab analysis report for 6.0% polymer asphalt (Graded as PG82-28)
Report Date: 10/1/08 Bituminous Tech. Lab No:
373908
Terminal: Mariani Asphalt Co.
Address: 500 North 19th St.
Tampa, FL 33605
Sample: 6.0% Polymer Date Tested: 9/16/08 – 9/19/08
Test Test Method Specification Test Results Original Binder
Absolute Viscosity, Poise
T202 Too viscous
Solubility, % soluble T44 99.0% minimum 99.89%
Spot Test Negative Negative
Flash Point, ˚C T48 230˚C minimum 316˚C+
Smoke Point, ˚C 280˚C
Softening Point, ˚F 204˚F
Rotational Viscosity, Pa.s, @135˚C
T316 3.0 maximum 5.75 Pa.s
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 1.0 minimum at 76˚C 1.0 minimum at 82˚C 1.0 minimum at 88˚C 1.0 minimum at 94˚C 1.0 minimum at 100˚C
3.32 kPa 2.17 kPa 1.47 kPa 1.05 kPa
0.8748 kPa
RTFOT Residue Mass Change, % T240 1.0 maximum -0.030%
Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 2.2 minimum at 76˚C 2.2 minimum at 82˚C 2.2 minimum at 88˚C 2.2 minimum at 94˚C
6.94 kPa 4.52 kPa 2.95 kPa 1.91 kPa
R28, PAV @100˚C Residue Dynamic Shear, kPa (G*/sin�, 10 rad/sec)
T315 5000 maximum at 31˚C 5000 maximum at 25˚C 5000 maximum at 19˚C 5000 maximum at 16˚C
947 kPa 1960 kPa 3840 kPa 5340 kPa
Creep Stiffness, S, @60 sec.
T314 300 maximum at -12˚C 300 maximum at -18˚C 300 maximum at -24˚C
105 231 380
Creep Stiffness, m-value, @60 sec.
T314 0.3 minimum at -12˚C 0.3 minimum at -18˚C 0.3 minimum at -24˚C
0.369 0.314 0.250
152
Table A-12: Gradations for F2C and its adjustments
Sieve Size (um) F2C F2G1 Adjustment F2G2 Adjustment
1 25000
3/4 19000 100 100 0 100 0
1/2 12500 95 90 -5 80 -15
3/8 9500 84 78 -6 71 -13
4 4750 52 52 0 52 0
8 2360 32 32 0 32 0
16 1180 21 21 0 21 0
30 600 15 15 0 15 0
50 300 9 9 0 9 0
100 150 6 6 0 6 0
200 75 5.2 5.2 0 5.2 0
Table A-13: Gradations for F4C and its adjustments
Sieve Size (um) F4C F4G1 Adjustment F4G2 Adjustment
1 25000
3/4 19000 100 100 0 100 0
1/2 12500 94 89 -5 79 -15
3/8 9500 89 80 -9 71 -18
4 4750 56 56 0 56 0
8 2360 30 30 0 30 0
16 1180 20 20 0 20 0
30 600 15 15 0 15 0
50 300 10 10 0 10 0
100 150 6 6 0 6 0
200 75 3.6 3.6 0 3.6 0
153
Table A-14: Volumetric properties of mixture design series 1 - 5
Property Symbol S1 S2 S3 S4 S5
Maximum theoretical density Gmm 2.543 2.276 2.253 2.550 2.563
Specific gravity of asphalt Gb 1.035 1.035 1.035 1.035 1.035
Bulk specific gravity of compacted mix Gmb 2.441 2.185 2.162 2.448 2.460
Asphalt content Pb 5.3 8.2 8.2 5.3 5.8
Bulk specific gravity of aggregate Gsb 2.725 2.346 2.311 2.745 2.776
Effective specific gravity of aggregate Gse 2.769 2.549 2.518 2.778 2.819
Asphalt absorption Pba 0.6 3.514 3.676 0.442 0.572
Effective asphalt content in the mixture Pbe 4.7 4.97 4.8 4.9 5.2
Percent VMA in compacted mix VMA 15.2 14.5 14.1 15.5 16.5
Percent air voids in compacted mix Va 4.0 4.0 4.0 4.0 4.0
Percent VFA in compacted mix VFA 74 72 72 74 76
Dust/asphalt ratio D/A 1.2 0.9 0.9 0.9 1.3
Table A-15: Volumetric properties of mixture design series 6 - 10
Property Symbol S6 S7 S8 S9 S10
Maximum theoretical density Gmm 2.603 2.589 2.494 2.313 2.554
Specific gravity of asphalt Gb 1.035 1.035 1.035 1.035 1.035
Bulk specific gravity of compacted mix Gmb 2.499 2.485 2.393 2.220 2.452
Asphalt content Pb 4.5 5.0 5.7 7.5 5.3
Bulk specific gravity of aggregate Gsb 2.781 2.775 2.689 2.389 2.729
Effective specific gravity of aggregate Gse 2.803 2.811 2.726 2.570 2.783
Asphalt absorption Pba 0.293 0.480 0.527 3.056 0.730
Effective asphalt content in the mixture Pbe 4.2 4.5 5.2 4.68 4.61
Percent VMA in compacted mix VMA 14.2 14.9 16.1 14.0 14.9
Percent air voids in compacted mix Va 4.0 4.0 4.0 4.0 4.0
Percent VFA in compacted mix VFA 72 73 75 71 73
Dust/asphalt ratio D/A 1.5 1.2 0.9 1.0 1.0
154
Table A-16: Volumetric properties of mixture design series 11 - 15
Property Symbol S11 S12 S13 S14 S15
Maximum theoretical density Gmm 2.420 2.555 2.571 2.570 2.557
Specific gravity of asphalt Gb 1.035 1.035 1.035 1.035 1.035
Bulk specific gravity of compacted mix Gmb 2.322 2.454 2.468 2.467 2.455
Asphalt content Pb 6.0 5.4 5.0 4.8 5.0
Bulk specific gravity of aggregate Gsb 2.604 2.701 2.763 2.752 2.764
Effective specific gravity of aggregate Gse 2.646 2.789 2.789 2.778 2.772
Asphalt absorption Pba 0.631 1.206 0.347 0.348 0.101
Effective asphalt content in the mixture Pbe 5.41 4.26 4.6 4.4 4.9
Percent VMA in compacted mix VMA 16.2 14.1 15.1 14.7 15.6
Percent air voids in compacted mix Va 4.0 4.0 4.0 4.0 4.0
Percent VFA in compacted mix VFA 75 72 74 73 74
Dust/asphalt ratio D/A 0.8 1.1 1.0 0.7 0.6
Table A-17: Volumetric properties of mixture design series 16 - 20
Property Symbol S16 S17 S18 S19 S20
Maximum theoretical density Gmm 2.550 2.535 2.445 2.567 2.539
Specific gravity of asphalt Gb 1.035 1.035 1.035 1.035 1.035
Bulk specific gravity of compacted mix Gmb 2.448 2.434 2.348 2.464 2.438
Asphalt content Pb 5.2 6.0 6.4 5.0 5.3
Bulk specific gravity of aggregate Gsb 2.750 2.748 2.572 2.756 2.757
Effective specific gravity of aggregate Gse 2.773 2.793 2.696 2.784 2.764
Asphalt absorption Pba 0.307 0.612 1.853 0.376 0.092
Effective asphalt content in the mixture Pbe 4.9 5.4 4.6 4.6 5.2
Percent VMA in compacted mix VMA 15.6 16.7 14.6 15.1 16.3
Percent air voids in compacted mix Va 4.0 4.0 4.0 4.0 4.0
Percent VFA in compacted mix VFA 74 76 73 74 75
Dust/asphalt ratio D/A 1.2 0.8 1.2 0.7 0.7
155
APPENDIX B CREEP COMPLIANCE TEST RESULTS
Table B-1: Creep compliance test results at -10˚C (1/GPa)
Mixtures for Gradation Effects
