Fhwa 1995 PDF

Publication No.
FHWA-SA-95-003
U.SDepartment February 1995

of Transportation
Federal Highway
Administration
BACKGROUND of SUPERPAVE
ASPHALT MIXTURE DESIGN
AND ANALYSIS
NATIONAL ASPHALT TRAINING CENTER

DEMONSTRATION PROJECT101
Innovation Through Partnerships

NOTICE
This document is disseminated under the sponsorship of the Department of Transportation in

the interest of information exchange. This publication does not constitute a standard,
specification, or regulation.
The United States Government does not endorse products or manufacturers. Trademarks or
manufacturers' names appear herein only because they are considered essential to the object
of this document.
Technical Report Documentation Pag
1, Report No. 1 2.Government Accession No. 3. Recipient's Catalog No.
I I
4. Title and Subtitle ( 5. Report Date

November 1994
Background of SUPERPAVE 6. Performing organization Code
Asphalt Mixture Design & Analysis
8. Performing Organization Report No.
7. Author(s)
R.B. McGennis, R.M. Anderson, T.W. Kennedy, M. Solaimanian
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Asphalt Institute
P.O. BOX14052 I 11,Contract or Grant No.
-
Lexinaton.- KY 40512-4052
i 13. T v ~ of
8
e Re~ortand Period Covered
12. Sponsoring Agency Name and Address Final Report
Federal Highway Administration
December 1992-November
Office of Technology Applications 'tf&i
400 Seventh St., SW 14. Sponsoring Agency Code
Washington, DC 20590 HTA-21
15. Supplementary Notes
FHWA Technical Representative - J.R. Bukowski
16. Abstract
This manual represents the first formal training document that embodies the complete series of SUPERPAVE
asphalt mixture design and analysis test equipment and procedures. These tests and procedures represent the
results df the SHRP 5-year research effort to investigate and improve asphalt cement technology. This manual
was developed under the FHWA's National Asphalt Training Center. Students attending the center utilize this
manual to obtain a better understanding of the underlying theory behind asphalt mixture design and analysis, as
well as how to perform each of the new procedures.
17. Key Words 18. Dlsfribution Statement

4sphatt mix, asphalt cement, gyratory compactor, indirect No restrictions. This document is available to the
ensile tester, level 1 design, mixture design, shear tester, public from the National Technical Information
SHRP, SUPERPAVE Service, Springfield, Virginia 22 161
I
9. Security Classif. (of this report) 1 20. Security Classif. (of this page) 1
121. No. of Pages 22. Price
Unclassified I
I
Unclassified
I I
corm DOT F 1700.7 (8-72) Reproduction of completed page authorized
FOREWORD
From October 1987 through March 1993, the Strategic Highway Research
Program (SWRP) conducted a $50 million research effort to develop new ways to specify,
test, and design asphalt materials. Near the end of SHRP, the Federal Highway
Administration assumed a leadership role in the implementation of SHRP research. An
essential part of FHWA' s implementation strategy was development of a nationally
accessible training center aimed at educating agency and industry personnel in the proper
use and application of the final SHRP asphalt products, collectively referred to as
SuperpaveTM.This project was administered by the FHWA's Office of Technology
Applications and designated Demonstration Project 101, the National Asphalt Training
Center (NATC).
The NATC resides at the Asphalt Institute's Research Center in Lexington,

Kentucky. While the day-to-day affairs of the NATC are directed by Institute personnel,
course development and technical direction were duties shared by a team of engineers
from the Asphalt Institute, the Pennsylvania State University, the University of Texas at
Austin, National Center for Asphalt Technology, Marathon Oil Company, and FHWA.
The objective of the educational program is to train students in the practical

applications of SHRP asphalt products. It is composed of two parts: Superpave asphalt
binder technology and Superpave asphalt mixture design and analysis,
This manual represents the textbook students use as a reference throughout the 40
hours of training in Superpave mixture design and analysis. Best efforts were made to
present the information in an easy to understand style. It was written for laboratory
technicians and engineers with no previous training in Superpave, but with some
knowledge in asphalt materials and mixture design. Other instructional aids consist of
provisional AASHTO test methods (when available) and a separate illustrated overview
document pertaining to Superpave gyratory compaction.
The training program consists of 40 hours of instruction. Of this 40 hours,

students receive 12 hours of classroom instruction, 16 hours of laboratory instruction, and
12 hours of group discussion of actual test results. By the end of the course, students will
be familiar with Superpave asphalt mixture test procedures and equipment. This course
emphasizes (but is not limited to) Superpave Level 1 design and analysis.
The training program and this manual do not present any information in English
units. Superpave test procedures were largely developed in SI or metric units. The
NATC team believed it would be counter productive and make learning more difficult if
material properties were shown in US.customary, as well as the original SI and metric
units. For example, it is easy for a student to understand and remember that the gyratory
compaction pressure is 600 kPa. To show an English conversion such as, "600 kPa (86
psi)," serves no purpose since students have no previous knowledge of typical U S .
customary units for this test parameter. The only exception to this is that some
Foreword
performance based testing software was developed (and remains) in U.S. customary units.
The NATC team has no control over these products but encourages the software
developers to assist the industry and this training effort by standardizing the units, in SI,
on test output.
Users of this manual will note that no references are cited throughout the text.
That is because as this manual was being prepared in late 1993, very few, if any, SHRP
research reports had been published. The authors were able to glean important
information from draft reports and verbally from researchers involved in the numerous
areas of the SHRP asphalt research program. The authors are indebted to the many
individuals who graciously shared their knowledge during the early phases of the NATC.
Users are strongly encouraged to obtain and study the reports cited in the bibliography for
the most complete information pertaining to Superpave.
As this edition was being prepared, Superpave was still in an emerging phase.
Many of the AASHTO test procedures were (and still are) under development. In
addition, Superpave testing equipment is only now becoming available. Consequently,
some of the information herein contained may be subject to change. Users of this manual
are resolutely encouraged to stay abreast of Superpave technology through the many
venues that have become available as a result of SHRP. The National Asphalt Training
Center and asphalt user-producer groups are two examples of forums that specifically
address Superpave technology.
Mr. John R. Bukowski of FHWA's Office of Technology Applications is the Contracting

Officer's Technical Representative for NATC. This manual was co-authored by:
Mr. Robert B, McGennis

Director of Research & Engineering Services
Asphalt Institute
Mr. R. Michael Anderson

Asphalt Systems Engineer-
Asphalt Institute
Dr. Thomas W. Kennedy

Engineering Foundations Professor of Civil Engineering
The University of Texas at Austin
Dr. Mansour Solaimanian

Research Engineer
The University of Texas at Austin
November 1994
Lexington, KY
SuperpaveTM is a registered trademark of the Strategic Highway Research Program

TABLE OF CONTENTS
I . HOW ASPHALT MIXTURES BEHAVE .............................................................................................1
ASPHALT BINDER BEHAVIOR .......................................................................................... 1
MINERAL AGGREGATE BEHAVIOR ................................................................................. 3
ASPHALT MIXTURE BEHAVIOR ........................................................................................... 7
Permanent Deformation ............................................................................................... 7
Fatigue Cracking............................................................................................................10
Low Temperature Cracking ....................................................................................... 13
CURRENT WAYS TO SPECIFY ASPHALT CEMENTS......................................................... 14
CURRENT ASPHALT MIXTURE DESIGN PROCEDURES...................................................16
11. SUPERPAVE TO THE RESCUE ...................................................................................................... 21

INTRODUCTION .................................................................................................................... 2 1
ASPHALT BINDERS .............................................................................................................. 21
MINERAL AGGREGATES .................................................................................................... 26
ASPHALT MIXTURES .......................................................................................................... 28
PUTTING IT ALL TOGETHER................................................................................................. 30
.
111 MATERIALS SELECTION ...................................................................................................... 3 3
INTRODUCTION ................................................................................................................. 3 3
ASPHALT BINDERS ................................................................................................................. 33
SUPERPAVE Weather Database .............................................................................. 34
Reliability ......................................................................................................................
34
Start with Air Temperature ........................................................................................ 35
Convert to Pavement Temperature ................................................................................ 36
Select Binder Grade....................................................................................................... 37
Effect of Loading Rate on Binder Selection ............................ . . ...............................38
Effect of Traffic Level on Binder Selection ..................................................................39
MINERAL AGGREGATE ...................................................................................................... 39
Consensus Properties ..................................................................................................... 40
Coarse Aggregate Angularity .......................................................................... 40
Fine Aggregate Angularity .......................................................................... 41
Flat, Elongated Particles ........... . ...................................................................42
Clay Content.................................................................................................. 44
Source Properties......................................................................................................... 45
Toughness ....................................................................................................... 45
Soundness...................................................................................................... 46
Deleterious Materials .....................................................................................-46
Gradation ..................................................................................................................... 46
ASPHALT MIXTURES .............................................................................................................. 50
Mixture Volumetric Requirements ................................................................................50
Dust Proportion ......................................................................................................... 51
Moisture Susceptibility ............................. . . .............................................................52
IV . ASPHALT MIXTURE VOLUMETRICS ........................................................................................... 53
INTRODUCTION ......................................................................................................................
53
COMPONENT DIAGRAM APPROACH ................................................................................... 53
SPECIFIC GRAVITY .............................................................................................................. 55
EXAMPLE CALCULATIONS ................................................................................................... 59
.
V SUPERPAVE GYRATORY COMPACTION...................................................................................... 65
INTRODUCTION .......................................................................................................................
65
TEST EQUIPMENT....................................................................................................................
65
iii
Table of Contents
SPECIMEN PREPARATION ....................................................................................................

68
OVERVIEW OF PROCEDURE .................................................................................................
-69
DATA ANALYSIS AND PRESENTATION............................................................................ 70
CALIBRATION AND STANDARDIZATION.......................................................................... 74
VI . VOLUMETRIC MIX DESIGN ...................................................................................................... 75

INTRODUCTION ................................................................................................................... 75
MATERIAL SELECTION ....................................................................................................... -76
Binder Selection............................................................................................................. 76
Aggregate Selection ..................................................................................................... 78
Coarse Aggregate Angularity ......................................................................... 80
Fine Aggregate Angularity .......................................................................... 80
Flat and Elongated Particles .......................................................................... 81
Clay Content (Sand Equivalent) .....................................................................82
SELECT DESIGN AGGREGATE STRUCTURE ........................ . .......................................-82
SELECT DESIGN ASPHALT BINDER CONTENT ................................................................. 103
EVALUATE MOISTURE SENSITIVITY.......... . ............. . ...............................................112
VII. PERFORMANCE TESTING (LEVELS 2 AND 3) ............................ . . . . .. . ..

... 115
INTRODUCTION ........................ . . .......................................................................................115
PERFORMANCE MODELS .................................................................................................. 1 1 6
TEST PARAMETERS ................ . .. .
.
................................................................................1 1 7
Test Temperatures .......................... .
.. ......................................................................117
.
Asphalt Binder Contents ...............................
. ..................................................... 118
SUPERPAVE SHEAR TESTER ......................... ..
.. ........................................................... 119
Testing Apparatus ................................................................................................... ,119
Test Control Unit .........................................................................................................120
Environmental Control Unit ......................... . . ........................................................ 120
Hydraulic System.................................. .
..
................................................................120
Specimen Preparation and Instrumentation.................................................................. 121
Test Procedures.... 123
Volumetric Test .............................................................................................. 123
Uniaxial Strain Test .................................................................................... 124
Repeated Shear Test at Constant Stress Ratio ............................... . . . . .........125
Repeated Shear Test at Constant Height ........................ . . . .....................127
Simple Shear Test at Constant Height ........................................................... 128
Frequency Sweep Test at Constant Height ......................................................129
INDIRECT TENSILE TESTER ..................... . ............. . . ...............................................1 3 0
Testing Apparatus .......................................................................................................... 131
Control Unit and Data Acquisition System ................................................................ 131
Load Measuring Device ............................................................................................... 132
Environmental Control Chamber ............................. . . ..............................................132
Specimen Preparation and Instrumentation .................................................................. 132
Test Procedures........... . . ........................................................................................... 133
IDT Creep Compliance and Strength (Low Temperature Cracking
Analysis) ........................................................................................................ 133
IDT Strength (Fatigue Cracking Analysis) ......................................................134
DATA ANALYSIS AND INTERPRETATION..................... . . ..........................................-135
APPENDIX A: SUPERPAVE ASPHALT BINDER SPECIFICATION..............
..............................139
APPENDIX B: SUPERPAVE ASPHALT MIXTURE GRADATION REQUIREMENTS.....................141

37.5 MM NOMINAL SIZE .........................................................................................................
141
.
25 MM NOMINAL SIZE..................................................
. . 142
19 MM NOMINAL SIZE ............................................. . .. ........................................... 1 4 3
Table of Contents
12.5 MM NOMINAL SIZE ....................................................................................................... 144

9.5 MM NOMINAL SIZE ....................................................................................................... 145
APPENDIX C: SUPERPAVE CONSENSUS AGGREGATE REQUIREMENTS.................................. 147

COARSE AGGREGATE ANGULARITY ............ . .............................................................147
FINE AGGREGATE ANGULARITY ...................................................................................... 147
FLAT AND ELONGATED PARTICLES ............................................................................... 148
CLAY CONTENT ..................................................................................................................... 148
APPENDIX D: VOLUMETRIC ANALYSIS OF COMPACTED HOT MIX ASPHALT (HMA) .........149

DESCRIPTION OF TERMS ..................................................................................................... 149
STANDARD CONVENTIONS ................................................................................................ 150
CALCULATIONS ..................................................................................................................... 151
APPENDIX E: OUTLINE OF STEPS IN SUPERPAVE LEVEL 1 MIX DESIGN .............................. 153

I . SELECTION OF MATERIALS................ . . ...................................................................153
I1. SELECTION OF DESIGN AGGREGATE STRUCTURE ..................................................153
I11. SELECTION OF DESIGN ASPHALT BINDER CONTENT ..................... . ..................154
IV . EVALUATION OF MOISTURE SENSITIVITY OF DESIGN ASPHALT
MIXTURE USING AASHTO T283 ..................................................................................... 154
APPENDIX F: TESTING REQUIREMENTS FOR SUPERPAVE ......................................................... 155

SUPERPAVE LEVEL 1.............................................................................................................. 155
SUPERPAVE LEVEL 2 ..............................................................................................................156
SUPERPAVE LEVEL 3 ..............................................................................................................157
BIBLIOGRAPHY .................................................................................................................................... 159

I.
HOW ASPHALT MIXTURES BEHAVE
Asphalt concrete (sometimes referred to as "hot mix asphalt" or simply "HMA")is a

paving material that consists of asphalt binder and mineral aggregate. The asphalt binder,
which can be asphalt cement or modified asphalt cement, acts as a binding agent to glue
aggregate particles into a cohesive mass. Because it is impervious to water, the asphalt
binder also functions to waterproof the mixture. When bound by the asphalt binder,
mineral aggregate acts as a stone framework to impart strength and toughness to the
system. Because HMA contains both asphalt binder and mineral aggregate, the behavior
of the mixture is affected by the properties of the individual components and how they
react with each other in the system.
ASPHALT BINDER BEHAVIOR
Asphalt binder alone is a very interesting and challenging construction material with
which to work. Its most important characteristic, which is both a strength and sometimes
a weakness, is its temperature susceptibility. That is, its measured properties are very
dependent on its temperature. That is why almost every asphalt cement and mixture
characterization test must be accompanied by a specified test temperature. Without
specifying a test temperature, the test result cannot be effectively interpreted. Asphalt
cement behavior is also dependent on time of loading. The same load applied for a
different duration will cause an asphalt to exhibit different properties. As with
temperature, asphalt cement tests must specify a loading rate. Because asphalt cement
behavior is dependent on temperature and duration of load, these two factors can be used
interchangeably (Figure 1-1). That is, a slow loading rate can be simulated by high
temperatures and fast loading rate can be simulated by low temperatures.
I. How Asphalt Mixtures Behave
1 hour
Figure 1-1. Asphalt Cement Time Temperature Dependency
Asphalt cement is sometimes referred to as a visco-elastic material because it

simultaneously displays both viscous and elastic characteristics (Figure 1-2). At high
temperatures, asphalt cement acts almost entirely as a viscous fluid. In other words, when
heated to a high enough temperature (e.g., > 100" C), it displays the consistency of a
lubricating fluid such as motor oil. At very low temperatures (e.g., c 0" C), asphalt
cement behaves mostly like an elastic solid. That is, it acts like a rubber band. When
loaded it stretches or compresses to a different shape. When unloaded, it easily returns to
its original shape. At intermediate temperatures, which also happen to be those in which
pavements are expected to function, asphalt cement has characteristics of both a viscous
fluid and an elastic solid.
There remains another important characteristic about asphalt cement. Because it is

composed of organic molecules, it reacts with oxygen from the environment. This
reaction is called "oxidation" and it changes the structure and composition of the asphalt
molecules. When an asphalt reacts with oxygen, a harder and more brittle structure
always results and that is the origin of the terms "oxidative hardening" or "age
hardening." Oxidation occurs more rapidly at high temperatures. That is why a
significant amount of hardening occurs during HMA production, when the asphalt cement
is necessarily heated to facilitate mixing and compaction. That is also why oxidation is
more of a concern when the asphalt cement is used in a pavement in a hot, desert climate.
I I I I
-30 25 60 135
Temperature, C
Figure 1-2. Visco-Elastic Behavior of Asphalt
Modified asphalt binders are produced to alter and improve the properties of the asphalt
to enhance the long term performance of pavements. While the modifier may affect many
properties, the majority of modifiers attempt to reduce temperature dependency and
oxidative hardening of asphalt cement and the moisture susceptibility of asphalt mixtures.
MINERAL AGGREGATE BEHAVIOR
A wide variety of mineral aggregate has been used to produce HMA. Some materials are
referred to as natural aggregate because they are simply mined from river or glacial
deposits and are used without further processing to manufacture HMA. These are often
called "bank-run" or "pit-run" materials. Processed aggregate can include natural
aggregate that has been separated into distinct size fractions, washed, crushed, or
otherwise treated to enhance certain performance characteristics of the finished HMA.
However, in most cases processed aggregate is quarried and the main processing consists
of crushing and sizing.
Synthetic aggregate consists of any material that is not mined or quarried and in many
cases represents an industrial by-product. Blast furnace slag is one example.
Occasionally, a synthetic aggregate will be produced to impart a desired performance
characteristic to the HMA. For example, light-weight expanded clay or shale is
sometimes used as a component to improve the skid resistance properties of HMA.
An existing pavement can be removed and reprocessed to produce new HMA. Reclaimed
asphalt pavement or " R A P is a growing and important source of aggregate for asphalt
pavements.
Increasingly, waste products are used as aggregate or otherwise disposed of in asphalt

pavements. Scrap tires and glass are the two most well known waste products that have
been successfully "landfilled in asphalt pavements. In some cases, waste products can
actually be used to enhance certain performance characteristics of HMA. In other cases,
it is considered sufficient that a solid waste disposal problem has been solved and no
performance enhancing benefit from the waste material is expected. However, it is hoped
that performance will not be sacrificed simply to eliminate a solid waste material.
Regardless of source, processing method, or mineralogy, aggregate is expected to provide

a strong, stone skeleton to resist repeated load applications. Cubical, rough-textured
aggregates provide more strength than rounded, smooth-textured aggregates (Figure 1-3).
Even though a cubical piece and rounded piece of aggregate may possess the same
inherent strength, cubical aggregate particles tend to lock together resulting in a stronger
mass of material. Instead of locking together, rounded aggregate particles tend to slide by
each other.
Cubical Aggregate Rounded Aggregate
Figure 1-3. Aggregate Stone Skeletons
When a mass of aggregate is loaded, there may occur within the mass a plane where
aggregate particles begin to slide by or "shear" with respect to each other (Figure I-4),
which results in permanent deformation of the mass. It is at this plane where the "shear
stress" exceeds the "shear strength" of the aggregate mass. Aggregate shear strength is of
critical importance in HMA.
shear plane
Before Load After Load
Figure 1-4. Shear Loading Behavior of Aggregate
Contrasting aggregate shear strength behavior can easily be observed in aggregate

stockpiles whereby crushed (i.e., mostly cubical) aggregates form steeper, more stable
piles than rounded aggregates. Engineers refer to the slope on stockpiles as the angle of
repose. The angle of repose of a crushed aggregate stockpile is greater than that of an
uncrushed aggregate stockpile (Figure 1-5).
angle of repose
Cubical Aggregate Rounded Aggregate
Figure 1-5. Stockpile Behavior of Cubical and Rounded Aggregate
Engineers explain the shearing behavior of aggregate (and many other) materials using
Mohr-Coulomb theory, named after the individuals who originated the concept. This
theory declares that the shear strength of an aggregate mixture is dependent on how well
the aggregate particles hold together in a mass (often called cohesion), the stress the
aggregates may be under, and the internal friction of the aggregate. The Mohr-Coulomb
equation used to express the shear strength of a material is:
I. How Aspha lt Mixtures Behave
where, z = shear strength of aggregate mixture,

c = cohesion of aggregate,
o = normal stress to which the aggregate is subjected, and
$ = angle of internal friction.
The Mohr-Coulomb shearing behavior of materials is shown in Figure 1-6.
Figure 1-6. Mohr-Coulomb Theory
A mass of aggregate has relatively little cohesion. Thus, the shear strength is primarily
dependent on the resistance to movement provided by the aggregates. In addition, when
loaded, the mass of aggregate tends to be stronger because the resulting stress tends to
hold the aggregate more tightly together. In other words, shear strength is increased. The
angle of internal friction indicates the ability of aggregate to interlock, and thus, create a
mass of aggregate that is almost as strong as the individual pieces.
A last consideration in understanding the shearing properties of aggregate is the concept

of dilatancy. When subjecting a mass of aggregate to shearing stresses, aggregate
particles must fracture or crawl up and over each other if movement is to occur. This
phenomenon is called dilation because it results in an enlargement or increased volume of
the mass of aggregate (Figure 1-7). Strong materials that are more densely packed and
have high internal friction tend to dilate more than weaker materials.
Before Shearing During Shearing
Figure 1-7. Dilation of Two Aggregate Particles When Sheared
To ensure a strong aggregate blend for HMA, engineers typically have specified
aggregate properties that enhance the internal friction portion of the overall shear
strength. Normally, this is accomplished by specifying a certain percentage of crushed
faces for the coarse portion of an aggregate blend. Because natural sands tend to be
rounded, with poor internal friction, the amount of natural sand in a blend is often
limited.
ASPHALT MIXTURE BEHAVIOR
While the individual properties of HMA components are important, asphalt mixture
behavior is best explained by considering asphalt cement and mineral aggregate acting as
a system. One way to understand asphalt mixture behavior is to consider the primary
asphalt pavement distress types that engineers try to avoid: permanent deformation,
fatigue cracking, and low temperature cracking.
Permanent Deformation
Permanent deformation is the distress that is characterized by a surface cross section that
is no longer in its proper position. It is called "permanent" deformation because it
represents an accumulation of small amounts of deformation that occur each time a load
is applied. This deformation cannot be recovered. Wheel path rutting is the most
common form of permanent deformation. While wheel path rutting can have many
causes (e.g., underlying HMA weakened by moisture damage, abrasion, traffic
densification), it has two principal causes.
In one case, the rutting is caused by too much repeated stress being applied to the native
soil (i.e., subgrade), subbase, or base below the asphalt layer (Figure 1-8). Although
stiffer paving materials will partially reduce this type of rutting, it is normally considered
more of a structural problem rather than a materials problem. It is often the result of too
thin a pavement section because there is simply not enough depth of cover on the
subgrade to reduce the stress from applied loads to a tolerable level. It may also be the
result of a subgrade that has been unexpectedly weakened by the intrusion of moisture.
The accumulated deformation occurs in the subgrade rather than in the overlying asphalt
layers.
weak subgrade or underlying layer \ subgrade

deformation
Figure 1-8. Rutting from Weak Subgrade
The other principal type of rutting (and that which is of most concern here) results from
accumulated deformation in the asphalt layers. This type of rutting is caused by an
asphalt mixture that is too low in shear strength to resist the repeated heavy loads to
which it is subjected (Figure 1-9). Sometimes the rutting occurs in a weak asphalt surface
course. In other cases, the surface course may not itself be prone to rutting, but may
simply conform to an underlying asphalt course that is too weak.
shear plane
Figure 1-9. Rutting from Weak Mixture

