Professional Documents
Culture Documents
Second Edition
1996
Freddy L. Roberts
Prithvi S. Kandhal
E. Ray Brown ,
National Center for Asphalt Technology
Auburn University, Alabama
Dah-Yinn Lee
Iowa State University
Ames, Iowa
Thomas W. Kennedy
University of Texas
Austin, Texas
First Edition,(1991)
First Printing-2/91
Second Printing-5/91
Third Printing-3/92
Fourth Printing-3/94
Second Edition (1996)
First Printing-11/96
Second Printing-I 0/97
Third Printing---4/00
Fourth Printing-3/03
This textbook has been prepared to assist the college professor in pro-
viding technical information on Hot Mix Asphalt (HMA) to engineering
students. In the past, professors have had to obtain information from many
sources to teach even a short course involving HMA. Many faculty who
have had n.o formal training in HMA have deleted the subject from their
_ Materials Courses completely or have reduced the amount of material cov-
ered so that it has little benefit to the student. This book should help alle-
viate these problems.
This book has been written to _cover most topics important to those
who plan to work in the HMA industry. It can be used for teaching under-
graduate as well as graduate courses. The primary purpose of the book is
to provide all information necessary for teaching Hot Mix Asphalt to
undergraduate students to eventually improve the quality of engineers in
the industry. A comprehensive undergraduate course should expose the
student to some of the material in Chapters 2, 3, 4, and 6. Obviously all
of this material cannot be presented to a student in one course. For exam-
ple, Chapters 4, 5, and 6 have been used as the basis for teaching three
graduate level courses.
This book is also useful to the practicing engineer who is to some
extent involved with Hot Mix Asphalt. This includes personnel that work
for the Federal Highway Administration, State Departments of
Transportation, Federal Aviation Administration, Corps of Engineers, Air
Force, Navy, County Engineers, City Engineers, Consulting Engineers,
and Contractors.
Much research is presently being performed on the subject of Hot Mix
Asphalt, under the Strategic Highway Research Program (SHRP). This
research effort alone is spending $50 million dollars over a period of five
years to study asphalt cement and asphalt mixtures. It is anticipated that
once this study nears completion, many additions and updates will need to
be made to this textbook. Hence it is expected that the second edition of
this book will be forthcoming within the next five years.
iii
PREFACE TO SECOND EDITION
The second edition of this book was published in 1996. The primary
purpose of the second edition was to include information on Stone Matrix
Asphalt, Superpave performance grade (PG) asphalt binders, Superpave
mix design, and asphalt modifiers. Other portions of the book were
reviewed and modified where necessary to bring each chapter up to date
with the latest technology.
The modifications were prepared by E. Ray Brown and Prithvi S.
Kandhal. The major modifications were reviewed by Jon Epps, Dale
Decker, and John Bukowski. Lois Flowers typed the revised manuscript
and made corrections as needed. Rajib Malli,!=k revised the index.
Much research is currently underway to validate the research from the
Strategic Highway Research Program. Other research is underway to eval-
uate laboratory and field wheel tracking devices. As this work is complet-
ed, it is anticipateq. that the textbook will again need to be revised. It is
expected that the third edition of the book will be initiated in approxi-
mately five years.
iv
ACKNOWLEDGMENTS
The authors thank all those individuals and organizations that provid-
,ed support for the preparation of this textbook. Those providing signifi-
. cant co.ntributions to the preparation of this book are identified below.
All funds for preparation and publication of this book were provided
by the National Asphalt Pavement Association (NAPA) Education
Foundation. General guidance was provided by the National Center for
Asphalt Technology (NCAT) Board of Directors and the NCAT
Applications Steering Committee.
All figures were prepared under the direction of Wally Ridgeway, who
also designed the book cover and dust jac~et. Others involved with prepar-
ing figures included Robert Stroud, Julia Freeman, Mike Cole, and Hugh
Smith.
l
Chapter 4-Paul Teng, John Carroll, Maqbool Khatri, and Dick Smith;
Chapter 5-Harold Von Quintus; Chapter 6-Mike Geller, Stuart Murray,
Jay Winford, Jay Gabrielson, and Ted Rapallo; Chapter 7-Leonard Wood
_____a_n_,d Cnarlesrlughes; Cnapter 8=Fred-Fir1n-and Marvin Traylor; and
Chapter 9-Donald Jordison and Mohammed Shahin.
The final manuscript was edited for grammar and accuracy by Wallace
Whatley, Ann Morton, Kim Toney, and Kristy Jones.
vi
TABLE OF CONTENTS
CHAPTER 1. INTRODUCTION ........................ 1
REFERENCES ....................................6
viii
Fatigue Cracking ......................... , .. 97
Low Temperature Cracking ..................... 98
Other Common Specification Criteria .............98
Selection of Superpave Binder Grades ............. 99
ix
Durability and Soundness ........................ 139
Particle Shape anal Surface Texture .................. 140
Cleanliness and Deleterious Materials ............... 143
Sand Equivalent Test ......................... 144
Clay Lumps and Friable Particles ................ 144 .
Plasticity Index ............................. 145
Specific Gravity ................................ 145
Specific Gravity and Absorption of
. Coarse Aggregate ........................ 147
Specific Gravity and Absorption of
Fine Aggregate .......................... 148
Size and Gradation .......... ,.,. .................. 153
Desired Gradation Curves ..................... 153
Maximum Aggregate Size ..................... 155
Typical Specification Bands and Tolerances ........ 157
Sieve Analysis ................, .............. 157
Aggregate Blending to Meet Specifications ......... 160
Surface Area .................................. 169
REFERENCES ...................................171
INTRODUCTION ...............................173
-------oBJEeTIVES-ANB-E-E-EM:EN-T:-S-0F-MIX-BESIGN-.-.-.-.-.-l-R.----
Objectives .................................... 181
Gradation and Maximum Density Curves ............ 185
Asphalt Film Thickness .......................... 187
Laboratory Compaction ......................... 189
X
Weight Volume Relationships ..................... 192
Definitions ................................ 192
Calculations ............................... 199
xi
CHAPTER 5. CHARACTERIZATION OF ASPHALT
MIXTURES ........................... .275
xii
CHAPTER 6. EQUIPMENT AND CONSTRUCTION ... 313
xiii
· COMPACTION ................................... 354
Mixture Factors Affecting Compaction .............. .354
Compaction Equipment ......................... 356
Steel-Wheeled Rollers ........................359
Pneumatic-Tired Rollers ..................... .360
Vibratory Steel-Wheeled Rollers ................ 363
Field Conditions Which Affect Compaction .......... .365
Rolling Patterns ............................... 366
Types of Density Specifications Used for
Construction of HMA ....................... 369
Density Specified as a Percent of
Laboratory Density ... _.................... 371
Density Specified as a Percent of
Theoretical Maximum Density .............. 375
Density Specified as a Percent of Control Strip ..... .376
Measurement of Density ......................... 378
Core Method ............................. .378
Nuclear Gauge Method ....................... 378
xv
HOT MIX ASPHALT RECYCLING ................. .439
Introduction ..................................439
Candidates for Recyding .........................440
Design Considerations for Recycled Mixtures ..........442
Production of Recycled Asphalt Mixtures- ............ .444
REFERENCES ...................................472
----S'ERVICEABI-Hfl/-PE-RF'ORMANGE-Gf>NGEP-'IS-.-.-;-.-.-A-7--7------
Pavement Roughness ........................... .479
Road Roughness Measurement Methods ............ .481
Smoothness Requirements for Construction ...........486
xvi
DISTRESSES IN HMA PAVEMENTS ... ,. ........... .489
Cracking ....................................494
Fatigue Cracking ...... _- ................... .494
Low Temperature (Thermal) Cracking ........... .495
Longitudinal Cracking ...................... .496
Block Cracking .............................498
Reflection Cracking ........................ .499
Slippage Cracking .......................... 501
Distortion ................................... .502
Rutting .................................. 502
Shoving and Corrugation ..................... 506
Disintegration ................................. 507
Ravelling ................................ .507
Wear Loss ................................. 509
Stripping ................................ .510
Potholes .................................. 517
Loss of Frictional Resistance ..................... .517
INDEX .............................................579
xviii
CHAPTER 1. INTRODUCTION
This textbook is a direct result of the concern of U.S. Hot Mix Asphalt
(HMA) contractors and the National Asphalt Pavement Association
(NAPA) about the lack of material on asphalt technology available to engi-
neering students in American universities and colleges and to technicians
in technical schools. In fact, two of the principal educational activities of
the National Center for Asphalt Technology (NCAT) have been directed
toward (1) providing the opportunity for professors currently on U.S. Civil
Engineering faculties to learn more about asphalt technology by attending
an annual summer course offered at NCAT, and (2) preparing a book suit-
able for use in college courses in asphalt technology.
Technology as used in this book means the systematic treatment of a
subject. The primary subject of this book is HMA. The topics covered
include manufacture of the asphalt and aggregates, design of HMA mix-
tures, characterizing HMA in terms of engineering properties, construc-
tion, performance, and maintenance and rehabilitation of aging HMA sur-
faces. This book has been prepared for use by engineers; however, it has
much to offer technicians in advancing their understanding of the topics
covered in the book.
This book is the first comprehensive treatment of the subject in many
years. Based on a cursory review of the NCAT library, it appears w'be the
first textbook in over 20 years. The last textbook that was published on
--------.1.B-.LJMA in rbe U.S. was in 1967 by Wallace and Martin (1) titled Asphalt
Pavement Engineering. Just prior to that, Bituminous Materials in Road
Construction was prepared by the staff of the Road Research Laboratory in
the United Kingdom in 1962 (2). In fact, the 1960s was the last decade in
which substantial printing of reference books dealing with asphalt materials
2 Hot Mix Asphalt Materials, Mixture Design and Construction
occurred. A sample of the books and book series printed in the 60s
includes the following:
While all four of these books and book sets were excellent as reference doc-
uments for the experienced asphalt technologist, they were not written for
use in the classroom as textbooks.
These comments should not lead the reader to believe that there is a
shortage of technical information on asphalt materials. In , the
Transportation Research Information Service (TRIS) maintained by the
Transportation Research Board (TRB) of the National Academy of
Sciences, there are tens of thousands of referen~es, papers, and research
reports on asphalt technology. Indeed, every year there are many papers
presented and published by the TRB, and the Association of Asphalt
Paving Technologists publishes a proceedings of their annual meeting,
which typically contains 20 to 30 of the most recent papers on asphalt
technology. In addition, the International Society for Asphalt Pavements
(ISAP) sponsors an international meeting and publishes a proceedings
every 5 years to focus on some aspect of the design and performance of
---asprratr_--pavemems:--Ad-ditiona±ly,-dre-Natrorra:1-Asphah-Paveme•-n-r-------
Association and the Asphalt Institute regularly publish and revise a whole
host of publications dealing with all aspects of asphalt pavement design,
construction, testing, and issues of particular concern to their individual
constituencies. However, with all this information available, little has been
Chapter !-Introduction 3
prepared specifically for the classroom. This book is targeted specifically for
that audience. Therefore, in preparing this textbook, the authors have
extracted the best current technology from the literature and put it into a
form suitable for classroom use.
REFERENCES
Historical Background
1. Natural Asphalts: These were laid down in geologic strata and occur
both as relatively soft asphalt material and also as a hard, friable,
black material in veins of rock formations, or impregnated in vari-
ous limestones, sandstone formations, and the like. The relatively
soft asphalt material, almost like heavy petroleum, is typified in the
Trinidad Lake deposit on the Island of Trinidad, in Bermudez Lake,
Venezuela, and in the extensive "tar sands" throughout western
Canada.
Hundreds of thousands of tons of asphalt have been removed
from the Trinidad Lake without its showing any appreciable sign of
loss (Z). As the material is removed, pressure deep in the earth forces
more of the heavy residue to the surface, where nature continues its
refining process. Trinidad asphalt contains about 50 to 57 percent
asphalt. The remainder is colloidal day of volcanic origin, plus some
inert organic matter. By the last two decades of the nineteenth cen-
tury, considerable amounts of asphalt from the Trinidad Lake
deposit wdre used for road building in the United States and else-
where in the world (1). The first sheet asphalt pavement using
Trinidad Lake asphalt was laid in 1876 on Pennsylvania Avenue in
Washington, D.C. Trinidad asphalt has a very low penetration
(3-10) and therefore was fluxed back for use in paving in the United
States. Normally, it is refined so .that the volcanic ash is between 3
and 5 percent. -
The hard variety of natural asphalts such as Gilsonite includes
asphaltites which are solid asphalts without impurities (silts, days,
etc.). Gilsonite is a black, brittle and hard asphalt material mostly
mined in Utah.
2. Petroleum Asphalts: These are colloidally dispersed hydrocarbons in
crude petroleum and are obtained by refining petroleum crudes.
Following the discovery of the refining process in the early 1900s
and the popularity of automobiles, large quantities of petroleum
asphalt were processed by the oil companies. Gradually these refined
asphalts became plentiful and of good quality. This forced the native
(natural) asphalts into a position of relative unimportance (1).
---AsphaltRefining--------------------'----
The material for this section on asphalt refining has largely been
obtained from papers by Corbett (1.) and Hodgson (j).
Almost all paving asphalt cement used today is obtained by processing
crude oils. Many refineries in the Unit~d States are located near water sup-
Chapter 2-Asphalt Refining, Uses, and Properties 9
Gasoline
Kerosene 6
Lt. Gas Oil 7 33
20
16
Bitumen
Residuum
30
I
Figure 2-2 Make-up of crude petroleum (after Corbett, 1)
The API gravity for water is 10. Asphalts have an API gravity of
approximately 5-10, whereas the API gravity of gasolines is about 55.
Low API gravity crudes (API less than about 25) yield relatively low
percentages of distillable overhead fractions and high percentages of
asphalt cement. On the other hand, the high API gravity crudes (API more
than about 25) yield relatively high percentages of overhead fractions and
low percentages of asphalt cement. The low gravity crudes are generally
known in the industry as heavy crudes, or as sour crudes if they contain
high sulfur content. The high gravity crudes are referred to as light crudes,
or sweet crudes if they contain low sulfur content. Therefore, a refiner
must choose the crude type depending on the types and amounts of end
pro uct esire to e pro uce , w 1c may e ctate by the marRe--t_ _ _ _ __
demand. Each crude. is identified by name or source along with its API
gravity. Table 2-1 lists some crude types and typical sources.
a distillation tower where the lightest components vaporize, rise to the top,
cool, condense, and are drawn off for further processing. At various heights
in the tower different fractions reach their boiling point and then, as the
temperature is reduced, condense on trays inside the tower. The interme-
diate or mid-tower components are drawn off at those levels and are treat-
ed to make jet fuel, kerosene, and diesel. The residuum from this distilla-
tion is usually fed into a vacuum distillation unit where heavier gas oils are
produced. The reduced pressure (typically 55 mm Hg, 1.0 psi) in the vac-
uum tower assists in flashing these oils at lower temperatures to prevent
thermal cracking of the asphalt cement. In cracking, asphalt molecules are
thermally broken apart. Often steam is added to the bottom of the tower
-----~·o_fur.ther-reduce..the.-P-r:ess.ur:e-fr0m-5-0-mm-Hg-to about 12 mm Hg, 0.24
psi. The bottom fraction from this unit is called vacuum processed, steam
refined asphalt cement. The grade of asphalt cement, whether it is mea-
sured by penetration or viscosity, is controlled by the amount of heavy gas
oil removed. All of this is done by a continuous flow operation.
12 Hot Mix Asphalt Materials, Mixture Design and Construction
~ - - - ~ - - Solvent
Stripping
Vacuum Towers STA
Residuum
tane) under supercritical conditions. This mixture is then fed into a sepa-
rator (SEP) at a predetermined controlled temperature (T1) and pressure,
to separate an asphaltene concentrate. Asphaltenes (first fraction) are then
recovered by stripping (STR) off the solvent used. The extract portion
from the preceding operation is then fed into a second separator where at
a controlled higher temperature (T2) a decrease in solubility results,
SEP Solvent
T, STA STA T
3
STA
Vacuum
·Residuum
SEP = Separator
STA = Stripping
SP = Softening Point
SCRUBBER
INCINERATOR
TO OIL
SKIMMER
___.!!PHAL T FLUX--1-.RS"
250·300"F
AIR
BLOWER
BLOWN ASPHALT
CRUDEA-C-E
(OR MIX)
1. Asphalt cements;
2. Emulsified asphalts; and
3. Cutback asphalts.
16 Hot Mix Asphalt Materials, Mixture Design and Construction
Asphalt Cements
ing, as specified in ASTM D3381. This grading is based on either the vi-s;-::--------
cosity of the original asphalt cement or on the viscosity of the asphalt
cement after aging in the rolling thin film oven (RTFO) test. Both vis-
cosities are measured at 140°F (60°C) and reported in poises. The viscosi-
ty grades based on original asphalt cements include AC-2.5, AC-5,
Chapter 2-Asphalt Re~ning, Uses, and Propertie 17
AC-10, AC-20, AC-30, and AC-40. The numerical values indicate vis-
cosity at 140°F (60°C) in hundreds of poises. The viscosity grades based on
the asphalt residue (AR) from the RTFO test include AR-1000, AR-2000,
AR-4000, AR-8000, and AR-16000. The numerical values indicate vis-
" cosity at 140°F (60°C) in poises. The viscosity tests and the viscosity grad-
ing systems will be discussed in detail later in this chapter.