Control G1 G2
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.044 0.055 0.040 0.058 0.048 0.058
2 0.048 0.062 0.045 0.062 0.053 0.062
5 0.055 0.069 0.049 0.068 0.059 0.068
10 0.061 0.076 0.055 0.073 0.064 0.073
20 0.067 0.081 0.061 0.078 0.071 0.079
50 0.078 0.099 0.069 0.086 0.085 0.093
100 0.091 0.111 0.079 0.091 0.096 0.098
Pr 0.382 0.347 0.299 0.322 0.296 0.352
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.044 0.069 0.076 0.077 0.054 0.091
2 0.049 0.076 0.087 0.084 0.060 0.101
5 0.057 0.087 0.104 0.093 0.069 0.116
10 0.064 0.098 0.116 0.104 0.078 0.128
20 0.072 0.109 0.123 0.118 0.088 0.145
50 0.091 0.126 0.156 0.134 0.109 0.178
100 0.105 0.149 0.184 0.151 0.130 0.196
Pr 0.387 0.378 0.380 0.377 0.356 0.296
Note: 1/GPa = 6.89×10-6/psi
156
Table B-2: Creep compliance test results at 5˚C (1/GPa)
Mixtures for Gradation Effects
Control G1 G2
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.093 0.120 0.099 0.117 0.125 0.131
2 0.117 0.148 0.127 0.138 0.152 0.163
5 0.158 0.198 0.172 0.183 0.200 0.216
10 0.198 0.246 0.221 0.230 0.249 0.266
20 0.268 0.313 0.297 0.295 0.308 0.322
50 0.377 0.449 0.421 0.402 0.430 0.437
100 0.516 0.598 0.557 0.529 0.555 0.540
Pr 0.317 0.334 0.371 0.343 0.357 0.416
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.092 0.126 0.077 0.140 0.116 0.182
2 0.109 0.145 0.095 0.164 0.140 0.223
5 0.138 0.185 0.121 0.198 0.187 0.288
10 0.171 0.220 0.156 0.246 0.227 0.344
20 0.209 0.268 0.187 0.304 0.297 0.407
50 0.292 0.364 0.269 0.424 0.367 0.535
100 0.388 0.476 0.335 0.512 0.440 0.658
Pr 0.428 0.452 0.489 0.303 0.299 0.433
Table B-3: Creep compliance test results at 25˚C (1/GPa)
Mixtures for Gradation Effects
Control G1 G2
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.692 0.699 0.744 0.839 0.794 1.108
2 0.899 0.962 1.045 1.146 1.075 1.547
5 1.420 1.400 1.547 1.662 1.546 2.285
10 1.954 1.819 2.107 2.257 2.192 3.023
20 2.535 2.420 2.854 2.997 2.905 3.885
50 3.626 3.259 4.159 4.359 4.206 5.282
100 4.733 4.028 5.604 5.778 5.509 6.534
Pr 0.338 0.358 0.269 0.265 0.304 0.357
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
Time (sec.) F2 F4 F2 F4 F2 F4
1 0.488 0.542 0.709 0.824 1.126 0.881
2 0.669 0.734 1.006 1.077 1.460 1.117
5 1.054 1.084 1.540 1.542 1.932 1.589
10 1.463 1.460 2.066 1.968 2.413 2.020
20 1.916 2.056 2.724 2.524 3.043 2.575
50 2.754 2.854 3.780 3.426 4.043 3.532
100 3.625 3.679 4.725 4.278 4.938 4.305
Pr 0.352 0.352 0.374 0.246 0.305 0.372
157
Table B-4: Creep compliance test results at 40˚C (1/GPa)
Mixtures for Gradation Effects
Control G1 G2
Time (sec.) F2 F4 F2 F4 F2 F4
1 3.994 2.266 1.734 1.670 1.976 1.616
2 5.019 2.727 2.242 2.196 2.565 2.195
5 6.800 3.588 3.294 3.400 3.597 2.965
10 8.073 4.468 4.178 4.377 4.585 3.696
20 9.383 5.543 5.055 5.374 5.812 4.706
50 11.049 7.884 7.134 7.139 8.153 6.305
100 12.721 9.516 8.817 8.764 10.921 7.950
Pr 0.291 0.325 0.268 0.312 0.298 0.306
Mixtures with SBS Polymer-modified Binder
P1 (3.0%) P2 (4.5%) P3 (6.0%)
Time (sec.) F2 F4 F2 F4 F2 F4
1 1.357 1.562 1.931 2.080 3.628 1.899
2 1.793 1.840 2.476 2.655 4.147 2.268
5 2.671 2.370 3.323 3.495 4.980 2.972
10 3.457 2.783 4.142 4.897 5.742 3.506
20 4.490 3.299 4.910 6.384 6.669 4.034
50 6.033 4.175 6.089 8.189 8.137 5.179
100 8.108 5.423 7.359 9.983 9.533 6.128
Pr 0.373 0.396 0.383 0.272 0.291 0.418
Creep Compliance for F2 Series at -10 Degree C
0.00
0.05
0.10
0.15
0.20
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2P1
F2P2
F2P3
Figure B-1: Creep compliance of F2 control and all polymer-modified levels at -10˚C.
158
Creep Compliance for F4 Series at -10 Degree C
0.00
0.05
0.10
0.15
0.20
0.25
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4P1
F4P2
F4P3
Figure B-2: Creep compliance of F4 control and all polymer-modified levels at -10˚C.
Creep Compliance for F2 Series at 5 Degree C
0.0
0.2
0.4
0.6
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2P1
F2P2
F2P3
Figure B-3: Creep compliance of F2 control and all polymer-modified levels at 5˚C.
Creep Compliance for F4 Series at 5 Degree C
0.0
0.2
0.4
0.6
0.8
0 50 100
Time (sec.)
Cre
ep
(1
/GP
a)
Control
F4P1
F4P2
F4P3
Figure B-4: Creep compliance of F4 control and all polymer-modified levels at 5˚C.
159
Creep Compliance for F2 Series at 25 Degree C
0
2
4
6
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2P1
F2P2
F2P3
Figure B-5: Creep compliance of F2 control and all polymer-modified levels at 25˚C.
Creep Compliance for F4 Series at 25 Degree C
0
2
4
6
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4P1
F4P2
F4P3
Figure B-6: Creep compliance of F4 control and all polymer-modified levels at 25˚C.
Creep Compliance for F2 Series at 40 Degree C
0
5
10
15
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2P1
F2P2
F2P3
Figure B-7: Creep compliance of F2 control and all polymer-modified levels at 40˚C.
160
Creep Compliance for F4 Series at 40 Degree C
0
3
6
9
12
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4P1
F4P2
F4P3
Figure B-8: Creep compliance of F4 control and all polymer-modified levels at 40˚C.
Creep Compliance for F2 Series at -10 Degree C
0.00
0.05
0.10
0.15
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2G1
F2G2
Figure B-9: Creep compliance of F2 control and modified gradation levels at -10˚C.
Creep Compliance for F4 Series at -10 Degree C
0.00
0.05
0.10
0.15
0.20
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4G1
F4G2
Figure B-10: Creep compliance of F4 control and modified gradation levels at -10˚C.