When an asphalt mixture ruts, it is evidence that the mixture has poor shear strength.
Each time a heavy truck applies a load, a small, but permanent, shear deformation occurs.
Shear deformation is characterized by a downward and lateral movement of the mixture.
With enough load applications a rut will appear. Rutted asphalt pavements pose a safety
hazard because the ruts will trap enough water to cause hydroplaning and ice
accumulation.
Asphalt pavement rutting from weak asphalt mixtures is a high temperature phenomenon.
That is, it most often occurs during the summer when high pavement temperatures are
evident. While this might suggest that rutting is solely an asphalt cement problem, it is
more correct to address rutting by considering the mineral aggregate and asphalt cement.
In fact, the previously described Mohr-Coulomb equation (T = c + o x tan @)can again be
used to illustrate how both materials can affect rutting.
In this case, z is considered the shear strength of the asphalt mixture. The cohesion term
(c) can be considered the portion of the overall mixture shear strength provided by the
asphalt cement. Because rutting is an accumulation of very small permanent
deformations, one way to ensure that asphalt cement provides its "fair share" of shear
strength is to use an asphalt cement that is not only stiffer but also behaves more like an
elastic solid at high pavement temperatures (Figure 1-10). That way, when a load is
applied to the asphalt cement in the mixture, it tends to act more like a rubber band and
spring back to its original position rather than stay deformed.
I stress
shear
(7)
shear
stress (z)
t "weak" binder
I normal stress (0) normal stress (0)
Figure 1-16. Contrasting Asphalt Binder Contribution

to Mixture Shear Strength
Another way to increase the shear strength of an asphalt mixture is by selecting an

aggregate that has a high degree of internal friction (+). This is accomplished by selecting
an aggregate that is cubical, has a rough surface texture, and graded in a manner to
develop particle-to-particle contact. Figure 1-11 shows the contrasting aggregate
contribution to mixture shear strength When a load is applied to the aggregate in the
mixture, the aggregate particles lock tightly together and function not merely as a mass of
individual particles, but more as a large, single, elastic stone. As with the asphalt
cement, the aggregate will act like a rubber band and spring back to its original shape
when unloaded. That way, no deformations (i.e., permanent) are accumulated.
shear shear
stress (T) stress (z)
t
I "weak aggregate
"strong" aggregate
normal stress (o) normal stress (o)
Figure 1-11. Contrasting Aggregate Contribution

to Mixture Shear Strength
While it is obvious that the largest portion of the resistance to permanent deformation of
the mixture is provided by the aggregate, the portion provided by the asphalt binder is
very important. Binders which have low shear characteristics due to composition or
temperature minimize cohesion and to a certain extent, the confining "normal" stress.
Thus the mixture begins to behave more like an unbound aggregate mass.
Fatigue Cracking
Like rutting, fatigue cracking is a distress type that most often occurs in wheel paths
where repeated heavy loads are applied. An early sign of Pdtigue cracking consists of
intermittent longitudinal wheel path cracks (i.e., in the direction of traffic). Fatigue
cracking is a progressive type of distress because at some point, the initial cracks will
join, which in turn, causes even more cracks to form. An intermediate stage of fatigue
cracking is sometimes called "alligator cracking" because the crack pattern resembles an
alligator's skin (Figure 1-12), In some extreme cases, the final stage of fatigue cracking is
disintegration when potholes form. A pothole forms when several of the pieces become
dislodged and removed under the action of traffic.
Figure 1-12, Alligator (Fatigue) Cracking
Engineers have long recognized that very stiff asphalt mixtures tend to have poor fatigue
properties when the pavement structure allows the asphalt mixture layer to deflect.
Stiffer materials, high deflection, and high stress levels translate to lower fatigue life.
While the mechanism of fatigue cracking is easy to understand, its cause often is not. It
cannot be addressed as just a materials problem. Fatigue cracking is usually caused by a
number of pavement factors that have to occur simultaneously. Obviously, repeated
heavy loads must be present. Some engineers believe that poor subgrade drainage,
resulting in a soft, high deflection pavement, is the principal cause of fatigue cracking.
Poorly designed and/or poorly constructed pavement layers that are also prone to high
deflections when loaded probably contribute to fatigue cracking. Thus, thin, very stiff
pavement layers, subjected to high deflections from repeated heavy loads are most
susceptible to fatigue cracking.
In many cases, fatigue cracking is merely a sign that a pavement has received the number
of load applications for which it was designed. Consequently, it is simply "worn out" and
in need of a planned rehabilitation. Assuming that the occurrence of fatigue cracking
coincides approximately with the design period, it may even not be considered a failure,
but rather the natural progression of a pavement design strategy. If the observed crachng
occurs much sooner than the design period, it may be a sign that the pavement received
more heavy loads, earlier than expected.
Consequently, the best ways to overcome fatigue cracking are:
adequately account for the anticipated number of heavy loads during design,
keep the subgrade dry using whatever means available,
use thicker pavements,
use paving materials that are not excessively weakened in the presence of
moisture, and
use paving materials that are resilient enough to withstand normal deflections.
In general, asphalt mixtures are unaffected and largely impervious to moisture. In some
extreme cases however, moisture vapor has been shown to strip asphalt cement from
mineral aggregate. While stripping of an underlying asphalt layer can manifest itself as
fatigue cracking in an upper asphalt layer, it is not normally considered a fatigue failure.
A more common instance of fatigue cracking being caused by a moisture weakened layer
is with an unbound base that has too many fine particles to allow for rapid drainage of
moisture. Unbound bases should be selected so that they do not trap moisture.
Only the last item, selection of resilient materials, can be addressed strictly from a
materials selection perspective. As a load is applied, horizontal tensile stresses occur
near the bottom of an asphalt layer (Figure 1-13}. Clearly, the material in this vicinity
must be very strong with sufficient tensile strength to withstand the applied tensile stress.
However, to overcome fatigue cracking, material in this vicinity also must be resilient. In
this context, resilient means that the material can withstand many load applications at
stress levels far less than the tensile strength, without cracking.
stresses
HMA must be
strong & resilient
Figure 1-13. Tensile Stresses at Bottom of HMA Layer
Thus, to overcome fatigue cracking from a materials perspective, HMA must be selected
so that it behaves like a soft elastic material. Since the tensile behavior of HMA is
strongly influenced by asphalt cement, this is accomplished by selecting an asphalt
cement that has upper limits placed on the elastic part of its overall stiffness. In effect,
soft asphalts have better fatigue properties than hard asphalts.
Low Temperature Cracking
As its name indicates, low temperature cracking is a distress type that is caused by
adverse environmental conditions rather than by applied traffic loads. It is characterized
by intermittent transverse cracks (i.e., perpendicular to the direction of traffic) that occur
at a surprisingly consistent spacing (Figure I- 14).
Low temperature cracks form when an asphalt pavement layer shrinks in cold weather.
As the pavement shrinks, tensile stresses build within the layer. At some point along the
pavement, the tensile stress exceeds the tensile strength and the asphalt layer cracks.
Thus, low temperature cracks occur primarily from a single cycle of low temperature.
Some engineers, however, also believe it is a fatigue phenomenon due to the cumulative
effect of many cycles of cold weather.
Figure 1-14. Low Temperature Cracking
Both groups agree that asphalt binder plays the central role in low temperature cracking.
In general, hard asphalt binders are more prone to low temperature cracking than soft
asphalt binders. Asphalt binders that are excessively oxidized, either because they are
unduly prone to oxidation or contained in a mixture left with too many air voids after
construction, or both, are more prone to low temperature cracking. Thus, to overcome
low temperature cracking engineers must use a soft binder, a binder that is not overly
prone to aging, and control in-place air void content so that the binder is not excessively
oxidized.
CURRENT WAYS TO SPECIFY ASPHALT CEMENTS
The current method to characterize asphalt cement consistency is by either penetration or

viscosity tests as shown in Figure 1-15. Both of these tests have been used to measure the
effect of temperature on asphalt behavior. This is done by measuring viscosity or
penetration at two temperatures and plotting the results as shown in Figure 1-16.
In this example, all three asphalts are the same viscosity grade because they are within
specified limits at 60' C. While Asphalts A and B display the same temperature
dependency, they have much different consistencies at a11 temperatures. Asphalts A and
C have the same consistency at low temperatures but remarkably different high
temperature consistency. Asphalt B has the same consistency at 60 C, but shares no
other similarities with Asphalt C. Because these asphalts share the same grade, they
might erroneously be expected to display the same characteristics during construction and
during hot and cold weather performance conditions.
Viscosity
vacuum
Penetration
0 sec 5 sec
Figure 1-15. Penetration and Viscosity Tests
[ Consistency
(pen or vis)
hardl Den
soft
Temperature, C
Figure 1-16. Temperature Susceptibility of Three Viscosity or

Penetration Graded Asphalts
Although viscosity is a fundamental measure of flow, it only provides information about

higher temperature viscous behavior, not about the low or intermediate temperature
elastic behavior needed to completely predict performance. Penetration describes only
the consistency at an intermediate temperature, 25" C. No low temperature properties are
directly measured in the current grading system. Often, viscosity and penetration tests do
not completely show the advantages or possible disadvantages of some modified asphalts.
Because of these deficiencies, many state highway agencies have amended standard test
procedures and specifications to better suit local conditions. In some locations, this
proliferation of tests and specifications has caused serious problems for asphalt suppliers
wishing to sell the same asphalt grades in several states. Often, states with very similar
performance conditions and materials will specify remarkably different asphalts. In the
current systems for specifying asphalt, tests are performed on unaged or "tank" asphalt
and on asphalt that has been laboratory aged to simulate construction aging. However, no
tests are performed on asphalts that have been aged to simulate in-service aging.
CURRENT ASPHALT MIXTURE DESIGN PROCEDURES
Most agencies currently use the Marshall mix design method. It is by far the most
common procedure used in the world to design HMA. This technique was developed by
Bruce Marshall, a former employee of the Mississippi State Highway Department. The
U.S. Army Corps of Engineers refined and added certain features to Marshall's approach
to the extent that it was formalized as ASTM D 1559, Resistance to Plastic Flow of
Bituminous Mixtures Using the Marshall Apparatus. The Marshall method entails a
laboratory experiment aimed at developing a suitable asphalt mixture by means of
stabilitylflow and densitylvoids analyses.
One of the strengths of the Marshall method is its attention to densitylvoids properties of
asphalt materials. This analysis ensures that the important volumetric proportions of mix
constituents are at their proper levels to achieve a durable HMA. Another advantage of
the Marshall method is that the required equipment is relatively inexpensive and very
portable, and thus, lends itself to remote quality control operations. Unfortunately, many
engineers believe that the impact method of laboratory compaction used with the
Marshall method does not simulate mixture densification that occurs under traffic in a
real pavement. Furthermore, the strength parameter used in this approach, Marshall
stability (Figure 1-17), does not adequately estimate the shear strength of HMA. These
two situations may result in asphalt mixtures prone to rutting. Consequently, there has
been a growing feeling among asphalt technologists that the Marshall method has
outlived its usefulness for modern asphalt mixture design.
test Marshall
specimen load / stability
breaking
J
head deformation
Figure 1-17. Marshall Stability
The Hveem mix design procedure was developed by Francis Hveem, once the Materials
and Research Engineer for the California Department of Transportation. Hveem and
others developed and refined the procedure over a long period. The procedure is outlined
in ASTM D 1560, Resistance to Deformation and Cohesion o f Bituminous Mixtures by
Means of Hveem Apparatus, and ASTM D 1561, Preparation of Bituminous Mixture Test
Specimens by Means of California Kneading Compactor. It is not commonly used
outside western states of the U.S.
The Hveem method also entails a density/voids and stability analysis. Mixture resistance
to swell in the presence of water is also determined. The Hveem method has two real
advantages. First, the kneading method of laboratory compaction is thought by most
engineers to better simulate the densification characteristics of HMA in a real pavement.
Second, the strength parameter, Hveem stability (Figure 1-18), is a direct measurement of
the internal friction component of shear strength. It measures the ability of a test
specimen to resist lateral displacement from application of a vertical load.
load
rubber
. .-.-
membrane
\
Figure 1-18. Hveem Stability
The disadvantage of the Hveem procedure is that the testing equipment, particularly the
kneading compactor and Hveem stabilometer, are somewhat more expensive than
Marshall equipment and not very portable. Furthermore, some important mixture
volumetric properties that are related to mix durability are not routinely determined as
part of the Hveem procedure. Some engineers believe that the method of selecting
asphalt content in the Hveem method is too subjective and may result in non durable
HMA with too little asphalt.
There are other mix design procedures in common use besides the Marshall and Hveem
procedures. For example, the Texas gyratory method is currently used by the state DOTS
in Texas, Oklahoma, and Colorado. This procedure retains volumetric design elements of
the Marshall method and the stability determination from the Hveem method. It is
differentiated from the others by its method of laboratory compaction, the Texas gyratory
compactor, which is thought by some engineers to be a suitable means of simulating
traffic densification. While the Texas gyratory design method eliminates some of the
disadvantages of the Marshall and Hveem methods, some believe that the operational
characteristics of the compactor need refining to be suitable for a wider variety of design
applications.
Increasingly, agencies are augmenting their customary mix design procedures with
em~iricalstrength testing. These tests are called empirical because their test outputs
simply result in a go or no go decision based on the experience of the agency with the test
calibrated to real pavements. One example of this type of testing is the Georgia Loaded
Wheel Tester (GALWT). The GALWT subjects HMA beam specimens to repeated
pneumatic stresses applied through a loaded wheel riding on a pressurized hose (Figure I-
I. How Asvhalt Mixtures Behave
19). After the required number of load applications, beam rutting is measured and the
mixture is either accepted or rejected.
rolling wheel
pressurized
beam test specimen /

Figure 1-19. Principle of the Georgia Loaded Wheel Tester
The advantage of empirical strength testing is that agencies can develop very clear
acceptkeject criteria, backed up by performance data from real pavements. This is also a
disadvantage however, because agencies have to expend considerable resources in
experimentation to achieve this experience. Even then the experience is only applicable
to the materials and environmental conditions tested. New products and materials require
additional experimentation. Furthermore, because empirical strength tests result in a
simple acceptlreject test result and no degree of performance is measured, they are
difficult to use for economic comparisons of alternate materials.
11.
SUPERPAVE TO THE RESCUE
INTRODUCTION
In 1987, the Strategic Highway Research Program (SHRP) began developing a new
system for specifying asphalt materials. The final product of the SHRP asphalt research
program is a new system referred to as Superpave which stands for Superior Performing
Asphalt Pavements, Superpave software is a computer program that assists engineers in
materials selection and mix design. However, the term "Superpave" refers to more than
just the conlputer program. Most important, it represents an improved system for
specifying component materials, asphalt mixture design and analysis, and pavement
performance prediction. The system includes test equipment, test methods, and criteria.
ASPHALT BINDERS
One portion of Superpave is a new asphalt binder specification with a new set of tests to
match. The document is called a binder specification because it is intended to function
equally well for modified as well as unmodified asphalts. A portion of the asphalt binder
specification is shown in Appendix A.
The new system for specifying asphalt binders is unique in that it is a performance based
specification. It specifies binders on the basis of the climate and attendant pavement
temperatures in which the binder is expected to serve. Physical property requirements
remain the same, but the temperature at which the binder must attain the properties
changes. For example, the high temperature, unaged binder stiffness (G*/sin 6) is
required to be at least 1.OO kPa. But this requirement must be achieved at higher
temperatures if the binder is expected to serve in a hot climate.
Performance graded (PG) binders are graded such as PG 64-22. The first number, 64, is
often called the "high temperature grade." This means that the binder would possess
II. Superpave to the Rescue
adequate physical properties at least up to 64' C. This would be the high pavement
temperature corresponding to the climate in which the binder is actually expected to
serve. Likewise, the second number (-22) is often called the "low temperature grade" and
means that the binder would possess adequate physical properties in pavements at least
down to -22' C. Additional consideration is given to the time of loading (open highway,
city streets, intersections, etc.) and magnitude of loads (heavy trucks).
Another key feature to binder evaluation in the Superpave system is that physical
properties are measured on binders that have been laboratory aged to simulate their aged
condition in a real pavement. Some binder physical property measurements are
performed on unaged binder. Physical properties are also measured on binders that have
been aged in the rolling thin film oven (RTFO) to simulate oxidative hardening that
occurs during hot mixing and placing. A pressure aging vessel (PAV) is used to
laboratory age binder to simulate the severe aging that occurs after the binder has served
many years in a pavement (Figure II-1).
Rolling Thin Film Oven (RTFO) Pressure Aging Vessel (PAV)
Figure 11-1. SHRP Binder Aging Techniques
Binder physical properties are measured using four devices:
dynamic shear rheometer,

rotational viscometer,
bending beam rheometer, and
0 direct tension tester.
The dynamic shear rheometer (DSR) is used to characterize the visco-elastic properties of
the binder. It measures the complex shear modulus (G*)and phase angle (6) by
subjecting a small sample of binder to oscillatory shear stresses while sandwiched
between two parallel plates (Figure 11-2).
Applied Stress
Position of
Oscillating Plate
Plate Fixed Plate

\ I
1 cycle
b
Figure 11-2. Dynamic Shear Rheometer
The DSR measures G* and 6 by measuring the shear strain response of the specimen to a
fixed torque as shown in Figure 11-3. In this figure, the shear strain response of a binder
specimen is "out of phase" with the applied stress by a certain time interval At. This time
interval represents the time lag in strain response. Phase lag is normally reported in
angular measurement by simply multiplying the time lag (At) by the angular frequency
(a)to arrive at a phase angle (6). For totally elastic materials there is no lag between
applied shear stress and shear strain response and 6 equals zero degrees. For totally
viscous materials, strain response is completely out of phase with applied stress and 6 is
90 degrees. Viscoelastic materials like asphalt binders posses phase angles between zero
and 90 degrees, depending on test temperature. At high temperatures, 6 approaches 90
degrees while at low temperatures 6 is nearly zero degrees. The binder specification uses
either G*/sin 6 at high temperatures (> 46" C) or G%in 6 at intermediate temperatures
(between 7" and 34" C) as a means of controlling asphalt stiffness.
Viscoelastic: 0 c 6 < 90 O
'L,
Shear complex shear modulus
Stress
Resulting
Shear
Strain
e
/ angular frequency
phase angle
Figure 11-3. Computation of G* and 6
By controlling stiffness at high temperatures, the binder specification ensures that asphalt
provides its fair share of the overall shear strength of the mixture in terms of high
temperature elasticity. Likewise, the specification ensures that the binder does not
contribute to fatigue cracking by limiting its stiffness at intermediate temperatures.
The rotational viscometer (RTV) characterizes the stiffness of the asphalt at 135O C,
where it acts almost entirely as a viscous fluid. It is a rotational coaxial cylinder
viscometer that measures viscosity by the torque required to rotate a spindle submerged in
a sample of hot asphalt (Figure 11-4) at a constant speed. The binder specification
requires that binders have a viscosity of less than 3 Pa-s. This ensures that the binder can
be pumped and otherwise handled during HMA manufacturing.
applied torque
from motor
sample chamber
Figure 11-4. Rotational Viscometer
The bending beam rheometer (BBR) is used to characterize the low temperature stiffness
properties of binders. It measures the creep stiffness (S) and logarithmic creep rate (m).
If. Superpave to the Rescue
These properties are determined by measuring the response of a small binder beam
specimen to a creep load at low temperatures (Figure 11-5). By knowing the load applied
to the beam and the deflection at any time during the test, the creep stiffness can be
calculated using engineering beam mechanics. The binder specification places limits on
creep stiffness and m-value depending on the climate in which the binder will serve.
Binders that have a low creep stiffness will not crack in cold weather. Likewise, binders
with high m-values are more effective in shedding stresses that build in asphalt
pavements as temperatures drop, again, ensuring that low temperature cracking will be
minimized.
Constant (Creep) Load
1 deflection
Figure 11-5. Bending Beam Rheometer
Some binders, particularly some polymer-modified asphalts, may exhibit a higher than
desired creep stiffness at low temperatures. However, may not crack because they retain
their ability to stretch without fracture at low temperatures. Consequently, the binder
specification allows a higher creep stiffness if it can be shown through the direct tension
test (DTT) that binders are sufficiently ductile at low temperatures, The output of the
DTT is tensile failure strain, which is measured on a small dog bone shaped specimen
that is stretched at low temperatures until it breaks (Figure 11-6). As with the BBR, the
DTT ensures that the binder's resistance to low temperature cracking is maximized.
strain
-
Figure 11-6. Direct Tension Tester
MINERAL AGGREGATES
SHRP researchers also believed that mineral aggregates played a key role in HMA
performance. While they did not develop any new aggregate lest procedures, they refined
existing procedures to fit within the Superpave system. Two types of aggregate
properties are specified in the Superpave system: consensus properties and source
properties.
Consensus properties are those which the SHRP researchers believed were critical in
achieving high performance HMA. These properties must be met at various levels
depending on traffic level and position within the pavement. High traffic levels and
surface mixtures (i.e., shallow pavement position) require more strict values for
consensus properties. Many agencies already use these properties as quality requirements
for aggregates used in HMA. These properties are:
coarse aggregate angularity,

fine aggregate angularity,
flat, elongated particles, and
clay content.
By specifying coarse and fine angularity, SHRP researchers were seeking to achieve
HMA with a high degree of internal friction and thus, high shear strength for rutting
resistance. Limiting elongated pieces ensures that the HMA will not be as susceptible to
aggregate breakage during handling and construction and under traffic. By limiting the
amount of clay in aggregate, the adhesive bond between asphalt binder and aggregate is
strengthened and otherwise enhanced.
Source properties are those which agencies often use to qualify local sources of aggregate.
The SHRP researchers believed that achieving these properties was important, but did not
specify critical values since they are so source specific. The source properties are:
toughness,
soundness, and
deleterious materials.
Toughness is measured by the LA abrasion test. Soundness is measured by the sodium or

magnesium sulfate soundness test. Deleterious materials are measured by the clay lumps
and friable particles test. These tests are already in common use by most agencies.
To specify aggregate gradation (Appendix B), SHRP researchers refined an approach

already in wide use by many agencies. It uses the 0.45 power gradation chart with control
limits and a restricted zone (Figure 11-7) to develop a design aggregate structure.
Percent Passing
nom max
,+escontrolpoint max size
+
B
size B
I Sieve Size, mm (raised to 0.45 power)
Figure 11-7. Superpave Gradation Limits, 12.5 mm Mixture
A Superpave design aggregate structure must pass between the control points while
avoiding the restricted zone. The maximum density gradation is drawn from the 100
IZ. Superpave to the Rescue
percent passing the maximum aggregate size through the origin. Maximum aggregate
size is defined as one size larger than the nominal maximum aggregate size. Nominal
maximum size is defined as one size larger than the first sieve size to retain more than 10
percent. The restricted zone is used by SHRP Superpave to avoid mixtures that have a
high proportion of fine sand relative to total sand and gradations that follow the 0.45
power line, which do not normally have adequate voids in the mineral aggregate (VMA).
In many instances, the restricted zone will discourage the use of fine natural sand in an
aggregate blend. It will encourage the use of clean manufactured sand. The design
aggregate structure approach ensures that the aggregate will develop a strong, stone
skeleton to enhance resistance to permanent deformation while achieving sufficient void
space for mixture durability.
ASPHALT MIXTURES
Two key features in the Superpave system are laboratory compaction and performance
testing. Laboratory compaction is accomplished by means of a Superpave Gyratory
Compactor (SGC). While this device shares some common traits with the Texas gyratory
compactor, it is a completely new device with new operational characteristics. Its main
utility is to fabricate test specimens. However, by capturing data during SGC
compaction, a mix design engineer can also gain insight into the compactibility of HMA.
The SGC can be used to design mixtures that do not exhibit tender mix behavior and do
not densify to dangerously low air void contents under the action of traffic.
The performance of HMA immediately after construction is influenced by mixture

properties resulting from hot mixing and compaction. Consequently, a short term aging
protocol was incorporated into the Superpave system. This was accomplished by
requiring that loose mixture specimens, prior to compaction by the SGC, be oven aged for
four hours at 135' C.
Perhaps the most important development to arise from the SHRP asphalt research
program was performance based tests and performance prediction models for HMA.
Output from these tests can be used to make detailed predictions of actual pavement
performance (Figure 11-8). In other words, test procedures and performance prediction
models were developed that will allow an engineer to estimate the performance life of a
prospective HMA in terms of equivalent axle loads (ESALs) or time to achieve a certain
level of rutting, fatigue cracking, and low temperature cracking.
Performance Prediction of:

Testing Rutting
Results Fatigue Cracking
Low Temperature Cracking
Figure 11-8. Superpave Pavement Performance Prediction
Two new performance based testing procedures were developed, the Superpave Shear
Tester (SST) and Indirect Tensile Tester (IDT). The output from these tests is input to
performance prediction models in Superpave to estimate actual pavement performance
(ag.,millimeters of rutting).
The SST is a testing device that performs the following six tests on HMA specimens:
volumetric test,
uniaxial strain test,
simple shear test at constant height,
repeated shear test at constant stress ratio,
frequency sweep test at constant height, and
repeated shear test at constant height (option).
The first two tests involve testing the specimen using confining pressure. To accomplish
this, the SST has a testing chamber capable of applying confining pressure by means of
compressed air. Test temperature is also carefully controlled by the testing chamber. The
SST has axial and horizontal hydraulic actuators with accompanying linear variable
differential transducers (LVDTs) to measure the response of test specimens to load. Tests
proceed by closed-loop feedback control. This means that the response of a specimen to
loading from one actuator is measured by an LVDT. The other actuator uses the signal
from this LVDT to respond as required. For example, in the simple shear test at constant
height, a shear stress is applied to the HMA specimen by the horizontal actuator, As the
specimen is sheared, it tends to dilate. The vertical LVDT senses this dilation as a change
in specimen height and a signal is sent to the vertical actuator to apply sufficient vertical
load to keep the specimen's height from changing. Thus, dilation is prevented.
11. Superpave to the Rescue
Tests using the SST are performed at a variety of temperatures to simulate actual
pavement temperatures. While a portion of the tests are aimed at fatigue cracking, the
SST's main utility is a means of designing against permanent deformation.
The IDT is used to measure creep compliance and tensile strength of HMA. This test
uses a single vertical actuator to load a test specimen across its diametral plane. It is used
to characterize HMA as a means of designing against fatigue and low temperature
cracking.
In the Superpave system, the results of SST and IDT testing are input into pavement
performance prediction models. Using these models, mix design engineers can estimate
the combined effect of asphalt binders, aggregates, and mixture proportions. The models
take into account the structure, condition, and properties of the existing pavement (if
applicable) and the amount of traffic to which the proposed mixture will be subjected
over its performance life. The output of the models is millimeters of rutting, percent area
of fatigue cracking, and spacing (in meters) of low temperature cracks. By using this
approach, the Superpave system accomplishes what no previous design procedure has;
namely, it joins material properties with pavement structural properties to predict actual
pavement performance. Thus, the benefit (or detriment) of new materials, different mix
designs, asphalt modifiers, and other products can finally be quantified in terms of cost
versus predicted performance.
PUTTING IT ALL TOGETHER
Because Superpave mixture design and analysis is more complex than those in current
use, the extent of its use depends on the traffic level or functional classification of the
pavement for which it is being used. Consequently, three levels of Superpave mixture
design were developed. Their extent of use and testing requirements are shown in Table
11-1.
Table 11-1. Superpave Mix Design Levels
Traffic, ESALs Design Level Testing Requirements'
ESALs 5 1o6 1 volumetric design
lo6 < ESALs 5 lo7 2 volumetric design + performance prediction tests
ESALs > lo7 3 volumetric design + enhanced performance
prediction tests
In all cases, moisture susceptibility must be evaluated using AASHTO T283.
While much of the resources in SHRP were devoted to developing the SST, IDT, their
protocols, and performance prediction models, volumetric mix design occupies a key role
in Superpave mix design. Volumetric design, which is all that is required by a Level 1
mixture design, entails fabrication of test specimens using the SGC and selecting asphalt
content on the basis of air voids, voids in the mineral aggregate (VMA), voids filled with
asphalt (VFA), and the ratio of dust to effective asphalt content. Consensus and source
aggregate properties must be achieved.
A Level 2 mixture design uses a volumetric mix design as a starting point. A battery of
SST and IDT tests are performed to arrive at a series of golno go performance
predictions.
A Level 3 mixture design encompasses most of the facets of Levels 1 and 2. Additional
SST and IDT tests are performed at a wider variety of temperatures. Level 3 design is the
only protocol that utilizes SST confined specimen testing. Because of the more
comprehensive range of tests and results, Level 3 design offers an enhanced and more
reliable level of performance prediction.
111.
MATERIALS SELECTION
INTRODUCTION
Superpave utilizes a completely new system for testing, specifying, and selecting asphalt
binders. While no new aggregate tests were developed, current methods of selecting and
specifying aggregates were refined and incorporated into the Superpave mix design
system. Superpave asphalt mixture requirements were established from currently used
criteria.
ASPHALT BINDERS
The new SHRP binder specification (a portion of which is shown in Appendix A) is

unique in that it is performance based and that binders are selected on the basis of the
climate in which they are intended to serve. The physical property ( e g , creep stiffness,
G*/sin 6, etc.) requirements are constant among all grades of binders.
What differentiates the various binder grades is the temperature at which the requirements
must be met. For example, a binder classified as a PG 64-22 means that the binder must
meet high temperature physical property requirements at least up to a temperature of 64"
C and low temperature physical property requirements at least down to -2Z0C.
Table 111-1 shows the current binder grades in the SHRP binder specification. In this
table, the PG 76 and 82 grades are used only to accommodate slow transient or standing
loads, or excessive truck traffic.
111. Materials Selection
A module in the Superpave software assists users in selecting binder grades. Superpave
contains three methods by which the user can select an asphalt binder grade:
By Geographic Area: An Agency would develop a map showing binder grade

to be used by the designer based on weather andlor policy decisions.
By Pavement Temperature: The designer would need to know design

pavement temperature.
By Air Temperature: The designer determines design air temperatures, which

are converted to design pavement temperatures.
Superpave Weather Database
Superpave software contains a database of weather information for 6500 reporting

stations in the US and Canada, which allows users to select binder grades for the climate
specific to project location. For each year a weather station has been in operation the
hottest seven-day period is determined and the average maximum air temperature for
those seven consecutive days is calculated. For all the years of record (stations with less
than 20 years of records were not used) a mean and standard deviation are calculated.
Likewise the coldest day of each year is identified and the mean and standard deviation
are calculated.
Reliability
As used in Superpave, reliability is the percent probability in a single year that the actual
temperature will not exceed the design temperature. SHRP binder selection is very
III. Materials Selection
flexible in that a different level of reliability can be assigned to high and low temperature
grades. Consider summer air temperatures in Topeka, Kansas, which has a mean seven-
day maximum of 36" C and a standard deviation of 2" C. Figure 111-1 shows the
frequency distribution for this data. In an average year there is a 50 percent chance the
seven-day maximum air temperature will exceed 36" C. However, only a two percent
chance exists that the temperature will exceed 40" C; hence, a design air temperature of
40" C will provide 98 percent reliability.
50 % reliability
36 40
7-Day Maximum Air Temperature
Figure 111-1. Distribution of Annual Seven-Day Maximum Air
Temperature for Topeka, KS
Start with Air Temperature
To see how the binder selection works assume that an asphalt mixture is designed for
Topeka. Figure 111-2 shows frequency distributions for high and low design air
temperatures. In a normal summer, the average seven-day maximum air temperature is
36" C with a standard deviation of 2" C. In a normal winter, the average coldest
temperature is -23" C. For a very cold winter the temperature is -31" C, with a standard
deviation of 4" C.
IZI. Materials Selection
Figure 111-2. Distribution of High and Low Design Air Temperatures

for Topeka, KS
Convert to Pavement Temperature
Superpave software calculates high pavement temperature 20 rnm below the pavement
surface and low temperature at the pavement surface. For a wearing course at the top of a
pavement section, the pavement temperatures in Topeka are 56" and -23" C for 50
percent reliability and 60" (56" + 2 standard deviations) and -3 1" C for 98 percent
reliability (Figure 111-3).
Figure In-3. Distribution of High and Low Design Pavement

Temperatures for Topeka, KS
In Superpave, the high pavement design temperature at a depth of 20 mm is computed by

the following formula:
where, Tzo,, = pavement temperature at a depth of 20 rnrn in "C,

Td, = maximum average high air temperature during the hottest seven-day period
in "C, and
lat = project latitude in degrees.
There are two possible ways to determine the low pavement design temperature in
Superpave. First, the low pavement design temperature simply can be assumed to be the
same as the low air temperature. This method was originally recommended by SHRP
researchers. This is a very conservative assumption because pavement temperature is
almost always warmer than air temperature in cold weather. The Topeka, Kansas
example above used this approach. The second method utilizes the following formula,
which was developed by Canadian SHRP researchers:
where, Tmi,= minimum pavement design temperature in "C,

Tair= minimum air temperature in average year in "C.
Using this approach for the Topeka example, the minimum pavement design temperature
would be 0.859 x -23" + 1.7" or 18" C. This method of computing minimum pavement
design temperature is gaining favor among asphalt technologists in North America.
However, the first method is still used by Superpave.
Select Binder Grade
For a reliability of at least 50 percent, the high temperature grade must be PG 58 for
Topeka. Selecting a PG 58 would actually result in a higher level of reliability, about 85
percent, because of the "rounding up" to the next standard grade. The next lower grade
only protects to 52" C, less than 50 percent reliability. The low temperature grade must
be a PG XX-28. As with high temperature grade, rounding to this standard low
temperature grade results in almost 90 percent reliability. For 98 percent reliability, the
needed high temperature grade is PG 64; the low temperature grade is PG XX-34.
Both of these low temperature grades utilize the Superpave approach that assumes low air
and low pavement temperatures are the same. Had the alternative approach been used,
the binder grades selected would have been PG 58-22 for minimum 50 percent reliability
and PG 58-28 for minimum 98 percent reIiability. The method of converting low air to
low pavement temperature has a profound effect on the binder selection process.
b
PG 64-34 (98 % minimum reliability)
Figure 111-4. Various Binder Grades for Topeka, KS
Manipulating temperature frequency distributions is not a task that the designer need
worry about. Superpave software handles the calculations. For any site, the user can
enter a minimum reliability and Superpave will calculate the required asphalt binder
grade. Alternately the user can specify a desired asphalt binder grade and Superpave will
calculate the reliability obtained.
Effect of Loading Rate on Binder Selection
SHRP binder selection by climate only assumes that a binder will be used in a mixture
subjected to fast moving loads. The loading rate used by the dynamic shear rheometer is
10 radians per second, which corresponds to a traffic speed of approximately 90
kilometers per hour. Much slower loading rates are experienced by pavements near
intersections, toll booths, etc. In some cases, loads are not moving but rather are
stationary. In these cases, a binder would have to exhibit a higher stiffness to overcome
the slower loading rate.
To accommodate these situations, Superpave requires that the high temperature grade be
increased by at least one or as many as two grades. For example, if a temperature based
selection resulted in a desired binder grade of PG 64-22, to account for slow transient
loads, the designer would select one grade higher binder, a PG 70-22. If standing loads
were anticipated, the designer would select a PG 76-22. Loading rate has no effect on the
selected low temperature grade. Pavement design temperatures of 76" or 82" C do not
correspond to any climate zone in North America. Specifying this grade is simply a
means of ensuring that the binder will have higher stiffness at 64" C, the actual high
pavement design temperature. Because the highest possible pavement temperature in
North America is about 70" C, two additional high temperature grades, PG 76 and PG 82,
were necessary to accommodate slow loading rates.
Effect of Traffic Level on Binder Selection
Superpave recommends that traffic level be considered when selecting binders. When the
design traffic level exceeds 10 million equivalent single axle loads (ESALs), the designer
is encouraged to "consider" increasing the high temperature grade by one grade. When
the design traffic level exceeds 30 million ESALs, the designer is required to increase the
high temperature grade by one grade. As with loading rate, there is no effect of traffic
level on low temperature grade. For the Topeka example where the temperature based
selection required a PG 58-28, a project with a very high number of ESALs would require
a PG 64-28.
MINERAL AGGREGATE
During SHRP, pavement experts were surveyed to ascertain which aggregate properties
were most important. There was general agreement that aggregate properties played a
central role in overcoming permanent deformation. Fatigue cracking and low temperature
cracking were less affected by aggregate characteristics. SHRP researchers relied on the
experience of these experts and their own to identify two categories of aggregate
properties that needed to be used in the Superpave system: consensus properties and
source properties. In addition, a new way of specifying aggregate gradation was
developed. It is called the design aggregate structure.
Consensus Properties
It was the consensus of the pavement experts that certain aggregate characteristics were
critical and needed to be achieved in all cases to arrive at well performing HMA. These
characteristics were called "consensus properties" because there was wide agreement in
their use and specified values. Those properties are:
coarse aggregate angularity,

fine aggregate angularity,
flat, elongated particles, and
. clay content.
There are required standards for these aggregate properties. The consensus standards are
not uniform. They are based on traffic level and position within the pavement structure.
Materials near the pavement surface subjected to high traffic levels require more stringent
consensus standards. They are intended to be applied to a proposed aggregate blend
rather than individual components. However, many agencies currently apply such
requirements to individual aggregates so that undesirable components can be identified.
Coarse Aggregate Angularity
This property ensures a high degree of aggregate internal friction and rutting resistance. It
is defined as the percent by weight of aggregates larger than 4.75 mm with one or more
fractured faces.
Many state DOTS have protocols to measure coarse aggregate angularity. These usually
involve manually counting particles to determine fractured faces. A fractured face is
defined as any fractured surface that occupies more than 25 percent of the area of the
outline of the aggregate particle visible in that orientation. One test method example is
the Pennsylvania DOT'STest Method No. 62 1, "Determining the Percentage of Crushed
Fragments in Gravel."
Table 111-2 outlines the required minimum values for coarse aggregate angularity as a
function of traffic level and position within the pavement.
Table 111-2. Superpave Coarse Aggregate Angularity Requirements
I Traffic, million I Depth from Surface I

<1 651- -1-
<3 751- 501-
< 10 85/80 601-
< 30 95190 80175
< 100 1001100 95/90
> 100 100/100 1001100
Note: "85180" means that 85 % of the coarse aggregate has one
fractured face and 80 % has two fractured faces.
Fine Aggregate Angularity
This property ensures a high degree of fine aggregate internal friction and rutting
resistance. It is defined as the percent air voids present in loosely compacted aggregates
smaller than 2.36 rnm. Higher void contents mean more fractured faces.
A test procedure currently promulgated by the National Aggregates Association is used to

measure this property. In the test, a sample of fine aggregate is poured into a small
calibrated cylinder by flowing through a standard funnel (Figure 111-5).
funnel
fine aggr sample
cylinder of known volume (V)
uncompacted voids =
v -WIG,,
v
Figure 111-5. Fine Aggregate Angularity Apparatus

By determining the weight of fine aggregate (W) in the filled cylinder of known volume
(V), void content can be calculated as the difference between the cylinder volume and
fine aggregate volume collected in the cylinder. The fine aggregate bulk specific gravity
(Gs,)is used to compute fine aggregate volume.
Table HI-3 outlines the required minimum values for fine aggregate angularity as a
function of traffic level and position within pavement.
Table 111-3. Superpave Fine Aggregate Angularity Requirements
Traffic, million Depth from Surface

ESALs < 100mm > 100 mm
< 0.3 - -
Note: Criteria are presented as percent air voids in loosely

compacted fine aggregate.
Flat, Elongated Particles
This characteristic is the percentage by weight of coarse aggregates that have a maximum
to minimum dimension of greater than five. Elongated particles are undesirable because
they have a tendency to break during construction and under traffic. The test procedure
used is ASTM D 4791, "Flat or Elongated Particles in Coarse Aggregate" and it is
performed on coarse aggregate larger than 4.75 mrn.
The procedure uses a proportional caliper device (Figure 111-6) to measure the
dimensional ratio of a representative sample of aggregate particles. In Figure 111-6, the
aggregate particle is first placed with its largest dimension between the swinging arm and
fixed post at position A. The swinging arm then remains stationary while the aggregate is
placed between the swinging arm and fixed post at position B. If the aggregate passes
through this gap, then it is counted as a flat or elongated particle.
1 :5 pivot point fixed post (B)
Figure 1114. Proportional Caliper Device to Measure Flat and

Elongated Particles
Two values are measured: percentage of flat particles and percentage of elongated
particles. Table 111-4 outlines the required maximum values for flat, elongated particles
in coarse aggregate.
Table 111-4. Superpave Flat, Elongated Particle Requirements
I Traffic, million I Percent
<3 10
< 10 10
< 30 10
< 100 10
> 100
- 10
Note: Criteria are presented as maximum
percent by weight of flat and elongated
1 particles. I
Clay Content
Clay content is the percentage of clay material contained in the aggregate fraction that is
finer than a 4.75 mm sieve. It is measured by AASHTO T 176, "Plastic Fines in Graded
Aggregates and Soils by Use of the Sand Equivalent Test."
In this test, a sample of fine aggregate is placed in a graduated cylinder with a

flocculating solution and agitated to loosen clayey fines present in and coating the
aggregate. The flocculating solution forces the clayey material into suspension above the
granular aggregate. After a period that allows sedimentation, the cylinder height of
suspended clay and sedimented sand is measured (Figure 111-7). The sand equivalent
value is computed as a ratio of the sand to clay height readings expressed as a percentage.
/- graduated
flocculating
d cylinder
reading
solution
suspended clay / reading
sedimented aggregate-
Figure 111-7. Sand Equivalent Test
Table III-5 outlines the required clay content values for fine aggregate.
Table 111-5. Superpave Clay Content Requirements
Source Properties
In addition to the consensus aggregate properties, pavement experts believed that certain
other aggregate characteristics were critical. However, critical values of these properties
could not be reached by consensus because needed values were source specific.
Consequently, a set of "source properties" were recommended. Specified values are
established by local agencies. While these properties are relevant during the mix design
process, they may also be used as source acceptance control. Those properties are:
toughness,
soundness, and
deleterious materials.
Toughness
Toughness is the percent loss of materials from an aggregate blend during the Los
Angeles Abrasion test. The procedure is stated in AASHTO T 96, "Resistance to
Abrasion of Small Size Coarse Aggregate by Use of the Eos Angeles Machine." This test
estimates the resistance of coarse aggregate to abrasion and mechanical degradation
during handling, construction, and in-service. It is performed by subjecting the coarse
aggregate, usually larger than 2.36 mm, to impact and grinding by steel spheres. The test
result is percent loss, which is the weight percentage of coarse material lost during the
test as a result of the mechanical degradation. Maximum loss values typically range from
approximately 35 to 45 percent.
Soundness
Soundness is the percent loss of materials from an aggregate blend during the sodium or
magnesium sulfate soundness test. The procedure is stated in AASHTO T 104,
"Soundness of Aggregate by Use of Sodium Sulfate or Magnesium Sulfate." This test
estimates the resistance of aggregate to weathering while in-service. It can be performed
on both coarse and fine aggregate. The test is performed by alternately exposing an
aggregate sample to repeated immersions in saturated solutions of sodium or magnesium
sulfate each followed by oven drying. One immersion and drying is considered one
soundness cycle. During the drying phase, salts precipitate in the permeable void space
of the aggregate. Upon re-immersion the salt re-hydrates and exerts internal expansive
forces that simulate the expansive forces of freezing water. The test result is total percent
loss over various sieve intervals for a required number of cycles, Maximum loss values
range from approximately 10 to 20 percent for five'cycles.
Deleterious Materials
Deleterious materials are defined as the weight percentage of contaminants such as shale,
wood, mica, and coal in the blended aggregate. This property is measured by AASHTO
T 112, "Clay Lumps and Friable Particles in Aggregates." It can be performed on both
coarse and fine aggregate. The test is performed by wet sieving aggregate size fractions
over prescribed sieves. The weight percentage of material lost as a result of wet sieving
is reported as the percent of clay lumps and friable particles. A wide range of maximum
permissible percentage of clay lumps and friable particles is evident. Values range from
as little as 0.2 percent to as high as 10 percent, depending on the exact composition of the
contaminant.
Gradation
To specify gradation, Superpave uses a modification of an approach already used by some

agencies. It uses the 0.45 power gradation chart to define a permissible gradation. This
chart uses a unique graphing technique to judge the cumulative particle size distribution
of a blend of aggregate. The ordinate of the chart is percent passing. The abscissa is an
arithmetic scale of sieve size in millimeters, raised to the 0.45 power. Figure III-8
illustrates how the abscissa is scaled. In this example, the 4.75 rnm sieve is plotted at
2.02 units to the right of the origin, This number, 2.02, is the sieve size, 4.75 rnm, raised
to 0.45 power. Normal 0.45 power charts do not show arithmetic abscissa labels such as
those in Figure 111-8. Instead, the scale is annotated with the actual sieve size as shown in
Figure 111-9.
Percent Passing
loo T
6o I Example:
4.75 mm sieve plots at (4.75)0-45= 2.02
20 /
Figure 111-8. Graphical Basis for 0.45 Power Chart
An important feature of this chart is the maximum density gradation. This gradation plots
as a straight line from the maximum aggregate size through the origin. Superpave uses a
standard set of ASTM sieves and the following definitions with respect to aggregate size
(Appendix B shows sieve sizes used by Superpave):
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 gradation (Figure 111-9) represents a gradation in which the
aggregate particles fit together in their densest possible arrangement. Clearly this is a
gradation to avoid because there would be very little aggregate space within which to
develop sufficiently thick asphalt films for a durable mixture. Figure III-9 shows a 0.45
ZIZ. Materials Selection
power gradation chart with a maximum density gradation for a 19 mm maximum

aggregate size and 12.5 rnm nominal maximum size.
( Percent Passing
I Sieve Size, mm Raised to 0.45 Power
Figure 111-9. Maximum Density Gradation for 19 mm Maximum Size
To specify aggregate gradation, two additional features are added to the 0.45 power chart:
control points and a restricted zone. Control points function as master ranges through
which gradations must pass. They are placed on the nominal maximum size, an
intermediate size (2.36 mm), and the dust size (0.075 mm).
The restricted zone resides along the maximum density gradation between the
intermediate size (either 4.75 or 2.36 mm) and the 0.3 mm size. It forms a band through
which gradations are not permitted to pass. Gradations that pass through the restricted
zone have often been called "humped gradations" because of the characteristic hump in
the grading curve that passes through the restricted zone. In most cases, a humped
gradation indicates a mixture that possesses too much fine sand in relation to total sand.
This gradation practically always results in tender mix behavior, which is manifested by
a mixture that is difficult to compact during construction and offers reduced resistance to
permanent deformation during its performance life. Gradations that violate the restricted
zone possess weak aggregate skeletons that depend too much on asphalt binder stiffness
to achieve mixture shear strength. These mixtures are also very sensitive to asphalt
content and can easily become plastic.
The term used to describe the cumulative frequency distribution of aggregate particle
sizes is the design aggregate structure. A design aggregate structure that lies between the
control points and avoids the restricted zone meets the requirements of Superpave with
respect to gradation. Superpave defines six mixture types (Table 111-6) as defined by their
nominal maximum aggregate size:
Table 111-6. Superpave Mixture Designations
I Superpave I Nominal Maximum I Maximum I