The ,third method of classifying asphalt binders is performance based
and w:as developed in the Strategic Highway Research Program (SHRP). It
is discussed later under the section titled "Superpave binder tests and
specification."
Emulsified Asphalts
Emulsified asphalt (also simply called emulsion) is a mixture of asphalt
cement, water, and emulsifying agent. Because the asphalt cement will
ASPHALT CEMENT
IN FORM OF
PERFECT
SPHERES
WA TER---tl:!'.'"
CONTAINING
EMULSIFYINGt:=:...=.-=:=====-==-=-=~-t:.=..;:==:.:..==::L..J
AGENT
Figure 2-8. Emulsified Asphalt
not dissolve in water, asphalt cement and water exist in separate phases as
shown in Figure 2-8. Hot asphalt cement and water containing the emul-
sifying agent (soap is one example) are passed under pressure through a col-
loid mill to produce extremely small (less than 5-10 microns) globules or
------drop-I-ets-of-aspha:lt-cement-whieh-a:re-st1-spefl:cl:ed in water (Figure 2-9). The
emulsifying agent imparts an electric charge to the surface of the droplets
which causes them to repel one another, and thus the globules do not coa-
lesce. Emulsified asphalts are also categorized as liquid asphalts because,
unlike asphalt cements, they are liquid at ambient temperatures.
18 Hot Mix Asphalt Materials, Mixture Design and Construction
Emulsions are made to reduce the asphalt viscosity for lower application
temperatures.
Two most commonly used types of emulsified asphalts are specified in
ASTM D977 and ASTM D2397:
I. Anionic-electro-negatively charged asphalt droplets; and
2. Cationic-electro-positively charged asphalt droplets'.
If the emulsifying agent is anionic, the asphalt droplets bear a negative
charge. If the emulsifying agent is cationic, the droplets bear a positive
charge.
Most mineral aggregates bear a positive or a negative or mixed charge
on the surface. Most siliceous aggregates, such as sandstone, quartz and
siliceous gravel, are negatively charged and th;refore are generally compati-
ble with the positively charged cationic emulsified asphalts. On the other
hand, some aggregates such as limestone bear a positive surface charge and
are therefore generally compatible with the negatively charged anionic emul-
sified asphalts. This happens because opposite charges attract one another.
HOT
ASPHALT
TANK COLLOID MILL
:==-=====OPUMP
WATER
+
- -
STORAGE STORAGE STORAGE
Cutback Asphalts
Tests on emulsionJ
Viscosity, Saybolt 20-100 - - 20-100 100+ 100+ 20-100 100+ 100+ 50+ 20-100 20-100
Furol at 77°F
iB:::
Viscosity, Saybolt - 75-400 75-400 - - - 20-100 100+ 100+ 50+ 20-100 20-100 ,l"
Furol at 122°F
Minimum Residu. 55 63 63 44 65 65 55 65 65 65 57 57
i
by distillation, o/o
Tests on residue
from distillation:
t
lf -
~
t)
Penetration at 77' F, 100-200 100-200 100-200 100-200 100-200 40-90 100-200 100-200 40-90 200+ 100-200 40-90
100 g, 5s ~-
l
Aooua,, I40"F, r - - 1200 - - - 1200 1200 1200 1200 - -
f
j.
::,
Chapter 2-Asphalt Refining, Uses, and Properties 21
Tests on emulsions:
Viscosity, Saybolt
Purol at 77°P - - - - 20-100 20-100
Viscosity, Saybolt
Purol at 122°P 20-100 100-400 50-450 50-450 - -
Distillation:
Oil distillate, % 3 3 12 12 - -
Minimum residue,% 60 65 65 65 57 57
Tests on residue
from distillation test:
Kinematic viscosity
at 140°F(60°C),cSt 70-140 250-500 800-1600 3000-6000
Kinematic viscosity
at 140°F(60°C),cSt 30-60 70-140 250-500 800-1600 3000-6000
11
Tests on residue
from distillation:
Viscosity at 140°F(60°C), P 300-1200 300-1200 300-1200 300-1200 300-1200
i Ductility at 77°F(25°C),
cm,minimum 100 100 100 100 100
Ii
II
Chapter 2-Asphalt Refining, Uses, and Properties 23
ASPHALT CEMENT
APPROX. FUROL VIS., 140" F, SEC., 15-30 35-70 125-250 400-800 1500-3000
APPROX. KINEMATIC VIS. 140"F, CS 30-60 70-140 250-500 800-1600 3000-6000
1. Consistency tests;
2. Durability tests;
3. Purity tests;
4. Safety tests; and
5. Other tests.
Consistency Tests
Consistency describes the degree of fluidity of asphalt cement at any
particular temperature. Since asphalt cement is a thermoplastic material, its
consistency varies with the temperature. Therefore, it is necessary to mea-
sure the consistency of different asphalt cements at the same temperature
and shear loading conditions if comparisons are to be made. The following
consistency tests are performed on asphalt cements:
TIMING
MARKS
FILLING
LINE
I,i1
Chapter 2-Asphalt Refining, Uses, and Properties 25
cement starts to flow, the time (in seconds) required for it to flow between
two timing marks is measured. The measured time (in seconds) is multi-
plied by the calibration factor for the viscometer tube to obtain the value
for viscosity in poises, which is the standard unit for measuring viscosity.
Manufacturers calibrate their viscometer tubes with standard oils and fur-
nish the calibration factor with each tube.
When testing a material such as asphalt cement of an unknown vis-
cosity; the shear stress is the same as that when calibrating the viscometer
tube. The only difference is the measured rate of shear, which
is inversely proportional to the time required to pass a liquid through the
tube. The following relationship can be used to determine the viscosity:
or
V=T(½)
2 2 T
1
where,
VI viscosity of standard material;
Tl time for standard material to pass through the tube;
vz viscosity of unknown material;
Tz time for unknown material to pass through the same
tube; and
V1
calibration factor for tube.
TI
Kinematic Viscosity at 275°F (135°C). ASTM D2170 is used to mea-
sure the kinematic viscosity at 275°F (135°C) using a Zeitfuchs
Cross-Arm viscometer (Figure 2-13). At this temperature the asphalt
cement is sufficiently fluid to flow through the capillary tube under gravi-
tational forces alone, and there is no need to apply any partial vacuum.
This temperature was selected because it approximates the mixing and lay-
down temperatures used in the construction of HMA pavements.
Asphalt cement is poured into the large opening of the viscometer
--------u-1'.l,til-it-r.eaches...the filling line Th.,._viscometer is then placed in a clear oil
bath medium to obtain the equilibrium temperature. A slight vacuum is
applied to the small opening or a slight pressure is applied to the large
opening to induce the flow of the asphalt cement over the siphon section
just above the filling line. Then the gravitational forces cause the asphalt
26 Hot Mix Asphalt Materials, Mixture Design and Construction
FILLING
LINE
ASPHALT
TIMING
MARKS
supplied with the viscometer tube. The unit of centistokes is used in kine-
matic viscosity at 275°F (135°C) because gravitational forces induce flow
and the density of the material affects the rate of flow. Absolute viscosity
(such as obtained at I 40°F or 60°C) can be calculated from kinematic vis-
cosity by multiplying it by the density of asphalt cement.
Example 2-1: An asphalt cement has a kinematic viscosity of 800 centis- ·
tokes. Its specific gravity is 1.03. What is its absolute viscosity in poises?
The penetration test is run in accordance with ASTM D5. The test
can also be run at other temperatures such as 32, 39.2, and l15°F (0, 4,
and 46°C). However, the needle load or penetration time or both are then
varied. At low temperatures, such as 39.2°F (4°C), the asphalt cement is
very stiff and therefore penetration of the needle is significantly lower if it
is loaded with a 100 g weight and is allowed to penetrate for 5 seconds. At
these temperatures, a 200 g weight and 60 seconds penetration time have
been used to obtain penetration values.
Thermometer
J-s,eelBall
~Sample
such as 60 and 39.2°F (16 and 4°C) rather than 77°F (25°C). The pulling
rate is usually lowered to 1 cm/minute when the test is run at 39.2°F (4°C)
to allow for more stretching prior to break.
Durability Tests
Asphalt cements undergo substantial short-term aging (hardening)
when they are mixed with hot mineral aggregates in a HMA mixing facil-
ity. Their long-term aging continues during the life of HMA pavements
which ·are subjected to environmental and other factors discussed later in
this chapter. The following two tests are generally used to approximate the
short-term hardening conditions which occur in normal HMA mixing
facilities:
Centimeter Scale
Asphalt Sample
Thin Film Oven Test. The thin film oven (TFO) test is conducted by
placing a 50 g sample of asphalt cement in a cylindrical flat-bottom pan
(5.5 inches or 140 mm inside diameter and 3/8 inch or 9.5 mm deep). The
asphalt cement layer in the pan is about 1/8 inch (3 mm) deep. The pan
containing the asphalt cement sample is transferred to a shelf in a ventilat-
ed oven maintained at 325°F or 163°C (Figure 2-17). The shelf rotates at
5 to 6 revolutions per minute. The sample is kept in the oven for 5 hours,
------arrd-rhen--transfer-red.-t-0-a-s-1:1:i1;al,le-e0fl-1Ca:ine-F-feF measuring penetration or
viscosity of the aged asphalt cement. The test method is described in
ASTM D 1754. The aged asphalt cement is usually required to meet spec-
ified minimum percentage retained penetration or maximum viscosity. A
loss or gain in weight of the test sample is also measured and reported.
30 Hot Mix Asphalt Materials, Mixture Design and Construction
325•F
Rolling Thin Film Oven Test. A variation of the thin film oven test
is in use by agencies in the western United States, although it has the same
purpose. Figure 2-18 shows the rolling thin film oven (RTFO) used in
ASTM D2872. A specified amount of asphalt cement is poured into a bot-
tle, which is placed in a rack in the oven maintained at 325°F (163°C). The
FAN
MOTOR
1725 RPM
SENSOR
rack rotates at a prescribed rate around a horizontal axis. The rotating bot-
tle continuously exposes fresh asphalt. The orifice of the sample bottle
passes in front of an air jet during each rotation. The vapors accumulated
in the sample bottle are purged by the heated air from the jet.
The RTFO achieves about the same degree of hardening (aging) as the
TFO but in less time (only 75 minutes). Also, it can accommodate a larg-
er number of samples .than the TFO. Some western states use the viscosity
of the aged residue (AR) of the asphalt cement from this test to grade their
paving asphalt cements.
Purity Tests
Refined asphalt cements consist of almost pure bitumen which, by def-
inition, is entirely soluble in carbon disulfide. Only very little amounts of
impurities are generally present in refined asphalt cements. To determine
the purity of asphalt cement, a solubility test (ASTM D2042) is conduct-
ed. An asphalt cement sample of known weight is dissolved in trichloroeth-
ylene (rather than carbon disulfide which is flammable and presents a safe-
ty hazard), and is then filtered through a glass fiber pad. The insoluble
material retained on the pad is washed, dried and weighed. The insoluble
material constitutes the impurity in the asphalt cement sample.
Specifications for paving asphalt cements generally require a minimum of
99.0 percent solubility in trichloroethylene. It is also desirable that the
.asphalt cement is free of moisture because it can cause the asphalt cement
to foam when it is heated above 212°F (I00°C).
Safety Tests
If asphalt cement is heated to a high enough temperature, it gives off
enough vapor to flash (ignite) in the presence of a spark or open flame.
Flash point indicates the temperature to which asphalt cement may be safe-
ly heated without the danger of instantaneous fl.ash in the presence of an
open flame. This temperature is below that of the fire point, which is the
temperature where a material combusts. Although the flash point of paving
asphalt cements is well above the temperatures normally used in HMA pro-
duction, it is necessary to measure and control it for safety considerations.
The Cleveland Open Cup method (ASTM D92) is generally used for
---determinmg-tl1:e~flash-p0i-rtt--e,f'-pav-i-ng-aspha:l-t-cements- (Figure-2-19). A
brass cup partially filled with asphalt cement is heated at a prescribed rate.
A small flame is passed over the surface of the cup periodically. The tem-
perature at which sufficient vapors are released to produce an instanta-
neous flash is reported as the flash point.
32 Hot Mix Asphalt Materials, Mixture Design and Construction
Thermometer
Other Tests
Historical Background
Halstead and Welborn have given a very good account of the history
of the development of asphalt specifications in Reference Q.
From 1878 to 1900 most of the asphalt cement used in the construc-
tion of sheet asphalt, penetration macadam, and mixed macadam (now
called asphalt concrete) was obtained from the Trinidad Lake. The Barber
Asphalt Paving Company was the primary supplier ofTrinidad Asphalt and
builder of asphalt pavements since the beginning of 1878. Bermudez Lake
asphalt found in Venezuela came into use during the 1890s and was a
major source by 1901. Like the Trinidad Lake asphalt, the Bermudez
asphalt had to be mined, and water and extraneous inorganic and organic
matter had to be removed. Both Lake asphalts were found to have remark-
able uniformity.
The first specification for asphalt cement in the United States was
------basecl-0n-t:he-app€a-i:a-11-G€----0-£.th€----Gr-114e-Tri.nidad.. asphalt and.on analytical
tests to determine amounts of bitumen (soluble in carbon disulfide) and
insoluble organic and inorganic matter. Such specifications were devised
merely to identify the source of asphalt cement at the exclusion of other
source materials. As the HMA industry grew and petroleum asphalts came
34 Hot Mix Asphalt Materials, Mixture Design and Construction
into use in the United States about 1900, it became evident that the ele-
ment of uncertainty in material requirements, especially the consistency,
m'ust be removed. The native asphalts were solid or semi-solid materials,
and it was necessary to add fluxing materials to reduce their consistency for
use as binders in HMA mixtures. Standardized methods were needed to
determine the consistency of the paving binders obtained from different
sources.
In 1888, H.C. Bowen invented the Bowen Penetration Machine, the
forerunner of the penetrometer, to determine consistency of the asphalt
cement. Previous to Bowen's invention, the method of testing the proper
degree of softening of the asphalt cement was by chewing. Even after the
invention of the Penetration Machine, the chewing method, crude as it
may now appear, served as a valuable check.An asphalt man generally prid-
ed himself on the fact that by chewing he could achieve approximately the
same results as those obtained by the machine (Q).
After several modifications of the initial Penetration Machine, by 1910
the penetrometer became the principal means of measuring and control-
ling consistency of semi....:solid asphalts at 77°F (25°C). The Bureau of
Public Roads (now the Federal Highway Administration), since 1901, and
the American Society for Testing and Materials (ASTM), since 1903, have
been instrumental in making the penetration test a standard for control-
ling the consistency of paving asphalt cements.
In 1918, the Bureau of Public Roads introduced the penetration grad-
ing system by developing various penetration grades suited to different cli-
matic conditions and applications. The American Association of State
Highway Officials (MSHO) published the standard specifications for
penetration graded asphalt cements in 1931.
The next major change in asphalt cement grading specifications was
initiated in the early 1960s by the FHWA, ASTM, MSHTO, industry,
and a number of state highway departments that wan~ed the asphalt
cements to be graded by viscosity at 140°F (60°C). The primary objectives
were: (a) to replace the empirical penetration test with a rational scientific
viscosity test, and (b) to measure the consistency at 140°F or 60°C (rather
than 77°F or 25°C), which approximates the asphalt pavement maximum
surface temperature on a hot summer day in most of the United States.
Various viscosity grades were developed to suit the different climatic con-
mons an app 1cat10ns. ---·
During the early 1960s when the viscosity grading system was being
developed, the California Department of Highways was developing a par-
allel aged residue (AR) viscosity grading with the cooperation of the Pacific
Coast User Producer Group. AR viscosity grading was based on the vis~
Chapt;r 2-Asphalt Relining, Uses, and Properties 35
cosity of the aged residue resulting from the rolling thin film oven (RTFO)
test rather than the viscosity of the original asphalt cement as recommend-
ed by the FHWA. The California Department of Highways had experi-
enced mix setting problems (tender mix) with some asphalt cements whose
original viscosity did not increase during the plant mixing phase as much
as that of other asphalt cements. Therefore, they wanted to grade the
asphalt after aging in the RTFO so that all asphalts after mixing would
behave about the same during construction.