161
Creep Compliance for F2 Series at 5 Degree C
0.0
0.2
0.4
0.6
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2G1
F2G2
Figure B-11: Creep compliance of F2 control and modified gradation levels at 5˚C.
Creep Compliance for F4 Series at 5 Degree C
0.0
0.2
0.4
0.6
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4G1
F4G2
Figure B-12: Creep compliance of F4 control and modified gradation levels at 5˚C.
Creep Compliance for F2 Series at 25 Degree C
0
2
4
6
8
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2G1
F2G2
Figure B-13: Creep compliance of F2 control and modified gradation levels at 25˚C.
162
Creep Compliance for F4 Series at 25 Degree C
0
2
4
6
8
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4G1
F4G2
Figure B-14: Creep compliance of F4 control and modified gradation levels at 25˚C.
Creep Compliance for F2 Series at 40 Degree C
0
5
10
15
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F2G1
F2G2
Figure B-15: Creep compliance of F2 control and modified gradation levels at 40˚C.
Creep Compliance for F4 Series at 40 Degree C
0
3
6
9
12
0 50 100
Time (sec.)
Cre
ep (
1/G
Pa)
Control
F4G1
F4G2
Figure B-16: Creep compliance of F4 control and modified gradation levels at 40˚C.
163
APPENDIX C TEST RESULTS FOR IDT AND DMT
Table C-1: Summary of resilient modulus and Poisson’s Ratio test results
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(°C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-1-2B 1625 1599 0.39 0.36 5
S-1-5B 1419 1522
1474 1537
0.28 0.34
0.32 0.34
S-1-2B 576 570 --- --- 25
S-1-5B 465 521
462 516
0.34 0.34
0.35 0.35
S-1-2B 165 215 --- ---
S-1
40 S-1-5B 183
174 217
216 0.35
0.35 0.32
0.32
S-2-11A 1279 1507 0.33 0.37
S-2-11B 1113 1180 0.25 0.28
S-2-12A 1283 1490 0.37 0.37 5
S-2-12B 1012
1172
1223
1350
0.29
0.31
0.36
0.34
S-2-11A --- --- --- ---
S-2-11B 439 549 0.27 0.30
S-2-12A 406 505 0.37 0.35 25
S-2-12B 438
428
519
524
0.36
0.33
0.32
0.32
S-2-11A 185 210 0.46 0.46
S-2-11B 168 205 0.20 0.27
S-2-12A 184 229 0.28 0.33
S-2
40
S-2-12B ---
179
---
215
---
0.31
---
0.35
S-3-8A 1035 1208 0.22 0.23
S-3-8B 1194 1357 0.26 0.29
S-3-9A 1194 1357 0.26 0.29 5
S-3-9B 1122
1136
1298
1305
0.32
0.27
0.32
0.28
S-3-8A --- --- --- ---
S-3-8B 467 605 0.28 0.35
S-3-9A 410 543 0.24 0.30 25
S-3-9B 475
451
604
584
0.35
0.29
0.38
0.34
S-3-8A 178 214 0.28 0.25
S-3-8B --- --- --- ---
S-3-9A 164 209 0.32 0.34
S-3
40
S-3-9B 218
186
263
229
0.38
0.33
0.33
0.31
164
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-4-13B --- --- --- ---
S-4-14A 1272 1464 0.28 0.33 5
S-4-14B 1089
1180
1313
1389
0.31
0.30
0.31
0.32
S-4-13B 367 447 0.28 0.25
S-4-14A --- --- --- --- 25
S-4-14B 414
390
524
485
0.23
0.26
0.25
0.25
S-4-13B --- --- --- ---
S-4-14A 94 122 0.22 0.23
S-4
40
S-4-14B ---
94
---
122
---
0.22
---
0.23
S-5-3A 2392 2620 0.33 0.33
S-5-3B 2732 2953 0.33 0.30
S-5-4A 2443 2768 0.31 0.28
S-5-5A 2845 3122 0.33 0.34
5
S-5-5B 2957
2674
3320
2958
0.34
0.32
0.36
0.33
S-5-3A 672 942 0.29 0.26
S-5-3B 833 1004 0.30 0.29
S-5-4A 717 912 0.38 0.32
S-5-5A 855 1079 0.27 0.34
25
S-5-5B 870
789
1261
1040
0.37
0.32
0.40
0.32
S-5-3A 245 329 0.29 0.31
S-5-3B 313 411 0.35 0.39
S-5-4A 253 305 0.40 0.35
S-5-9A 296 181 0.25 0.34
S-5
40
S-5-9B 194
237
339
336
0.33
0.33
0.47
0.37
S-6-5A 3312 3505 0.30 0.27 5
S-6-6B 2955 3134
3170 3337
0.17 0.24
0.16 0.21
S-6-5A 1204 1378 0.38 0.38 25
S-6-6B 1361 1282
1550 1464
0.35 0.37
0.31 0.34
S-6-5A 279 451 0.31 0.30
S-6
40 S-6-6B 387
333 564
508 0.31
0.31 0.32
0.31
165
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-7-4B 2264 2536 0.20 0.20
S-7-5A 2303 2564 0.25 0.25 5
S-7-5B 1430
1999
1570
2223
0.16
0.20
0.15
0.20
S-7-4B 718 977 0.27 0.26
S-7-5A 681 957 0.29 0.31 25
S-7-5B 486
629
641
858
0.16
0.24
0.13
0.23
S-7-4B 196 298 0.34 0.35
S-7-5A 167 276 0.31 0.33
S-7
40
S-7-5B 122
162
197
257
0.18
0.28
0.24
0.31
S-8-4A 1952 2178 0.17 0.18
S-8-4B 1723 1929 0.22 0.21 5
S-8-5B 1805
1827
2015
2041
0.23
0.21
0.22
0.20
S-8-4A 542 804 0.23 0.27
S-8-4B 590 785 0.41 0.40 25
S-8-5B 407
513
613
734
0.17
0.27
0.19
0.29
S-8-4A 175 280 0.35 0.39
S-8-4B 136 220 0.36 0.40
S-8
40
S-8-5B 145
152
232
244
0.34
0.35
0.42
0.40
S-9-4A 1797 1985 0.22 0.22
S-9-5A 1991 2165 0.22 0.22 5
S-9-5B 1538
1776
1715
1955
0.19
0.21
0.20
0.21
S-9-4A 690 836 0.24 0.22
S-9-5A 523 675 0.21 0.21 25
S-9-5B 483
566
621
711
0.19
0.22
0.20
0.21
S-9-4A 227 375 0.28 0.34
S-9-5A 172 275 0.22 0.24
S-9
40
S-9-5B 161
187
258
303
0.19
0.23
0.21
0.26
166
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-10-5B 2070 2270 0.26 0.24
S-10-6A 1993 2210 0.19 0.20 5
S-10-6B 2344
2136
2580
2353
0.20
0.22
0.19
0.21
S-10-5B 584 780 0.31 0.30
S-10-6A 612 805 0.26 0.26 25
S-10-6B 651
616
824
803
0.23
0.27
0.18
0.25
S-10-5B 175 445 0.40 0.46
S-10-6A 177 269 0.28 0.37
S-10
40
S-10-6B 170
174
293
336
0.19
0.29
0.25
0.36
S-11-3A 1936 2146 0.24 0.25
S-11-3B 2017 2213 0.19 0.19 5
S-11-4B 2036
1996
2222
2193
0.21
0.22
0.19
0.21
S-11-3A 557 801 0.35 0.35
S-11-3B 486 742 0.23 0.24 25
S-11-4B 553