I Designation
I
Size, mm I I
Size, mm
25 mm 25 37.5
19 mm 19 25
12.5 rnrn 12.5 19
9.5 rnm 9.5 12.5
Figure III- 10 illustrates the control points and restricted zone for a 12.5 mrn Superpave
mixture. Appendix B shows numerical gradation limits and gradation charts for the six
Superpave mixtures.
I Percent Passing
.075 .3 2.36 4.75 9.5 12.5 19.0

Sieve Size, mrn (raised to 0.45 power)
Figure 111-10. Superpave Gradation Limits, 12.5 mm Mixture

Superpave recommends, but does not require, mixtures to be graded below the restricted
zone. It also recommends that as project traffic level increases, gradations move closer to
the coarse control points. Furthermore, the Superpave gradation control requirements
were not intended to be applied to special purpose mix types such as stone matrix asphalt
or open graded mixtures.
ASPHALT MIXTURES
Asphalt mixture design requirements in Superpave consist of:
mixture volumetric requirements,

dust proportion, and
moisture susceptibility.
Specified values for these parameters are applied during the Level 1 mixture design
phase.
Mixture Volumetric Requirements
Mixture volumetric requirements consist of air voids, voids in the mineral aggregate and
voids filled with asphalt. Air void content is an important property because it is used as
the basis for asphalt binder content selection. In Superpave, the design air void content is
four percent.
Superpave defines voids in the mineral aggregate (VMA) as the sum of the volume of air
voids and effective (i.e., unabsorbed) binder in a compacted sample. It represents the
void space between aggregate particles. Specified minimum values for VMA at the
design air void content of four percent are a function of nominal maximum aggregate
size. Table 111-7 shows Superpave VMA requirements.
Table 111-7. Superpave VMA Requirements
I Nominal Maximum
I I
Aggregate Size
I
Minimum VMA, %
I
Voids filled with asphalt (VFA or Pf,) is defined as the percentage of the VMA
containing asphalt binder. Consequently, VFA is the volume of effective asphalt binder
expressed as a percentage of the VMA. The acceptable range of design VFA at four
percent air voids is a function of traffic level as shown in Table 111-8.
Table 111-8. Superpave VFA Requirements
Traffic, ESALs Design VFA, %

< 3 x 105 70 - 80
< 1 x 106 65 - 78
< 3 x 106 65 - 78
< 1 107 65 - 75
Dust Proportion
Another mixture requirement is the dust proportion. This is computed as the ratio of the
percentage by weight of aggregate finer than the 0.075 mm sieve to the effective asphalt
content expressed as a percent by weight of total mix. Effective asphalt content is the
total asphalt used in the mixture less the percentage of absorbed asphalt. Dust proportion
is used during the mixture design phase as a design criterion. An acceptable dust
proportion is in the range from 0.6 to 1.2, inclusive for all mixtures.
Moisture Susceptibility
The moisture susceptibility test used to evaluate HMA for stripping is AASHTO T 283,
"Resistance of Compacted Bituminous Mixtures to Moisture Induced Damage." This test
is not a performance based test but serves two purposes. First, it identifies whether a
combination of asphalt binder and aggregate is moisture susceptible. Second, it measures
the effectiveness of anti-stripping additives.
In the test, two subsets of test specimens are produced. Specimens are compacted to
achieve an air void content in the range from six to eight percent with a target value of
seven percent. Test specimens should be sorted so that each subset has the same air void
content. One subset is moisture conditioned by vacuum saturation to a constant degree of
saturation in the range from 55 to 80 percent. This is followed by an optional freeze
cycle. The final conditioning step is a hot water soak. After conditioning both subsets
are tested for indirect tensile strength. The test result reported is the ratio of tensile
strength of the conditioned subset to that of the unconditioned subset. This ratio is called
the "tensile strength ratio" or TSR. Superpave requires a minimum TSR of 80 percent.
Table 111-9 outlines the current test parameters in AASHTO T 283.
Table 111-9. Test Parameters for AASHTO T283
Test Parameter Test Requirement

Short-Term Aging Loose mix1: 16 hrs at 60" C
Compacted mix: 72-96 hrs at 25' C
Air Voids Compacted Specimens 6 to 8 %
Sample Grouping
Saturation
I
Average air voids of two subsets should be
eoual
55 to 80 %
I
Swell Determination None
Freeze Minimum 16 hrs at - 18" C (optional)
Hot Water Soak 24 hrs at 60" C
Strength Property Indirect tensile strength
Loading Rate 5 1 m d m i n at 25" C
precision Statement I None
Short-term aging protocol of AASHTO T 283 does not match short-term aging
protocol of Superpave. Suggest using T283 procedure of 16 hours at
60" C .
IV.
ASPHALT MIXTURE VOLUMETRICS
INTRODUCTION
A factor that must be taken into account when considering asphalt mixture behavior is the
volumetric proportions of asphalt binder and aggregate components, or more simply,
asphalt mixture volumetrics. The developers of Superpave felt that the volumetric
properties of asphalt mixtures were so important that a volumetric mixture design
protocol was developed. The following section describes volumetric analysis of HMA,
which plays a significant role in most mixture design procedures, including the Superpave
system.
COMPONENT DIAGRAM APPROACH
The model used to describe HMA mass and volume properties is the component diagram.
It considers a compacted sample of HMA with its constituent air voids, asphalt cement,
and mineral aggregate shown as discrete components (Fig IV-I). The compacted sample
is assumed to consist of a unit volume (e.g., one cubic meter, one cubic centimeter, etc.)
with known mass (e.g., kilograms or grams). The component diagram is particularly
suited to metric units because in this system, density and specific gravity are numerically
the same since the density of water is very nearly 1.000 gram per cubic centimeter and its
specific gravity is 1.000 at 25' C. It is a tool commonly used for many civil engineering
applications because it represents a convenient model to track distinct masses and
volumes in non-homogeneous construction materials.
ZV. Asphalt Mixture Volumetrics
VOLUME MASS
I
--
air Mass air = 0

VMI
-
Unit rota1
Volume Mass
Bulk
rol aggr
Figure IV-1. Component Diagram of Compacted Sample of HMA
The component diagram provides a clear definition of density, that is, the mass of a unit
volume of compacted material. Since the model consists of several distinct materials, the
density of the entire sample is often called its bulk density. It is determined by dividing
the total mass of the sample by its total volume.
For a given asphalt content, the maximum theoretical density is the mass of aggregate
and asphalt divided by the volume of only these two components. In other words, the
volume of air voids is not included. Maximum theoretical density (or specific gravity) is
an extremely useful property because it can be used as a reference to calculate several
other important properties such as air void content.
AsphaZt content is the mass concentration of asphalt binder. It is expressed as percent by

total mass of mixture or percent by total mass of aggregate. Most agencies use percent by
mass of mixture. Effective asphalt content is the mass concentration of asphalt binder
that is not lost to absorption. Absorbed asphalt content is the mass concentration of
asphalt binder absorbed by the aggregate. It is normally reported as a percentage of the
mass of aggregate.
The volume concentration of air within the compacted sample is the air void content. Air
voids are always expressed as a percentage of total volume of mixture.
IV. As~haltMixture Volumetrics
The intergranular space occupied by asphalt and air in a compacted mixture is called the
voids in the mineral aggregate or VMA. In the component diagram, the sum of the
volume of air and volume of effective asphalt, expressed as a percent of total volume, is
the VMA. The volume of absorbed asphalt is usually not considered to be part of the
VMA.
Not shown on the diagram is the percentage of voidsfilled with asphalt or VFA. This
property is the percentage of the VMA that contains asphalt. While it could be computed
by dividing the volume of asphalt by the volume of the VMA, it is normally computed by
the following formula.
VFA = [(VMA - Volume of Air)NMA] x 100 %
Although contrary to physical laws, the model shows mass and volume on the same
diagram, with the same scale. Another deceptive feature of the component diagram is
that it is not well suited for considering secondary weights and volumes such as absorbed
asphalt. Furthermore, narrow reliance on the physical model sometimes inhibits a more
fundamental understanding of volumetric properties such as VMA. Even with these
flaws, the component diagram is still the best way to define and illustrate determination
of the properties of compacted HMA.
Note that when calculating HMA properties during mix design, engineers seldom work
from a sketch of a component diagram. They normally use well established formulas,
originally derived from a component diagram, to arrive at the various properties of
interest. Appendix D contains a list of all the formulas used to compute compacted mix
properties.
SPECIFIC GRAVITY
In order to use the component diagram, it is necessary to be able to convert between mass
and volume. Specific gravity is the tool employed for this purpose. Specific gravity is
the ratio of the mass of a given volume of a substance to the mass of an equal volume of
water, both at the same temperature. It is a unique material property that allows for two
important determinations.
IV. Asphalt Mixture Volumetrics
First, specific gravity is used to determine density by:
D = G x 1.000
where, D = density of material in grams per cubic centimeter (g/cm3),

G = specific gravity of material, and
1.000 = density of water in grams per cubic centimeter.
The terms "density" and "specific gravity" are often interchanged, which suggests they
have the same meaning. In fact, in metric units, they have the same numerical value.
While this usage is technically incorrect, context most often conveys the intended
meaning. This equation offers the most precise meaning of each.
Second, knowing the mass and specific gravity of a material, the volume of the material
can be determined by:
where, V = volume of material,

M = mass of material, and
G = specific gravity of material, and
1.000 = density of water (1.000 g/cm3).
Use of this equation is best understood by the following example.
Consider an object placed on a scale and found to weigh 75 kilograms. This object is
known to have a specific gravity very nearly that of water, or 1.000. Using these values
in the above equation indicates the object has a volume of about 75,000 cubic centimeters
(i.e., 175 kg x 1,000g/kg]/[1.000 x 1.000 g/cm3] = 75,000 cm3).
This example is also useful to illustrate the fact that different specific gravities must often
be considered. The conditions of the example were somewhat obscure with respect to the
precise meaning of the specific gravity used.
While the object may be a homogeneous material, it is more likely a composite of several
materials. Consequently, the conditions of the example needed to be more precise and
should have specified bulk specific gravity. Bulk specific gravity is least determinate
since it considers the object in whole or "bulk" form and is blind to the contributions of
the object's individual components. A volume determined from a bulk specific gravity
must be assumed to include the total volume and not unique component volumes.
In the case of mineral aggregate, bulk, effective, and apparent specific gravities are
usually determined. Bulk specific gravity (AASHTO T84 and T85) is determined by
measuring the dry weight and bulk volume of an aggregate sample (Figure IV-2). The
bulk volume includes the solid aggregate volume plus the volume of surface pores
holding water. The bulk volume is measured on the aggregate in a saturated surface dry
(SSD) condition.
Bulk Vol
Bulk Volume = solid volume +
water permeable pore volume
" S S D Level
water permeable pore volume
Figure IV-2. Bulk Specific Gravity of Aggregate
Apparent specific gravity (also measured using AASHTO T84 and T85) is determined by
measuring the dry weight and apparent volume of an aggregate sample (Figure IV-3).
The apparent volume only includes the volume of the solid aggregate and does not
include the volume of any surface pores.
ZV. Asphalt Mixture Volumetrics
-.---
"" - App Vol C
volume of surface pores
Figure IV-3. Apparent Specific Gravity of Aggregate
Effective specific gravity is measured on asphalt mixtures (AASHTO T209) of known

asphalt content. It the context of bulk and apparent specific gravity, it is computed using
the dry weight of aggregate and the effective volume of the aggregate. The aggregate
effective volume includes the volume of the solid aggregate and the volume of surface
pores filled with water but not asphalt (Figure IV-4). Aggregate effective specific gravity
is not directly measured in the same manner as bulk and apparent specific gravity.
Instead, it is calculated by knowing the maximum theoretical specific gravity of a mixture
and the asphalt content.
+ volume of water permeable pores not

r[9&!\
k~,cri&~ filled with asphalt
f@$j volume of water permeable pores not

filled with asphalt
6$ 4 ~
asphalt coating
Figure IV-4. Effective Specific Gravity of Aggregate

Only bulk and effective specific gravities are used during mix design volumetric
calculations. Volumes calculated with each of these would have different meanings and
thus, numeric values. The wide array of asphalt, aggregate, and mixture specific gravities
are often confusing to those new to asphalt technology. Careful attention to the meaning
of each, and the desired HMA property will clarify the analysis.
EXAMPLE CALCULATIONS
A sample of compacted HMA is known to have the following properties at 25" C:
Mix Bulk Specific Gravity = 2.329

Aggregate Bulk Specific Gravity = 2.705
Aggregate Effective Specific Gravity = 2.731
Asphalt Binder Specific Gravity = 1.015
Asphalt Content = 5.0 percent by mass of total mix
The air void content, VMA, VFA, maximum theoretical specific gravity, absorbed asphalt
content, and effective asphalt content should be determined. Figure IV-5shows these
known items on a component diagram. The required calculations are with the following
steps.
G,, = 2.329
P, = 5.0 %
by total mass of mix
Figure IV-5. Known Items for Example

Step 1: Determine Density
Thus, the sample in the component diagram is assumed to have a mass of 2.329 grams
and occupy one cubic centimeter. Note that an alternate approach could use 2,329
kilograms and one cubic meter.
Steu 2: Determine Mass
Thus, in the component diagram, the asphalt binder has a mass of 0.1 16 g and the
aggregate 2.213 g. The air was assumed to be without mass.
Stev 3: Determine Volumes
Vtotal-,ph = MaMhlGasph x 1.OW g1cm3= 0.1 16 gl(1.015 x 1.000 g/cm3)= 0.1 14 cm3
Vbulk-aggr = Maggi%ulk-aggr X 1.000 g l ~ m=

3 2.2 13 gf(2.705 X 1.ooO g/~m3)
= 0.8 18 cm3
Veffmaggr= Maggr/Geff-, x 1.000 glcm3 = 2.213 gl(2.73 1 x 1.000 glcm3) = 0.810 cm3
Vabsqasph = Vt,ulk-aggr - Veffmaggr= 0.818 cm3 - 0.810 cm3 = 0.008 cm3
Vefgasph = Vtotal-asph - Vabs-asph = 0.114 cm3 - 0.008 cm3= 0.106 cm3
Vair = Vtotal - (Veff-asph + Vbulk-aggr) = 1.000 cm" (0.106 cm3 + 0.8 18 cm3)= 0.076 cm3
VMA = V,ir + Veff-asph = 0.076 cm3 + 0.106 cm3= 0.182 cm3

IV. Amhalt Mixture Volurnetrics
Percent volume concentration by total mix of each component is calculated by:
Percent volume air voids = (0.076 cm3/1.000cm3) x 100 % = 7.6 %
Percent volume of total asphalt = (0.1 14 cm3/1.000cm3)x 100 % = 11.4 %
Percent volume of effective asphalt = (0.106 cm3/1.000 cm3) x 100 % = 10.6 %
Percent volume of bulk aggregate = (0.818 cm3/1.000cm3) x 100 % = 81.8 %
Percent VMA = (0.182 cm3/1.000 cm3) x 100 % = 18.2 %
An alternative procedure to compute VMA is:
Percent VMA = 100 % - % volume of bulk aggregate = 100 % - 8 1.8 % = 18.2 %
Step 4: Calculate Effective Asphalt Content and Absorbed Asphalt Content
Mass of effective asphalt = Vefi-asph x Gmph x 1.000 &m3 = 0.106 cm3 x 1.015 x 1.000 glcm3 = 0,108 g
Mass of absorbed asphalt = Vabs-asph x Gaqh x 1.OOO g/cm3 = 0.008 em3 x 1.Ol5 x 1,000 &m3 = 0.008 g
Effective Asphalt Content = (Meffmasph/ Mtotal ) x 100 % = (0.108 g / 2.329) x 100 % = 4.6 %
Absorbed Asphalt Content = (Mabs-asph / Maggr ) x 100 % = (0.008 g / 2.213 g) x 100 % = 0.4 %
Step 5: Calculate VFA
Percent VFA = Veff-asph / Volume of VMA x I 00 % = (0.106 cm3/0.182cm3) x LOO % = 58.2 %
Step 6: Calculate Maximum Theoretical Specific Gravity
Grnax-theo = [(Masph + Maggr)l(Veff-asph + Vbulk-aggr)l/l-oOOgtcm3
= [(0.116 g + 2.213 g)/(0.106 cm3 + 0.818cm3)]/l.000g/cm3= 2.521

N. Asphalt Mixture Volumetrics
Answer: Air Void Content = 7.6 %

VMA = 18.2 %
VFA = 58.2 %
Absorbed Asphalt Content = 0.4 %
Effective Asphalt Content = 4.6 %
Max. theo. sp. grav. = 2.521
All of the computed masses and volumes are shown on a component diagram in Figure
IV-6.
VOL (cm3)
Figure IV-6. Computed Masses and Volumes for Example
The conditions of this example stated that the asphalt content was five percent by mass of
total mix." Although this is the most common method of expressing asphalt content,
some agencies express asphalt content as percent by mass of aggregate. Had asphalt
content been expressed in this way, the weight of asphalt and aggregate would have been
calculated by:
5/100 = Masph / Maggr and 2.329 = Masph + Maggr
Solving these equations simultaneously yields:

N.Asphalt Mixture Volumetrics
The analysis would continue from this point as before. This illustrates the importance of
clearly stating the basis for asphalt content.
The example used the volume of effective asphalt and air voids to compute VMA. In
effect, the aggregate bulk specific gravity was used to compute VMA. This is the
approach currently used by most agencies. Superpave also uses this convention. In the
Superpave mix design procedure, VMA criteria are based on aggregate bulk specific
gravity. Use of other aggregate specific gravities to compute VMA means that the VMA
criteria no longer apply and the mixture does not meet the requirements of Superpave.
SUPERPAVE GYRATORY COMPACTION
INTRODUCTION
In selecting a method of laboratory compaction, SHRP researchers had several goals.

Most important, they desired a device that would realistically compact trial mix
specimens to densities achieved under actual pavement climate and loading conditions.
The device needed to be capable of accommodating large aggregates. Furthermore, it was
desired that the device afford a measure of compactibility so that potential tender mixture
behavior and similar compaction problems could be identified. A high priority for SHRP
researchers was a device that was well suited to mixing facility quality control and quality
assurance operations. No compactor in current use achieved all these goals.
Consequently, a new compactor was developed, the Superpave Gyratory Compactor
(SGC).
The basis for the SGC was a large Texas gyratory compactor modified to use the
compaction principles of a French gyratory compactor. The Texas device accomplished
the goals of achieving realistic specimen densification and it was reasonably portable. Its
6-inch sample diameter (ultimately 150 mm on an SGC) could accommodate mixtures
containing aggregate up to 50 mm maximum (37.5 nominal) size. SHRP researchers
modified the Texas device by lowering its angle and speed of gyration and adding real
time specimen height recordation. In fact, a considerable amount of this phase of SHRP
mixture research was conducted on a modified Texas gyratory compactor loaned to SHRP
by the Texas DOT.
TEST EQUIPMENT
The SGC is an mechanical device comprised of the following system of components:
reaction frame, rotating base, and motor,

V. Superpave Gyratoiy Compaction
loading system, loading ram, and pressure gauge,

height measuring and recordation system, and
mold and base plate.
Figure V-1 shows a generic SGC.
control and data

reaction acquisition pane1
frame loading
/ ram
mold
rotating
base
Figure V-1. Superpave Gyratory Compactor
The reaction frame provides a non-compliant structure against which the loading ram can
push when compacting specimens. The base of the SGC rotates and is affixed to the
loading frame. It supports the mold while compaction occurs. Reaction bearings are
used to position the mold at an angle of 1.25 degrees, which is the compaction angle of
the SGC. The electric motor drives the rotating base at a constant speed of 30 revolutions
per minute.
A hydraulic or mechanical system applies a load to the loading ram, which imparts 600
kPa compaction pressure to the specimen. The loading ram diameter nominally matches
the inside diameter of the mold, which is 150 rnm. A pressure gauge with digital signal
conditioning measures the ram pressure during compaction, As the specimen densifies
during compaction, the pressure gauge signals the loading system to adjust the position of
the loading ram so that a constant compaction pressure is maintained throughout the
compaction process.
Specimen height measurement is an important function of the SGC. By knowing the

mass of material placed in the mold, the diameter of the mold, and the specimen height,
V. Superpave Gyratory Compaction
an estimate of specimen density can be made at any time throughout the compaction
process. Specimen density is computed by dividing the mass by the volume of the
specimen. The specimen volume is calculated as the volume of a smooth-sided cylinder
with a diameter of 150 rnrn and the measured height. Height recordation is variously
accomplished by measuring the position of the ram before and during the test. The
vertical change in ram position identically equals the change in specimen height. The
specimen height signal is processed through a serial port connection which is connected
to a personal computer, printer, or other device to record height (i.e., density)
measurements throughout the compaction process. By this method, a compaction
characteristic is developed as the specimen is compacted (Figure V-2).
Percent of Maximum
Theoretical Density
10 100 1000 -
Log Gyrations
Figure V-2. Compaction Characteristic of SGC
The SGC uses a mold (Figure V-3) with an inside diameter of 150 mrn and a nominal
height of 250 m. A base plate fits in the bottom of the mold to afford specimen
confinement during compaction.
ram pressure
0.6 MPa
Figure V-3. SGC Mold Configuration and Compaction Parameters
SPECIMEN PREPARATION
Compaction specimens are required to be mixed and compacted under equiviscous

temperature conditions corresponding to 0.170 Paas and 0.280 Pa-s, respectively. Figure
V-4shows a typical temperature-viscosity chart for an asphalt binder. Mixing is
accomplished by a mechanical mixer. After mixing, loose test specimens are subjected to
four hours of short term aging in a forced draft oven maintained at a constant 135' C.
During short term aging, loose mix specimens are required to be spread into a thickness
resulting in 21 to 22 kg per cubic meter and stirred every hour to ensure uniform aging.
The compaction molds and base plates should also be placed in an oven at 1 3 5 O C for at
least 30 to 45 minutes prior to use.
Viscosity, Paws
Temperature, C
Figure V-4. Temperature-Viscosity Relationship