Figure 2-20 shows the grading systems used in the U.S. during 1985
(.Z). Several states were using more than one grading system. A detailed dis-
cussion of the three grading systems follows. Performance based asphalt
binder grading system developed in the Strategic Highway Research
Program (SHRP) will be presented later.
C.=:J AC Grading
- Penetration Grading
AR Grading
More Than One Method
85-100, are most commonly used for paving in the U.S. Other require-
ments, such as ductility at 77°F (25°C) and retained penetration after thin
film oven (TFO) test, are also listed in Table 2-6.
Cl)
Cl)
w
z
LL
A
LL
77
TEMPERATURE (°F)
Figure 2-21. Temperature Susceptibility of Penetration Graded
Asphalt Cements
38 Hot Mix Asphalt Materials, Mixture Design and Consuuction
Viscosity. 140°F {60°C), P 250: 50 500: 100 1000 :t:200 2000 :t 400 4000 ± 800
Viscosity, 275°F (135°C), min, cSt 80 110 150 210 300
Pene1rat1on, 77°F (25°C), 100 g, 5 s. min 200 120 70 40 20
Flash pomt. Cleveland open cup, min, °F ( C)
0
325 (163) 350 (177) 425 {219) 450(232) 450 (232)
Solubility in trichloroethylene, min,% 99.0 99.0 99.0 99.0 99.0
Tests on residue !rem thin-film oven test:
Viscosity, 140°F (60°C). max, P 1250 2500 5000 10 000 2o ooo
Ductility, 77°F (25°C), 5 cm/min, min. cm 1QQA 100 50 20 10
"'II ductiUty is less than 100, material win be accepted if ductility at 60°F (15,5°C) is 100 minimum at a pull rate ol 5 cm/min,
~~:1~!~:~\~~~ ~2s50
1 65 40 25 20 20
=e~t~~~ 1~;j ~).m~fn 40 45 50 S2
D1.1cti!ity, 77°F (25°C), 5 cm/min, min, cm ~00 8 100 8 75 75 75
Testsononginal asphalt:
Flash point. Cleveland open cup, min. °F (0 q 400 (205) 425(219) 440 (22i} 450(232) 460(238)
Solub1l1ty in trichloroethy!ene, min, 't. 99.0 99.0 99.0 99.0 99 0
"Thin-film even test may be used but the rolling thin-film oven test shall be the referee method.
8 lf ductility is less than 100, matenal wm be accepted II ductility at 60°F (15.5°C) is 100 minimum at a ;,1.111 rate of 5 cm/min.
AC-30, and AC-40. ASTM Table 3 pertains to the AR (aged residue) vis-
cosity grading system, which is discussed later. AASHTO Specification
M226 on viscosity graded asphalt cements is similar to ASTM D3381 with
minor differences.
The poise is the standard unit of measurement for viscosity. The lower
the number of poises, the less viscous the asphalt cement. AC-2.5 (asphalt
cement with a viscosity of 250 poises at 140°F or 60°C) is "softer" than
AC-40 (asphalt cement with a vis~osity of 4000 poises at the same tem-
perature). AC-20 grade is most commonly used for paving in the United
----5tates;-AG=3fl-grade-wa-s---reeently-added-to-the-specifi:cations-and-is-tlsed-by------
some southeastern states. Other test requirements such as penetration at
77°F (25°C), viscosity at 275°F (135°C), viscosity at 140°F (60°C) and
ductility at 77°F (25°C) of the residue from thin film oven test (TFOT),
and flash point are also given in the specifications. The penetration at 77°F
Chapter 2-Asphalt Refining, Uses, and Properties 39
(25°Cj controls the consistency of the asphalt cement near the average ser-
vice temperature, and the viscosity at 275°F (135°C) controls its consis-
tency near the mixing and compacting temperatures. Together these spec-
ified requirements also control the temperature susceptibility of the asphalt
cement as shown in Figure 2-22. The primary difference between ASTM
Tables 1 and 2 (within Table 2-7) is that the asphalt cements in Table 2 are
less temperature susceptible because the specified minimum penetration
values at 77°F (25°C) and the minimum viscosity values at 275°F (135°C)
are higher than those in Table 1.
!
U)
CJ)
I
I
I
I
w
z :,-60 PEN
LI..
LI..
i=
~
>-
I-
U)
0
u
CJ) 1600 T
> I
I
I
I
I
I
I
I
V)
II.I
V)
5 Pen. AC-40
a.
0
: 100
40
50
T
C, AC-20
ea
II.I
:::)
a 50
Pen.
60
T
in 70 Pen. AC-10
~
l-
o
LL.
l-
85
100
Pen.
T
AC-5
a::
G:
~::::
120
l~O
T
PenJ AC-2.5
u 10
b
(l)
r
200
300 T
I-
ii> 5
0
u
1. It is highly regional.
2. It requires more testing equipment (RTFO).
3. Testing time is longer because aged residue from the RTFO test
---------cm=u=s=tc-1b""'e,_o-=-1b~t='ainea first.
4. Detecting contamination is a problem because there are no consis-
tency requirements on original asphalt cement.
5. Aged residue may not relate to hardening that occurs in the drum
mix process.
42 Hot Mix Asphalt Materials, Mixture Design and Construction
Rheological Properties
5. Syneresis; and
6. Separation.
2
-- - -
en
w
en
0 104 . ···----· --···--
--- --- -- ----·-
--
.. --- ------
a..
lJ..
0
0
¢
...
9
8
7
6
5
----- __.
.,,,,,---.--------
,,,:::.---- - -
.-•-- --·
-----------
I- 4 . I
<(
>-
I-
3
ci5 I
0
u
en
'/
/
ASPHALT
ASPHALT
I ----
2 ------ ---
> ASPHALT 3-----
ASPHALT 4-·-·-·-·-
ASPHAL T 5 -··· - · · · -
ASPHALT 6-·-·-·-·-·-
IXING
. . 60 80
TIME - MONTHS
Figure 2-24. Viscosity at 140°FVersus Time in Months (after Kandhal
et al, B)
44 Hot Mix Asphalt Materials, Mixture Design and Construction
T
~Y=-- (Eq.1)
a+bT
or
T
- =a+ bT (Eq. 2)
~y
where,
l:!.Y = change in test property (such as penetration and viscosity)
with time T or the difference between the zero-life value
and the value at any significant time;
T = time;
a =constant, the intercept of the Eq. 2 line on the ordinate;
b =slope of the Eq. 2 line; and
lib = the ultimate change (limiting value of change) of the prop-
erty at infinite time.
Both percent retained penetration and aging index have been used to
evaluate relative aging of asphalt cements of different grades and/ or from
different sources.
The thin film oven (TFO) test and rolling thin film oven (RTFO) test
generally simulate the asphalt cement hardening which occurs in the pug-
mill of a HMA batch facility. Because the asphalt cements from different
sources age differently in service after pugmill mixing, the tests on TFO or
RTFO residue are not completely reliable to predict the long-term asphalt
aging or durability. There is an urgent need to develop an accelerated lab-
_ _ _ _ _ _o_r_a_to_ry_,__a-ogi_gg procedure for asphalt cements that can closely simulate the
hardening that takes place in the pavement under certain climatic condi-
tions (15). Field observations indicate that the asphalt cement does not age
as rapidly in a drum HMA facility as it does in a batch facility because of
the presence of water vapor in the drum, which reduces oxidation.
., 46 Hot Mix Asphalt Materials, Mixture Design and Construction
tent, age of the pavement, traffic, and climatic conditions also influence
this general relationship between penetration and cracking.
70
60 -BEAVER CO.
-------LYCOMING CO.
LL -·-WASHINGTON CO.
I'- -···-LEBANON CO.
I'- 50 X CRACKING FIRST OBSERVED
,, I-
<(
z ' ' ' ·...
0 40 ',~
I-
<(
et::
I- ~
w 30
z
w
Q..
·--·---
20.
·--~"::: ::.::.::.it~.::=::.-:..- -
10
0 20 40 60 80 100 120 140
LIFE OF PAVEMENTS, MONTHS
Figure 2-25. Penetration Versus Time in Months (after Kandhal, 12)
that is, the ratio of shear stress to shear strain rate is constant. At low tem-
peratures, the ratio of shear stress to shear strain rate is not a constant, and
the asphalt cements behave like non-Newtonian liquids. Unlike the empir-
ical tests of penetration and ductility, viscosity is a fundamental consisten-
cy measurement in absolute units that is generally not affected by changes
in test configurations or geometry of the sample (.25). Standard ASTM test
methods which use capillary viscometers are available to determine asphalt
cement viscosity at 140°F or 60°C (ASTM D2171) and 275°F or 135°C
(ASTM D2170) on a routine basis. Large vacuum capillary viscometers
have been used to test asphalt cement viscosity at 77°F (2£). However, the
method needs to be refined further. The cone and plate viscometer is used
in ASTM D3205 to measure viscosity in the range from 103 to 10 10 poises
and is therefore suitable for use at temperatures where viscosity is in the
range indicated. The shear rate may vary between approximately 103 to 10 2
sec1, and the method is suitable for either Newtonian or non-Newtonian
asphalt cements. This method is not used routinely at the present time.
Asphalt cement viscosity at 140°F (60°C) has some influence on the
performance of HMA pavements during hot summer days when the pave-
ment surface temperatures are near 140°F (60°C). A low viscosity at 140°F
(60°C) can induce flushing and/or rutting if other factors are identical.
This is especially true for mixes that have a relatively high asphalt content.
Therefore, proper viscosity grade selection appropriate to prevailing cli-
matic conditions must be made. For example, use of AC-10 in the south-
ern part of the United States will most likely result in an increase in flush-
ing and/or rutting ofHMA pavements when compared to AC-20. On the
other hand, an unusually high viscosity (or stiffness) at low service tem-
peratures (such as -20°F or -29°C) in the northern portion of the United
States) will result in non-load associated transverse shrinkage cracking.
It has been observed that aging of the HMA pavement results in pro-
gressively higher viscosity with time, as shown in Figure 2-24. However,
vis~osity of the aged asphalt cement alone does not necessarily control .the
pavement performance. Aging indexes based on viscosity at 77°F (25°C)
were found to be more meaningful to indicate comparative aging and pave-
ment performance of test pavements in Pennsylvania. Ravelling was
------,obs&FV-e-d--GH1-t.h~™-pa-¥emts-w.h.en...this-agingJndex exceeded 12 (11).
PI= 20 - 500A
1 + 50A.
Chapter 2-Asphalt refining, Uses, and Properties 51
Note that the temperatures used in the equation are in degrees centigrade.
The lower the PI value of an asphalt cement, the higher its tempera-
ture susceptibility. Most paving asphalt cements have a penetration index
between + 1 and -1. Asphalt cements with a PI below -2 are highly tem-
perature susceptible, usually exhibit brittleness at low temperatures, and
are very prone to transverse cracking in cold climates.
L-X
PVN = - - (-1.5)
L - M
where,
The viscosity values of L and M can be read from the graph (Figure
2-26) developed by McLeod by plotting penetration at 77°F (25°C)
against viscosity at 275°F (135°C) for "typical" asphalts. This graph con-
tains two lines with the assigned PVN values of 0.0 and -1.5, which are
used. However, the following equations (based on least squares line) can be
used to calculate more accurate values of L and M:
....
~ 5000 1-...-,PV---'~---+--4-<<1--E-."·~,.........,Haa:::+:--+=:.;;;;:::1-+-+-+-+·;;.;;i··1---+--+-+-++t-++1 C
~ 3000 ~"o.-"-,;;;::--+--F.i
.. f--M--+A---+:.'"4·::r--llt++<::::.._.+--1-+-fwi<···l----+--+-+-t-H+H
Cl) 2000 i-0. ~ , , I .. ···· ···· ....
0 1
I- ,- 'o. ,""'~~'-F-<81-+.=l--+t;;r--+.-····+--t---····+·--HH-+:;;;a:::1---+-+-+--t-H-++I
rau I000---a ,,, r--... ,,r-. = ..
----: •.o - . .
.... ..... . ..
~ I 00 1--4n,50 --· ..
>-
I-
Cl)
0 --,so12ou - ·· :\ ,~ ';1:,, _
u
Cl)
3 0 ,..300/40r • .. ···· ···· ::: ', .. ,....-~--+-+-+++11
20 --aoo11000· ·· ···· ···· ··· · ~,.,. "'o
> I 1 •• •··· •••• ••• I~♦.;.~...
10 : I i:-: - ; -frH,-Hf-- 1-1:-:-:-:11----1--1 -1~--
Where Vis the viscosity in centistokes at 275°F and Pis the penetration at
770F.
L-X
PVN = - - (-1.S)
L - M
The lower the PVN value of an asphalt cement, the higher its temper-
ature susceptibility. Most paving asphalt cements have a PVN between
+0.5 to -2.0.
Although PVN can readily be determined from the specification tests,
some asphalt technologists believe that the temperature susceptibility
obtained in the higher temperature range (77-275"F) cannot be extrapo-
lated to be applicable at temperatures below 77°F. One notable difference
between PI and PVN is that the PI changes on aging (during mixing and
subsequently in service), whereas the PVN remains substantially the same
(22).
below 140°F (60°C). The numerical differences between the VTS values of
different asphalt cements are not large. The VTS value ranged from 3.36
to 3.98 for 52 asphalts sampled in the United States (3.Q). Very little use of
VTS is found in the literature.
CD
15
0
1/)
Elastic Behavior
Delayed or Retarded
Elastic Behavior
DI
E
-
~
0
CTI
Cl)
..
1/)
1/)
..
Q) (I) At Short Loading Times• S = E
S (t, T) = cr/£
where,
S = stiffness, in psi or kg/ c~;
· cr = axial stress;
£ = axial strain;
t = time of loading; and
T = temperature.
1. Indirect Methods: Van der Poel conducted extensive creep and dynamic
tests using well compacted (approximately 3 to 5 percent air voids)
Chapter 2-Asphalc Refining, Uses, and Properties 57
10•-----------------------
.E
u
0, 0
..>::
Cl) I0 4 1-----:i.-----,,,.'--~-~------I------I
....
.2
X
~
-0
II)
II)
1031---..,C....--IIC--..,C....------1------1----....,..-1
Cl)
-
.E
U)
10----.i...i...i.i.......i..--...i..i.~.i......i...i..i..i.....i.--i....1..1.......i......i........
10 10 2 10 3 10 4 10~ 10•
2
Stiffness of Asphalt, kg/cm
.. •
r~ .
i~ •
fw;I~
I
.IA ;.
~
;
r~ 1 .
!: ::;~:l
!!i2 1 ~
a
!f; l C
0 , . . 11 ......L
]~! ..
iI .
r•o:!ti
~ i~
a:0.1-.::
..
~,,id
1 .
r~··
m~~
!i.1i::
_J
OlflllCll,l.'nl,l,JN3,,I
10 3
8
6
Cv = 0.90
4
o.aa'\..
.-0~
o.~§, '\
2
if;-_ 0 0,t:S!
r---......0.8~
~ (\
2
10 -0.74 -...... ...
- .....- ........ ' "
"'-
8 o,.,zg ... ' "
-
~
-----... ~~ ~"""
~ ~~
\\
"
::::--.......-........:-
---------
en 2 Cv=0.60~ ~
10
- ~ ~ ..._,~
__ ,
8 -- -......
-......
,~~
~ ,,
.......
~
........_,
6
4 '
2
I
----------1-Ge-... ----1----+110 I0 2 3
I Qe-------110~ 10 9
sbih kg/cm2
Figure 2-30. a Function of sbit and Cv (after Heu.kelom
smix / sbit as
and Klomp, JJ)
Hot Mix Asphalt Materials, Mixture Design and Construction
60
neous results are possible when the R&B softening point is used for waxy
asphalt cements.
McLeod (2.i.) further modified the Heukelom method for determining
the stiffness of asphalt cement. The major difference is the use of pen-vis
number (PVN) by McLeod in lieu of penetration index PI (pen/pen) used
by Heukelom. Figures 2-31 and 2...:32 are used to determine asphalt stiff-
ness (28). McLeod's method requires measurement of asphalt cement pen-
etration at 77°F (25°C) and viscosity at 275°F (135°C). Stiffness moduli
135 400
w
a: 130
::,
~ 125 ii!:
a:
w 120 w
~ 115
a:
::,
w
I-
110 ~ ~
105 "'
W
ffia..