532
766
770
0.30
0.29
0.27
0.29
S-11-3A 180 301 0.14 0.19
S-11-3B 166 271 0.24 0.27
S-11
40
S-11-4B 167
171
267
280
0.29
0.22
0.30
0.25
S-12-3B 2427 2738 0.24 0.24
S-12-4A 1953 2148 0.19 0.18 5
S-12-4B 2103
2161
2388
2425
0.20
0.21
0.24
0.22
S-12-3B 579 862 0.22 0.23
S-12-4A 510 676 0.25 0.27 25
S-12-4B 532
541
765
768
0.20
0.22
0.23
0.23
S-12-3B 112 171 0.25 0.24
S-12-4A 125 197 0.24 0.26
S-12
40
S-12-4B 142
126
231
200
0.30
0.26
0.34
0.28
167
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-13-3B 2315 2564 0.21 0.22
S-13-5A 2702 2979 0.15 0.17 5
S-13-5B 2359
2459
2657
2733
0.23
0.20
0.24
0.21
S-13-3B 903 1081 0.29 0.29
S-13-5A 1059 1167 0.22 0.17 25
S-13-5B 831
931
1038
1095
0.18
0.23
0.20
0.22
S-13-3B 238 385 0.25 0.30
S-13-5A 292 487 0.26 0.34
S-13
40
S-13-5B 228
253
376
417
0.29
0.27
0.34
0.33
S-14-3B 1993 2170 0.23 0.21
S-14-4B 2561 2783 0.28 0.29 5
S-14-5A 1974
2176
2198
2384
0.20
0.23
0.20
0.23
S-14-3B 625 997 0.25 0.33
S-14-4B 942 1131 0.34 0.27 25
S-14-5A 600
722
904
1011
0.20
0.26
0.25
0.28
S-14-3B 154 247 0.29 0.38
S-14-4B 327 518 0.35 0.39
S-14
40
S-14-5A 146
209
240
335
0.17
0.27
0.24
0.34
S-15-4B 2937 3222 0.33 0.33
S-15-5A 2311 2550 0.25 0.25 5
S-15-5B 2277
2509
2459
2744
0.18
0.25
0.17
0.25
S-15-4B 889 1088 0.37 0.30
S-15-5A 674 911 0.26 0.25 25
S-15-5B 807
790
1137
1045
0.24
0.29
0.25
0.27
S-15-4B 265 422 0.34 0.33
S-15-5A 206 336 0.33 0.39
S-15
40
S-15-5B 254
242
367
375
0.36
0.34
0.34
0.35
168
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-16-3A 2153 2344 0.24 0.23
S-16-5A 2016 2257 0.18 0.19 5
S-16-5B 2084
2084
2322
2308
0.18
0.20
0.19
0.20
S-16-3A 746 985 0.31 0.30
S-16-5A 601 815 0.22 0.22 25
S-16-5B 622
656
869
890
0.19
0.24
0.24
0.25
S-16-3A 216 355 0.23 0.28
S-16-5A 198 314 0.21 0.26
S-16
40
S-16-5B 209
208
334
334
0.19
0.21
0.27
0.27
S-17-3A 2147 2403 0.31 0.33
S-17-3B 1928 2089 0.17 0.16 5
S-17-5A 1855
1977
2066
2186
0.18
0.22
0.19
0.23
S-17-3A 886 1029 0.35 0.31
S-17-3B 753 976 0.20 0.20 25
S-17-5A 659
766
789
931
0.22
0.26
0.18
0.23
S-17-3A 285 411 0.31 0.31
S-17-3B 265 396 0.29 0.31
S-17
40
S-17-5A 247
266
401
403
0.28
0.29
0.32
0.31
S-18-4A 2631 2835 0.24 0.25
S-18-4B 2551 2787 0.19 0.21 5
S-18-5A 2696
2626
2923
2848
0.26
0.23
0.28
0.25
S-18-4A 1702 1809 0.48 0.43
S-18-4B 1343 1475 0.19 0.19 25
S-18-5A 1347
1464
1540
1608
0.28
0.31
0.28
0.30
S-18-4A 807 919 0.38 0.35
S-18-4B 783 936 0.25 0.28
S-18
40
S-18-5A 640
743
968
941
0.31
0.31
0.47
0.37
169
Table C-1: Summary of Resilient Modulus and Poisson’s Ratio Test Results (Continued)
MR (ksi) Poisson’s Ratio
Total Instantaneous Total Instantaneous Mix Temp.
(˚C)
Sample
ID Total Avg. Inst. Avg. Total Avg. Inst. Avg.
S-19-3A 2102 2324 0.13 0.13
S-19-4A 2040 2243 0.10 0.10 5
S-19-5B 2076
2073
2255
2274
0.12
0.12
0.12
0.12
S-19-3A 907 1025 0.21 0.22
S-19-4A 713 891 0.13 0.14 25
S-19-5B 735
785
880
932
0.27
0.20
0.25
0.20
S-19-3A 224 353 0.27 0.27
S-19-4A 174 274 0.16 0.17
S-19
40
S-19-5B 209
202
324
317
0.30
0.25
0.30
0.25
S-20-3B 1838 2046 0.15 0.15
S-20-4A 2070 2281 0.17 0.17 5
S-20-4B 2077
1995
2283
2203
0.19
0.17
0.19
0.17
S-20-3B 493 736 0.19 0.19
S-20-4A 653 1004 0.20 0.20 25
S-20-4B 588
578
907
882
0.28
0.22
0.29
0.23
S-20-3B 141 224 0.24 0.25
S-20-4A 132 205 0.26 0.28
S-20
40
S-20-4B 122
131
199
210
0.34
0.28
0.35
0.29
Note: 1 ksi = 6.89 MPa
170
Table C-2: Summary of dynamic modulus testing results
Dynamic Modulus (psi) at Frequency (Hz) Mixture
Temperature
(°C) 25 Hz 10 Hz 5 Hz 1 Hz 0.5 Hz
5 2718556 2505315 2334144 1940728 1763329
25 1363593 1177199 1010382 667755 533823 S1
40 514862 448082 347376 189291 146889
5 2195185 2043910 1928686 1643632 1513252
25 837880 684018 598500 413070 368702 S2
40 370582 349512 275622 156980 130454
5 1665338 1523653 1383432 1041750 926707
25 1015881 840265 708900 471784 393277 S3
40 479986 363429 289464 170541 139570
5 1688377 1515004 1375835 1058085 923347
25 758494 618952 519722 328501 274068 S4
40 321230 239316 189747 124783 102381
5 1940695 1813721 1700122 1423737 1298946
25 1046943 899545 784929 548407 459125 S5
40 437124 332714 265866 155747 128673
5 3219817 2887077 2676451 2166454 1936833
25 1402727 1165921 966038 593581 459123 S6
40 475793 369106 277443 144513 126079
5 2657447 2445081 2254771 1791554 1498947
25 1137414 959075 793326 472220 358789 S7
40 424137 301779 227473 133711 104263
5 2106279 1898941 1711824 1276268 1093710
25 797849 655478 520685 298806 246089 S8
40 276775 198701 158492 95001 73974
5 2279604 2125885 1963997 1563898 1391163
25 1131702 936634 795165 505711 397558 S9
40 431144 327847 254922 159677 129964
5 2771753 2490818 2286261 1808405 1601989
25 1116202 908742 747045 442313 338010 S10
40 390597 273649 203734 121979 95205
171
Table C-2: Summary of dynamic modulus testing results (Continued)
Dynamic Modulus (psi) at Frequency (Hz) Mixture
Temperature
(°C) 25 Hz 10 Hz 5 Hz 1 Hz 0.5 Hz
5 2460067 2216427 2012981 1560490 1367085
25 1012301 824140 672212 392622 306538 S11
40 322355 228212 177715 105194 83242