Three specimen sizes are used. If specimens are to be used for volumetric determinations
only, use sufficient mix to arrive at a specimen 150 mm in diameter by approximately 115
mm height. This requires approximately 4500 grams of aggregate. In this case, the test
specimen produced is tested without trimming, Alternatively, to produce specimens for
performance testing, approximately 5500 grams of aggregate is used to fabricate a
specimen that is 150 mm in diameter by approximately 135 mm height. In this case,
specimens will have to be trimmed to 50 mm before testing in the SST or IDT. At least
one loose sample should remain uncompacted to obtain a maximum theoretical specific
gravity using AASHTO T 209. For performing AASHTO T283, test specimens are
fabricated to a height of 95 rnm, which requires approximately 3500 grams of aggregate.
OVERVIEW OF PROCEDURE
After short term aging the loose test specimens are ready for compacting. The compactor
is initiated by turning on its main power. The vertical pressure should be set at 600 kPa
(+ 18 kPa). The gyration counter should be zeroed and set to stop when the desired
number of gyrations is achieved. Three gyration levels are of interest:
design number of gyrations (Ndesign),

initial number of gyrations (Niitial), and
maximum number of gyrations (N,,,imu,).
Test specimens are compacted using Nmaximum

gyrations. The relationship between Nbiw,
Nmaximum, and Ninitial are:
The design number of gyrations (Ndesign) ranges from 68 to 172 and is a function of the
climate in which the mix will be placed and the traffic level. The average design high air
temperature is provided by Superpave software and represents the average seven-day
maximum air temperature for project conditions. The range of values for Ndesign,
Nmaxlmum, and Ninitidis shown in Table V-1.
Table V-1. Superpave Gyratory Compaction Effort
Design Average Design High Air Temperatur

ESALs <3gC 39 - 40C 41 - 42C 43 - 44C
(millions) N,,, Ndes Nmax I Nini Ndes Nmax I Nini Ndes Nmax Nini Ndes Nmax
< 0.3 7 68 104 7 82 127

0.3 - 1 7 76 117 8 93 146
1-3 7 86 134 8 105 167
3 - 10 8 96 152 9 119 192
10-30 8 109 174 9 135 220
30 - 100 9 126 204 10 153 253
> 100 9 142 233 10 172 288
After the base plate is in place, a paper disk is placed on top of the plate and the mold is
charged in a single lift. The top of the uncompacted specimen should be slightly rounded.
A paper disk is placed on top of the mixture.
The mold is placed in the compactor and centered under the ram. The ram is then
lowered until it contacts the mixture and the resisting pressure is 600 kPa (+ 18 kPa).
The angle of gyration (1.25" k 0.02") is then applied and the compaction process begins.
When NmXi,,, has been reached, the compactor should automatically cease. After the
angle and pressure are released, the mold containing the compacted specimen is then
removed. After a suitable cooling period, the specimen is extruded from the mold.
The bulk specific gravity of test specimens should be measured using AASHTO T 166.
Maximum theoretical specific gravity should be measured using AASHTO T 209.
DATA ANALYSIS AND PRESENTATION
Superpave gyratory compaction data should be analyzed by computing the estimated bulk
specific gravity, corrected bulk specific gravity, and percentage of maximum theoretical
specific gravity for each desired gyration. The example specimen compaction
information in Table V-2 illustrates this analysis.
Table V-2. Example Specimen Compaction Information
Specimen No. 1: Total Mass = 4869 g

G, (measured) = 2.563
No. of Gyrations Height, mm Gnb Gmb % Gm,
(estimated) (corrected)
8 (Nmi) 127.0 2.170 2.218 86.5
50 118.0 2.334 2.385 93.1
100 115.2 2.391 2.444 95.4
109 W d e J 114.9 2.398 2.45 1 95.6
150 113.6 2.425 2.478 96.7
174 (Nrnad 113.1 2.436 2.489 97.1
Gmb(measured) 2.489 - -
Project conditions for this mixture are such that N, = 174, Ndes= 109, and Nini = 8
gyrations. During compaction, the height was measured after each gyration and recorded
for the number of gyrations shown in the first column. The Gmb(estimated) values were
determined by:
where, V,,, = volume of specimen in mold during compaction (cm3),

d = diameter of mold (150 mrn), and
h, = height of specimen in mold during compaction (mm).
G,, (estimated) = K,1vm

1.000~
/ cm3
where, Gmb(estimated)= estimated bulk sp grav of specimen during compaction,

W, = mass of specimen (g).
To illustrate this determination, consider the specimen conditions at 50 gyrations. The

specimen height is measured as 118.0 mm. The estimated volume of the specimen at 50
gyrations is:
Thus, the Gmb(estimated)at 50 gyrations is:
4867.8g / 2085.2cm3
G,, (estimated) = = 2.334
1.OOOg I cm3
This calculation assumes that the specimen is a smooth-sided cylinder, which of course, it
is not. The volume of the specimen is slightly less than the volume of a smooth-sided
cylinder because of surface irregularities. That is why the final estimated Gmbat 174
gyrations, 2.436, is different than the measured Gmb after 174 gyrations, 2.489. To correct
for this difference, the estimated Gmbat any given number of gyrations is corrected by a
ratio of the measured to estimated bulk specific gravity at Nmimum using the following
formula.
G , (measured)
C=
G , (estimated)
where, C = correction factor,

Gmb(measured)= measured bulk specific gravity after Nmimum,and
Gmb(estimated)= estimated bulk specific gravity at Nma,imum.
The estimated Gmbat all other number of gyrations can then be corrected by using the
correction factor in the following formula.
G,, (corrected) = C x G,, (estimated)
where, Gmb(corrected)= corrected bulk sp grav of specimen at any gyration, N,

C = correction factor, and
Gmb(estimated)= estimated bulk sp grav at any gyration, N.
In this example, this ratio is 2.48912.436or 1.022. Percent G, is computed as the ratio
Gmb(corrected) to G, (measured).
If this example had been for the purpose of mix design, a companion specimen would
have been compacted and average percent G,, values resulting from the two specimens
would have been used for further analysis. A densification plot for this example showing
two specimens and an average is shown in Figure V-5.
II I
I
I I I
10 100 1000
Number of Gyrations
Figure V-5. Densification Plot for Example Specimens
Design parameters are established on the basis of air void content at Ndesign,
Ninitial,
and
Nm,im,,. The following table shows the criteria and observed average values considering
the average of the two specimens in the example.
Table V-3. Densification Values and Criteria
No. of gyrations Criterion for %Gmm Observed %Gmm
Nini (8) < 89.0 87.1

CALIBRATION AND STANDARDIZATION
A critical calibration item is the height measurement system. This is normally

accomplished by means of a dummy specimen of known dimensions. The loading ram
can be calibrated by means of a proving ring or load cell of suitable accuracy. The speed
of gyration can be checked by accurately timing the rotation over a known number of
rotating base revolutions.
Calibration of the angle of gyration is another critical calibration item. This is

accomplished by various means that are compactor dependent. One method of calibrating
the angle involves the use of a digital protractor that directly reports angular deviation
from a fixed datum. Another method uses precise dial gauge measurements collected
with the mold at various orientations. The measurements are used to calculate the angle
of gyration. In any case, the angle should be checked while the mold contains a specimen
under loaded conditions.
VI.
VOLUMETRIC MIX DESIGN
INTRODUCTION
Volumetric mix design plays a central role in Superpave mixture design. The best way of
illustrating its steps is by means of an example. This section provides the Superpave
Level 1 mixture design test results for a project that was constructed in 1992 by the
Wisconsin Department of Transportation on IH-43 in Milwaukee. The information
presented follows the logical progression of testing and data analysis involved in a Level
1 mixture design and encompasses the concepts outlined in previous sections. There are
four major steps (see Appendix E for an outline of the major steps in Level 1 mix design)
in the testing and analysis process:
1. selection of materials (aggregates, binders, modifiers, etc.),

2. selection of a design aggregate structure,
3. selection of a design asphalt binder content,
4. evaluation of moisture sensitivity of the design mixture.
Selection of materials consists of determining the traffic and environmental factors for the
paving project. From that, the performance grade of asphalt binder required for the
project is selected. Aggregate requirements are determined based on traffic level and
layer depth. Materials are selected based on their ability to meet or exceed the established
criteria.
Selection of the design aggregate structure is a trial-and-error process. This step consists
of blending available aggregate stockpiles at different percentages to arrive at aggregate
gradations that meet Superpave requirements. Three trial blends are normally employed
for this purpose. A trial blend is considered acceptable if it possesses suitable volumetric
properties (based on traffic and environmental conditions) at a predicted design binder
content. Once selected, the trial blend becomes the design aggregate structure.
VZ. VolumetricMix Design
Selection of a design asphalt binder content consists of varying the amount of asphalt
binder with the design aggregate structure to obtain acceptable volumetric and
compaction properties when compared to the mixture criteria, which are based on traffic
and environmental conditions. This step is a verification of the results obtained from the
previous step. This step also allows the designer to observe the sensitivity of volumetric
and compaction properties of the design aggregate structure to asphalt content. The
design aggregate structure at the design asphalt binder content becomes the job-mix
formula.
Evaluation of moisture sensitivity consists of testing the designed mixture by AASHTO

T283 to determine if the mix will be susceptible to moisture damage.
MATERIAL SELECTION
For the IH-43 project, design ESALs are determined to be 18 million in the design lane.
This places the design in the traffic category from 10 to 30 million ESALs. Traffic level
is used to determine design requirements such as number of design gyrations for
compaction, aggregate physical property requirements, and mixture volumetric
requirements. The traffic level also determines the level of mixture design required. For
18 million ESALs and higher, a Superpave Level 3 design is required. Consequently, the
design process requires a Level 1 design to determine mixture volumetric properties,
followed by performance testing and analysis required by Level 3.
The mixture in this example is an intermediate course mixture. It will have a nominal
maximum particle size of 19.0 mm. It will be placed at a depth less than 100 rnm from
the surface of the pavement.
Binder Selection
Environmental conditions are determined from weather station data stored in the
Superpave weather database. The project is near Milwaukee, which has 2 weather
stations:
VI. Volumetric Mix Design
Table VI-1. Project Environmental Conditions and Binder Grades

Weather Station Min. Pvmt. Max. Pvmt. Binder Design Air
Temp. ("C) Temp. ("C) Grade Temp. ("C)
Low Reliability (50%)
Milwaukee Mt. Mary -26 51 PG 52-28 32
Milwaukee WSO AP -25 51 PG 52-28 31 .
Paving Location
-26 51 PG 52-28 32
1 (Assumed)
High Reliability (98%)
Milwaukee Mt. Mary -32 55 PG 58-34 36
Milwaukee WSO AP -33 54 PG 58-34 34 1
Paving Location
(Assumed) -33 55 PG 58-34 35
Low and high reliability level binders are shown. Reliability is the percent probability
that the actual temperature will not exceed the design pavement temperatures listed in the
binder grade. In this example, the designer chooses high reliability for all conditions.
Thus, a PG 58-34 binder is needed. The average Design High Air Temperature is 35OC.
Having determined the need for a PG 58-34 binder, the binder is selected and tested for
specification compliance. Test results are indicated in Table VI-2.
Table VI-2. Binder Specification Test Results
Test Property 1)
Test Result 11 Criteria
Original Binder
Flash Point n/a 304C 230C minimum
Rotational Viscosity 135C 0.575 P a s 3 Pass maximum
Rotational Viscosity 175C 0.142 Pa+s n/a
Dynamic Shear Rheometer G*/sin 6 @ 58C 111.42 kPa 11 1.00 kPa minimum
RTFO-aged
" Binder
Mass Loss nla 0.14% 1.00% maximum
Dynamic Shear Rheometer G*/sin 6 @ 58C 2.41 kPa 2.20 kPa minimum
PAV-aged Binder
Dynamic Shear Rheometer G*sin 6 @ 16C 1543 kPa 5000 kPa maximum
Bending Beam Rheometer Stiffness @ -24C 172.0 MPa 300.0 MPa
maximum
Bending Beam Rheometer m-value @ -24C 0.321 0.300 minimum
Comparing the test results to specifications, the designer verifies that the asphalt binder
meets the requirements of a PG 58-34 grade. Specification testing requires only that
rotational viscosity be performed at 135C. Additional testing was performed at 175C to
establish laboratory mixing and compaction temperatures. Figure VI-1 illustrates the
temperature-viscosity relationship for this binder. Based on these test results, the mixing
temperature range is selected between 165C and 172C. The compaction temperature
range is selected between 151C and 157C.
PG 58-34 Binder
Figure VI-1. Temperature-ViscosityRelationship for Project Binder
Aggregate Selection
Next, the designer selects the aggregates to use in the mixture. For this example, there
are 5 stockpiles of materials consisting of three coarse materials and two fine materials.
It is assumed that the mixing facility will have at least 5 cold feed bins. If fewer cold feed
bins are available, fewer stockpiles will be used. The materials are split into
representative samples, and a washed sieve analysis is performed for each aggregate.
These test results are shown in the section on selecting design aggregate structure.
The bulk and apparent specific gravities are determined for each aggregate. These
specific gravities are used in trial binder content and VMA calculations. Test results are
indicated in Table VI-3.
Table VI-3. Aggregate Specific Gravities
Aggregate Bulk Sp. Gravity Apparent Sp. Gravity

#1 Stone 2.703 2.785
112" Chip 2.689 2.776
318" Chiz, 2.723 2.797
Manuf. Sand 2.694 2.744
Screen Sand 2.679 2.731
In addition to sieve analysis and specific gravity determination, Superpave requires that
consensus aggregate tests be performed to assure that the aggregates selected for the mix
design are acceptable. The four tests required are: coarse aggregate angularity, fine
aggregate angularity, thin and elongated particles, and clay content. In addition, the
specifying agency can select any other aggregate tests deemed important. These tests can
include items such as soundness, toughness, and deleterious materials among others.
Superpave consensus aggregate criteria are intended to be applied to combined aggregate

gradations rather than individual aggregate components. However, some designers find it
useful to perform the aggregate tests on the individual aggregate components. This step
allows the designer to use the test results in narrowing the acceptable range of blend
percentages for the aggregates. It also allows for greater flexibility if nlultiple trial blends
are attempted. The test results from the components can be used to estimate the results
for a given combination of materials. The drawback to this procedure is that it takes
more time to perform this additional testing. For this example, the aggregate properties
are measured for each stockpile as well as for the aggregate trial blends.
Coarse Aggregate Angularity
This test is performed on the coarse aggregate particles of the aggregate stockpiles. The
coarse aggregate particles are defined as particles larger than 4.75 mm. Test results are
Table VI-4. Coarse Aggregate Angularity
Note that this test is not performed on the two fine aggregates, even though they have
some small percentage retained on the 4.75 mm sieve. The manufactured sand has 4.5%
retained and the Screen Sand has 10.5% retained on the 4.75 mm sieve.
Table VI-4 also indicates criteria for fractured faces based on traffic (18 million ESALs)
and depth from the surface (< 100 mm). The criteria change as the traffic level and layer
position (relative to the surface) change. The criteria are also based on the test results
from the aggregate blend rather than individual materials. Thus, even though the #1
Stone appears to be below the minimum criteria, it can be used as long as the selected
blend of aggregates meets the criteria in Table VI-4.
Fine Aggregate Angularity
This test is performed on the fine aggregate particles of the aggregate stockpiles. The fine
aggregate particles are defined as particles smaller than 2.36 mm. Test results are
Table VI-5. Fine Aggregate Angularity
Aggregate % Air Voids (Loose) Criterion

Manufactured Sand 62% 45% min
I Screen Sand I 36% I I
Note that this test is not performed on the three coarse aggregates, even though they have
a small percentage passing the 2.36 millimeter sieve. The #1 Stone has 1.9%passing, the
112"Chip has 2.6% passing, and the 318" Chip has 3.0% passing the 2.36 rnm sieve.
Table VI-5 also indicates criterion for fine aggregate angularity based on traffic and depth
from the surface. Even though the Screen Sand appears to be below the minimum
criterion, it can be used as long as the selected blend of aggregates meets the criterion in
Table VI-5.
Flat and Elongated Particles
This test is performed on the coarse aggregate particles of the aggregate stockpiles. The
coarse aggregate particles are defined as particles larger than 4.75 mrn. Test results are
Table VI-6. Plat and Elongated Particles
Aggregate I %Thin/Elongated I Criterion

#1 Stone
112" Chiu
0%
0% 10% max I
I 3/8" Chip I 0% I 1
Note that this test is not performed on the two fine aggregates, even though they have
some small percentage retained on the 4.75 mm sieve. The manufactured sand has 4.5%
retained and the Screen Sand has 10.5% retained on the 4.75 mm sieve. Table VI-6 also
indicates the criterion for percentage of flat and elongated particles, which is based on
traffic only. The criterion changes as the traffic level changes. In this case, the
aggregates are cubical and not in danger of failing the criterion.
Clay Content (Sand Equivalent)
This test is performed on the fine aggregate particles of the aggregate stockpiles. The fine
aggregate particles are defined as particles smaller than 4.75 mm. Test results are
indicated in Table VI-7
Table VI-7. Clay Content (Sand Equivalent)
I Ageregate I Sand Eauivalent I Criterion 1

( Manufactured Sand I 47 1 45 rnin I
L Screen Sand 1 70 I I
14ote that this test is not performed on the three coarse aggregates, even though they have
some small percentage passing the 4.75 mm sieve. The #1 Stone has 2.1 % passing, the
112" Chip has 3.1% passing, and the 318" Chip has 4.8% passing the 4.75 mm sieve.
Table VI-7 also indicates the criterion for clay content (sand equivalent) based on traffic
only. The criterion changes as the traffic level changes. The criterion is also based on the
test results from the aggregate blend rather than individual materials. Both fine
aggregates are above the minimum requirement, so there is reasonable expectation that
the blend will also meet the clay content requirement. Once all of the aggregate testing is
complete, the material selection process is complete. The next step is to select the design
aggregate structure.
SELECT DESIGN AGGREGATE STRUCTURE
To select the design aggregate structure, the designer establishes trial blends by
mathematically combining the gradations of the individual materials into a single
gradation. The blend gradation is then compared to the specification requirements for the
appropriate sieves. Gradation control is based on four control sieves: the maximum
sieve, the nominal maximum sieve, the 2.36 mm sieve, and the 75 micron sieve.
The nominal maximum sieve is one sieve size larger than the first sieve to retain more
than ten percent of combined aggregate. The maximum sieve size is one sieve size
greater than the nominal maximum sieve.
The restricted zone is an area on either side of the maximum density line. For a 19.0mm
nominal mixture, it starts at the 2.36 mm sieve and extends to the 300 micron sieve.
The minimum and maximum values required for the control sieves change (as does the
restricted zone) as the nominal size of the mixture changes. Table VI-8 indicates the
gradation requirements for this example.
Table VI-8. Gradation Criteria for 19.0 mm Nominal Mixture
Gradation I Sieve Size, I Minimum, Maximum, 1

I Control Item I mm I % % I
I Control 1 19.0 1 90.0 100.0 I
I Points I 2.36 I 23.0 49.0 I
0.075 2.0 8.0
2.36 34.6 34.6
Restricted 1.18 22.3 28.3
Zone 0.600 16.7 20.7
Any proposed trial blend gradation has to pass between the control points established on
the four sieves. In addition, it has to be outside of the area bounded by the limits set for
the restricted zone.
Figure VI-2 indicates the gradation requirements for a 19.0 mm nominal mixture.
Superpave 19.0 mm Nominal Gradation
Sieve Size (mm) Raised to 0.45 Power)
Figure VI-2. Gradation Requirements for 19.0 mm Nominal Mixture
Any number of trial blends can be attempted, but three is the standard number of blends.
Trial blending consists of varying stockpile percentages of each aggregate to obtain blend
gradations meeting the gradation requirements for that particular mixture. For this
example, three trial blends are used: an intermediate blend, a coarse blend, and a fine
blend. The intermediate blend is combined to produce a gradation that is not close to
either the gradation limits for the control sieves, or the restricted zone. The stockpile
percentages and combined gradation for Trial Blend 1 are indicated in Table VI-9 and
Figure VI-3. The coarse blend is combined to produce a gradation that is close to the
minimum criteria for the nominal maximum sieve, the 2.36 mm sieve, and the 75 micron
sieve. The stockpile percentages and combined gradation for Trial Blend 2 are indicated
in Table VI-10 and Figure VI-4. The fine blend is combined to produce a gradation that is
close to the maximum criteria for the nominal maximum sieve, and the restricted zone.
The stockpile percentages and combined gradation for Trial Blend 3 are indicated in
Table VI- 11 and Figure VI-5.
Table VI-9. Trial Gradation for Intermediate Blend

#l 1/2" 3/8" Mfg Scr.
Stone chip chip sand sand
Blend 1 25.0% 15.0% 22.0% 18.0% 20.0%
Blend 2 30.0% 25.0% 13.0% 17.0% 15.0%
Blend 3 10.0% 15.0% 30.0% 31.0% 14.0%
Blend 1 Blend 2 Blend 3
Sieve Gradation Gradation Gradation
IH-43 Trial Blend 1- Intermediate Blend

19.0 mm Nominal Mixture
0.075 0.300 2.36 19.0 25.0

Figure VI-3. Trial Blend 1 - Intermediate Blend

VZ. Volumetric Mix Design
Table VI-10. Trial Gradation for Coarse Blend

#1 1 3/8" Mfg Scr.
Stone chip chip sand sand
Blend 1 25.0% 15.0% 22.0% 18.0% 20.0%
Blend 2 30.0% 25.0% 13.0% 17.0% 15.0%

Blend 3 10.0% 15.0% 30.0% 31.0% 14.0%
Blend 1 Blend 2 Blend 3
Sieve Gradation Gradation Gradation

25.4mm 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
IH-43 Trial Blend 2 - Coarse Blend

0.075 0.300 2.36 19.0 25.0

Figure VI-4. Trial Blend 2 - Coarse Blend

Table VI-11. Trial Gradation for Fine Blend

#I 1/2 " 3/8 " Mfg Scr.
Stone chip chip sand sad
Blend 1 25.0% 15.0% 22.0% 18.0% 20.0%
Blend 2 30.0% 25.0% 13.0% 17.0% 15.0%
Blend 3 10.0% 15.0% 30.0% 31.0% 14.0%

Blend 1 Blend 2 ~ l3 ~ ~ d
Sieve Gradation Gradation &adation
25.4 mm
19.0 mm
12.5 mm
9.5 mm
4.75 mm
2.36 mm
1.18 mm
600 pm
300 pm
150 pm
75 pm
IH-43 Trial Blend 3 - Fine Blend

0.075 0.300 2.36 19.0 25.0

Sieve Size (mm) Raised to 0.45 Power
-
Figure VI-5. Trial Bknd 3 Fine Blend
All three of the trial blends are shown graphically in Figure VI-6. Note that all three trial
blends pass below the restricted zone. This is not a requirement. Superpave allows but
does not recommend blends that plot above the restricted zone.
IH-43 Trial Gradations

.-
0.075 0.300 2.36 19.0 25.0

Figure VI-6. IH-43 Trial Blends
Once the trial blends are selected, a preliminary determination of the blended aggregate
properties is necessary. This can be estimated mathematically from the aggregate
properties (Tables VI-3 to VI-7). Estimated values are indicated in Table VI-12.
Tablc VI-12. Estimated Aggregate Blend Properties
Property Criteria Trial Blend 1 Trial Blend 2 Trial Blend 3
Coarse Ang. 95%190%rnin. 96%/92% 95%/92% 97%/93%
Fine h g . 45% min.
ThinlElongated
Sand Equivalent 45 min.
Combined G,b n/a 2.699 2.697 2.701
Combined G,, nla 2.768 2.769 2.767
Values for coarse aggregate angularity are shown as percentage of one or more fractured
faces followed by percentage of two or more fractured faces. Based on the estimates, all
three trial blends are acceptable. When the design aggregate structure is selected, the
blend aggregate properties will need to be verified by testing.
The next step is to evaluate the trial blends by compacting specimens and determining the
volumetric properties of each trial blend. For each blend, a minimum of two specimens
will be compacted using the SGC. The trial asphalt binder content is determined for each
trial blend by estimating the effective specific gravity of the blend and using the
calc:ulations shown below. The effective specific gravity (G,) of the blend is estimated
by:
The: factor, 0.8, can be adjusted at the discretion of the designer. Absorptive aggregates
ma!y require values closer to 0.6 or 0.5. The blend calculations are shown below:
Blend 1: Gse= 2.699 + 0.8~(2.768- 2.699) = 2.754