~ ~
w
1:1:! 1- 100 <t 1-
::,~
~w ~ ~
~ / z
ffi~ 90 ,;r / ~
~
~~ 85 ~ / >-
z<(
I-~
W I-
80 ~
;s-i? .,.6u / /
~ w 75 ~~ ~
~~ 10 I~ "s / 20 ~·
z w
~
LP~ .,. . / ~
w~
!;!;! I-
,... !1:- ~
o,3"
~
..,,..3' 15 0
I- a: 60 ii.~- . .,. , ;!;
~ 55
~ @ 50
e ""~
Js,(i) ,,",
,, ,? 10 f3
8 3
<(
ii!:~ 45 R:-(,ij O >
w .J ~ 6
~~ 40 ,!/, 5
I-
fa1-
ffiw 35 -9"' ,? 4
z
lt 30 _p '.3
3
0
0
25 /.!ff GlYEt:I ~
a:
~ <t PENETRATION AT 25"C=90 1-
~ 20 PEN-VIS NUMBER=-1 ·0 2 w
zw
~ I5Il:::IEli a..
------'W----t-------J=R.9M-GRAP't1--------
~ IO TEMPERATURE DIFFERENCE=20"C
~ 5 A~
O BASE TEMPERATURE•25•20=45"C
Figure 2-31. Relationship Between Penetration, PVN, and Base
Temperature for Asphalt Cements (after McLeod, 2!l)
Chapter 2-Asphalt Refining, Uses, and Properties 61
Despite the foregoing limitations, the indirect methods can be used for
initial estimates of stiffness modulus if direct testing is not possible.
2. Direct Methods: At high temperatures, viscosity can be used as a measure
of asphalt cement stiffness. At 140°F (60°C) and higher, capillary vis-
cometers are suitable because of the Newtonian behavior of asphalt cement
in these temperature ranges. However, at low temperatures, a viscometer
must be able to quantify viscosity as a function of both temperature and
shear stress or rate. This complicated factor has resulted in the develop-
ment of a large number of devices for measurement of shear-dependent
viscosities (25).
In 1974, Schweyer (16) presented a pictorial overview of many exper-
imental viscometers that have been used to study the viscoelastic behavior
of asphalts. These included (a) rotational types, which utilize coaxial cylin-
ders or a cone and plate; (b) rheometers, a specialized capillary type in
which a piston is used to drive the asphalt through a capillary tube; and (c)
miscellaneous types such as the sliding-plate microviscometer. Schweyer
and others have done considerable work on the use of the capillary
rheometer (32) and the development of several generations of the con-
stant-stress rheometer Ufl), which has been used to examine both asphalt
cement stiffness and viscosity at low temperatures.
The methods most widely used to determine the asphalt cement vis-
osity-adow tempera:tnres-are-proba:b1y--clte-cone and plate viscometer or
the sliding-plate microviscometer. Both were adopted as ASTM standard
test methods and are capable of examining viscosities over a wide range of
shear rates. ASTM has recently adopted a capillary tube method (D4957)
for measuring the apparent viscosity of non-Newtonian bitumens.
62 Hot Mix Asphalt Materials, Mixture Design and Construction
10 1001,00010,000
VISCOSITY, CP
I \
I \
I \
,BASE TEMl:ER.4TURE, \
•c ABOVE BASE T ~ T/ 'C BELOW BASE TEMPE{fATURE
,'9.?, ~, ,, ~~}.P,.~?..~~.'f,.,~.. ~,?., :A.~~,,,., . ,~
DIFFERENCE IN 'C BETWEEN ANY SPECIFIED TEMPERATURE
/ AND BASE TEMPERATURE \
I \
I \
I \
I \
I \
I
I \
I \
\
I \
I 2 4 2
\
/ I kg/cm =14-2psi=9-81 xi 0 N/m \
I \
I \
I
FREQUENCY,CY/SEC
\
'ioo _ 10 .. _ 1' I SEC I MIN
- - - - - , · 864 --2- - - - - - . . - 2 - - - - - - - - - - - - - - - - - - - - - - - - - - - , - - - - - -
24A 24A 24~ 2 4 • 24U 248
0.001 Q.01 O,I I 1,000 10,000 100,000 1,000,000
LOADtlG TME, SEC
ORIGINAL RHEOMETER
APPLIED 0
LOAD
MODIFIED RHEOMETER
MODIFIED
~~-1'-BACK-PLATE
r---,_...._, CLAMP
MODIFIED
BACK-PLATE
Cracking. Over the years, engineers have been able to categorize crack-
ing under two broad groups: load associated and nonload associated,
although most pavement cracks can be described according to their geom-
etry such as longitudinal, transverse, polygon (also alligator and map) and
block, or by the mechanism that causes the cracking such as slippage,
--~hrinkage..a.nclr.eflec.tio.n_(.ii)_._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _~
I. Load-associated cracking
The principal class of load-associated cracking has been described as
fatigue cracking (alligator cracking), the phenomenon of fracture under
Chapter 2-Asphalt Refining, Uses, and Properties 65
I
I
, I
,,
I
w
0:: I
w
,, •
I
> CORR. COEFF. =- 0.92 I
w , 1-1 , I
I,
,, ,
Cl) I
, I
I
, ,,
,
2-1,
. ,
I
I
, I
, I
(!) w , I
, I
z I- I
,,
I
~
:J w I
I
,,
I I
..J
w a
0
2-2 ,
I
> , ,,
I I
~,, ,
::I
<
0:: I
I
I
1-2/
I ••
u.. / 3
0 ,,
I
z
I- I-
:r:
, I
w S2 ,'7
I- ...J
(/)
,
I
X I
W, , I
,,
I
I
I 6 ,
r
I
I
.,, •'
,, ,,
I
I
,
w
z , ,I
0
z ,,
6 8 10 12 14 16 18
AIR VOID CONTENT, PERCENT
Figure 2-34. Air Void Content Versus Extent of Ravelling
(after Kandhal, 43)
This means that asphalt cements which become hard through aging
will demonstrate poor fatigue life (develop fatigue cracking) in thin HMA
surfacings.
l
Chapter 2-Asphalt Refining, Uses, and Properties 67
mixes which have high stiffness modulus at low temperatures are very
prone to cracking. Mix stiffness at low temperatures is primarily dependent
on the stiffness of the asphalt cement. Therefore, high asphalt cement stiff-
ness at low temperatures is the predominant cause of this type of cracking.
Other factors such as thickness of the HMA layer and subgrade soil type
influence the frequency or extent of these cracks.
Laboratory research supported by field observations indicates that
asphalt cement consistency, largely reported in terms of penetration at
77°F (25°C), and temperature susceptibility (PI or PVN) are the major
asphalt cement properties influencing low temperature cracking.
Increasing the asphalt cement penetration is suggested as the most direct
designer-controlled variable for minimizing cracking. However, care must
be exercised to avoid asphalt cements that are highly temperature suscepti-
ble (that is, asphalt cements that could become extremely stiff at low tem-
peratures regardless of the initial penetration). Table 2 in MSHTO
Specifications M226 for Viscosity Graded Asphalt Cements generally pro-
vides an asphalt cement with lower stiffness at lower temperatures than
that provided in Table 1. However, Table 2 is not adequate for most states
in the northern portions of the United States and Canada. More stringent
specifications to control the temperature susceptibility of asphalt cements
are needed to minimize or eliminate low temperature cracking.
Many researchers have recommended limiting stiffness values of
asphalt cement (40, 46, 1ZJ or HMA (111) for a particular temperature
range based on field experiments. The Pennsylvania Department of
Transportation modified Table 2 of AASHTO Specifications for Viscosity
Graded Asphalt Cements in 1977, based on a limiting stiffness modulus of
275 kg/cm2 (3900 psi) at the minimum pavement temperature of -10°F
(-23°C) and 20,000 sec loading time (1Z). The limiting stiffness modulus
was established based, on the performance of six test pavements. Figure
2-35 shows two test pavements in adjacent lanes. One developed extensive
transverse cracks. during the first winter and the other was free of cracks.
Using the limiting stiffness of 275 kg/cm2 and McLeod's nomograph
- - - - - - - < , - • · - k 'ethod..~~ roioiro11m_allo:wable...EY1Sls_w.e.r.e_determined for va,rious pen-
Specified Minimum
MinimumPVN Viscosity at 135°C
Penetration at 25°C Allowable centistokes
60 -0.80 390
65 -0.95 330
70 -1.10 290
75 -1.25 250
It can be seen that higher temperature susceptible asphalt cements can
be used if their penetration values get higher in accordance with those list-
ed above. At higher penetration values, the temperature viscosity lines are
shifted relatively lower at 77°F (25°C), thus the asphalt cement does not
exceed the limiting stiffness modulus at -10°F (-23°C) in spite of steeper
slopes (higher temperature susceptibility) of these lines.
r.
L
Chapter 2-Asphalt Refining, Uses, and Properties 69
stiffness of lxl0 9 N/m2 (145,000 psi) at 0.5-h loading time (that is, equiv-
alent to approximately 6xl0 8 N/m2 at 2.0-h loading time; 5xl0 8 N/m2 at
10,000-s loading time). Readshaw (5JJJ based the British Columbia
asphalt cement specifications on a critical asphalt cement stiffness of 2xl08
N/m 2 at 2.0 h loading time, a value arrived at largely from the St. Anne
Test Road data. Fromm and Phang (1§J suggested l.4xl0 8 N/m2 (that is,
20,000 psi) at 10,000-s loading time as the critical asphalt cement stiff-
ness.
Gaw (if!) has recommended that the low-temperature asphalt cement
specifications be based on a specific nomographic cracking temperature
selected from Figure 2-36, at a performance level appropriate to the antic-
ipated minimum ambient temperature or pavement surface temperature
for the region. The asphalt cement specification limits are represented by a
single line drawn on a logarithm plot of asphalt cement penetration at
25°C (77°F) and 5°C (41 °F) similar to Figure 2-36. Candidate asphalt
cements for which the penetration at 25 and 5°C appear on, or to the right
u
0
300
10
(\J
.,_ 200 I
<1: I
I
z 150
0
.,_
<1: 100
.,_
0:::
80 I
I
w I
I
z -25 I
w 60 I
I
a.. I
I
I
I- I
I
_J 40 I
<1: I
I
I I
a..
CJ)
I
I
I
I
<1: I
I
-0- - 3 4 6 8. 10 15~2~0___ 30 40
2
ASPHALT PENETRATION AT 5°C
· Figure 2-36. Nomograph for Predicting Cracking Temperatures in °C
from asphalt Penetration (after Gaw, 1!l)
70 Hot Mix Asphalt Materials, Mixture Design and Construction
of, the specification line are considered acceptable. Both unaged and aged
asphalt cement penetrations yield similar nomographic cracking tempera-
tures, and either can be used to establish the specification limits. This
approach requires the penetration measurement at one additional temper-
ature (that is, at 5°C).
L
Chapter 2-Asphalt Refining, Uses, and Properties 71
usually in the presence of moisture between the aggregate surface and the
asphalt cement in a HMA pavement or mixture. Strength of the mixture is
derived from the cohesional resistance of binder and grain interlock and fric-
tional resistance of the aggregate. The cohesional resistance is only fully avail-
able if a good bond exists between the binder and the aggregate (52). If the
bond is poor, failure occurs at the binder-aggregate interface and may result
in premature failure of the mixture and the HMA pavement.
There may be as many as five different mechanisms by which stripping
of asphalt cement from an aggregate surface may occur. Those five mecha-
nisms include detachment, displacement, spontaneous emulsification, pore
pressure, and hydraulic scouring (.53). The stripping phenomenon is dis-
cussed in detail later in the book.
Stripping is a complex problem depending on many variables, includ-
ing the type and use of mix, asphalt cement characteristics, aggregate char-
acteristics, environment, traffic, construction practice, and the use of anti-
strip additives; however, the presence of moisture is the common factor to
all stripping (23, 51, 5..i).
The chemistry of both the asphalt and aggregate at the as-
phalt-aggregate interface plays an important and primary role in the strip-
ping phenomenon. The effect of asphalt cement characteristics especially
the physical properties, although secondary in role, is discussed here. The
most frequently referenced relationship between the characteristics of the
asphalt cement in a paving mixture and the tendency of the mix to strip
relates stripping resistance to the viscosity of the asphalt binder in service.
Binders of high viscosity have generally been observed to resist displace-
ment by water much better than those of low viscosity. Low viscosity, how-
ever, is desirable during mixing operations because a low viscosity fluid has
more wetting power than one of high viscosity. Moreover, high viscosity
asphalt cements cannot be used in many instances because of other con-
siderations such as low-temperature cracking in cold regions and potential
reduction in fatigue life of the surface courses. There is a need to under-
stand the fundamentals of aggregate-asphalt adhesion so that the problem
can be minimized by other means rather than increasing the asphalt
cement viscosity, which is not effective in all cases and which may result in
otl:rerp-e-rformance-probtem. ------· - -
72 Hot Mix Asphalt Materials, Mixture Design and Construction
1. Tests such as penetration and ductility are empirical and are not
directly related to HMA pavement performance. Any relationship
between the empirical tests and performance has to be gained by
experience as discussed previously.
2. Tests are conducted at one standard temperature such as penetration
at 77°F (25°C) and viscosity at 140°F {60°C) in spite of different
climatic conditions at project sites or geographical regions.
3. The entire range of pavement temperatures typically prevalent at a
project site is not considered. For example, there is no test method
or specification requirements for asphalt cement stiffness at low
temperatures to control thermal cracking.
4. Current test methods and specifications consider short term aging
of asphalt cements during HMA production and construction by
conducting thin film oven (TFO) or rolling thin film oven (RTFO)
tests. However, long term aging of asphalt cements in service is not
considered. As discussed previously, asphalt cements continue to
stiffen during service thereby affecting the HMA performance in
terms of fatigue cracking and low temperature cracking.
5. Asphalt cements can have different temperature and performance
characteristics within the same grading system whether it is a pene-
tration grading or a viscosity grading system.
6. Current tests and specifications have been developed for asphalt
cements based on experience. They are not suitable or applicable to
modified asphalt binders which are gaining. in popularity. See
asphalt additives and modifiers in Chapter 7.
_I_
Chapter 2-Asphalt Refining, Uses, and Properties 73
Rolling Thin Film Oven (RTFO). The rolling thin film oven
(RTFO) test, AASHTO T240 and ASTM D2872, was described earlier in '
this chapter. Like the thin film oven (TFO) test, the RTFO simulates the
asphalt binder aging during the manufacture and construction of HMA
I
Table 2-8. Superpave Asphalt Binder Testing Equipment and Purpose
8 8-: ....,
e:... ::r, '"O
• 'Tj~ ~ el "'
~
>-g::io...,::ro
>;rlrl(bo-8
8"' 0 0 : 3 ... , 0
Equipment Purpose Performance Parameter I .8. ~ ~ e:. 5.. g
.... ;:I (b " '
.... o' ;:I A- .....
I>,) .... ~ .... 8 ....,
Rolling Thin, Film Oven (RTFO) Simulate binder aging (hardening) during Resistance to aging (durability ) during s,. 8
(b (b
::r~
... • (b ::r ::r
I
HMA production and construction construction .... P..8--::;i ~ (b
! s,. ~ rl (b rl :;q
I>,) ::rP-~....,
;:i::t"o,-., 'Tl
.... . ;:I A-
Pressure Agjg Vessel (PAV) Simulate binder aging (hardening) during Resistance to aging (durability) during ::n
< g.. g:
;:I
s,. e: • 0
o,-.•p,.o>-t~
HMA service life service life
5~~ e; g:ie; ::r:
~
i:: o::r ~ "'
.... p,.'"O
..., 0-
. Rootfoo,l lnmct<t (RV) Measure binder properties at high con- Handling and pumping " ' ~ o e:.,.P-h
o
ri I>,) "' rl i:: (")
...a (fQ ;:I 0- ::i. ~
2:::
x·
i);-
struction temperatures s. s· g s·~ A- "Cl
::r
ri (£Q 8 Ft" ...,
o- !'¾-
A-p.,o>-t@:~ 2:::
'::I") ;:I ~ A- .... I>,)
Dynamic Sh~ar Rheometer (DSR) Measure binder properties at high and Resistance to permanent deformation 0 A- Vi:: ;:I i:: ~
::! 0::: ~- (b (£Q ~
i
intermediate service temperatures (rutting) and fatigue cracJcing I ~'--'P-8 ?,....._
.--o 1--t..._.,,e,. 2:::
0
...., rl (b
"Ti r-t rn }:::: S
:=;· ~-
Bending Beajn Rheometer (BBR) Measure binder properties at low service Resistance to thermal cracking 0 ~;:I0::ip,.n .... 0 ~
0
rl"'rlCtQ,-.,0 t;)
temperatures I ~ 0 e:... I>,) ~ g
r-t a o 0a.~ i::s· ~-
!,l
'-< ~
~
V
Direct Tension Tester (DTT) Measure binder properties at low service Resistance to thermal cra~king '--I I>,) ? ;:;.; e:... A.