5 2351471 2090847 1889063 1427932 1233845
25 927331 737812 588777 324985 263962 S12
40 278916 184464 152604 84398 67535
5 3011584 2818300 2624445 2136791 1924084
25 1328278 1125221 953667 612433 485643 S13
40 559182 419748 327764 187856 148362
5 2728458 2499581 2320051 1889681 1703802
25 1218686 1040795 882519 571472 456110 S14
40 554386 414528 324182 189306 142947
5 2572339 2395084 2216750 1775124 1584504
25 1023437 842892 697684 456682 350667 S15
40 400128 285613 214864 132733 102079
5 2677688 2489353 2278723 1827880 1634091
25 1115825 944472 787322 485782 378116 S16
40 385569 281676 220259 131109 101240
5 2526217 2251277 2064508 1640875 1461569
25 1056637 906080 770487 500564 401294 S17
40 409877 311574 248031 153194 120265
5 3453820 3194201 3048829 2708039 2554885
25 1916335 1719572 1566964 1229368 1088818 S18
40 1029854 860866 740421 509969 423155
5 2513563 2344507 2175441 1749294 1564667
25 1144054 935766 779194 466181 357925 S19
40 441512 320771 244152 142058 108477
5 2739617 2560305 2366486 1898011 1689792
25 1048483 842779 696163 407042 309066 S20
40 390325 276400 207707 120487 95873
Note: 1 psi = 6.89 KPa
172
Table C-3: Summary of phase angle testing results
Phase Angle (degree) at Frequency (Hz) Mixture
Temperature
(°C) 25 Hz 10 Hz 5 Hz 1 Hz 0.5 Hz
5 6.9 8.2 9.4 11.8 13.2
25 16.9 19.2 21.6 26.9 30.0 S1
40 28.5 30.4 32.8 35.2 37.6
5 6.5 7.5 8.5 10.4 11.7
25 16.5 18.3 20.2 25.0 28.6 S2
40 29.2 29.1 30.9 32.7 34.7
5 16.1 19.0 20.9 25.2 27.8
25 22.9 25.4 27.7 31.0 32.2 S3
40 31.4 32.0 31.3 29.0 28.8
5 9.5 11.3 12.7 16.5 18.5
25 20.2 22.8 25.2 28.9 31.1 S4
40 30.3 31.5 31.5 27.8 28.1
5 7.4 8.3 9.3 11.2 12.1
25 14.4 17.5 19.1 22.1 23.7 S5
40 22.8 27.9 29.5 30.7 31.2
5 7.2 8.8 10.3 13.3 14.9
25 19.5 22.1 24.9 30.0 32.8 S6
40 35.9 35.1 36.7 36.4 36.2
5 9.6 10.2 11.7 14.3 17.4
25 21.8 24.6 27.2 32.3 35.1 S7
40 32.0 34.7 36.1 33.9 34.6
5 11.7 13.2 14.8 18.7 20.8
25 27.0 28.3 31.1 34.7 37.2 S8
40 35.1 35.2 33.7 29.9 29.4
5 9.0 10.5 11.9 15.0 16.8
25 19.0 21.0 23.6 28.5 31.2 S9
40 30.4 31.4 32.7 29.7 30.3
5 9.2 11.0 12.5 15.7 17.6
25 22.2 24.5 27.1 31.8 34.3 S10
40 32.7 35.8 36.7 32.2 31.9
173
Table C-3: Summary of phase angle testing results (Continued)
Phase Angle (degree) at Frequency (Hz) Mixture
Temperature
(°C) 25 Hz 10 Hz 5 Hz 1 Hz 0.5 Hz
5 9.1 11.8 13.3 16.6 18.4
25 22.2 25.5 28.0 32.6 34.8 S11
40 34.7 37.0 35.6 30.9 30.2
5 8.7 12.7 14.5 18.4 20.7
25 23.0 27.6 30.5 35.8 38.6 S12
40 37.2 38.4 35.2 31.0 29.6
5 7.6 9.3 10.5 13.1 14.6
25 18.6 20.6 22.8 27.3 30.0 S13
40 28.7 30.4 32.0 32.7 34.9
5 9.6 9.8 10.9 13.2 14.5
25 15.4 21.2 23.2 27.2 29.5 S14
40 27.1 28.7 30.4 32.4 33.3
5 6.6 10.3 11.5 14.0 15.5
25 22.6 23.2 25.7 30.7 33.3 S15
40 33.6 34.7 36.1 32.2 32.4
5 4.0 9.7 11.4 14.1 15.5
25 17.1 23.0 25.2 30.0 32.4 S16
40 32.6 33.7 34.2 32.7 33.8
5 5.3 10.8 11.9 14.3 15.6
25 17.9 20.8 22.7 26.4 28.6 S17
40 28.4 29.2 30.4 30.3 31.9
5 3.5 5.9 6.3 7.3 7.8
25 11.1 11.9 13.0 15.1 16.3 S18
40 17.4 19.0 20.4 23.1 24.8
5 8.5 9.6 10.9 13.6 15.2
25 21.1 23.5 26.1 31.7 34.6 S19
40 32.7 34.4 36.0 35.4 36.6
5 8.5 10.1 11.3 14.2 15.8
25 21.2 24.3 26.8 32.3 35.6 S20
40 34.8 36.2 37.4 35.0 35.1
174
REFERENCES
1. AASHTO, “Guide for Design of Pavement Structures”, American Association of State Highway and Transportation Officials (AASHTO), Washington, D.C., 1993.
2. AASHTO TP 31, “Standard Test Method for Determining the Resilient Modulus of Bituminous Mixtures by Indirect Tension”, American Association of State Highway and Transportation Officials (AASHTO), Washington, D.C., 1996.
3. Aglan, H.A., “Polymer Modifiers for Improved Performance of Asphalt” In Asphalt Science and Technology, Usmani, Marcel Dekker, 1997, pp. 197-216.
4. American Society for Testing and Materials. 1994 Annual Book of ASTM Standards, Vol. 04.04, Rubber, Natural and Synthetic-General Test Methods, 1994.
5. Pradhan, M. M. and J. D. Armijo, “Assessing Effects of Commercial Modifiers on Montana Asphalts by Conventional Testing Methods” Transportation Research Record, No. 1417, 1993, pp. 84-92.
6. Asphalt Institute and the Heritage Group, “The Bailey Method: Achieving Volumetrics and HMA Compactability” Asphalt Institute Educational Course and Seminar Materials, Lexington, KY, 2005.
7. Asphalt Institute, “Superpave Level I Mix Design” Asphalt Institute, Lexington, KY, Report No. SP-2, 1995.
8. Asphalt Institute, “Superpave Mix Design” Superpave Series No. 2 (SP-02), Asphalt Institute, Lexington, KY, 2001.
9. ASTM D 3497, “Standard Test Method for Dynamic Modulus of Asphalt Mixtures”, ASTM, Philadelphia, PA., 1997.
10. ASTM D 4123, “Standard Test Method for Indirect Tension Test for Resilient Modulus of Bituminous Mixture”, ASTM International, West Conshohocken, PA, 1982.
11. Bahia, H.U. and D.A. Anderson, “Strategic Highway Research Program Binder Rheological Parameters: Background and Comparison with Conventional Properties” Transportation Research Record, No. 1488, 1995, pp. 32-39.