Blend 2: Gse= 2.697 + O.gx(2.769 - 2.697) = 2.755
Blend 3: Gse= 2.701 + 0.8~(2.767- 2.701) = 2.754
The: volume of asphalt binder (Vba)absorbed into the aggregate is estimated using the
following equation:
Ps x (1- Va) 1 1
Vbn =
Pb Ps x(---)
Gsb Gse
(-+-I
Gb Gse
where Vba = volume of absorbed binder, cm3/cm3of mix

pb = percent of binder (assumed 0.05),
PS = percent of aggregate (assumed 0.95),

Gb = specific gravity of binder (assumed 1.02),
v, = volume of air voids (assumed 0.04 crnVcm3 of mix)
Blend 1: 0.95 x (1- 0.04) - 1 - -1 1 = 0.017 1cm3/cm3 of mix

vba = 0.05 0.95 X(2.699 2.754
(-+1.02
-I 2.754
Blend 2: Vba = 0.95 x (1- 0.04) x (- 1 - -1 1 = 0.018 1cm3/cm3 of mix

(=+El
1.02 2.755
2.697 2.755
Blend 3: 0.95 x (1- 0.04) 1 1 ) = 0,0165 cm3km3 of mix

Vba =
0.0s 0.95 X ( ~ - ~
(-+-)
1.02 2.754
The volume of the effective binder (Vbe)can be determined from the equation below:
where S, = the nominal maximum sieve size of the aggregate blend (in inches)
Blend 1-3: Vbe= 0.08 1 - 0.0293 1x [ln(0.75)] = 0.089 cm3/cm3of mix
Finally, the initial trial asphalt binder (Phi) content is calculated from the following
equation:
CTb x (Vbe + Vba)

Pbi = x 100
(Gbx (Vbe + Vba)) + Ws
where Phi = percent (by weight of mix) of binder

W, = weight of aggregate, grams
(-+-)
G b Gse
Blend 1: Ws = 0.95 x (1 - 0.04) = 2.315

,0.05 0.95 .
Pbi =
(0.0g9 0.0171)
+
x 100 = 4.46% (by mass of mix)
(1.02 x (0.089 + 0.0171)) + 2.3 15
Blend 2: 0.95 x (1- 0.04)

Ws = = 2.316
,0.05 0.95 ,
pbl = (Oaog9 O.Olgl)

+
(1.02 x (0.089 + 0.0171))+ 2.316
Blend 3: 0.95 x (1 - 0.04)

Ws = = 2.315
,0.05 0.95 .
Pbi = (Oeog9 0.0165)

+

(1.02 x (0.089 + 0.0171)) + 2.315
Next, a minimum of two specimens for each trial blend are compacted using the SGC.
Two specimens are also prepared for determination of the mixture's maximum theoretical
specific gravity (G,,). An aggregate weight of 4500 grams is usually sufficient for the
compacted specimens. An aggregate weight of 2000 grams is usually sufficient for the
specimens used to determine maximum theoretical specific gravity (G,,). AASHTO
T 209 should be consulted to determine the minimum sample size required for various
mixtures.
VI. Volumetric Mix Desinn
Specimens are mixed at the appropriate mixing temperature, which is 165C to 172C for
the selected PG 58-34 binder. The specimens are then short-term aged by placing the
loose mix in a flat pan, in a forced draft oven at 135OC, for 4 hours. Next, the specimens
are brought to compaction temperature range (151C to 157C) by placing them in
another oven for a short time (generally less than 30 minutes). Finally, the specimens are
then removed and either compacted or allowed to cool loose (for G, determination).
The number of gyrations used for compaction is determined based on the design high air
temperature of the paving location (35C) and the traffic level. Table VI- 13 indicates the
number of gyrations required.
Table VI-13. Gyratory Compactive Effort

Design iverage Design Hi h Air Temperatur
ESALs 39" - 40C 41" - 42C 43" - 44C
(millions) Nini Ndes Ntnt Ndes Nmax Nini Nde, Nmax Nini Ndes Nmax
c 0.3 7 74 114 7 78 121 7 82 127

0.3 - 1 7 83 129 7 88 138 8 93 146
1-3 8 95 150 8 100 158 8 105 167
3 - 10 8 106 169 8 113 181 9 119 192
-
10 30 9 121 195 9 128 208 9 135 220
30 - 100 9 139 228 9 146 240 10 153 253
> 100 10 158 262 10 165 275 10 172 288
From Table VI-13, the number of gyrations for initial compaction, design
compaction, and maximum compaction are determined:
= 8 gyrations
Ninitial
NdeSign
= 109 gyrations
= 174 gyrations
Nmaximum
The equations used to develop the information in Table VI-13, which describes the
relationship among Ndesign,
Ninitial,
and Nmaximum
are shown below:
VI, Volumetric Mix Design
Each specimen will be compacted to the maximum number of gyrations, with specimen
height data collected during the compaction process. This is illustrated for Trial Blend 1
in Table VI-14. SGC compaction data reduction is accomplished as follows.
During compaction, the height of the specimen is continuously monitored. Knowing the
mass of the mix, the fixed diameter of the mold (150 mm), and the measured height at
any gyration, the specimen specific gravity can be estimated [Gmb(est)in Table VI-141 at
any gyration throughout the compaction process. This is accomplished by dividing the
mass of the specimen by volume of the specimen, which is represented by the volume of
a smooth-sided cylinder of known diameter and (measured) height.
After compaction is complete, the specimen is extruded from the mold and allowed to
cool. Next, the measured bulk specific gravity [Gmb(meas)in Table IV- 141 of the
specimen is determined by AASHTO T166. The G, of each blend is also determined
by AASHTO T209 [G,,(meas) in Table VI-141.
A comparison of the specimen's estimated bulk specific gravity [Gmb(est)]and measured

bulk specific gravity [Gmb(meas)]at itm, shows a difference between these two
parameters. In Table VI-14, these two values are 2.436 and 2.489, respectively. The
assumption that was used to estimate the bulk specific gravity was that during
compaction, the volume of the specimen could be represented by a smooth-sided cylinder,
which of course, it is not. The actual volume of the specimen is slightly smaller due to
the presence of surface voids surrounding the perimeter of the specimen. Thus, the
estimated bulk specific gravity of the specimen at any given gyration must be corrected
by a factor that is the ratio of the measured to estimated bulk specific gravity of the
specimen at N,,. In Table VI-14, this ratio is 2.48912.436 or 1.022. This corrective step
is indicated in Table VI-14 in the column labeled Gmb(cOrr).In this step, each value in
the column labeled Gmb(est)is multiplied by the correction factor, 1.022, to arrive at the
values in the column labeled Gmb(cOrr).
Coarser aggregate mixtures, or mixtures lean in asphalt binder, tend to have larger
differences between estimated and measured bulk specific gravity at NmaXimum.
Finer
aggregate mixtures, or mixtures rich in asphalt binder, tend to have smaller differences
between these two parameters. That is because fine, high asphalt content mixtures more
closely approximate the "smooth-sided cylinder" assumption.
The final step is to report %Gmmfor each specimen. This is computed by dividing the
corrected bulk specific gravity of the specimen by the measured value for G,. The
average %G,, for the duplicate specimens is also reported. The average %G,, is used
as the basis for comparison among the trial mixtures.
The SGC data reduction for the three trial blends is shown in Tables VI-14, VI-15, and
VI-16. The most important points of comparison are %Gm at Ninitial,
Ndesign,
and
which are highlighted in these tables. Figures VI-7 to VI-9 illustrate the
Nmaximm
compaction plots for data generated in these tables. The figures show %Gmmversus the
logarithm of the number of gyrations.
Table VI-14. Densification Data for Trial Blend 1

3,, (meas) = 2.563
I Specimen 2 AVG
Gyrations Ht, mm Gmb(est) Gmb(corr) %Gm Ht, mm Gmb(est) Gmb(con) %G,,
5 130.3 2.154 2.209 86.2
8
10
15
20
30
40
50
60
80
100
109
125
150
174
Gmb(meas)
IH-43, 19.0 mm Nominal, 4.4% AC, Trial Blend 1
I Specimen 2
A,,,
75 1 1
1 10 loo 1000
Number of Gyrations
Figure VI-7. Densification Curves for Trial Blend 1


G,,,,,,
(meas) = 2.565
I Specimen 1 Specimen 2
-
AVG
Gyrations Ht, mrn Gmb(est) Gmb(corr) %G,,
5 131.7 2.090 2.158 84.2
Ht, mm G,b(est) G,b (corr) %G,,
132.3 2.098 2.159
-
%G,
84.2
8
10
15 88.1
20 89.2
30 90.7
40 91.9
50 92.7
60 93.4
80 94.5
100 95.4
109
125
150
174
Gmb(meas)
75 1 1
1 10 100 1000
Number of Gyrations
I

3- (meas) = 2.568
1 Swecimen 1 Specimen 2 AVG
Gyrations Kt, mm Gmb(est) Gmb(corr) %G,,
5 129.5 2.136 2.188 85.2
8
10
15
20
30
40
50
60
80
100
109
125
150
174
GmbWas)

1
Number of Gyrations
The average %Gmmis determined for Ni,itial (8 gyrations), Ndesign (109 gyrations), and
NmaXi,,, (174 gyrations) for each trial blend. This data is extracted directly from Tables
VI- 14 to VI-16. Table VI- 17 indicates these values for Trial Blends 1,2, and 3.
Table VI-17. Determination of %Gm at Nini,Ndesrand N,, for Trial Blends
Trial Blend % G- @ Nini % Gmm@ Ndes % Gmm@ N-

1 87.1 96.2 97.6
2 85.6 95.7 97 -4
3 86.3 95.2 96.5
The percent of air voids and voids in the mineral aggregate (VMA) are determined at
Nksign. The percent air voids is calculated as follows:
%Air Voids = 100 - %Gmm@ Ndesign
Blend 1: %Air Voids = 100 - 96.2 = 3.8%

Blend 2: %Air Voids = 100 - 95.7 = 4.3%
Blend 3: %Air Voids = 100 - 95.2 = 4.8%
The percent voids in the mineral aggregate is calculated as follows:
%Gmm @ Ndesx G m m x Ps
%VMA= 100 - ( 1
Gsb
96.2% x 2.563 x 0.956) - 12m74

Blend 1: %VMA = 100 - ( -
2.699
95.7% x 2.565 x 0.956

Blend 2: %VMA = 100 - ( ) = 13.0%
2.697
95.2% x 2.568 x 0.956

Blend 3: %VMA = 100-( ) = 13.5%
2.701
Table VI-18. Compaction Summary of Trial Blends
Blend %AC %Gm, @ N=8 %G,, @ %G,, @ %Air %VMA

N=174 N= 109 Voids
1 4.4% 87.1% 97.6% 96.2% 3.8% 12.7%
2 4.4% 85.6% 97.4% 95.7% 4.3% 13.0%
3 4.4% 86.3% 96.5% 95.2% 4.8% 13.5%
Table VI-18 indicates the compaction summary of the trial blends. The central premise in
Superpave Level 1 mix design is that the correct amount of asphalt binder is used in each
trial blend so that each blend achieves exactly 96% of G, or 4% air void content at
Ndesign.Clearly, this did not happen for any of the three IH-43 trial blends. Because the
trial blends exhibit different air void contents at Ndesign,
their other volumetric and
compaction properties cannot be properly compared. For example, Trial Blend 1
contained slightly too much asphalt to achieve 4 % air voids at Ndesign.Instead, it had
only 3.8% air voids. The VMA of Trial Blend 1 is also too low. The designer must ask
the question, "If I had used less asphalt in Trial Blend 1 to achieve 4% air voids at Ndesignr
would the VMA and other required properties improve to acceptable levels?'
Providing an answer to this question is an important step in Level 1 mix design. To

answer this question, an estimated asphalt binder content to achieve 4% air voids (96%
G, at Ndesign)
is determined for each trial blend using the following empirical formula.
= estimated percent binder

where Pb,estimated
Phi = initial (trial) percent binder
V, = percent air voids at Ndesign
Blend 1: Pb,estimated
= 4.4 - ( 0 . 4 ~ ( 4
- 3.8)) = 4.3%
Blend 2: Pb,estimated
= 4.4 - ( 0 . 4 ~ ( -44.3)) = 4.5%
Blend 3: = 4.4 - ( 0 . 4 ~ ( -44.8)) = 4.7%

Pb,estimated
The volumetric (VMA and VFA) and mixture compaction properties are then estimated at
this asphalt binder content using the equations that follow. These steps are solely aimed
at answering the question, "What would have been the trial blend properties if I had used
the right amount of asphalt to achieve 4% air voids at Ndesign?' It is by these steps that a
proper comparison among trial blends can be accomplished.
For VMA:
where: %VMAinitial = %VMAfrom trial asphalt binder content

C = constant (either 0.1 or 0.2)
Note: C = 0.1 if V, is less than 4.0%
C = 0.2 if V, is greater than 4.0%
Blend 1 : = 12.7 + (O.lx(4.0- 3.8)) = 12.7%

%VMLstimated
Blend 2: = 13.0 + ( 0 . 2 ~ ( 4 .-04.3)) = 13.0%
%VMAestimakd
Blend 3: = 13.5 + ( 0 . 2 ~ ( 4 .-04.8)) = 13.3%
%VMLStimakd
For VFA:
(12.7 - 4.0)
Blend 1: %VFAestirnated = 100% X = 68.5%
12.7
(13.0 - 4.0)
Blend 2: %VFAestirnated = 100%X = 69.2%
13.0
Blend 3:
Blend 1: estimated @ Nini = 87.1 - (4.0 - 3.8) = 86.9%

%Gmm
Blend 2: @ Nini = 85.6 - (4.0 - 4.3) = 85.9%
%Gmmestimated
Blend 3: @ Nini= 86.3 - (4.0 - 4.8) = 87.1%
%Gmmestimatea
Blend 1: %Gmm = 97.6 - (4.0 - 3.8) = 97.4%

estimated @ Nmax
Blend 2: %Gmm = 97.4 - (4.0 - 4.3) = 97.7%

estimated @ Nmax
Blend 3: %Gmm,,tim,d@ N, = 96.5 - (4.0 - 4.8) = 97.3%
Tables VI- 19 and VI-20 indicate the estimated volumetric and mixture compaction
properties for the trial blends at the asphalt binder content that should result in 4.0% air
voids at Ndesign.
Table VI-19. Estimated Mixture Volumetric Properties @ Naesig,
Blend Trial %AC Est. %AC %AirVoids %VMA %VFA

1 4.4% 4.3% 4.0% 12.7% 68.5%
2 4.4% 4.5% 4.0% 13.0% 69.2%
3 4.4% 4.7% 4.0% 13.3% 70.1%
Estimated properties are compared against the mixture criteria. For the design traffic and
nominal maximum particle size, the volumetric and densification criteria are as follows:
% Air Voids 4.0%

% VMA 13.0% (19.0 rnrn nominal mixture)
% VFA 65% - 75% (10-30 x lo7ESALs)
% Gmm @ Ninitial less than 89%
%G,, @ Nmaximumless than 98%
Table VI-20. Estimated Mixture Compaction Properties
Blend Trial %AC Est. %AC %G,, @ N=8 %G,, @ N=174

1 4.4% 4.3% 86.9% 97.4%
Finally, there is a required range on the dust proportion. This criteria is constant for all
levels of traffic. It is calculated as the percent by mass of the material passing the 0.075
rnrn sieve (by wet sieve analysis) divided by the effective asphalt binder content
(expressed as percent by mass of mix). The effective asphalt binder content is calculated
as follows:
Gse - Gsb
Pbe,estimated = -(Ps x Gb) X ( ) + Pb,estimated
Gse x Gsb
2.754 - 2.699
Blend 1: Pbe,estimated = -(95,7 X 1.02) X ( ) + 4.3 = 3.6%
2.754 x 2.699
2.755 - 2.697
Blend 2: Pbe,estimated= 495.5 X 1.02)X ( ) + 4.5 = 3.7%
2.755 x 2.697
2.754 - 2.701
Blend 3 : Pbe,estimated = 495.3 X 1.02) X ( ) + 4.7 = 4.0%
2,754 x 2.701
Dust Proportion is calculated as follows:
p.075
DP =
P b e , estimated
Blend 1:
Blend 2:
Blend 3:
The dust proportion must be between 0.6 and 1.2. Table VI-2 1 indicates the results.
Table VI-21. Dust Proportion of Trial Blends
I Blend I Dust Proportion I Criterion I

Trial Blend 1 0.86
Trial Blend 2 0.78 0.6 - 1.2
Trial Blend 3 0.88
After establishing all the estimated mixture properties, the designer can observe the
values for the trial blends and decide if one or more are acceptable, or if further trial
blends need to be evaluated.
Blend 1 is unacceptable based on a failure to meet the minimum VMA criteria. Blend 2
is acceptable, but the VMA is at the minimum. Blend 3 has acceptable VMA as well as
meeting the criteria for VFA, dust proportion, and the densification criteria. From this
data, Trial Blend 3 is selected as the design aggregate structure.
SELECT DESIGN ASPHALT BINDER CONTENT
Once the design aggregate structure is selected, Trial Blend 3 in this case, specimens are
compacted at varying asphalt binder contents. The mixture properties are then evaluated
to determine a design asphalt binder content.
A minimum of two specimens are compacted at each of the following asphalt contents:
estimated binder content

estimated binder content k 0.5%, and
estimated binder content + 1.0%.
For Trial Blend 3, the binder contents for the mix design are 4.2%, 4.7%, 5.2%, and
5.7%. Four asphalt binder contents are a minimum in Superpave Level 1 analyses.
A minimum of two specimens are also prepared for determination of maximum

theoretical specific gravity at the estimated binder content. Specimens are prepared and
tested in the same manner as the specimens from the "Select Design Aggregate Structure"
section.
Tables VI-22 to VI-25 indicate the test results in tabular form for each trial asphalt binder
content. Figures VI- 10 to VI- 13 illustrate the densification curves for each trial asphalt
binder content. Figure VI-14 illustrates the average densification curves for each trial
asphalt binder content.
Table VI-22. Densification Data for Blend 3,4.2% Asphalt Binder

;
,,(meas) = 2.582
Specimen 1 Specimen 2 AVG
Gyrations 3,mm Gmb(est) Gmb(corr) %Gm Ht, mm Gmb(est) G* (corr) %G,,
5 131.3 2.116 2.167 83.9 131.0 2.136 2.186 84.7
8 128.8 2.172 2.223 86.1
10 127.4 2.196 2.248 87.1
15 125.5 2.229 2.281 88.4
20 124.2 2.253 2.306 89.3
30 122.4 2.286 2.340 90.6
40 121.1 2.310 2.364 91.6
50 120.1 2.330 2.385 92.4
60 119.4 2.342 2.397 92.9
80 118.3 2.365 2.421 93.8
100 117.4 2.383 2.439 94.5
109 117.1 2.389 2.445 94.7
125 116.6 2.400 2.456 95.1
150 115.9 2.413 2.470 95.7
174 115.4 2.424 2.481 96.1
Gmb(meas) 2.481
IH-43, 19.0 mm Nominal, Blend 3,4.2% AC
I Average 1
10 100
Number of Gyrations
Figure VI-10. Densification Curves for Blend 3,4.2% Asphalt Binder

3- (meas) = 2.562
Specimen 2 AVG
Gyrations Ht, mm Gmb(est) Gmb(corr) %Gmm Ht, mm Gmb(est) G,,b (corr) %G,,
5 130.4 2.152 2.199 85.8
8
10
15
20
30
40
50
60
80
100
109
125
150
174
Gmb(meas1
IH-43, 19.0 mm Nominal, Blend 3,4.7% AC
I -Specimen 2 (
I Average I
75 1
1 10 100 1000
Number of Gyrations
VI. Volumetric Mh Design

f,,(meas) = 2.542
I Specimen 1 Specimen 2 AVG
Gyrations Ht, mm Gmb(est) Gmb(corr) %G,, It, mm Gmb(est) Gmb(corr) %Gm %Gmm
5 132.0 2.148 2.187 86.0 132.6 2.142 2.183 85.8 85.9
8
10
15
20
30
40
50
60
80
100
109
125
150
174
Gmb(meas)
I IH-43, 19.0 mm Nominal, Blend 3,5.2% AC
75
1 10 100 1000
Number of Gyrations

3- (meas) = 2.523
Specimen 1 Saecimen 2 AVG
Gyrations Ht, mm Gmb(est) G,b(corr) %Gm Ht, mm G,b(est) Gb, (corr) %G,,
5 130.4 2.170 2.204 87.4
8 128.6 2.201 2.236 88.7
10 127.4 2.222 2.257 89.5
15 125.4 2.256 2.292 90.8
20 124.0 2.282 2.318 91.9
30 122.4 2.313 2.349 93.1
40 120.5 2.349 2.386 94.6
50 119.4 2.371 2.408 95.5
60 118.9 2.381 2.419 95.9
80 117.6 2.406 2.444 96.9
100 116.7 2.426 2.464 97.7
109 116.2 2.453 2.474 98.1
125 115.4 2.452 2.491 98.7
150 114.9 2.463 2.502 99.2
174 114.3 2.476 2.515 99.7
Gmb(meas) 2.515
IH-43, 19.0 mm Nominal, Blend 3, 5.7%AC
10 100
Number of Gyrations
-
IH-43, 19.0 mm Nominal, Blend 3
10 100
Number of Gyrations
Figure VI-14. Average Densification Curves for Blend 3, Varying
Asphalt Binder Content
Mixture properties are evaluated for the selected blend at the different asphalt binder
contents, by using the densification data at Ninitid
(8 gyrations), Ndesign
(109 gyrations),
and Nmmimum
(174 gyrations). Tables VI-26 and VI-27 indicate the response of the
mixture's compaction and volumetric properties with varying asphalt binder contents.
Table VI-26. Summary of Blend 3 - Mix Compaction Properties

-
Table VI-27. Summary of Blend 3 Mix Volumetric Properties at Ndesign
%AC %Air Voids %VMA %VFA Density (kg/m3)

4.2% 5.5% 13.4% 59.3% 244 1
4.7% 3.9% 13.2% 70.1% 246 1
5.2% 3.O% 13.4% 77.9% 2467
5.7% 1.9% 13.6% 86.2% 2476
The volumetric properties are calculated at the design number of gyrations (Ndesign)
for
each trial asphalt binder content. From these data points, the designer can generate
graphs of air voids, VMA, and VFA versus asphalt binder content.
The design asphalt binder content is established at 4.0% air voids. In this example, the
design asphalt binder content is 4.7% - the value that corresponds to 4.0% air voids at
= 109 gyrations. All other mixture properties are checked at the design asphalt
Ndesign
binder content to verify that they meet criteria. The design values for the 19.0 mm
nominal mixture (Trial Blend 3) are indicated in Table VI-28.
IH-43 Intermediate Course, Blend 3
I % Asphalt Binder 1
Figure VI-15. Air Voids versus Asphalt Binder Content
14.5
14.3
14.1
13.9
13.7
13.5
13.3
13.1
12.9
12.5
12.7
3.7
1 4.2 4.7 5.2 5.7 6.2
% Asphalt Binder
Figure VI-16. VMA versus Asphalt Binder Content
3.7
c 4.2 4.7 5.2 5.7 6.2
% Asphalt Binder
Figure VI-17. VFA versus Asphalt Binder Content

Table VI-28. Design Mixture Properties at 4.7% Binder Content
I Mix Pro~ertv I Result I Criteria 1