Vl ;:I,....._ i:: ~ ()
temperatures I 9'-<~~
,..,1:i::fl-~
;:I .... ;:I '"O
0
~
:::C
gj;3">~&;~ 0
a.
~@ if9 ~ 0
::,
Chapter 2-Asphalt Refining, Uses, and Properties 75
The RTFO test provides an aged asphalt binder for further testing by
the dynamic shear rheometer (DSR), and allows the determination of the
mass of volatiles lost from the binder during the test. The amount of
volatiles lost indicates the amount of aging that may occur during HMA
production and construction. Some asphalt binders gain weight, rather
than lose weight, during the RTFO aging because of the oxidative prod-
ucts formed during the test.
Pressure Aging Vessel (PAV). Pressure aging has been used in asphalt
research and for aging rubber products for many years. A method involv-
ing the pressure aging was developed by Dr. D. Y. Lee at Iowa State
University for long term aging of asphalt cements, and is currently a stan-
dard method for aging rubber products (ASTM D454 and ASTM D572
in Reference 5.fi). Pressure aging of asphalt binders has the following advan-
tages: (a) limited loss of volatiles, (b) the oxidative process is accelerated
without resorting to high temperatures, (c) an adequate amount of asphalt
binder can be aged at one time for further testing, and (d) the test is prac-
tical for routine laboratory testing.
The pressure aging vessel (PAV) was developed in SHRP to simulate
the asphalt binder aging that occurs during 5-10 years of in-service HMA
pavements (i2). Since the asphalt binder in the HMA pavement has
undergone short-term aging during production and construction, the PAV
is used to age RTFO residue. Figure 2-37 shows a schematic of the PAV
and related accessories.
air
pressure
temperature
probe
/ asphalt
~
pressure vessel sample rack sample pan
The RTFO residue is exposed to high air pressure and temperature for
20 hours in the PAV to simulate the effect oflong term aging. A cylinder
of dean, dry compressed air with a pressure regulator, release valve, and a
slow-release bleed valve is used to supply and regulate air pressure. The
PAV composed of stainless steel must be able to operate under the pressure
(2070 kPa) and temperature conditions (90°, 100°, or 110° C) of the test.
The vessel accommodates a sample rack on which ten sample pans can be
placed for aging. The PAV is placed in an oven for 20 hours to maintain
the desired temperature inside the vessel. PAV equipment which incorpo-
rates the pressure aging chamber and the oven built as one unit, is also
available.
RTFO-aged asphalt binder is heated until fluid, stirred and pla~ed in
the PAV pan which can accommodate a 50-gram sample. The sample rack
containing sample pans is placed in an unpressurized PAV preheated to the
desired test temperature depending on the design climate. The lid of the
PAV is secured quickly to minimize heat loss. When the PAV reaches with-
in 2°C of the desired temperature, a pressure of 2070 kPa is applied using
the valve on the air cylinder. After 20 hours, the air pressure is released
slowly (over a period of 8-10 minutes) using the bleed valve. A quick
pressure release may cause foaming in the binder samples. The sample
pans are then placed for 30 minutes in an oven maintained at 163°C to
remove entrapped air from the samples. The samples are then stored for
further testing (such as dynamic shear rheometer, bending beam rheome-
ter, and direct tension tester).
torque
spindle
control
keys spindle extension
ermo;;
container
1000 cP = 1 Pa.s
Therefore, the Brookfield viscosity output in cP is multiplied by 0.001
to obtain viscosity in Pa.s.
Most agencies use equiviscous temperatures for mixing and com-
paction during HMA mix design (see Chapter 4). This can be accom-
plished by measuring rotational or Brookfield viscosity at one more tem-
perature other than 135°C; and plotting the temperat~re-viscosity curve.
sented by arrows 1 and 2 in Figure 2-41), that is, they simultaneously act
like elastic solids and viscous liquids. When viscoelastic materials are
loaded, part of the deformation is elastic (recoverable) and part is viscous
(non-recoverable). The DSR gives a complete picture of the behavior of
asphalt binders at pavement service temperatures by measuring both G*
ando.
Viscous
Part
Viscous
Part
G"
G"
G'
Elastic Part
Elastic Part
Asphalt A Asphalt B
Viscous Behavior
Elastic Behavior
_ _ _ ____,F.....igur.e...2d:l.-Yisco_us__andElastkBehavior of Asphalt Binders (i6)
Figure 2-40 shows two asphalt binders with the same complex modu-
lus G* (indicated by length of the diagonal) but with different phase
angles ◊. However, Asphalt B has a larger elastic component compared to
80 Hot Mix Asphalt Materials, Mixture Design and Construction
Asphalt A. When both asphalt binders are loaded, Asphalt B will display
more elastic (recoverable) deformation and less viscous or non-recoverable
(permanent) deformation compared to Asphalt A. This example demon-
strates that complex modulus (G*) alone is not sufficient to characterize
asphalt binders, phase angle (3) is also needed.
Applied Stress
Osei!::~~))
Position of
Oscillating Plate B
C
1 cycle
v•mo
Applied
:,_v•
I,
Shear
Stress I
Ii time lag = 8
Resulting ~
Shear
Strain
2T
nr
'Ymax
' er
h
where,
8 = time lag
Resulting
Shear
Strain--------·--------
time
j
Figure 2-44. Stress-Strain Response of a Viscoelastic Material
In constant stress rheometers used for Superpave binder testing, the con-
stant stress used in testing should be such that the resulting shear strain is
Chapter 2-Asphalt Refining, Uses, and Properties 83
Rutting Parameter:
and thus the more resistant to rutting the asphalt binder will be. The lower
the o value, the more elastic the asphalt binder.
The increased elasticity makes the asphalt binder more resistant to per-
manent deformation. The high temperature measurements for rutting are
made on the original and RTFO aged samples. The RTFO aged samples
represent the. asphalt binder in the early stages of its life immediately after
placement and before long-term oxidative aging begins. Since long term
oxidative aging increases the stiffness of an asphalt binder, it continues to
increase the binder's resistance to rutting. Testing of the asphalt binder
prior to long term oxidative hardening, therefore, is critical with respect to
rutting. The specification for DSR testing of the original asphalt binder
was added as a safeguard against those asphalt-binders whose RTFO aged
residues do not indicate the actual aging during HMA production and
placement in the field.
The SHRP researchers considered rutting as a stress controlled, cyclic
loading phenomenon in determining the rutting parameter chosen for
specification purposes. With each traffic loading cycle, work is being done
to deform .the HMA pavement surface. A part of this work is recovered by
elastic rebound of the surface while some is dissipated in the form of per-
manent deformation and heat. In order to minimize permanent deforma-
tion (rutting), the amount of work dissipated during each loading cycle
must be minimized. Mathematically, the work dissipated per loading cycle
at a constant stress can be expressed as follows (§fl}:
W =1txcr 2 1 ]
C O [G* /sino
where
o
G* complex modulus,
= phase angle. j
1
This equation indicates that the work dissipated per loading cycle is ·,
mversely proportional to G*/smo. For this purpose, the Gorysino_p_ar_am--e--- - - - - - - -
ter was chosen as a Superpave asphalt binder specification. It can be seen
from the preceding equation that the work dissipated per loading cycle can
be decreased by either increasing the value of G* and/or decreasing the
value of o. This appears logical because increasing the G* of an asphalt
Chapter 2-Asphalt Refining, Uses, and Properties 85
binder makes it stiffer and thus more resistant to rutting. Also, decreasing
the O of an asphalt binder makes the asphalt binder more elastic, again
increasing its resistance of rutting.
Fatigue Parameter:
where E is the strain and the other variables are as previously described.
This equation indicates that as G* and/or o are increased, more work will
be dissipated per traffic loading cycle. As G* decreases, the asphalt binder
becomes less stiff and thus able to deform without building up large stress-
es. Additionally, asphalt binders with low o values will be more elastic and
thus can regain their original condition without dissipating work. Work
can be dissipated in several different ways; heat dissipation, plastic flow,
cracking, and crack propagation can all dissipate work. However, all of •
these contribute to pavement distresses. The lower the amount of energy
dissipated per cycle, the lower the likelihood of fatigue cracking or any
other damage phenomena to occur. The G*sino parameter was, therefore,
chosen for Superpave specification purposes to limit the total amount of
energy dissipated thereby minimizing fatigue cracking @. PAV aged
samples are tested in the DSR for determining the asphalt binder's ability
to resist fatigue cracking. As the asphalt binder ages during its service life,
it becomes stiffer thus causing it to become more susceptible to fatigue
cracking. PAV aged samples are, therefore, used to test the asphalt binder
to simulate the most critical in-service condition.
nere are tlrree-s-ep-a:rate--specification--requirements for the DSR test
parameters. When the unaged asphalt binder is tested, the G*/sino value
must be a minimum of 1 kPa. When the RTFO residue is tested for the
same asphalt binder, the G*/sino value must be a minimum of 2.2 kPa.
These two minimum limits were established for rutting purposes as previ-
86 Hot Mix Asphalt Materials, Mixture Design and Construction
ously discussed. The third DSR specification requirement is that the G*sin
8 value of the PAV aged asphalt binder must be no more than 5000 kPa.
This maximum specification limit was established to address fatigue crack-
ing in HMA pavements.
Each of these specification limits was established by consensus of the
Federal Highway Administration Asphalt Binder Expert Task Group
(ETG). This group consists of persons from industry, specifying agencies,
and academia. The group fir;st established the minimum limit of 1 kPa for
original asphalt binders. This was done by referencing the conventional
viscosity data for paving asphalt cements at 60°C. This pavement service
temperature was thought to be representative of a moderate climate where
the use of AC-10 viscosity graded asphalt cement (viscosity of 1000 Poises)
would be common. When unaged AC-10 asphalt cements are tested in
the DSR at 10 rad/second, their G*/sin8 values are approximately 1 kPa.
Therefore, _the minimum limit of 1 kPa for the unaged asphalt binder was
considered reasonable from a rutting standpoint because AC-10 asphalt
cements were providing reasonable service in moderate climates.
The minimum limit of 2.2 kPa for the G*/sin8 value of RTFO aged
asphalt binders was established based on the 1 kPa value of the unaged
asphalt binder. RTFO test data on asphalt cements indicated that, on the
average, the aging index (viscosity after RTFO/viscosity before RTFO) for
asphalt cements ranged from 2 to 2.5. In other words, asphalt binders
become from 2 to 2. 5 times stiffer when aged in the RTFO. Therefore, the
average of the range, 2.2 was used to establish the 2.2 kPa minimum limit
for RTFO aged asphalt binders.
The original specificatioti. for fatigue cracking called for a maximum
limit of 3000 kPa. However, when the test results of the 42 asphalt binders
used to set the limits were investigated, half would not meet the 3000 kPa
maximum criterion. The ETG, therefore, deemed the 3000 kPa maximum
a
limit to be too restrictive. It was decided to use higher value of 5000 kPa
_J
until field validation data could be gathered and analyzed. Since about 15
percent of the 42 asphalt binders would not meet this limit, the value of
5000 kPa appeared reasonable at the time.
/ Deflection
1-----1■1/ Transducer
r
,,,,.-- Air Bearing
Control and
Data Acquisition Load Cell
/ Fluid
r--'--'--~/ Bath Loading
Frame
Supports
theory of bending prismatic beams would apply (52). The BBR uses a
blunt-nosed shaft to apply a midpoint load to the asphalt beam resting on
two end supports. The load cell mounted on the loading shaft is enclosed
in an air bearing to eliminate frictional resistance. A transducer affixed to
the shaft monitors deflections of the beam. Loads are applied by pneu-
matic pressure. The data acquisition system records the load and deflec-
tion test results and calculates two parameters: (a) creep stiffness, S(t)
which is a measure of how the asphalt binder resists the constant (creep)
loading, and (b) m-value which is a measure of the rate at which the creep
stiffness changes with loading time.
Asphalt binder beams measuring 125 mm in length, 6.25 mm in
width, and 12.5 mm in height are formed bya,pouring heated binder into
aluminum or silicon.e molds. The beam specimens are demolded after
cooling the mold assembly in a freezer for 5-10 minutes. The beam is then
kept in the test bath for 60±5 minutes before testing. After pre-loading
procedures (iZ), a 100-gram load is applied to the beam for a total of 240
seconds. The deflection of the beam is recorded during this loading peri-
od as shown in Figure 2-46. Load and deflection versus time plots are also
generated on the computer screen. After 240 seconds, the BBR software
performs all calculations to determine creep stiffness and m-value as fol-
lows.
S(t) = pp
4 bh 3 8(t)
where,
S(~ =' creep stiffness at time, t = 60 seconds
P = applied constant load, 100 g (980 mN)
L distance between beam supports, 102 mm
b beam width, 12.5 mm
h beam thickness, 6.25 mm
8(t) = deflection at time, t = 60 seconds
Deflection
simulates stiffness
/ after 2 hours at
/ 10 Clower temp
60
Log Time, s
Log Creep
Stiffness, S
/ slope = m-vaiu,
60
Log Loading Time, s
where,
S(t) = the asphalt binder stiffness,
t = time,
A, B, and C = constants.
Chapter 2-Asphalt Refining,_?ses, and Properties 91
Since the time dependency of asplialt binders varies, the shape of the
master stiffness curve as well as the stiffness at 2. hours loading time are
important. Therefo~e, the slope of the master stiffness curve is also used
for specification purposes (§2). The stiffness, S(t), is a measure of the ther-
mal stresses developed in the HMA pavement as a result of thermal con-
traction. The slope of the stiffness curve, m, is a measure of the rate of
stress relaxation by asphalt binder flow as discussed previously (52). The
effect of these two specification parameters, S(t) and m-value, on thermal
cracking is analogous to the effect of G* and 3 on rutting and fatigue
cracking. As S(t) increases, the thermal stresses developed in the pavement
due to thermal shrinking also increase, and thermal cracking becomes
more likely. On the other hand, as the m-value decreases, the rate of stress
relaxation decreases. In other words, as the slope of the asphalt binder stiff-
ness curve flattens, the ability of the HMA pavement to relieve thermal
stresses by flow decreases. This again would increase the propensity of
thermal cracking in the pavement. Accordingly, the Superpave binder
specification requires a maximum limit of S(t) at 60 seconds and a mini-
mum .m-value. The numerical limits for these two parameters were cho-
sen as S(60)=300 MPa maximum, and m=0.300 minimum. These limits
were established based on data from previous studies as well as the data
obtained by SHRP (52). A recent report (§5) based on six asphalt cement
binders used in Pennsylvania in 1976 (15) has offered some insight into
the field validation of these two parameters. The six test sections, which
used AC-20 asphalt cements from different sources, were monitored for
thermal cracking for a period of 7 years. Samples of the original asphalt
binders were retained and tested in 1995 using the Superpave binder
equipment. The results indicated that the S and m-values chosen in
Superpave binder specifications were generally reasonable. The low tem-
perature behavior of all but one of the AC-20 asphalt cements could be
explained by these values (§5).
Physical hardening is a phenomenon which was observed by the
SHRP researchers. The phenomenon is believed to be caused by delayed
volume shrinkage and results in significant time-dependent hardening
(@. How this phenomenon affects the HMA pavement performance is
not known. Therefore, no specification limit could be established for
physical hardening (@. However, the results of the physical hardening
---------tc-h.a.t-occm-o:v.er-2--tio.urs_ofiso..thermaLs.to.rage_can_be reported according
to Superpave binder specification. To minimize the effect of physical hard-
ening on the BBR test data, a standard conditioning time of 60±5 minutes
was selected as mentioned before.
For the BBR test results to be valid, the measurements must be made
92 Hot Mix Asphalt Materials, Mixture Design and Construction
t
C)
Load.
-~-- change in length (AL)
failure strairilcr) = =============~~=======--------
effective gauge length (Le)
The DTT measures the low temperature ultimate tensile strain of the
asphalt binders, typically in the 0° to -36°C range. Like the BBR test, the
asphalt binder is subjected to RTFO and PAV aging before the DTT test.