12. Barksdale, R. D., J. Alba, N. P. Khosla, R. Kim, P. C. Lambe, and M. S. Rahman, «Laboratory Determination of Resilient Modulus for Flexible Pavement Design», NCHRP Program 1-28 Project, TRB, Washington, D.C., June 1997.
175
13. Birgisson, B. and B. Ruth, “Development of Tentative Guidelines for the Selection of Aggregate Gradations for Hot-Mix Asphalt” American Society for Testing and Materials, STP 1412, 2001.
14. Birgisson, B., R. Roque and G. Page, “Evaluation of Water Damage Using Hot Mix Asphalt Fracture Mechanics” Journal of the Association of Asphalt Paving Technologists, 2003, Vol. 72, pp. 424-462.
15. Birgisson, B., R. Roque, J. Kim, L.V. Pham, «The Use of Complex Modulus to Characterize the Performance of Asphalt Mixtures and Pavements in Florida», Final Report, BC-354-22, The Florida Department of Transportation, Tallahassee, Florida, 2004.
16. Birgisson, B., R. Roque, G.C. Page, “Performance-Based Fracture Criterion for Evaluation of Moisture Susceptibility in Hot-Mix Asphalt” Transportation Research Record: Journal of the Transportation Research Board, No. 1891, TRB, National Research Council, Washington, D.C., 2004, pp. 55-61.
17. Brule, B., Y. Brion and A. Tanguy, “Paving Asphalt Polymer Blends: Relationships between Composition, Sturcture and Properties” Proceedings, the Association of Asphalt Paving Technologists, Vol. 57, 1988, pp. 41-64.
18. Buttlar, W. G. and R. Roque, “Development and Evaluation of the Strategic Highway Research Program Measurement and Analysis System for Indirect Tensile Testing at Low Temperatures”, Transportation Research Record: Journal of the Transportation Research Board, No. 1454, Washington, D.C., 1994, pp. 163-171.
19. Button, J.W., “Summary of Asphalt Additive Performance at Selected Sites” Transportation Research Record, No. 1342, 1992, pp. 67-75.
20. Carpenter, S.H. and S. Shen, “Dissipated Energy Approach to Study Hot-Mix Asphalt Healing in Fatigue” Transportation Research Record, 2006, pp.178-185.
21. Carpenter, SH. and T. VanDam, “Laboratory Performance Comparisons of Polymer-Modified and Unmodified Asphalt Concrete Mixtures” Transportation Research Record 1115, TRB. Washington, DC: National Research Council, 1987, pp. 62–74.
22. Chen, J. and C. Huang “Fundamental Characterization of SBS-Modified Asphalt Mixed with Sulfur” Journal of Applied Polymer Science, Vol. 103, pp. 2817–2825, 2007.
23. Chen, J., M. Liao and C. Lin, “Determination of Polymer Content in Modified Bitumen” Materials and Structures, Vol. 36, November 2003, pp. 594-598.
176
24. Chen, J., M. Liao and M. Shiah, “Asphalt Modified by Styrene-Butadiene-Styrene Triblock Copolymer: Morphology and Model” Journal of Materials in Civil Engineering, May/June, pp. 224-229, 2002.
25. Chowdhury, A., J. Grau, J. Button and D. Little, “Effect of Aggregate Gradation on Permanent Deformation of Superpave HMA” the 80th Annual Meeting of Transportation Research Board, Washington, D.C., 2001.
26. Collins, J.H., M.G. Bouldin, R. Gellens and A. Berker, “Improved Performance of Paving Asphalts by Polymer Modifications” Proceedings, Association of Asphalt Paving Technologists, Vol. 60, 1991, pp. 43-79.
27. Di Benedetto, H., F. Olard, C. Sauzeat and B. Delaporte B., “Linear Viscoelastic Behavior of Bituminous Materials: From Binders to Mixes”, International Journal of Road Materials and Pavement Design, Vol. 5, Special Issue, 2004, pp. 163-202.
28. Di Benedetto, H., M. N. Partl, L. Francken and De La Roche St Andre, “Stiffness Testing for Bituminous Mixtures”, Journal of Materials and Structures, Vol. 34, 2001, pp. 66-70.
29. Diani, E., M.D. Via, and M.G. Cavaliere, “Chemistry and Technology of SBS Modifier” Asphalt Science and Technology, Marcel Dekker Inc., New York, pp. 279-295, 1997.
30. Drescher, A., D.E. Newcomb, and W. Zhang, “Interpretation of Indirect Tension Test Based on Viscoelasticity”, Transportation Research Record, No. 1590, 1997, pp. 45-52.
31. Ekingen, E. R., “Determining Gradation and Creep Effects in Mixtures Using the Complex Modulus Test” Master’s Thesis, Department of Civil Engineering, University of Florida, 2004.
32. Federal Highway Administration, “Background of Superpave Asphalt Mixture Design and Analysis” FHWA Report Number FHWA-SA-95-003, 1995.
33. Federal Highway Administration, “Superpave Asphalt Mixture Design: Illustrated Level I Lab Methods”, FHWA Report Number FHWA-SA-95-004, 1995.
34. Fonseca, O. A. and M. W. Witczak, “A Prediction Methodology for the Dynamic Modulus of In-Place Aged Asphalt Mixtures”, Journal of the Association of Asphalt Paving Technologists, Vol. 65, 1996, pp. 532-559.
35. Foster, C.R., “Development of Marshall Procedures for Designing Asphalt Paving Mixtures” National Asphalt Pavement Association, NAPA IS84, 1982.
36. Fuller, W.B. and S. Thompson, “The Laws of Proportioning Concrete” Journal of Transportation Division, ASCE, Vol. 59, 1907, pp. 67-143.
177
37. Goode, J.F., and L.A. Lufsey, “A New Graphical Chart for Evaluating Aggregate Gradations” Proceedings, AAPT, Volume 31, 1962, pp. 176-207.
38. Hines, M.L., “Asphalt Cement Performance Improved by Styrelf-Laboratory and Field Data” Koch Materials Company, 1993.
39. Hondoros, G., “Evaluation of Poisson’s Ratio and the Modulus of Materials of a Low Tensile Resistance by the Brazilian (Indirect Tensile) Test with Particular Reference to Concrete”, Australian Journal of Applied Science, Vol. 10, No. 3, 1959, pp. 243-268.
40. Huang, Y., «Pavement Analysis and Design». Appendix A: Theory of Viscoelasticity. Prentice Hall, Englewood Cliffs New Jersey 07632, 1993.
41. Huber G.A. and T. Shuler, “Providing Sufficient Void Space for Asphalt Cement: Relationship of Mineral Aggregate Void and Aggregate Gradation”, Effects of Aggregates and Mineral Fillers on Asphalt Mixtures performance” ASTM SPT 1147, Philadelphia, 1992.
42. Huffman, J.E., “The Use of Ground Vulcanized Rubber in Asphalt, Asphalt Pavement Construction: New Materials and Techniques” American Society for Testing and Materials, Special Technical Publication 724, 1980, pp. 3-12.
43. Hveem, F.N., “Gradation of Mineral Aggregate for Dense-Graded Bituminous Mixtures” Proceedings, AAPT, Vol. 11, 1940, pp. 315-339.
44. Jones, S.D., K. Mahboub, R. Anderson and H. Bahia, “Applicability of Superpave to Modified Asphalts – A Mixture Study” Transportation Research Record 1630, TRB, National Research Council, Washington, D.C., 1998, pp. 42-50.
45. Kandhal, P. and R. Mallick, “Effect of Mix Gradation on Rutting Potential of Dense-Graded Asphalt Mixtures” Transportation Research Record, No. 1767, 2001, pp. 146-151.