% Air Voids 4.0% 4.0%
%VMA 13.2% 13.0%min.
%VFA 70.0% 65% - 75%
Dust Pro~ortion 0.88 0.6 - 1.2
%Gmm@ Ninitial=8 87.1% less than 89%
%Gm, @ Nmwimum= 174 97.5% less than 98%
EVALUATE MOISTURE SENSITIVITY
The final step in the Level 1 mix design process is to evaluate the moisture sensitivity of
the design mixture. This step is accomplished by performing AASHTO T 283 testing on
the design aggregate blend at the design asphalt binder content. Specimens are
compacted to approximately 7% air voids. One subset of three specimens are considered
control specimens. The other subset of three specimens is the conditioned subset. The
conditioned subset is subjected to vacuum saturation followed by an optional freeze
cycle, followed by a 24 hour thaw cycle at 60' C. All specimens are tested to determine
their indirect tensile strengths. The moisture sensitivity is determined as a ratio of the
tensile strengths of the conditioned subset divided by the tensile strengths of the control
subset. Table VI-29 indicates the moisture sensitivity data for the mixture at the design
asphalt binder content.
VI. Volumetric Mix: Design
Table VI-29. Moisture Sensitivity Data for Blend 3 at 4.7% Design

ialt B lder Content
SAMPLE
Diameter, rnrn D
Thickness, mm t
Dry mass, g A
SSD mass, g B
Mass in Water, g C
Volume, cc (B-C) E
Bulk Sp Gravity (AIE) F
Max Sp Gravity G
% Air Voids(lOO(G-F)/G) H
Vol Air Voids (HEI100) I
l
Saturated
SSD mass, g B'
l
Mass in water, g C'
Volume, cc (B1-C') E'
Vol Abs Water, cc (B1-A) J'
% Saturation (lOOJ1/I)
% Swell (lOO(E1-E)/E)
Conditioned
.
Thickness, mm t"
SSD mass, g B"
Mass in water, g C"
Volume, cc (B1'-C") E"
Vol Abs Water, cc (B"-A) J"
% Saturation (100J"A)
ky
% Swell (100(E"-E)E)
Load, N
Str. (2OOOP/(tDn)) ii
Wet Str. (2000P"/(t"Dn;)) St,
Average Dry Strength (kPa)
Average Wet Strength (kPa)
%TSR
The minimum criteria for tensile strength ratio 80%. The design blend (82.6%) exceeded
the minimum requirement. The Superpave Level 1 Mix Design is now complete for the
intermediate mixture for M-43. Additional performance prediction testing is required as
described under the Level 3 testing process.
VII.
PERFORMANCE TESTING (LEVELS 2 AND 3)
INTRODUCTION
In the Superpave mixture design and analysis system, performance tests are used only in
situations involving moderate to high traffic. This means that they are required only for
Levels 2 and 3 mixture designs. Performance testing utilizes new equipment and
procedures to ensure that Superpave mixtures exhibit acceptable amounts of the distress
types that were considered by SHRP researchers: permanent deformation, fatigue
cracking, and low temperature cracking. Two performance test devices were developed:
the Superpave Shear Tester (SST) and the Indirect Tensile Tester (IDT). The extent of
use of performance testing for Levels 2 and 3 mix design are shown in Table VII- 1 for a
new two layer HMA system, which is the most new layers considered by Superpave.
Table VII-1, Performance Tests, Levels 2 and 3 (New Construction)
Design Performance Distress Mode

I I
Level Permanent ~eformation' Fatigue Cracking

Simple shear test at constant Simple shear test at constant Indirect tensile creep compl
height at Tefl(PD). height at Teff(FC). at 0, -lo0, -20' C.
2 Frequency sweep test at Frequency sweep test at Indirect tensile strength at
constant height at TEff(PD). constant height at Teff(FC). -10" C.
Indirect tensile strength at Binder creep stiffness (S)
Teff(FC). and creep rate (m).
Uniaxial strain test at 4O,

20, 40" C. Indirect, tensile creep
compliance and strength at
40' C. - lo0,4O, and 20" C. 00, -to0, -zoo C.
Simple shear test at constant
height at 4O, 20, 40" C.
To check r tertiary flow, Level 2 and 3 require repeated shear test at constant stress ratio at T,
VZI. Perfomance Testing
If an overlay is being designed, Superpave does not attempt to predict fatigue cracking or
low temperature cracking. Only permanent deformation is considered. Consequently, the
extent of use of performance testing for asphalt mixtures used for overlays is shown in
Table VII-2,
Table VII-2. Performance Tests, Levels 2 and 3 (Overlay Construction)
Design Level Permanent ~eformation'

Simple shear test at constant height at T,ff(PD)
2
Frequency sweep test at constant height at T,dPD)
I I Frequency sweep test at constant height at 4O, 20, and 40C

I
I I Uniaxial strain test at 4', 20, and 40' C
I
II
3
I Volumetric test at 4O, 20, and 40 C
II
II Simple shear test at constant height at 4O, 20, and 40" C
' To check for tertiary flow, Level 2 and 3 require repeated shear test at constant stress
I ratio at T, I
PERFORMANCE MODELS
While much attention was focused in SHRP on the new test equipment and testing
protocols, a key component of performance testing are performance models. These are
prediction algorithms that accept performance test results and output predicted pavement
performance. The models account not only for the new asphalt mixture being designed,
but also the characteristics of the in-place pavement. The use of performance testing and
performance prediction models represents an important new tool for engineers in
designing and managing pavements.
Performance prediction is accomplished by the Superpave software using four

components:
material property model,

environmental effects model,
pavement response model, and
pavement distress model.
VII. Per$ormance Testing
Performance test results (i.e., SST and IDT) are used as input to the material property
model to determine non-linear elastic, viscoelastic, plastic, and fracture properties. The
environmental effects model calculates pavement temperature as a function of depth and
material thermal characteristics. The pavement response model uses a two-dimensional,
axisymmetric finite element approach to predict stresses and strains within the layered
system. It uses output from the material property and environmental effects models to
predict these responses of the pavement system to traffic and environmental loads.
Output from the pavement response and material property models are used by the distress
models to estimate rutting and fatigue and low temperature cracking. Figure IV-1 shows
the performance prediction approach of Superpave.
Project Data: layer info, traffic, climate
Rutting
Environmental
.Fatigue Cracking
Effects Model
Thermal
I Material Property Model

I II
1 Performance Test Results
I
Figure VII-1. Superpave Performance Prediction System
TEST PARAMETERS
Test Temperatures
Level 3 testing provides a more reliable prediction of pavement performance because it

involves performance testing over a wider range of temperatures (i.e., 4*, 20, and 40"C).
This allows use of the environmental effects model to more accurately predict pavement
performance.
Level 2 testing involves performing tests at an effective temperature (Teff). While a less
accurate performance prediction results in Level 2 testing, the testing is greatly
VII. Performance Testinn
simplified. Because permanent deformation and fatigue cracking occur at different

temperatures, two effective temperatures are used: T,ff(PD) and Teff(FC). Teff(PD)is the
single temperature at which the predicted permanent deformation would be identical to
that predicted by a multiple temperature analysis. Teff(FC)is the single temperature at
which an equal amount of fatigue damage would occur to that measured by considering
each season separately throughout the year. Both temperatures are computed by
Superpave software and are a function of project mean annual air temperature, layer
depth, and user selected reliability.
While tertiary flow is a permanent deformation type of distress, it is treated separately by

Superpave. Tertiary flow occurs when an asphalt mixture densifies to a very low air void
content, normally less than about two to three percent air voids. In this condition, the
mixture exhibits extreme plastic flow with very few load applications. Figure VII-2
illustrates the concept of tertiary flow.
Log Plastic Strain
ary Flow
-- -
Log Number of Load Applictions
Figure VII-2. Tertiary Rutting
The tertiary flow analysis using the repeated shear test at constant stress ratio is
conducted at a control temperature (T,). The control temperature is computed by
Superpave software and depends on project weather and traffic conditions.
Asphalt Binder Contents
Superpave performance testing requires that specimens be tested at multiple asphalt

binder contents. For tests concerned with permanent deformation, fatigue cracking, and
VII. Performance Testing
low temperature cracking, these consist of binder contents that result in three, four, and
six percent air voids at Ndesign.However, performance test specimens are fabricated using
less gyrations than Ndesignin order to achieve test specimens containing approximately
seven percent air voids.
For the tertiary flow analysis, only the binder content resulting in three percent air voids
is used. Two test specimens are required for a given test condition (i.e., test temperature
and binder content). They are fabricated to achieve three percent air voids. Appendix F
shows a visual representation of specimen requirements for Levels 1,2, and 3 testing.
SUPERPAVE SHEAR TESTER
The SST (Figure VII-3) is a closed-loop feedback, servo hydraulic system that consists of
four major components: the testing apparatus, the test control unit and data acquisition
system, the environmental control chamber, and the hydraulic system.
environmental n
control chamber
\
control and testing
data acquisition apparatus
hydraulic system
Figure VII-3. Superpave Shear Tester Components
Testing Apparatus
The testing apparatus includes a reaction frame and shear table. It also serves to house
the various coniponents that are driven by other system components such as
temperaturelpressure control, hydraulic actuators, and input and output transducers. The
reaction frame is extremely rigid so that precise specimen displacement measurements
can 'be achieved without worrying about displacements from frame compliance. The
shear table holds specimens during testing and can be actuated to impart shear loads.
Test Control Unit
The test control unit consists of the system hardware and software. The hardware
interfaces with the testing apparatus through input and output transducers, and it consists
of controllers, signal conditioners, and a computer and its peripherals. The software
consists of the algorithms required to control the testing apparatus and to acquire data
during a test.
Linear variable differential transducers (LVDTs) are affixed to specimens and measure
the response of specimens to applied testing loads. The LVDTs make it possible for the
system to operate in a closed loop feedback mode, which means that LVDT signals are
used to control applied testing loads.
Environmental Control Unit
The environmental control unit is required to control the temperature and air pressure
inside the testing chamber at a constant level. The unit is capable of providing
temperatures within a wide range from l o to 80' C. Air pressure and the rate of pressure
change within the chamber is precisely controlled. Air pressure is normally applied at a
rate of 70 kPa per second up to a maximum value of 840 kPa. This is achieved by storing
compressed air in separate storage tanks that can be emptied into the testing chamber at
the required rate. Air pressure provides specimen confinement for two of the six tests.
Hydraulic System
The hydraulic system provides the required force to load specimens for different testing
conditions. A hydraulic motor powers two actuators, each with a capacity of
approximately 32 MV. The vertical actuator applies an axial force to test specimens. The
horizontal actuator drives the shear table, which imparts shear loads to the specimen.
VII. Performance Testinn
Specimen Preparation and Instrumentation
The first step in specimen preparation is to trim test specimens to a thickness of 50 mm.
For the three tests that require no confining pressure, the specimen is glued between two
platens.
A gluing device (Figure VII-4) is used to squeeze the specimen between the platens while
the glue cures. An epoxy-type glue such as Devcon Plastic Steel is employed for this
purpose. The gluing device rigidly holds the platens and specimen to ensure that the
platen faces are parallel.
After the glue has cured, four screws are affixed to the side of the specimen using a gap
filling variety of cyanoacrylate glue. These screws are used to affix the bracket that holds
the horizontal LVDT (Figures VII-5 and VII-6). Axial LVDTs are affixed to the platens.
platens
Figure VII-4. SST Gluing Device

VII. Peq4ormance Testing
screws affixed
to specimens
scre& affixed
to specimens
Figure VIM. Specimen Instrumentation for Unconfined SST Tests
(Side View)
screws
affixed to
specimen
axial
LVDT
horizontal
LVDT
Figure VII-6. Specimen Instrumentation for Unconfined SST Tests

(Front View)
A different specimen configuration is used for confined tests. Test specimens are still
placed between platens, however, no glue is used. A rubber membrane surrounds the
specimen. A radkd LVDT is affixed by a collar that surrounds the perimeter of the
specimen (Figure VII-7). Axial LVDTs are affixed to the platens.
VII. Pe$ormance Testing
radial LVDT rubber membrane
axial
LVDT
Figure VII-7. Specimen Instrumentation for Confined SST Tests

(Front View)
Test Procedures
Six tests are performed using the SST:
volumetric test,
uniaxial strain test,
repeated shear test at constant stress ratio,
repeated shear test at constant height (not required by Superpave),
simple shear test at constant height, and
frequency sweep test at constant height.
The volumetric and uniaxial strain tests use confining pressure in their protocol. These
two tests are performed only for a Level 3 mixture design. Levels 2 and 3 design use
repeated shear at constant stress ratio, simple shear at constant height, and frequency
sweep at constant height tests. The repeated shear test at constant height is a stand-alone
test that can be used for rut depth estimation and it is not a part of the Superpave mixture
design and analysis system. A brief description of each test follows.
Volumetric Test
The volumetric test is one of two tests that uses confining pressure. It is performed at
three temperatures and pressures as indicated below.
Table VII-3. Volumetric Test Parameters
Temperature, "C Pressure, kPa

4 830
20 690
40 550
The test is performed by increasing the confining stress at a rate of 70 kPa per second up
to the values shown and measuring the circumferential strain by means of the radial
LVDT, Figure VII-8 shows the change in confining pressure versus time during the
volumetric test at 20 C.
Air Pres, kPa
30 sec
0 10 20 30 40 50 60 70 80
Time, sec
Figure VII-8. Co&ning Pressure in Volumetric Test at 20" C
Uniaxial Strain Test
The uniaxial strain test also uses confining pressure. In this test, axial stress is applied to
the test specimen and the specimen tries to increase its circumference. The radial LVDT
senses this change in circumference and air pressure is applied so that the circumference
remains constant. This approach uses the signal from the radial LVDT as feedback for
the purpose of applying confining pressure to prevent radial deformation. Three axid
stress levels are used depending on the test temperature as shown in Table VII-4.
VII. PerJomance Testing
Table VII-4. Uniaxial Strain Test Parameters
Temperature, "C Axial Stress, kPa

4 655
20 550
40 345
Confining pressure is measured throughout the test. Axial deformation is measured on

both sides of the specimen by the vertical LVDTs. Axial load is also measured. Radial
deformation is also measured although it should be relatively small. Figure VII-9 shows
the application of axial stress during the test.
I Axial Stress, kPa
Confining Stress, kPa

variable magnitude
circumference constant
10 20
Time, sec
Figure VII-9. Stress Applications in Uniaxial Strain Test at 20" C
Repeated Shear Test at Constant Stress Ratio
The repeated shear test at constant stress ratio is performed for either Level 2 or 3 mix
design. It is a screening test to delineate an asphalt mixture that is subject to tertiary
rutting. This form of rutting occurs at low air void contents and is the result of bulk
mixture instability.
VIZ. Pe$ormance Testing
In this test, repeated synchronized haversine shear and axial load pulses are applied to the
specimen. A load cycle consists of 0.7-second, which is comprised of 0.1-second load
application followed by 0.6-second rest period. Test specimens are subjected to a varying
number of load cycles in the range from 5000 to 120,000, depending on the traffic level
and climate conditions or until accumulated permanent strain reaches five percent. The
ratio of axial to shear stress is maintained constant in the range from 1.2 to 1.5. The
magnitude of stresses are selected to simulate actual in-place stresses that will be
encountered by the mixture. Suggested stress values are shown in Table VII-5.
Table VII-5. Suggested Stress Values for Repeated Shear Test at

Constant Stress Ratio
Asphalt Content
High Medium Low
Base I
Shear Stress, Axial Stress, I
Shear Stress, Axial Stress, I
Shear Stress, Axial Stress,
Condition kPa Pa kPa kPa kPa kPa
Weak 84 119 63 98 49 56
I
I
I
Strong 98 175 84 105 56 91
In Table VII-5, a weak base is considered any unstabilized granular material while a
strong base is considered an existing pavement or stabilized layer. The test temperature
used is called the control temperature (T,) for permanent deformation. It is computed by
Superpave as a function of the project traffic conditions and climate. The test is typically
performed at high asphalt contents corresponding to three percent air voids, which is the
extreme condition for tertiary rutting.
During the test axial and shear loads and deformations are measured and recorded.
Figure VII-10 shows typical stress pulses in the test.
1 Shear y s s , kPa
I Axial S F s s , kPa variable milgni tide

to keep axiaVshear
r\ stress between 1.2-1.5
I Time, sec
Figure VII-10. Stress Pulses in Repeated Shear Test

at Constant Stress Ratio
Repeated Shear Test at Constant Height
This test is performed as an option to Levels 2 or 3 design to estimate rut depth and is not
a required by Superpave. A haversine shear load is applied to achieve a controlled shear
stress level of 68 kPa. When the repeated shear load is applied, the test specimen seeks to
dilate. The signal from the axial LVDT is used as feedback by the vertical actuator to
apply sufficient axial load to keep the specimen from dilating.
A load cycle consists of 0.7-second, which is comprised of 0. l-second shear load

application followed by 0.6-second rest period. Test specimens are subjected to 5000
load cycles or until the permanent shear strain reaches five percent.
The test temperature used is T, which is the seven-day maximum pavement

temperature at 50 mm depth. During the test, axial and shear loads and deformations are
measured and recorded. Figure VII- 11 shows typical stress pulses in the test.
VZZ. Pe$ormance Testing
( Shear Stress, kPa
Axial Stress, kPa variable magnitude
height constant
0.1 0.7 1.4

I Time, sec
Figure VII-11. Stress Pulses in Repeated Shear Test at Constant Height
Simple Shear Test at Constant Height
This test is performed in Levels 2 or 3 design. A controlled shearing stress is applied to a

test specimen. As the test specimen is sheared, it seeks to dilate, which increases its
height. The vertical actuator uses the signal from the axial LVDT to apply sufficient axial
stress to keep the specimen height constant. The test is performed at different stress
levels and temperatures depending on whether a Level 2 or 3 design is being performed.
The following tables outline test parameters.
Table VII-6. Simple Shear Test Parameters
Mix Design Level Temperature, "C Shear Stress, kPa
In this table, Teff(FC)is the effective pavement temperature for fatigue cracking. It is
computed by Superpave as a function of climate, depth of mixture in pavement, and
designer selected reliability level in the same manner as TCE(PD).Figure VII-12 shows
the application of stresses during the test. During the test axial and shear loads and
deformations are measured and recorded.
Shear Stress, kPa
Axial Stress, kPa

variable magnitude
10 20 30
Time, sec
Figure VII-12. Stress Applications in Simple Shear Test at 20 C
Frequency Sweep Test at Constant Height
This test is performed in Levels 2 or 3 design. A repeated sinusoidal shearing load is

applied to the specimen to achieve a controlled shearing strain of 0.005percent. One
hundred cycles are used for the test at each of the following loading frequencies: 10,5,2,
1,0.5,0.2,0.1,0.05,0.02, and 0.01 Hz.
As the test specimen is sheared, it seeks to dilate, which increases its height. The vertical
actuator uses the signal from the axial LVDT to apply sufficient axial stress to keep the
specimen height constant. The test is performed at different temperatures depending on
whether a Level 2 or 3 design is being performed. The following table outlines test
parameters.
--
Table VII-6. Frequency Sweep Test Parameters
I Mix Design Level I Temperature, OC I
During the test axial and shear loads and deformations are measured and recorded.
Figure VII-13 illustrates the application of shearing strains and axial stresses during the
test.
100 cycles
Shear Strain, %
Axial Stress, kPa ,variable magnitude
Time
Figure VII-13. Shear Strain and Axial Stress Applications in

Frequency Sweep Test at Constant Height
INDIRECT TENSILE TESTER
The IDT is a device that measures the creep compliance and strength of asphalt mixtures
using indirect tensile loading techniques at intermediate to low temperatures (< 20" C).
Indirect tensile testing involves applying a compressive load across the diametrical axis of
a cylindrical specimen (Figure VII-14). The mechanics of the test are such that a nearly
uniform state of tensile stress is achieved across the diametrical plane.
load
test compression tension
\
stress distribution across

diametral plane
load
Figure VII-14. Indirect Tensile Test
The IDT device has four components: the testing apparatus, the test control unit and data
acquisition system, load measuring device, and the environmental control chamber.
Testing Apparatus
The testing apparatus consists of a closed-loop electrohydraulic, servohydraulic, or

mechanical screw system capable of resolving static loads as low as 5 N. A rigid loading
frame is also necessary so that precise displacement measurements can be made without
frame compliance contributing to displacement measurements.
Control Unit and Data Acquisition System
The reaction of specimens to load can be captured by means of a multi-channel strip chart
recorder or an analog to digital data acquisition device.
VIZ. Performance Testing
Load Measuring Device
Applied loads are measured and controlled by means of an electronic load cell. The load
cell resides between the loading piston and loading platen. It accurately measures the
load applied to the test specimen.
Environmental Control Chamber
The environmental chamber controls test specimen temperature during the test. It must
be able to accurately control temperature in the range from -20" to 20" C and have
sufficient room to accommodate at least three test specimens and the loading frame.
Specimen Preparation and Instrumentation
The first step in specimen preparation is to trim test specimens to a thickness such that
their thickness to diameter ratio is greater than 0.33. For a 150 mm diameter specimen,
the minimum specimen thickness is 50 mm. Specimens must also be trimmed so that
they posses smooth, parallel surfaces onto which measurement gauges can be mounted.
The response of test specimens to load is measured by LVDTs mounted to the face of the
specimen (Figure VII- 15). Two LVDTs are mounted at right angles on each side of the
,
specimen for a total of four mounted LVDTs. The LVDTs are mounted very close to the
surface of the specimen, in no case greater than 6 mm.
test specimen 6 m m max
\
Front View
' LVDTs /
Side View
Figure VII-15. Specimen Instrumentation for IDT

Test Procedures
Two tests are performed using the IDT:
IDT Creep Compliance and Strength at Low Temperatures and

0 IDT Strength at Intermediate Temperatures.
A brief description of each test follows.
IDT Creep Compliance and Strength (Low Temperature Cracking Analysis)
This test is used to analyze mixtures for low temperature cracking. It is performed at
three temperatures for both levels of mixture design. These temperatures are 0,-lo0,
and -20" C.
In the first phase of the test, a static creep load of fixed magnitude is placed on the
specimen (Figure VII-16). The magnitude of the load should be that which produces
between 50 and 750 horizontal microstrain in the test specimen during the 100 seconds,
which is the duration of the creep phase of the test. Vertical and horizontal deformations
are measured on both sides of the specimen throughout the test.
0
Deformation, microstrains
100
Time, sec
Figure VII-16. Load Controlled Creep Portion of IDT Test

VIZ. Peq4ormance Testing
At the conclusion of the creep loading period, the specimen is loaded until failure (peak
load) by applying additional load at a rate of 12.5 mrn per minute. Vertical and horizontal
movements and load are measured. Measurements are taken until the load has decreased
to a value of at least 10 percent less than peak load. Figure VII-17 shows the controlled
deformation portion of the test.
Vertical Load
100
Deformation
A
Time. sec
Figure VII-17. Deformation Controlled Failure Portion of IDT Test
In Level 2 mixture design, test specimens are tested for creep compliance at 0, -lo0, and
-20" C with tensile strength measured only at -10" C . Level 3 mixture design requires
that creep compliance and tensile strength be measured at all three temperatures.
IDT Strength (Fatigue Cracking Analysis)
This test is used to analyze mixtures for fatigue cracking resistance. It is performed at
various temperatures ranging from -10" to 20" C. Levels 2 and 3 use different
temperatures at which to acquire data as shown in Table VII-7.
VII. Pet$ormance Testing
Table VII-7. Indirect Tensile Strength Test P rrameters