A small dog-bone shaped specimen is loaded in tension at a constant rate
(I mm/min.) as shown in Figure 2-48 until it breaks. The failure strain in
the specimen is the change in length (LlL) divided by the effective gauge
length (Le). The failure strain in the DTT corresponds to the stress where
the load on the specimen reaches its maximum value, and not necessarily
the load when the specimen breaks. The asphalt binder must exhibit a fail-
ure strain of at least 1.0 percent to meet the Superpave binder specification.
At this time (1996), the DTT equipment is being modified for a bet-
ter control of the test temperature and a more accurate measurement of the
failure strain.
I Original Binder
I Original Binder J·
Flash Point Terµp, T48: Minimum °C 230 s
Viscosity, ASTM D 4402: b .fl
Maximum, BPa•s (3000 cP), 135 g_
Test Temp,l°C
Dynamic Sµea~, TP5: c ...,,a°
G•/sin o, ~nimum, 1.00 kPa 70 76 82
Test Tern ,rature @ 10 rad/s, °C I·
I Rolling TI1in Film Oven (T 2-i0J or Thin Film Oven <T 179) Residue
Mass Loss, Maipmum, % 1.00
Dynamic Sheat, TP5:
o,
G•/sin Minimum, 2.20 kPa 70 76 82
Test Temp:@ 10 !"Jd/sec, °C
Pres.~ure Aging \'essel Residue (!'Pl J
PAV Aging Temperature, 0 cd 100(110) 100( I 10) 100(110)
Dynamic Shear, TP5:
G•sin o, Maximum, 5000 kPa 34 31 28 25 22 19 37 34 31 28 22 40 37 34 31 28
Test Temp @ 10 rad/sec, °C
Physic-JI Hardening e Report
Creep Stiffness, TPl: f
S, Maximum, 300 MPa
m-value. Minimum, 0.300 0 -6 -12 -18 -24 -30 0 -6 -12 -18 -24 0 -6 -12 -18 -24
Test Temp. @ 60 sec, °C
Direct Tension, TP3: r \D
V\
Failure Strain, Minimum, 1.0% 0 -6 -12 -18 -24 -30 0 -6 -12 -18 -24 0 -6 -12 -18 -24
Test Temp @ i.0 mm/min, °C
I
\D
Table 2-9. ~erformance Graded Asphalt Binder Specification (continued} 0\
l
Notes: '
a. Pavement teriperaturcs can be estimated from air temperatures using an algorithm contained in the
Superpave™ sof;tware program or may be provided by the specifying agency, or by following the procedures
as outlined in PPX.
b. This require~ent 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
~uality LI of unmodffied a,pl,ak cement production, measurement of the vOCOSty of. the origm,I
asphalt cementlmay be substituted for dynamic shear measurements 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 capill ry or rotational viscometer (AASHTO T 201 or T 202). i
d. The PAV agipg temperature is based on simulated climatic conditions and is one of three temperatures
9Q°C, 100°C or /110°C. The PAV aging temperature is 100°C for PG 64- and above, except in desert climates,
where it is 110°C.
I
~
~
e. Physical Harening - TP 1 is performed on a set of asphalt beams according to Section 13.1 of TP 1 , tJ
except the con~tioning time is extended to 24 hrs ± 10 minutes at 10°C above the minimum performance ~-
temperature. re 24-hour stiffness and m-value are reported for information pu~s only.
~
p..
f. If the creep ~ness is below 300 MPa, the direct tension test is not required. If the creep stiffness Q
between 300 _3rd 600 MPa the direct tension failure strain requirement can be used in lieu of the creep
stiffness requ~ment. The m-value requirement must be satisfied in both cases.
AASHTO Perfo.rmance Graded Binder Specification (MPl)
Ig.
::,
Chapter 2-Asphalt Refining, Uses, and Properties 97
Perfonnance Grade
PAV A in Tern ,C
Dynamic Shear, TP5:
G*sin 6, Maximum, 5000 kPa
Test Tern @ 10 rad/sec, C
Physical Hardening e Specification requirement
Creep Stiffness, TPI / to control fatigue cracking
S, Maximum, 300 MPa
m-value, Minimum, 0.300
Test Tern , @60 sec, C.
Direct Tension, TP3:
Failure Strain, Minimum, 1.0%
Test Temp @ 1.0 mm/min, C
Flash and Fire Points by Cleveland Open Cup", is used to determine the
flash point of unaged asphalt binders.
The specification also contains a maximum viscosity requirement of 3
Pa.sat 135°C for original binders to ensure pumping and handling of the
asphalt binders (especially modified asphalts) at the HMA facility.
A maximum mass loss of 1.00 percent applies to RTFO residues of all
PG grade binders. This requirement ensures that excessive aging
(volatilization) of the asphalt binder does not occur during HMA produc-
tion and construction.
-31 -23
~31 -23
imum reliability, the high temperature grade must be PG 58. Since there
is no PG 56 grade it has to be "rounded up" to PG 58, which actually
results in a probability of about 85 percent rather than 50 percent. On the
other hand, the low temperature grade for at least 50% reliability must be
-28°C which gives almost 90% reliability rather than 50%. For 98 per-
cent reliability, a PG 64 grade is needed for high temperature and a PG-34
grade for low temperature resulting in a PG 64-34 grade.
The preceding binder selection process based on climate only assumes
that the H.MA pavements will be subjected to fast moving traffic. The
DSR tests to determine the rutting factor G*/sino are conducted with a
loading rate of 10 radians per second which corresponds to a traffic speed
of about 90 kilometers per hour. However, .significantly slower loading
rates are experienced by HMA pavements at some locations such as inter-
sections and toll booths. There are cases such as bus stops where loads are
also stationary. In such cases of slower loading rates an asphalt binder must
exhibit a higher stiffness to minimize rutting. To meet these situations, the
high temperature grade should be increased by one grade for slow transient
loads and by two grades for stationary loads. For example, if a PG 64-22
grade was selected based on climate for fast moving traffic, a PG 70-22
grade would be used for slow transient loads, and a PG 76:-22 grade would
be used for stationary loads. The selected low temperature grade is not
affected by the loading rate. Since the highest possible pavement temper-
ature in North America is about 70°C, it was necessary to add two high
temperature grades, PG 76 and PG 82, to accommodate slow traffic load-
ing rates (5§).
Most state highway agencies are planning at the present time (1996)
to adopt the Superpave asphalt binder system beginning in 1997.
ASPHALT CHEMISTRY
Origin of Asphalt
Asphalt Cement
I
Asphaltenes Maltenes
(Petrolenes)
I
Resins Oils
"Sol:" "Sol" asphalt cements typify a system in which the resins keep the
asphaltenes highly "peptized" (or dispersed) in the oily phase. "Sol" asphalt
---cements largely exnioit Newtoman flow charactensncs. Accordmg to
Rostler (Z.l), nitrogen bases in asphalt cement are primarily responsible for
its "sol" characteristics.
"Gel·" "Gel" asphalt cements typify a system in which resins are not very
effective in "peptizing" asphaltenes. Excessive presence of paraffins in rela-
tion to nitrogen bases also tends to reduce solubility, leading to increased
"gel" characteristics, and suggesting increased separation of the dispersed
and dispersing phases (Zl). "Gel" asphalt cements exhibit largely
non-Newtonian behavior.
The degree to which an asphalt cement acts as a "sol" or as a "gel"
influences its rheological properties, aging characteristics, and its perfor-
mance in a H.MA mixture.
The compatibility and relationships of different components in the
microscopically homogeneous mixture controls the overall behavior of
asphalt cement rather than the quantitative amount of any single compo-
nent according to Petersen (@. Various components of asphalt cement
interact to form a balance or compatible system, which gives asphalt
cement its unique viscoelastic properties as a binder. It has long been rec-
ognized that this compatibility, or lack of compatibility, is related to
asphalt quality. Oliensis's spot test (Z2) was an attempt to determine the
"stability" of the asphalt components when dissolved in a solvent. Many
agencies adopted the Oliensis spot test as a specification test for homo-
geneity of asphalt cements and therefore as an indicator of asphalt cement's
durability in service. However, as was shown by Heithaus and Fink (Z3),
the spot test did not distinguish between "good" and "bad" asphalt
cements. This test did serve a good purpose when first introduced because
it identified some nondurable asphalt cements obtained from the cracking
of petroleum in certain refining processes.
Analytical Procedures
The two most frequently used fractionation methods are: (1) Rostler
and Sternberg's chemical precipitation method, and (2) Corbett's selective
adsorption-desorption (chromatographic) method. However, in all cases
the "generic" fractions are still complex mixtures and not defined chemical
species; thus the same generic fraction from different asphalt cements may
have a quantitatively different effect on the physical properties.
I ASP~ALT I
n-Pentane
Precipitate Solution
NITROGEN
ASPHALTENES
BASES (85%)
PARAFFINS
Asphaltenes (A)
--------80-per-€@B-t-sHlfuFie-a€i-cl.-is-us@d-t-0-r-emo-v-e-0-r-pr-e-cip-i-tat-e-p0lar-ni-tr-0gen------
compounds (quite aromatic in nature) from the maltene solution .. This
fraction is called "nitrogetl bases" and is considered to be the peptizer for
the strongly associated asphaltenes in the solvating phase.
Chapter 2-Asphalt Refining, Uses, and Properties 109
Paraffins (P)
This is the final and oily constituent of the maltenes which is nonre-
active to the fuming acid. According to Rostler, paraffins are believed to act
as "gelling agents."
Many asphalt technologists with an engineering background have dif-
ficulty relating the results from this analytical procedure using sulfuric acid
of different strengths to the performance of asphalt cement. However,
Rostler has suggested the following relationships concerning the quality
and durability of asphalt cement:
Compatibili'ty R.atio = N
p
~ n-Heptane
Precipitate PETROLENES
(MALTENES)
NAPHTHENE Benzene
AROMATICS (elute)
POLAR 1. Methanol-benzene
AROMATICS 2. Trichloroethylene
(elute)
components virtually unaltered for further analysis (ZQ). The most polar
and least soluble asphalt cement components (asphaltenes) are first precip-
itated by n-heptane which is a nonpolar solvent. The heptane solution of the
remaining maltene fraction is then introduced into a chromatographic
column. Alumina in the column adsorbs the remaining three components.
Different solvents of increasing polarity are then used to sequentially
desorb (wash out) fractions of increasing polarity. Corbett's method sepa-
rates and quantifies the following four components:
_ _ _ _A_sphalten~e~s~(A~)'..__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
Saturates (S)
Saturates are the first fraction to emerge from the column when elut-
ed (desorbed) with n-heptane. This fraction is not adsorbed by the alumi-
na in the column because it lacks polar chemical functional groups.
Saturates are equivalent to paraffins in the Rostler analysis. Often sulfur is
found incorporated in molecules of this fraction. Saturates are liquid at
ambient temperatures and hardly change with time. They have a negative
contribution to temperature susceptibility of asphalt cement.
They are the final fraction to emerge from the column when eluted
with a highly polar mixture of alcohol and benzene (or toluene). Alcohol
helps to debond this strongly adsorbed and most polar component of the
maltenes from the absorbent (alumina). Its molecular structure comprises
a higher percentage of condensed aromatic ring systems and functional
groups containing heteroatoms than the other fractions of maltenes. It is
solid or semi-solid at ambient temperatures. It has been related to ductili-
ty of asphalt cement and is also one of the aging fractions in asphalt
cement.
Legend
--Asphalt A
--Asphalt B
I
/
I
I
I
I
L
MMS SMS
0 ~ JO I~ 20
Elution Time ( lnutesl
the Strategic Highway Research Program (SHRP) has shown that the large
molecular size (LMS) fraction is composed predominantly of a collection
(conglomerate) of smaller, highly associated polar molecules which thus
appear in the chromatogram as molecules of "apparent" large molecular
size. Thus, this technique has the potential of characterizing the strongly
associating molecular components that play a major role in determining
the rheological properties and aging characteristics of asphalt cement relat-
ed to pavement performance.
The IEC analytical technique was also used to isolate a fraction termed
"amphoterics". This fraction is believed to be the key building block of
asphalts. The term amphoterics is used to describe an asphalt molecule
which has both an acid and basic group in the molecule, but not at the
same site. Although amphoterics constitute only 10 to 15 percent of
asphalts, they are largely responsible for the viscoelastic properties of
asphalts based on the test data obtained in SHRP.
Asphalt chemistry is very complex. Researchers have debated in the
past and continue to debate on this very interesting subject. Reference Zfl.
contains SHRP research data on asphalt chemistry.
REFERENCES
42. Pink, H.S., R.E. Merz and D.S. Bosniack. Asphalt Rheology: Ex-
perimental Determination of Dynamic Moduli at Low Temper-
atures. Proceedings AA.PT, Vol. 49, 1980.
43. Kandhal, P.S. and W.C. Koehler. Pennsylvania's Experience in the
Compaction of Asphalt Pavements. ASTM, Special Technical
Publication 829, 1984.
44. Finn, F.N., K. Nair and J.M. Hilliard. Minimizing Premature
Cracking in Asphaltic Concrete Pavement. TRB, NCHRP Report
195, 1978.
45. Kandhal, P.S. Low Temperature Shrinkage Cracking of Pavements
in Pennsylvania. Proceedings AA.PT, Vol. 47, 1978.
46. Fromm, H.J. and W.A. Phang. Temperature Susceptibility Control
in Asphalt Cement Specifications. HRB, Highway Research Re-
cord 350, 1971.
47. Kandhal, P.S. Evaluation of Low Temperature Pavement Cracking
on Elk County Research Project. TRB, Transportation Research
Record 777, 1980.
48. McLeod, N.W. The Case for Grading Asphalt Cements by Pen-
etration at 77°F. Proceedings Canadian Technical Asphalt As-
sociation, Vol. 20, 1975.
49. Burgess, R.A., 0. Kopvillem, and F.D. Young. Ste. Anne Test
Road-Relationship Between Predicted Fracture Temperatures
and Low Temperature Field Performance. Proceedings, AA.PT, Vol.
40, 1971.
50. Readshaw, E.E. Asphalt Specifications in British Columbia for
Low Temperature Performance. Proceedings AA.PT, Vol. 41, 1972.
51. Kandhal,·P.S. Evaluation of Baghouse Fines in Bituminous Paving
Mixtures. Proceedings AA.PT, Vol. 50, 1981.
52. Majiq.µ.deh, K. and F.N. Brovold. State-of-the-Art: Effect ofWater
on Bitumen-Aggregate Mixtures. HRB, Special Report 98, 1968.
53. Kandhal, P.S. Field and Laboratory Evaluation of Stripping in
Asphalt Pavements: State-of-the-Art-Report. Transportation
Research Record 1454, 1994.
54. Kandhal, P.S. Moisture Susceptibility of HMA Mixes:
-------~Identification_.of...Eroblem.-an.cl.Recomm.ended Solutions. NAPA
Quality Improvement Publication 119, December 1992.
55. Kandhal, P.S., C.W. Lubc5ld, and F.L. Roberts. Water Damage to
Asphalt Overlays: Case Histories. Proceedings AA.PT, Vol. 58,
1989.
118 Hot Mix Asphalt Materials, Mixture Design and Construction
INTRODUCTION
Igneous rocks are primarily crystalline and are formed by the cooling
of molten rock magma as it moves toward or on the surface of the earth.
Igneous rocks are classified based on size of the crystal grains and on com-
position as either acidic or basic. The crystal grains are classified as coarse
if they are larger than 2 mm and fine when smaller than 0.2 mm.
Classification of igneous rocks based on composition is shown below:
Gravel is formed from the breakdown of any natural rock. Gravel par-
ticles are found in existing or ancient waterways, and the particles are usu-
ally smooth and typically rounded or subrounded by wear as the material
is moved along the waterway by the action of water. Gravel is usually
required to be crushed prior to use in HMA.
Sands consist primarily of the most resistant final residue of the dete-
rioration of natural rocks. The predominant i.ineral in sand is normally
quartz. The grain size normally ranges from a No. 8 (2.36 mm) sieve down
to dust size particles (No. 200 or 75 µm sieve). Many sand deposits often
contain silt and/or clay particles in various quantities and may be required
to be washed prior to use in HMA. Many northern states have a great deal
of glacial sands and gravels which have less clay than most river gravels and
sands.
AGGREGATE PRODUCTION
Quarry Operations
Screening
Deck + lln.
II in. and ¾in.I
Minus ¾in.
Cone
Crusher
passes over the grizzley is fed into a jaw crusher (primary crusher) that
breaks the rock down to a size that can be handled by other crushers.