46. Kennedy, T. W., “Characterization of Asphalt Pavement Materials Using the Indirect Tensile Test”, Journal of the Association of Asphalt Paving Technologists, Vol. 46, 1977, pp. 132-150.
47. Kennedy, T., H. Torshizi and D. Jones IV, “Mix Design Procedures and Considerations for Polymer-Modified Asphalt Compatibility and Stability” Center for Transportation Research, Research Report 492-1F, Project 3-9-87/1-492, 1992.
48. Khattak, M.J. and G.Y. Baladi, “Engineering Properties of Polymer-Modified Asphalt Mixtures” Transportation Research Record, No. 1638, 1998, pp. 12-22.
178
49. Khattak, M.J. and G.Y. Baladi, “Fatigue and Permanent Deformation Models for Polymer-Modified Asphalt Mixtures” Transportation Research Record, No. 1767, 2001, pp. 135-145.
50. Kim, B., R. Roque, and B. Birgisson, “Laboratory Evaluation of the Effect of SBS Modifier on Cracking Resistance of Asphalt Mixtures,” Journal of the Transportation Research Board, TRR 1829, Washington, D.C., 2003, pp. 8-15.
51. Kim, J., “Accurate Determination of Dissipated Creep Strain Energy and Its Effect on Load - and Temperature - Induced Cracking of Asphalt Pavement” PhD Dissertation, University of Florida, 2005.
52. Kim, Y. R., Y. Seo, M. King, and M. Momen, “Dynamic Modulus Testing of Asphalt Concrete in Indirect Tension Mode”, Transportation Research Record, No. 1891, 2004, pp. 163-173.
53. King, G.N., H. Muncy, and J. Prudhomme, “Polymer Modification: Binder’s Effect on Mix Properties” Proceedings, AAPT, Vol 55, 1986, pp. 519-540.
54. King, G.N., H.W. King, O. Harders, W. Arand and P. Planche, “Influence of Asphalt Grade and Polymer Concentration on the Low Temperature Performance of Polymer Modified Asphalt” Proceedings, AAPT, Vol. 62, 1993, pp. 1-22.
55. King, M., M. Momen, and Y. R. Kim, “Typical Dynamic Moduli Values of Hot Mix Asphalt in North Carolina and Their Prediction”, Transportation Research Board, 2005 84th Annual Meeting Compendium of Papers CD-ROM, Washington, D.C.
56. Lacroix, A., A. M. Khandan, and Y. R. Kim, “Predicting the Resilient Modulus of Asphalt Concrete from the Dynamic Modulus”, Transportation Research Record: Journal of the Transportation Research Board, No. 2001, 2007, pp. 132-140.
57. Lalwani, S., A. Abushihada and A. Halasa, “Reclaimed Rubber-Asphalt Blends Measurement of Rheological Properties to Access Toughness, Resiliency, Consistency and Temperature Sensitivity” Proceedings, AAPT, Vol. 51, 1982, pp. 562-579.
58. Loulizi, A., G. W. Flintsch, I. L. Al-Qadi and D. Mokarem, “Comparing Resilient Modulus and Dynamic Modulus of Hot-Mix Asphalt as Material Properties for Flexible Pavement Design”, Transportation Research Record: Journal of the Transportation Research Board, No. 1970, Washington, D.C., 2006, pp. 161-170.
59. Lu. X, U. Isacsson and J. Ekblad, “Low-temperature properties of styrene–butadiene–styrene polymer modified bitumens” Construction and Building Materials Vol.12, Issue 8, 1998, pp. 405–414.
179
60. McGennis, R.B., S. Shuler, and H.U. Bahia, “Background of Superpave Asphalt Binder Test Methods” FHWA, Report No. FHWA-SA-94-069, July 1994.
61. National Cooperative Highway Research Program, Project 1-37A-2002 Mechanistic-Empirical (M-E) Pavement Design Guide (Draft), NCHRP, Washington, D.C., 2004.
62. National Cooperative Highway Research Program, Project 9-29 Simple Performance Tester for Superpave Mix Design, “First Article Development and Evaluation”, NCHRP Report 513, Washington, D.C., 2003.
63. Nijboer L.W., “Plasticity as a Factor in the Design of Dense Bituminous Road Carpets” Elsevier Publishing Company Inc., New York, NY, 1948.
64. Papazian, H. S., “The Response of Linear Viscoelastic Materials in the Frequency Domain with Emphasis on Asphaltic Concrete”, (1st) International Conference on the Structural Design of Asphalt Pavements, 1962, University of Michigan, pp. 454-463.
65. Pellinen, T. K. and M. W. Witczak, “Stress Dependent Master Curve Construction for Dynamic (Complex) Modulus”, Journal of the Association of Asphalt Paving Technologists, Vol. 71, 2002, pp. 281-309.
66. Pellinen, T. K., “Investigation of the Use of Dynamic Modulus as an Indicator of Hot-Mix Asphalt Performance”, Ph.D. Dissertation, Arizona State University, 2001.
67. Ping, W. V. and Y. Xiao, «Evaluation of the Dynamic Complex Modulus Test and Indirect Diametral Test for Implementing the AASHTO 2002 Design Guide for Pavement Structures in Florida », Final Report, The Florida Department of Transportation, BC-352-12, Tallahassee, FL, 2007.
68. Ping, W.V. and Y. Xiao, “Empirical Correlation of Indirect Tension Resilient Modulus and Complex Modulus Test Results for Asphalt Concrete Mixtures” International Journal of Road Materials and Pavement Design, Volume 9, 2008, pp. 177-200.
69. Quintus, H., J. Mallela and M. Buncher, “Quantification of Effect of Polymer-Modified Asphalt on Flexible Pavement Performance” Transportation Research Record: Journal of the Transportation Research Board, No. 2001, 2007, pp. 141-154.
70. Roberts, R.L., P.S. Kandhal, E.R. Brown, D. Lee, and T.W. Kennedy, «Hot Mix Asphalt Materials, Mixture Design, and Construction» National Center for Asphalt Technology (NCAT), National Asphalt Pavement Association (NAPA), Research and Education Foundation, Lanham, Maryland, 1996.
71. Roque, R. and W. G. Buttlar, “The Development of a Measurement and Analysis System to Accurately Determine Asphalt Concrete Properties Using the Indirect Tensile
180
Mode”, Journal of the Association of Asphalt Paving Technologists, Vol. 61, 1992, pp. 304-332.
72. Roque, R., B. Birgisson and S. Kim, “Development of Mix Design Guidelines for Improved Performance of Asphalt Mixtures” Florida Department of Transportation, Final Report, BD545, RPWO#16, 2006.
73. Roque, R., B. Birgisson, Z. Zhang, B. Sangpetngam, and T. Grant, “Implementation of SHRP Indirect Tension Tester to Mitigate Cracking in Asphalt Pavements and Overlays” Final Report, Florida Department of Transportation, BA-546, 2002.
74. Roque, R., B. Birgisson, C. Drakos, and B. Dietrich “Development and Field Evaluation of Energy-Based Criteria for Top-down Cracking Performance of Hot Mix Asphalt” Journal of the Association of Asphalt Paving Technologists, Vol. 73, 2004, pp. 229-260.
75. Roque, R., B. Birgisson, B. Sangpetgnam, Z. Zhang, “Hot Mix Asphalt Fracture Mechanics: A Fundamental Crack Growth Law for Asphalt Mixtures” Journal of the Association of Asphalt Paving Technologists, Vol. 71, 2002, pp. 816-828.