Mix Design Level Temperature, "C
2 TedFC)
3 - 10,4,20
In this test, the specimen is loaded at a constant deformation rate of 50 mm per minute of
vertical ram movement. The specimen is loaded until failure, which is the peak load.
Load and deformation are measured throughout the test. Figure VII-18 shows the load
and deformation characteristics of this test.
Vertical Load
Deformation
0
Time, sec
Figure VII-18. Load and Deformation Characteristics of IDT Fatigue

Cracking Test
DATA ANALYSIS AND INTERPRETATION
The data collected from performance testing is used by the performance prediction
models in Superpave to predict pavement performance for various combinations of
asphalt binder and mineral aggregate. Performance plots such as those shown in Figures
VII-19, 20, and 21 are used to select a mixture that offers the desired level of
performance. In these figures, Materials A, B, and C might be three entirely different
materials. If so, the performance prediction would be considered part of an analysis
procedure. This methodology is suited to evaluating the performance effects of aggregate

types and proportions, asphalt and mixture modifiers, or any other potentially innovative
HMA ingredient.
For the materials represented in Figures VII- 19,20, and 2 1, no material meets all the
distress criteria at the design number of ESALs. However, if distress such as fatigue and
low temperature cracking were of most concern, Material C would be a clear choice since
it meets the specified performance values. Unfortunately, Material C would exhibit
significant rutting after relatively few load applications. Both Materials A and B meet the
rutting criterion but they fail the cracking criteria. Because fatigue life is greatly affected
by pavement thickness, it may be possible to slightly increase the layer thickness and so
that Material B would meet the fatigue cracking criterion.
Alternatively, Materials A, B, and C might be the same aggregate blend with varying
asphalt content. Material A has the lowest asphalt content while Material C has the
highest asphalt content. Material B has a median value of asphalt content. In that case,
the performance prediction would be considered a design procedure and three additional
design plots would be useful (Figures VII-22). These design plots would define the range
of asphalt contents meeting performance standards. In this example, an asphalt content
approximately two-thirds between B and C would optimize pavement performance.
This type of information would also be useful in establishingjob control tolerances.
Rut Depth, mm
Design ESALs
10 20 30
ESALs x lo6
-
Figure VII-19. Predicted Performance Permanent Deformation
VII. Pe$ormance Testing
Ipercent Area (Fatigue Cracking)

Design ESALs A
i
Figure VII-20. Predicted Performance - Fatigue Cracking
1 crack Spacing, m t
i
Design Life
L
5 10
Years
-
Figure VII-21. Predicted Performance Low Temperature Cracking
Fatigue Low Temp

Rut Depth, mm Cracking, % Crack Spacing, m
I I
A B c A B C A B C
% Asphalt % Asphalt % Asphalt
Figure VII-22. Design Chart

APPENDIX A: SUPERPAVE ASPHALT BINDER
SPECIFICATION
Performance Graded Binder Specification
Performance Grade PG 52 P G 58 PG 64 PG 70
Average 7-day Maximum Pavement 4 2

Design Temperature, "C a
>-lo >-I6 >-22 >-28 >-34 2-40 >-46 >-I6 >-22 2-28 >-34 2-40 >-I6 >-22 >-28 >-34 >-40 >-lo >-I6 >-22 2-21
Minimum Pavement Design
Temperature, "C a
Original Binder
Hash Point Temp, T48: Minimum "C 230
Viscosity, ASTM D 4402: b 135

Maximum, 3 P a s (3000 cP),
Test Temp, OC
Dynamic Shear, TP5: 52 58 64 70

G*/sin 6, Minimum, 1.00 kPa
Test Temperature @ 10 radh, "C
Rolling Thin Film Oven (T24O) or Thin Film Oven (T179) Residue
Mass Loss, Maximum, % 1.OO
Dynamic Shear, TP5: 52 58 64 70

G*/sin 8, Minimum, 2.20 kPa
Test Temp @ 10 radlsec, "C
Pressure Aging Vessel Residue (PP1)
PAV Aging Temperature, "cd

I 90
I 100
I 100
Dynamic Shear, TP5:

G*sin6~Maximum*5000kPa 25 22 19 16 13 10 7 25 22 19 16 13 28 25 22 19 16 34 31 28 25
Test Temp @ 10 radlsec, "C
IPhysical Hardening
Creep Stiffness, TPI:
(
S, Maximum, 300 MPa

m-value, Minimum, 0.300 0
Test Temp, @ 60 sec, "C
Direct Tension, TP3:

Failure Strain, Minimum, 1.0%
Test Temp @ 1.0 mmlmin, O C
Notes:
a. Pavement temperatures can be estimated from air temperatures using an algorithm contained in the SUPERPAVE software program or may be provided by
the specifying agency, or by following the procedures as outlined in PPX.
b. This requirement may be waived at the discretion of the specifying agency if the supplier warrants that the asphalt binder can be adequately pumped and
mixed at temperatures that meet all applicable safety standards.
c. For quality control of unmodified asphalt cement production, measurement of the viscosity of the original asphalt cement may be substituted for dynamic
shear measurmeents of G*/sin 6 at test temperatures where the asphalt is a Newtonian fluid. Any suitable standard means of viscosity measurement may be
used, including capillary or rotational viscometry (AASHTO T 201 or T 202).
d. The PAV aging temperature is based on simulated climatic conditions and is one of three temperatures 90" C, 100" C or 110" C. The PAV aging
temperature is 100" C for PG 58- and above, except in desert climates, where it is 110" C.
e. Physical Hardening - TP 1 is performed on a set of asphalt beams according to Section 13.1, except the conditioning time is extended to 24 hrs + 10
minutes at 10' C above the minimum performance temperature. The 24-hour stiffness and m-value are reported for information purposes only.
f. If the creep stiffness is below 300 MPa, the direct tension test is not required. If the creep stiffness between 300 and 600 MPa the direct tension failure
strain requirement can be used in lieu of the creep stiffness requirement. The m-value requirement must be satisfied in both cases.
APPENDIX B: SUPERPAVE ASPHALT MIXTURE
GRADATION REQUIREMENTS
37.5 MM NOMINAL SIZE
I 1 I Restricted Zone
Boundary
Sieve, mm Control Points Minimum Maximum
50 1 100.0

Appendix B: Superpave Asphalt Mixture Gradation Requirements
25 MM NOMINAL SIZE
I I I Restricted Zone I
Boundary
Sieve, mm Control Points I
Minimum Maximum
100 -
90 --
80 --
ba
--
70
'1 60 --
2
5a 50 --
*B 40--
30 --
20 --
10 --
I
0
.075 .3 2.36 4.75 25 37.5
Sieve Size Raised to 0.45 Power
19 MM NOMINAL SIZE
Restricted Zone
Boundary
Sieve, mm Control Points Minimum I Maximum
0 ' .
.075 .3 2.36
19 25
12.5 MM NOMINAL SIZE
Restricted Zone
Sieve. mm Control Points

9.5 MM NOMINAL SIZE
I Restricted Zone
I I I Boundary
I Sieve, mm I Control Points I Minimum ( Maximum

APPENDIX C: SUPERPAVE CONSENSUS
AGGREGATE REQUIREMENTS
COARSE AGGREGATE ANGULARITY
Coarse Aggregate Angularity:

Traffic, million Depth from Surface
ESALs < 100 rnrn ,> 100 mm
< 0.3 551- -I-
>
- 100 I 100/100 1001100
Note: "85180 denotes that 85 % of the coarse aggregate has one
I fractured face and 80 % has two fractured faces. I
FINE AGGREGATE ANGULARITY
Fine Aggregate Angularity:

Traffic million, Depth from Surface
ESALs < 100 mm > 100 mm
< 0.3 - -
<1 40 -
<3 40 40
< 10 45 40
< 30 45 40
< 100 45 45
> 100
- 45 45
Note: Criteria are presented as percent air voids in loosely
compacted fine aggregate.
Appendix C: Superpave Consensus Aggregate Requirements
FLAT AND ELONGATED PARTICLES
Flat. Elongated Particles

I Traffic, million I Percent
ESALs
< 0.3 -
CLAY CONTENT
I ClayTraffic,
Content
I
I
I ESALs
million
I
Sand Equivalent, minimum
APPENDIX D: VOLUMETRIC ANALYSIS OF
COMPACTED HOT MIX ASPHALT (HMA)
DESCRIPTION OF TERMS
Term Identifier Description

Air Voids Paor V, total volume of the small air pockets
between coated aggregate particles;
expressed as a percentage of the bulk
volume of the compacted paving mixture
Voids in the VMA the volume of inter granular void space
Mineral Aggregate between the aggregate particles of a
compacted paving mixture that includes
the air voids and effective asphalt content;
expressed as a percentage of the total
volume of the compacted paving mixture
Effective Asphalt Pbe the total asphalt content of the paving
Content mixture less the portion of asphalt binder
that is absorbed by the aggregate particles;
expressed as a percentage of the total
weight of the compacted paving mixture
Voids Filled with Pfaor VFA the portion of the VMA that contains
Asphalt asphalt binder; expressed as a percentage
of the total volume of mix or VMA
Aggregate Bulk the ratio of the mass in air of a unit
Specific Gravity volume of aggregate, including permeable
and impermeable voids, to the mass of an
equal volume of water, both at the same
temperature
Aggregate Effective the ratio of the mass in air of a unit
Specific Gravity volume of aggregate, excluding voids
permeable to asphalt, to the mass of an
temuerature
Asphalt Binder the ratio of the mass in air of a given
Specific Gravity volume of asphalt binder to the mass of an
temuerature
Appendix D: Volumetric Analysis of Compacted HMA
the ratio of the mass in air of a given

Specific Gravity volume of compacted HMA to the mass of
an equal volume of water, both at the
same temperature
the ratio of the mass of a given volume of
Maximum Specific HMA with no air voids to the mass of an
Gravity of the Mix equal volume of water, both at the same
temperature
the volume of asphalt binder that has been
STANDARD CONVENTIONS
The following conventions are used to abbreviate binder, aggregate, and mixture
characteristics.
Specific Gravity ( G I Gxy
x- b =binder
s = aggregate (i.e., stone)
m = mixture
y- b = bulk
e = effective
a = _apparent
m = maximum theoretical
Mass (P) or Volume (V) Concentration: P,, or Vx,
x- b = binder
s = aggregate (i.e., stone)
a = _air
y- e = effective
a = absorbed
(note: standard conventions do not apply to Vba and Pf,)

CALCULATIONS
The following equations are necessary to compute the volumetric properties of
compacted HMA:
Bulk Specific Gravity of the combined aggregate (Gsb):
where, Pi = percent by mass of each component aggregate in blend

(note: PI + P2 + Pg = 100)
Gi = Gshof each component aggregate in blend
Effective Specific Gravity (Gse):
where, Ph = percent by mass total mix of asphalt binder in mix,

Gh = specific gravity of asphalt binder, and
G, = maximum theoretical specific gravity of mixture at Pb.
Maximum Theoretical Specific Gravity (G-):
Percent Absorbed Asphalt (Pba):
where Pba= percent absorbed asphalt by total mass of aggregate

Percent Effective Asphalt Content (Pbe):
where, Pb, = percent effective asphalt by total mass of mix,

Pb = percent asphalt content in mix by mass of total mix, and
P, = percent aggregate content in mix by mass of total mix.
Voids in the Mineral Aggregate (VMA):
where, Gmb= bulk specific gravity of compacted mix.
Percent Air Voids (Pa):
(Note: Pa is sometimes abbreviated V,)
Percent Voids Filled with Asphalt (Pf,):
VMA - P,
VMA
(Note: Pf, is often called VFA)
Dust Proportion:
where, P.075= percent by mass of total aggregate passing 0.075 mm sieve, and
Pbe= percent effective asphalt content by mass total mix.
(Note: The 0.075 rnm sieve is often called the 75 micron sieve).
APPENDIX E: OUTLINE OF STEPS IN
SUPERPAVE LEVEL 1 MIX DESIGN
I. SELECTION OF MATERIALS
A. Selection of Asphalt Binder
1. Determine project weather conditions using weather database
2. Select reliability
3. Determine design temperatures
4. Verify asphalt binder grade
5. Temperature-viscosity relationship for lab mixing and compaction
B. Selection of Aggregates
1. Consensus properties
a. Combined gradation
b. Coarse aggregate angularity
c. Fine aggregate angularity
d. Flat and elongated particles
e. Clay content
2. Agency and Other properties
a. Specific gravity
b. Toughness
c. Soundness
d. Deleterious materials
e. Other
C. Selection of Modifiers
11. SELECTION OF DESIGN AGGREGATE STRUCTURE
A. Establish Trial Blends

1. Develop three blends
2. Evaluate combined aggregate properties
B. Compact Trial Blend Specimens

1. Establish trial asphalt binder content
a. Superpave method
b. Engineeringjudgment method
2. Establish trial blend specimen size
3. Determine Ninitial & Ndesign & Nmnximum
4. Batch trial blend specimens
5. Compact specimens and generate densification tables
6 . Determine mixture properties (Gmm& Gmb)
Appendix E: Outline of Steps in Superpave Level I Mix Design
C. Evaluate Trial Blends

Determine %Gmm @ Ninitlal & Ndesip & Nmaximum
Determine %Air Voids and %VMA
Estimate asphalt binder content to achieve 4% air voids
Estimate mix properties @ estimated asphalt binder content
Determine dust-asphalt ratio
Compare mixture properties to criteria
D. Select Most Promising Design Aggregate Structure for Further Analysis
111. SELECTION OF DESIGN ASPHALT BINDER CONTENT

A. Compact Design Aggr Structure Specimens Multiple Binder Contents
1. Batch design aggregate structure specimens
2. Compact specimens and generate densification tables
B. Determine Mixture Properties versus Asphalt Binder Content

1. Determine %Gmm @ Ninitial & Ndesign & Nmaximum
2. Determine volumetric properties
3. Determine dust-asphalt ratio
4. Graph mixture properties versus asphalt binder content
C. Select Design Asphalt Binder Content

1. Determine asphalt binder content at 4% air voids
2. Determine mixture properties at selected asphalt binder content
3. Compare mixture properties to criteria
IV. EVALUATION OF MOISTURE SENSITIVITY OF DESIGN

ASPHALT MIXTURE USING AASHTO T283
APPENDIX F: TESTING REQUIREMENTS FOR
SUPERPAVE
SUPERPAVE LEVEL 1
Design Aggregate Structure
I Blend 1 Blend 2 Blend 3
Rep 1
Rep 2
x1
Design Binder Content I

I - 0.5% Est. Optimum +0.5% +1.0%
Rep 1
Rep 2
x1
AASHTO T283
Total Specimens = 20
Wet
x1
Appendix F: Testing Requirements for Superpave
SUPERPAVE LEVEL 2
I
Simple Shear & Frequency Sweep
at T,(PD) and Teff(FC)
I high
- a/c medium a/c low a/c
Rep 1
Rep 2
x1
Repeated Shear Constant Indirect Tensile Strength at T,&C)

Stress Ratio at T, high a/c medium a/c low a/c
I high a/c
Rep 1
Rep 1
Rep 2
Rep 2 x1
x1
Indirect Tensile Creep Compliance test strength at -10 C only

and Strength at 0, -10, and -20 C /
I - a/c
high medium a/c low a/c
Rep 1 Total Specimens = 23
Rep 2
Rep 3
x1
Appendix F: Testing Requirements for Superpave
SUPERPAVE LEVEL 3
I Simple Shear & Frequency Sweep at 4,20,40 C
I high a/c medium a/c low a/c
Rep 1
k
Rep 2
\ use same specimen
at each of three temps
I high a/c medium a/c low a/c r
Rep 1
Rep 2
x I
Indirect Tensile Strength at -10,4,20 C
I high a/c
I high a/c medium a/c low a/c
Rep 1
I
Rep 2
Rep 1
high a/c
-
xl
Indirect Tensile Creep Compliance

and Strength at 0, -10, and -20 C
medium a/c low a/c
- use different specimens
at each of three temps
Total Specimens = 59
1. R. J. Cominsky, G. A. Huber, T. W. Kennedy, and R. M. Anderson, "The Superpave
Mix Design Manual for New Construction and Overlays," Strategic Highway
Research Program, SHRP-A-407, 1994. This is an important manual that describes
the details of Superpave mix design and analysis, It is essential for those who are
interested in or will be intimately involved in mix designs of the future.
2. T . W . Kennedy, G. A. Huber, E. T. Harrigan, R. J. Cominsky, C. S. Hughes, H. L.

Von Quintus, and J. S. Moulthrop, "Superior Performing Asphalt Pavements
(Superpave): The Product of the SHRP Asphalt Research Program," Strategic
Highway Research Program, SHRP-A-410, 1994. This report provides a broad
overview of the Superpave system including binder and mixture analysis. It explains
special approaches necessary in Superpave for handling modified binder and mixture
systems.
3. G. A. Huber, "Weather Database for the Superpave Mix Design System," Strategic
Highway Research Program, SHRP-A-648A, 19%. This report describes how
weather station data is used for the purpose of Superpave binder selection. It lists
binder selection weather database parameters for 5313 United States and 1515
Canadian weather stations. This is an important information source for those
interested in understanding weather based Superpave binder selection.
4. R. J. Cominsky, R. B. Leahy, and E. T. Harrigan, "Level One Mix Design: Materials

Selection, Compaction, and Conditioning," Strategic Highway Research Program,
SHRP-A-408, 1994. This report describes the background of many features found in
Superpave Level 1 mix design. It relates important information pertaining to
aggregate selection, gyratory compaction, aging, and moisture sensitivity.
5. J. D'Angelo, J. R. Bukowski, and T. Harman, "Superpave Level 1 Mixture Design

Workshop, Illinois #9402B," Federal Highway Administration, Office of Technology
Applications, Demonstration Project No. 90, July 1994. This report, developed by
FHWADTA, provides very useful, practical Level 1 mix design information. It also
provides a good list of standard AASHTO and ASTM test procedures forpe$orming
Superpave mix design. It is available through F H W m T A .
6. D. A. Anderson, D. W. Christensen, H. U. Bahia, R. Dongre, M. G. Sharma, C. E.

Antle, and J. Button, "Binder Characterization and Evaluation, Volume 3: Physical
Characterization," Strategic Highway Research Program, SHRP-A-369, 1994. This
report explains the science and engineering behind the Superpave method of binder
characterization. It is a very important resource for those who are interested in
understanding more about asphalt binder rheology, particularly as generalized by the
Superpave system.
Bibliography
7. R. L. Lytton, J. Uzan, E. G. Fernando, R. Roque, D. Hiltunen, and S. M. Stoffels,

"Development and Validation of Performance Prediction Models and Specifications
for Asphalt Binders and Paving Mixes," Strategic Highway Research Program,
SHRP-A-357, 1993. This report explains the methodology and theory of the
pavement per3cormance prediction algorithms in Superpave Levels 2 and 3 mix design.
It is a must for those who intend to use this important new tool for asphalt mix design
and analysis.
8. R. B. Leahy and H. L. Von Quintus, "Validation of Relationships Between

Specification Properties and Performance," Strategic Highway Research Program,
SHRP-A-409, 1994. This report outlines the two-stage process and results used to
validate Superpave binder and mixture tests and specifications.
9. E. T. Harrigan, R. B. Leahy, and J. S. Youcheff, "The Superpave Mix Design System

Manual of Specifications, Test Methods, and Practices," Strategic Highway Research
Program, SHRP-A-379, 1994. This report lists the Superpave binder and mixture
tests as produced by SHRP researchers. It is a valuable tool for practitioners who
desire to know the exact protocols recommended by SHRP researchers. However,
users are cautioned that many changes have occurred to these test procedures as they
have been adopted by AASHTO.
10. The Asphalt Institute, "Mix Design Methods for Asphalt Concrete and Other Hot-Mix
Types," Manual Series No. 2 (MS-2), Sixth Edition, 1993. While this manual is
mostly devoted to Marshall and Hveem methods of mix design, its chapter on
volumetric properties of compactedpaving mixes remains one of the best available
sources on this important subject.
SHRP publications are available through the Transportation Research Board,

Printed By
Federal Highway Administration
Office of Technology Applications
US. Department of Transportation
400 Seventh Street, SW
Washington, D.C. 20590

Fhwa 1995 PDF

Uploaded by

Copyright:

Available Formats

Fhwa 1995 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fhwa 1995 PDF

Uploaded by

Copyright:

Available Formats

What is the purpose of this document?

What is the purpose of this document?

What are some important SHRP publications mentioned in the bibliography section?

What are some important SHRP publications mentioned in the bibliography section?

Publication No.

U.SDepartment February 1995

NATIONAL ASPHALT TRAINING CENTER

Innovation Through Partnerships

This document is disseminated under the sponsorship of the Department of Transportation in

4. Title and Subtitle ( 5. Report Date

FHWA Technical Representative - J.R. Bukowski

17. Key Words 18. Dlsfribution Statement

The NATC resides at the Asphalt Institute's Research Center in Lexington,

The objective of the educational program is to train students in the practical

The training program consists of 40 hours of instruction. Of this 40 hours,

Mr. John R. Bukowski of FHWA's Office of Technology Applications is the Contracting

Mr. Robert B, McGennis

Mr. R. Michael Anderson

Dr. Thomas W. Kennedy

Dr. Mansour Solaimanian

SuperpaveTM is a registered trademark of the Strategic Highway Research Program

11. SUPERPAVE TO THE RESCUE ...................................................................................................... 21

SPECIMEN PREPARATION ....................................................................................................

VI . VOLUMETRIC MIX DESIGN ...................................................................................................... 75

VII. PERFORMANCE TESTING (LEVELS 2 AND 3) ............................ . . . . .. . ..

APPENDIX B: SUPERPAVE ASPHALT MIXTURE GRADATION REQUIREMENTS.....................141

12.5 MM NOMINAL SIZE ....................................................................................................... 144

APPENDIX C: SUPERPAVE CONSENSUS AGGREGATE REQUIREMENTS.................................. 147

APPENDIX D: VOLUMETRIC ANALYSIS OF COMPACTED HOT MIX ASPHALT (HMA) .........149

APPENDIX E: OUTLINE OF STEPS IN SUPERPAVE LEVEL 1 MIX DESIGN .............................. 153

APPENDIX F: TESTING REQUIREMENTS FOR SUPERPAVE ......................................................... 155

BIBLIOGRAPHY .................................................................................................................................... 159

HOW ASPHALT MIXTURES BEHAVE

Asphalt concrete (sometimes referred to as "hot mix asphalt" or simply "HMA")is a

ASPHALT BINDER BEHAVIOR

Figure 1-1. Asphalt Cement Time Temperature Dependency

Asphalt cement is sometimes referred to as a visco-elastic material because it

There remains another important characteristic about asphalt cement. Because it is

Figure 1-2. Visco-Elastic Behavior of Asphalt

MINERAL AGGREGATE BEHAVIOR

Increasingly, waste products are used as aggregate or otherwise disposed of in asphalt

Regardless of source, processing method, or mineralogy, aggregate is expected to provide

Cubical Aggregate Rounded Aggregate

Figure 1-3. Aggregate Stone Skeletons

Before Load After Load

Figure 1-4. Shear Loading Behavior of Aggregate

Contrasting aggregate shear strength behavior can easily be observed in aggregate

Cubical Aggregate Rounded Aggregate

Figure 1-5. Stockpile Behavior of Cubical and Rounded Aggregate

where, z = shear strength of aggregate mixture,

The Mohr-Coulomb shearing behavior of materials is shown in Figure 1-6.

Figure 1-6. Mohr-Coulomb Theory

A last consideration in understanding the shearing properties of aggregate is the concept

Before Shearing During Shearing

Figure 1-7. Dilation of Two Aggregate Particles When Sheared

ASPHALT MIXTURE BEHAVIOR

weak subgrade or underlying layer \ subgrade

Figure 1-8. Rutting from Weak Subgrade

Figure 1-9. Rutting from Weak Mixture