Suppose for this simplified crushing operation that four aggregate sizes will
be produced: 1 inch-3/4 inch (25.4 mm-19 mm), 3/4 inch-3/8 inch
(19 mm-9.5 mm), 3/8 inch-No. 4 (9.5 mm-4.75 mm), and minus No. 4
(4.75 mm). The next operation after the jaw crusher is to screen the aggre-
gate into various sizes. The first screen deck separates aggregate into three
sizes by using a 1-:--inch (25.4 mm) and a 3/4-inch (19 mm) screen. The
material larger than 1 inch (25.4 mm) is fed to the cone crusher for addi-
tional crushing. The material smaller than 1 inch (25.4 mm) but larger
than 3/4 inch (19 mm) is stockpiled. The material smaller than 3/4 inch
(19 mm) is fed to a second screen deck for further separation. The second
screen deck uses three screen sizes 3/4 inch (19 mm), 3/8 inch (9.5 mm),
and No. 4 (4.75 mm). The material larger than 3/4 inch (19 mm) is fed
back to the cone crusher for further crushing. The other three material sizes
are stockpiled as shown. This is a very simplified crushing operation that
identifies methods that are normally used to crush the aggregate and sepa-
rate it into sizes . Most operations are much more complicated than this
but, regardless of the complexity, the operation is almost entirely made up
of crushers and screening decks. Many operations also have the capability
to wash the aggregate as needed at certain points in the operation.
During the operation of a crushing facility, it is absolutely essential
that the properties of the final product be consistent. This requires that the
parent rock being fed into the crusher has consistent physical properties
such as specific gravity and absorption. When rock is removed from vari-
ous locations in the quarry, the physical properties of the aggregate may
vary substantially, thus producing problems in maintaining consistent air
voids in the HMA. Quality control during the crushing operation must
ensure that physical properties of the aggregate do not vary excessively.
Ideally, the awount of material being fed to the crushing operation should
be approximately constant. If the rate of material feed is increased or
decreased, the amount of overflow on particular screen sizes changes, espe-
cially when operating near full capacity for individual screens. For instance,
increasing the rate of flow of materials through the crusher normally results
in more overflow of finer aggregate over individual screen sizes. This over-
flow results in a stockpile with a finer aggregate gradation. The feed rate
must exceed some minimum amount for the finer aggregate to begin to
------<oJ-:v.ei:flo:w-the...scteen,_ _ _ _ _ _ _ _ _ __
The primary crusher provides the initial mechanical reduction of the
pit run material down to maximum sizes ranging from 8 inches to 1 inch.
Secondary and tertiary crushing refer to the additional stages of reduction
that are required to reduce the aggregate to some desired size.
126 Hot Mix Asphalt Materials, Mixture Design and Construction
IMPACTORS
•
VERTICAL
>
IMPACT
PULVERIZERS
•
BRADFORD
BREAKERS •
HAMMERMILLS
• • •
RING-TYPE
GRANULATORS • • •
SINGLE-ROLL
CRUSHERS • • •
• •
-. •
GYRATORIES
J)(W ..CROSRERS"
--.- ----·- · · - -
CONE
• • •
Figure 3-3. Display of the Reduction Methods Employed by Different
Types of Crushers (1)
Chapter 3-Aggregates 127
most effective system for reducing the size of larger particles of very hard
and abrasive rocks.
The types of crushers for a particular application must be carefully
selected to produce a material meeting the customer's specifications while
operating at minimum cost. Obviously the type of rock being processed
has a significant influence on the choice of crushing equipment to be used.
AGGREGATE SAMPLING
Prior to conducting any tests on aggregates, samples must be obtained
from some source using proper sampling techniques (ASTM D75). The
samples may be randomly selected or may be selected to be representative
(average of the material being sampled), depe~~ing on the purpose of sam-
ples. Usually for mix design, representative samples are taken, and for qual-
ity control, random samples are taken. When only one or two samples are
used to characterize a mixture property, representative samples are
required. However, when many samples are taken to evaluate mix quality,
random samples are normally taken. If proper samples are not taken, all
tests conducted on the aggregate are meaningless, and this may result in the
design or construction of a HMA mixture that has poor quality, resulting
in unsatisfactory performance. ,
gate also segregates on the belt. In this case a random sample is taken at
one point in time from the belt. A representative sample is taken by com-
bining a number of random samples throughout some desired time peri-
od, such as one day for belt samples, or by taking samples from various
locations in a stockpile and combining these samples. Representative sam-
ples are almost always used to evaluate stockpiles. Again, samples must be
taken in a way to minimize the effect of stockpile segregation. The aggre-
gate at the bottom edge of a stockpile is usually coarser than the rest of the
stockpile. The preferred method of sampling a stockpile is to climb the side
of the pile to some point away from the top or bottom of the pile, remove
the surface aggregate and obtain a sample from underneath the surface. It
is sometimes necessary to place a form above the place being sampled to
prevent the aggregate from flowing into the area being sampled. Samples
should be taken at several locations throughout the stockpile and com-
bined for a representative sample.
After taking the required number and amount of samples and bring-
ing them back to the laboratory, each sample must be reduced to the
appropriate size for testing. Samples can be reduced with a sample splitter
or by quartering (ASTM C702). Both methods work successfully, but one
of these must be utilized if an accurate sample is to be obtained for tests.
The importance of sampling and splitting to test size cannot be overem-
phasized since all test results are dependent on the use of proper tech-
mques.
Aggregate Mineralogy
Common Minerals Found in Aggregates. A majority of natural aggre-
gates are composed of a combination of minerals. Among the most impor-
130 Hot Mix Asphalt Materials, Mixture Design and Construction
For brief descriptions, see "Standard Descriptive Nomenclature of Constituents of Natural Mineral
Aggregates" (ASTM C294).
~
I (")
::,'
-
j Rock Quartz Orthoclase Plagioclase Augite Hornblende Biotite Muscovite Epldote Rock Glass
I Igneous Rocks
C"'
n
\,I,)
i
v>
-
~
dranite 30 41 8 - - 3 3 1
rliorite
dabbro
8
0.5 -
7 30
44
3
28
27
9
4
2
0.1
-
5
1
-
-
fJ (JQ
tjiabase - - 44 46 - - - - 2 ~ ~
32 45 3 - 0.7 3 2 2 0.4 n (JQ
~yolite 8,,,
achyte
~ndesite
Basalt
3
0.6
-
42
-
-
1
48
36
2
14
35
6
3
-
0.5
-
-
-
-
-
8
3
-
0
13
21
~
~
I Metamorphic Rocks ::s
n
I .
guartztte
Feldspathic-
84 3 - - I 2 2 2 - ""
e.
~quartzite 46 27 1 - - 2 5 1 - n0
ornblende-
i:neiss 10 16 15 3 45 3 1 2 - .g
Granite-
~eiss
B otite-
37 32 3 - - 7 11 2 - ...
0
"'r:t.
hist 34 13 3 - I 38 3 2 - 0
::s
Mica-schist 37 16 I - - 13 26 2 -
slate 29 4 - - - 55 2 - g,
~arble 3 0.2 0.2 - - - Calcite=% - ""g:,
Ainphibolite 3 1 8 - 70 I 0.2 12 -
Sandstone 79 5 0.3
Sedimentary Rocks
- - 0.2 1 - -
8-
Feldspathic-
sandstone 35 26 2 - - 0.6 2 1 -
~
Calcareous-
sandstone 46 3 ·2 - - - Calcite= 42 - -
Chert 93 - - - - Calcite = 1 - -
Limestone 6 - - - Dolomite= 8 Calcite= 83 - -
Dolomite 5 - - - Dolomite = 82 Calcite = 11 - -
....
.......,
132 Hot Mix Asphalt Materials, Mixture Design and Construction
Igneous:
Granite, syenite, Good Good Good Good Possible
diorite
Felsite Good Good Question- Fair Possible
able
Basalt, diabase, Good Good Good Good Seldom
gabbro
Periodotite Good Fair Question- Good Possible
able
Sedimentary:
Limestone,dolomite Good Fair Good Good Possible
Sandstone Fair Fair Good Good Seldom
Chert Good Poor Poor Fair Likely
Conglomerate, Fair Fair Good Good Seldom
breccia ,_
Shale Poor Poor Poor Good Possible
Metamorphic:
Gneiss, schist, Good Good Good Good Seldom
Quartzite Good Good Good Good Seldom
Marble Fair Good Good Good Possible
Serpentinite Fair Fair Good Fair to poor Possible
Amphibolite Good Good Good Good Seldom
Slate Good Good Good Poor Seldom
properties for common aggregate types, and Table 3-4 summarizes desir-
able properties of rocks for HMA aggregates. Most aggregates are com-
posed of several minerals, often with variable compositions. Even with
aggregates of uniform mineralogy, the properties may be altered by oxida-
tion, hydration, leaching, weathering and foreign coatings. Therefore,
mineralogy alone cannot provide a basis for predicting the behavior of an
aggregate in service. Petrographic examinations are useful, and past perfor-
mance of similar aggregates under similar environmental and loading con-
ditions can be helpful in evaluating aggregates.
In addition to the properties listed in Table 3-3 and Table 3-4, min-
eral composition of coarse aggregates also affects the skid resistance of
HMA. Quartz and feldspar are harder and more polish resistant minerals
which are normally found in igneous rocks such as basalt and gabbro. On
the other hand, calcite and dolomite, which occur in limestones, are exam-
ples of soft minerals. Limestones that have a high percentage of soft mate-
rials tend to polish more rapidly than most other aggregate types. The acid
insoluble residue test (ASTM D3042) has been used to measure .the
amount ofh:arcler materials present in carbonate aggregates. Many highway
agencies specify a minimum of 10 percent acid insolubles to assure accept-
able frictional properties. Other agencies use a polishing test (ASTM
D3319 or E660 and E303) or petrographic examination (ASTM C295) to
evaluate the polishing potential of an aggregate.
Chapter 3-Aggregates 133
Igneous
Granite Fair Fair Fair Fair
Syenite Good Fair Fair Fair
Diorite Good Fair Fair Good
Basalt (trap rock) Good Good Good Good
Diabase (trap rock) Good Good Good Good
Gabbro (trap rock) Good Good Good Good
Sedimenta,y
Limestone, dolomite Poor Good Good Fair
Sandstone Fair Good Good Good
Chert Good Fair Poor Good
Shale Poor Poor Fair Fair
Metamo,phic
Gneiss Fair Fair Good Good
Schist Fair Fair Good Fair
Slate Good Fair Fair Fair
Quartzite Good ...EaiL________ ...-Good Good
Marble Poor Good Fair Fair
Se,pentine Good Fair Fair Fair
tAggregates that are hydrophilic (water-loving) tend to strip more readily since water more easily replaces an asphalt film.
Freshly crushed aggregates with many broken ionic bonds also tend to strip more readily.
134 Hot Mix Asphalt Materials, Mixture Design and Construction
'lillm
I - Important
N - Not Important
U - Importance Unknown
138 Hot Mix Asphalt Materials, Mixture Design and Construction
Particle shape and surface Friction; skid-resistance; ASTM D3398 Minimum percent of crushed
texture compaction; mix stability ASTMD479! particles; ASTM D692;
ASTMC1252 AASHTOM283
Resistance to polishing Skid and wear resistance ASTM D33!9 Maximum polish value;
ASTM E303 Minimum acid insoiubles
ASTM E660
ASTM D3042
Methods that are available include ASTM C131 for coarse aggregate
smaller than 1.5 inch (38 mm) and ASTM C535 for testing aggregates
larger than 0.75 inch (19 mm) and up to 3 inch (76 mm) maximum size.
ASTM Cl31 is performed by mixing 5,000 g of aggregate blended to meet
one of several gradings, along with a charge of six to twelve steel balls, each
approximately 2 inches (51 mm) in diameter and approximately 400 g in
weight. This mix is tumbled inside a steel drum for 500 revolutions at a
speed of 30-33 rpm. This is the method that is normally used for HMA
aggregates. ASTM C535 is performed with 10,000 g of blended aggregate,
twelve steel balls, and 1,000 revolutions. After tumbling, the fines that pass
a No. 12 (1.68 mm) sieve are weighed, and the percentage loss by weight
of original sample is calculated as the Los Angeles abrasion. Initially, none
of the material would pass a No. 12 sieve. Hence if 40 percent of the mate-
rial passed a No. 12 sieve after the test, the L.A. abrasion would be 40.
Typical test values range from 10 percent for extremely hard igneous
rocks to 60 percent for soft limestones and sandstones. Obviously, a num-
ber of aggregates will exceed this typical range. The maximum wear for
coarse aggregate used in high-type HMA is typically limited by specifica-
tions to 40 percent by some agencies to as much as 60 percent by other
agencies. ASTM D 1073 does not specify a maximum abrasion loss for fine
aggregates for any purpose.
The Los Angeles abrasion test is primarily a measure of the resistance
of coarse aggregate to degradation by abrasion and impact; however, field
observations do not show a good relationship between L.A. abrasion loss
and performance. This test does not appear to be fully satisfactory for use
with slags, cinders, and other lightweight aggregates. Experience has shown
that many aggregates such as slag and soft limestones provide excellent per-
formance even though the Los Angeles abrasion is high. One item that
usually must be dealt with when producing HMA with aggregate having a
high L.A. abrasion is the generation of dust during handling of aggregates
and production of HMA. A high dust content will likely produce envi-
ronmental problems as well as mixture control problems.
sodium or magnesium sulfate. During this test salt crystals grow in the aggre-
gate pores and cause particles to disintegrate in some aggregates. Presumably
this crystal formation simulates that of ice crystals and hence to some extent
simulates the effect of freezing and thawing. An oven-dried sample of the
aggregate is separated into specified sizes and immersed in a saturated solu-
tion of sodium or magnesium sulfate. The aggregate remains in the solution
at a constant temperature for approximately 18 hours. The sample is then
removed from the solution, dried to constant weight at 230 ± 9°F (110±5°C)
and cooled. This cycle is typically repeated five times, after which the sample
is washed to remove the salt and is dried. The loss in weight for each size frac-
tion is determined by sieving, and the weighted average percent loss for the
entire sample is computed (see ASTM C88 for sample calculation).
A number of tests to evaluate the effects of freezing and thawing have
been proposed. Typically the procedures are similar to that of ASTM C88,
except that different sized samples are immersed in water or 0.5 percent
water-alcohol solution and then alternately subjected to freezing and
thawing for the required number of cycles (16 to 50). MSHTO T103 is
one such procedure.
When specifications for aggregates contain a requirement for sound-
ness by the sodium or magnesium sulfate test, the specifications generally
state both the number of cycles and the maximum allowable percentage of
loss. For example, ASTM D692 specifies a maximum of 12 percent loss
after 5 cycles when using sodium sulfate and 18 percent loss when using
magnesium sulfate since the action of the magnesium sulfate is more
severe. The sulfate soundness test has been an accepted method of testing
for many years. Despite this acceptance, it has been widely criticized, and
numerous reports have appeared which describe its inability to accurately
predict field performance for specific aggregates. Many engineers believe
that freezing and thawing cycles do not create problems with aggregates in
HMA. The aggregate is initially dried during production and hence should
have little or no moisture in the pores immediately after production.
(Aggregates with high absorption values probably do have some free mois-
ture in the pores.) During HMA production, the aggregates are coated
with a film of asphalt binder which should prevent the aggregate from
absorbing a significant amount of moisture during the life of the mixture.
Since there is little or no moisture in the aggregate, freezing and thawing
should not be a significant problem.
------------------------------
- - - - - - - - -----------
des exhibit greater interlock and internal friction, and hence result in
greater mechanical stability than do rounded particles. On the other hand,
mixtures containing rounded particles, such as most natural gravels and
sands, have better workability and require less compactive effort to obtain
the required density. This ease of compaction is not necessarily an advan-
tage, however, since mixtures that are easy to compact during construction
may continue to densify under traffic, ultimately leading to rutting due to
low voids and plastic flow.
The surface texture, like particle shape, also influences the workability
and strength of HMA. A rough, sandpaper-like surface texture, such as
found on most cru,shed stones, tends to increase strength and requires
additional asphalt cement to overcome the loss of workability, as compared
to a smooth surface found in many river gravels and sands. Voids in a com-
pacted mass of rough-textured aggregate are also usually higher, providing
additional space for asphalt cement. Smooth-textured aggregates may be
easier to coat with an asphalt film, but the asphalt cement usually forms
stronger mechanical bonds with the rough-textured aggregates.