76. Ruth B.E., R. Roque, and B. Nukunya, “Aggregate Gradation Characterization Factors and Their Relationships to Fracture Energy and Failure Strain of Asphalt Mixtures” Journal of the Association of Asphalt Paving Technologists, Vol. 71, 2002, pp. 310-344.
77. Schapery, R.A., “Viscoelstic Behavior and Analysis of Composite Materials” Report MM 72-3, Mechanics and Materials Research Center, Texas A&M University, College Station, Texas, 1972.
78. Scofield, L.A., “The History, Development, and Performance of Asphalt Rubber at ADOT” Special Report, Report Number AZ-SP-8902, Arizona Department of Transportation, Materials Section, Phoenix, AZ, December 1989.
79. Shen, S. and S.H. Carpenter, “Application of the Dissipated Energy Concept in Fatigue Endurance Limit Testing” Transportation Research Record, 2005, pp. 165-173.
80. Shih, C., “Evaluation of Asphalt Additives for Improved Cracking and Rutting Resistance of Asphalt Paving Mixtures” PhD Dissertation, University of Florida, 1996.
81. SHRP-LTPP Protocol P07, “Test Method for Determining the Creep Compliance, Resilient Modulus, and Strength of Asphalt Materials Using the Indirect Tensile Test Device”, LAW PCS, SHRP, National Research Council, Washington, D.C., August 2001.
82. Terrel, R.L. and J.L. Walter, “Modified Asphalt Pavement Materials - The European Experience” Proceedings, AAPT, 1986, Vol. 55, pp. 482-518.
181
83. The Asphalt Institute, “Mix Design Methods for Asphalt Concrete and Other Hot Mix Types” MS-2, May 1984.
84. The Asphalt Institute, “Mix Design Methods for Asphalt Concrete and Other Hot Mix Types” Manual Series No. 2 (MS-2). (6th ed.), Lexington, KY, 1993.
85. The Asphalt Institute, “Performance Graded Asphalts Binder Specification and Testing” Superpave Series No.1 (SP-1), Lexington, KY, 1994.
86. Transportation Research Board, “Superior Performing Asphalt Pavements (Superpave): The Product of the SHRP Asphalt Research Program”, TRB Report Number SHRP-A-410, 1994.
87. Usmani, A.M., “Polymer Science and Technology” Asphalt Science and Technology, Marcel Dekker Inc., New York, pp. 307-336, 1997.
88. Vallerga, B.A., and W.R. Lovering, “Evolution of the Hveem Stabilometer Method of Designing Asphalt Paving Mixtures” Proceedings, Association of Asphalt Paving Technologists, Vol. 54, 1985, pp. 243 - 265.
89. Vavrik W.R., W.J. Pine, G. Huber, and S.H. Carpenter, “The Bailey Method of Gradation Evaluation: The Influence of Aggregate Gradation and Packing Characteristics on Voids in the Mineral Aggregate” Journal of the Association of Asphalt Paving Technologists, 2001, Vol. 70, pp. 132-175.
90. Verhaeghe, B.M.J.A., F.C. Rust, R.M. Vos and A.T. Visser, “Properties of Polymer- and Fiber-Modified Porous Asphalt Mixes” 6th Conference on Asphalt Pavements for Southern Africa. Proceedings, Vol. 1, Cape Town, South Africa, October 1994, pp. III-262-280.
91. Villiers C., “Sensitivity of Superpave Mixtures for Development of Performance-Related Specifications” PhD Dissertation, University of Florida, 2004.
92. Warren, R.S., R.B. McGennis, And H.U. Bahia, “Superpave Asphalt Binder Test Methods — An Illustrated Overview” FHWA, Report No. FHWA-SA-94-068, July 1994.
93. Wegon, V. and B. Brule, “The Structure of Polymer Modified Binders and Corresponding Asphalt Mixtures” Proceedings, AAPT, Vol. 68, 1999, pp. 64-88.
94. White, T.D., “Marshall Procedures for Design and Quality Control of Asphalt Mixtures” Proceedings, AAPT, Volume 54, 1985, pp. 265-284.
182
95. Witczak, M. W. and O. A. Fonseca, “Revised Prediction Model for Dynamic (Complex) Modulus of Asphalt Mixtures”, Transportation Research Record: Journal of the Transportation Research Board, No. 1540, Washington, D.C., 1996, pp. 15-23.
96. Witczak, M. W., I. Hafez, H. Ayres, & K. Kaloush, “Comparative Study of MSHA Asphalt Mixtures Using Advanced Dynamic Material Characterization Tests” Maryland Department of Transportation, Maryland Port 532 Administration. Department of Civil Engineering, University of Maryland of College Park, MD, 1996.
97. Witczak, M. W., “A comparison of the Dynamic (Complex) Modulus Test (E*) and Indirect Diametral Test (MR) for AC Mixtures”, 1999.
98. Witczak, M. W., R. Bonaquist, H. Von Quintus, and K. Kaloush, “Specimen Geometry and Aggregate Size Effects in Uniaxial Compression and Constant Height Shear Test”, Journal of the Association of Asphalt Paving Technologists, Vol. 69, 2000, pp. 733-782.
99. Witczak, M. W., K. Kaloush, T. Pellinen and M. El-Basyouny, «Simple Performance Test for Superpave Mix Design», NCHRP Report 465, Washington, D.C., 2002.
100. Witczak, M. W., T. K. Pellinen and M. M. El-Basyouny, “Pursuit of the Simple Performance Test for Asphalt Concrete Fracture/Cracking”, Journal of the Association of Asphalt Paving Technologists, Vol. 71, 2002, pp. 767-778.
101. Witczak, M. W., “Laboratory Determination of Resilient Modulus for Flexible Pavement”, NCHRP Research Results Digest, No. 285. TRB, Washington, D.C., January 2004.
102. Yoder, E.J. and M.W. Witczak, “Principles of Pavement Design” Wiley, New York, 1975.
103. Zhang, W., A. Drescher, and D. E. Newcomb, “Viscoelastic Analysis of Diametral Compression on Asphalt Concrete”, Journal of Engineering Mechanics, Vol. 123, ASCE, 1997, pp. 596-603.
104. Zhang, Z., “Identification of Suitable Crack Growth Law for Asphalt Mixtures Using the Superpave Indirect Tensile Test (IDT)” Ph.D. Dissertation, University of Florida, Gainesville, FL, 2000.
105. Zhang, Z., R. Roque, and B. Birgisson, “Identification and Verification of a Suitable Crack Growth Law for Asphalt Mixtures” Journal of the Association of Asphalt Paving Technologists, Vol. 70, 2001, pp. 206-241.
183
BIOGRAPHICAL SKETCH
Yuan Xiao was born and raised in Hubei, China. He began his college education at Shanghai
JiaoTong University in China and received a Bachelor of Science degree in Mechanical
Engineering in 1999. He then became a graduate student at the Institute of Refrigeration and
Cryogenics in Shanghai JiaoTong University and received a Master’s degree in spring, 2002.
After his graduation from SJTU, Yuan Xiao enrolled in the Department of Mechanical
Engineering at Florida State University in fall, 2002. He received a Master of Science degree in
2004 and then joined the transportation materials group at the Department of Civil Engineering
in Florida State University. He worked as a graduate research assistant with his doctoral advisor,
Dr. Wei-Chou Virgil Ping, from January 2005.
Yuan Xiao was involved in research projects related to the experimental testing and
analytical modeling of flexible pavement materials and design. He plans to complete the Doctor
of Philosophy Degree in Civil Engineering at Florida State University in spring 2009.