The combined effects of particle shape and surface texture of an aggre-
gate can be determined using the particle index test of ASTM D3398. The
equipment required for this test is simple, consisting basically of a cylin-
drical steel mold 6 inches (152 mm) in diameter (diameter ranges from 2
inches or 51 mm for particles passing No. 50 sieve to 8 inches or 203 mm
for particles larger than 1 inch or 25 mm) by 7 inches (178 mm) high, and
a steel rod 5/8 inch (16 mm) in diameter by 24 inches or 610 mm long
(the tamping end is rounded to a hemispherical tip). Tamping rods weigh
between 34 g for the finest size fraction to 2,204 g for the largest size frac-
tion. The procedure requires that a clean, washed, oven-dried, single-size
aggregate fraction be used. The mold should be filled in three equal layers
with each layer receiving ten tamps with the tamping rod. Each tamp con-
sists of a drop with the tamping rod from 2 inches above the surface of the
layer being compacted. The weight of the contents of the mold in each case
is determined and the corresponding percentage of voids is calculated using
the bulk specific gravity of each aggregate fraction. This procedure is
repeated using the same material but applying 50 blows on each of the
. three layers. The particle index is then calculated using the following equa-
tion:
Ia= l.25V10 - 0.25V50 - 32.0
the rutting potential of HMA based on the work done at the National
Center for Asphalt Technology (12, 1.3.). It has been included in the
Superpave mix design system. In this method, a 100-cm3 cylinder is filled
with fine aggregate of prescribed gradation by allowing the sample to fl.ow
through the orifice of a funnel into the calibrated cylinder. Excess mater-
ial is struck off and the cylinder with aggregate is weighed. Uncompacted
void content of the sample is then computed using this weight and the
bulk dry specific gravity of the aggregate. Three variations of the method
are used. Method A uses a graded sample of specified gradation. Method
Buses three individual size fractions: Nos. 8 to 16 (2.36 to 1.18 mm),
Nos. 16 to 30 (1.18 to 0.6 mm), and Nos. 30 to 50 (0.6 to 0.3 nim); and
the mean void content is determined. Method C uses the gradation as
received. Method A is used in the Superpave mix design system.
Flat and elongated particles impede compaction and thus may prevent
the development of satisfactory strength in HMA. AASHTO M283
requires that the portion of aggregate retained on the 3/8-inch (9.5 mm)
sieve shall not contain more than 15 percent of particles by weight that
have a ratio oflength to width or width to thickness that exceeds 5: 1. The
fl.at and elongated particles in coarse aggregates can be determined follow-
ing ASTM D4791, using a proportional caliper device. The shape of the
fine aggregate is also very important, and research is being performed to
develop tests that accurately quantify the particle shape and thus determine
the suitability of the fine aggregate blend for use in HMA.
..,
~
--:-400
- PARTICLE INDEX-15
,I I I I I I I
.,..0 PARTICLE INDEX-11.5 I
sf" I I I I I
,--- 300
PAR 111..LI,. ll'IUc..-.-t:J
I j
-< I I I
,->-
,-t-
~- 2900 I
-~
-Iii I
I
COMP. TEMP.-280"F
. I/
. ,
-..J I/ / J
-..J-
< 1o'o V VMA-15% lj j
,,,.
-~
-<
:::E ,_
- r.f... II-'
AIR VOIDS-4%
I I I I
I -86 88 90 92 94 95 98 10011021 I I
I I %LAB. COMP. SPEC. GRAV. I I I I I I
Clay Lumps and Friable Particles. Clay and friable particles (ASTM
C142) may be present in the form oflumps. The harm caused by these
lumps depends on their survival as lumps in HMA. These lumps in the fin-
ished HMA can break down from freezing and thawing or wetting and
drying and cause stripping, ravelling, or otherwise affect the durability of
the HMA. These particles can normally be removed during the aggregate
crushing operation at the grizzley or in. a washing operation. Some of the
particles break down when processed through the HMA facility. Natural
sands and gravels are the most likely source of these lumps. If the lumps
remain intact during placement of the HMA, surface pitting is likely to
occur as the lumps on or near the surface deteriorate under rolling or traf-
fic. If the total percent of day lumps or friable particles in the HMA is very
low (less than 1 percent), .the surface problems will be minimal, and per-
formance will be satisfactory. Some pitting and popouts may occur, but
will
--~o_th.e.r.:wi.s_e_th_e_p_erformance not be adversel;i~affc=e=ct=e=d~._____________ ..J._
In this test for day lumps and friable particles, aggregates are first
washed and dried to remove material passing the No. 200 (75 µm) sieve
and then separated into different sizes. Test samples are weighed and
soaked in water for 24 hours. Any particles that can be broken with the fin-
il
Chapter 3-Aggregates 145
gers after soaking and removed by wet seiving are classified as clay lumps
or friable particles, and the percent of this material is calculated by weight
of the total test sample. Specifications normally limit the amount of clay
lumps and friable particles to a maximum of 1 percent.
Plasticity Index. The Plasticity Index (PI) (ASTM D43 I 8) is the dif-
ference between the liquid limit and the plastic limit of the material pass-
ing the No. 40 (0.42 mm) sieve. PI is a measure of the degree of plasticity
of fines (material passing the No. 200 sieve) and can indirectly indicate the
amount and type of plastic fines. ASTM D1073 and D242 limit the PI of
the aggregate fraction passing the No. 40 sieve of both the fine aggregate
and the mineral filler to four or less
Specific Gravity
weight
specific graviry =
volume(unit weight of water)
When working in the metric system the unit weight of water is 1.0 gram
per ml. Hence the equation for specific gravity becomes:
weight
specific graviry =
volume
_ _ _ _w.he.n_tb_e_w.eigh_t is in grams and the volume is in ml. The!:<:; .tJe four dif-
ferent aggregate specific gravities used for HMA based on the method used
to define the volume of the aggregate particles:
Referring to Figure 3-9, the apparent specific gravity includes only the
volume of the aggregate particle. It does not include the volume of any
pores or capillaries that become filled with water after a 24-hour soaking.
The bulk specific gravity includes the overall volume of the aggregate par-
ticle, as well as the volume of the pores that become filled with water after
a 24-hour soaking. The effective and bulk-impregnated specific gravity
includes the overall volume of the· aggregate, plus the pores that become
filled with water after a 24-hour soaking, min4~ the volume of the larger
pores that absorb asphalt. The volume of solids ~) contains internal pores
that are impermeable to both water and asphalt. Since the volume of the
impermeable pores cannot be accurately determined, for all practical pur-
poses they are considered as part of the volume of the solids. Reducing the
aggregate size by crushing exposes some of these internal voids and thus
increases the apparent specific gravity of the aggregate.
Air voids in the compacted HMA appears throughout the mix as small
pockets of air between the asphalt coated aggregate particles. The choice of
an aggregate specific gravity to be used in asphalt mix calculations can have
a substantial effect upon the calculated amount of air voids in the com-
pacted HMA. The effective specific gravity of the aggregate in the mixture
depe1J.ds upon the degree to which the aggregate absorbs asphalt. When the
apparent specific gravity is used in the calculations, the voids between the
asphalt filrp. and the aggregate (pores not filled with asphalt) are included
in the calc'½llated voids. If bulk specific gravity is used, it is assumed that no
asphalt is absorbed by any of the water permeable pores; hence, the mea-
sured voids include the voids between the aggregate coated particles minus
the volume of the absorbed asphalt. Except in rare cases, neither the bulk
or apparent specific gravity provides a correct measure of air voids in the
HMA mixture. The concept of effective specific gravity more nearly
describes the true case for asphalt absorption when calculating the voids in
a compacted asphalt-paving mix. In this case, the calculated voids are truly
those voids between the asphalt coated particles and do not include the
----vo_1._,.ds in ti:ie pores between--rhe-asphal-r-a-nd-aggrega:1:e:-T-he-e:ff-e-c-t-ive-speci.f-~-----
ic gravity is equal to or higher than the bulk specific gravity and equal to
or lower than the apparent specific gravity. When the aggregate absorption
is zero, all four types of aggregate specific gravity are theoretically equal.
The bulk and apparent specific gravities of coarse and fine aggregates
Chapter 3-Aggregates 147
Vs = Volume of solids
Vpp = Volume of water permeable pores
Vop = Volume cif pores absorbing asphalt
Vpp - Vop = Volume of waler permeable pores not absorbing asphalt
Ws
Apparent specific gravity
where,
A= oven-dried weight of aggregate, g;
B= saturated surface-dry (SSD) weight of aggregate, g; and
C= submerged weight of aggregate in water, g.
G =
(7)
where,
G = average specific gravity;
- - - - - - - - ~ G1-,-G.z.,---¥--5:P.ecific_gJ:aY.icy. __v.aluedorfraction 1, 2, ... n; and
P 1, P2 , ••• Pn= weight percentages of fraction I, 2, ... n.
This same equation can be used to calculate the average specific grav-
ity of an aggregate blend in which case G 1, G 2 , •.. Gn are the specific gravi-
150 Hot Mix Asphalt Materials, Mixture Design and Construction
ties of each of the aggregates I, 2, ... n and P1, P2 , ••• Pn are the weight per-
centages of each aggregate 1,2, ... n in the blend. This equation can be used
to determine average bulk specific gravity or average apparent specific grav-
ity, but the specific gravity type must be the same for all aggregates being
blended.
where,·
G,a = apparent specific gravity; and
G,b = bulk specific gravity.
The volume of the external pores is equal to the volume of absorbed water
when the aggregate is in SSD condition. Assume for this example that the
weight of dry aggregate is I 00 grams. Then ·
The water content in the saturated surface dry condition is 2.0 percent as
_ ___,given in the problem. The aggregate must be in the SSD condition to mea-
sure all of the external pores
G.=2.607
This problem is worked two ways. The equation for determining the aver-
age specific gravity of a blend of aggregate is:
G= P1+Pz
Pi Pz
-+-
Gl Gz
G =0.500 + 0.500
0.500 + 0.500
2.500 3.000
Simplifying
1 1
G=----=--
0.200 + .167 .367
------Hefl:€e--------------
G = 2.725
152 Hot Mix Asphalt Materials, Mixture Design and Construction
The problem will now be worked by simply dealing with weights and
volumes. The two aggregates are combined by equal proportions by
weight. Assume the weight of aggregate # 1 is 100 grams. Hence, the
weight of aggregate #2 is also 100 grams.
.~ ~• weight of aggregate 1
volume o, aggreg...e 1 =
specific gravity of aggregate 1
100
=--
2.500
= 40.000
100
=--
3.000
= 33.333
Hence, the total weight of the blended aggregate is 200 grams and the total
volume of the blended aggregate is 73.333 ml. It follows that
G . = weight = 200
.sb volume 73.333
and
The average of the two specific gravities is 2.75 which is not the correct
specific gravity of the blend.
where d is the diameter of the sieve size in question, P is the total percent
passing or finer than the sieve, and D is the maximum size of the aggre-
gate. Studies by Fuller and Thompson showed that a maximum density can
be obtained for an aggregate when n=0.5.
200
100
100 00 30 ,. ." I
,,
I
,-- V
~
I I 90
90
~
,
/
I I :
I :
80
~ , II ;
I: I I :
70 70
I
/ :
"'ti I / I !
~ so~
0. I / I j
...z ,_ __ , I/
/"'4 !'--Uniformly j
~
a::
,0 Gop-groded> _
/ I
graded
""
-- --
UJ
0.
;l 40~
/ I
... ,,v I
I
'
4
...
0
I / I
3o~ I : 3
Well-graded or / I
20
10
~
~ I
I Dense grad~~
-~ I/
.I
I
I
:
20
0~
200
----
SIEVE NO.
100 00 30 ,. . ~,-:-
i" t" ¼" It"
SQUARE OPENINGS
0
.., 00
00 00
"' "'
00
,., ,.,
..,
:,0 :,0
zo zo
10 10
0
o
••
No.200JOO .50 JO ic 11211
.
'"
SIEVE SIZES
mally limited to about one-half of the lift thickness. In recent years, there
has been an increase in the use of large stone mixes to minimize rutting
potential of HMA. The introduction oflarger stone sizes increases the vol-
ume concentration of the aggregate. This factor contributes to a reduction
of both design asphalt content and cost of the mix. These large stone mix-
tures are more resistant to rutting than the smaller aggregate size mixtures.
However, the use of a maximum aggregate size greater than 1 inch (25.4
mm) often results in harsh mixes that tend to segregate during placement.
Therefore, special attention must be given to mix design, mat thickness,
material handling, mixing, and paving procedures when these larger max-
imum aggregate size mixtures are used.
the measured amount of material passing the No. 200 (75 µm) sieve is
most likely significantly lower than the actual amount in the aggregate
mixture. For clean aggregate, a dry sieve analysis may be accurate enough
since the amount of material passing the No. 200 (75 µm) sieve is low.
Example 3-3. Assume that an aggregate blend was split into two sam-
ples. Determine the percent passing each sieve size using the dry and
washed gradation methods. A dry gradation was conducted on the first
sample which weighed 2012 grams (Table 3-9). A washed gradation was
performed on the second sample which weighed 2074 grams. The weights
and calculations for the washed gradation shown in Table 3-10.
The steps in a dry sieve analysis include splitting the sample to the cor-
rect size, determining the total dry weight, pouring it over a set of sieves
and shaking, weighing the material collected on each sieve size, determin-
ing the percent retained on each sieve, and determining the percent pass-
ing each sieve. When a washed sieve analysis is conducted to obtain the
washed gradation, the material is washed over a No. 200 (75 µm) sieve to
remove the dust. The material is then dried, weighed, poured over a nest
of sieves, vibrated, and the amount retained on each sieve determined.
1:11
C
j sot---+-+-+--+--+--+----r--1---.--+.+'-----+----leo
:.
i
i 40
-- -·---··--·---·- '~.'iot~;;o--.~oi=:f..re==i's=====•~==::;3;;';;,.-::;;;,====a..l,•.::;;_,~:::..=--=-:i,_g__________
Sieve Sizes
Dense Mixtures -
CT'
n
~
~
&
w
Sieve S ,:e Mix Designation and Nominal Maximum Sire or Aggregate
I
>I k~
2 inch 1½ inch 1 inch %Inch ½ Inch "Inch
No.4
(4.75 mm) No.8
No.16
(1.18mm) ~ ere,
~
~
(50 mm) (37.5 mm) (25.0 mm) (19.0 mm) (12.5 mm) (9.5 mm) (Sand Asphalt) (2.36 mm) (Sheet Asphalt)
""'
Grading or Total Aggregate (Coarse Plus Fine Plus Filler IC Required) t,
Amounts Finer than Each Laboratory Sieve (Square Opening), Percent by Weight ~
~ ....01
\C)
"'
160 Hot Mix Asphalt Materials, Mixture Design and Construction
Sieve opening, In •
~g ......
.,,.,
"'"'0
q
00
00
t-Ol
qq
100°
00 00
I J I/
90
80
:~~::o~e~ol: tw:lily lo produce, I I I
,/I I
-- - IP --
g' 70
-.; -'V /
i 60
Crl!lcal m!xlures ~ become V
~ 50 readily unstabla with silght
------
excess of asphalt or waler / V
.~ 40
a.
30 --
L--- i---
~ .4"ar~h mbcfw;e Inclined lo
segregate, nol critical
20 '---;;;, --- ~
i:;.. i:::-
, L I I
Porous· mixture, ~: <lack te11slle strength
when mixed with fuel oil, best
I I
10
r:-- , r■sull1 with heavy bilu~n I
0
200 100 80 50 40 30 20 16 108
Sieve number
(19 mm-9.5 mm), 3/8 inch-No. 4 (9.5 mm-4.75 mm), and minus 4
(4.75 mm) prior to hauling and stockpiling. Another reason for blending
aggregates is that it is often more economical to combine naturally occur-
ring and processed materials to meet specifications than to use all processed
materials.
0.45 Power Gradation Chart
80
f
Ill
60
...
a.
e
C
40
:.
p = Aa + Bb + Cc + ... (9)
where,
p = the percent of material passing a given sieve for the
combined aggregates A, B, C, .....
A, B, C, ... = the percent of material passing a given sieve for each
aggregate A, B, C, ...
a, b, c, ... = proportions (decimal fractions) of aggregates A, B,
C, ... to be used in the blend, a+ b + c... = 1.00
100
I
90 I
I
I
80
t!)
z 70
in
(J)
< 60
11..
I
E
I-
z ✓
ILi ✓
u
a:;
✓
✓
ILi
11.. 30 /A
✓
--- --
_../
10
o._..=-=;:;;;;L~--1L...-~....1....---1..--1_---1...J.....----1..--L..---'--__.__ __,
200 100 80 50 40 30 20 16 108 4 3/8 3/4 1-1/2 3 In.
SIEVE NUMBER
- -0 ~Jt---""-l-49---0.42-0-CUl~l-----2..00-4~7..6--L2.7 - . 25.4 76.1
·'\"T-"'• I I I I I I I I
SIEVE OPENING, mm
a
Example 3-4. Determine blend of the two aggregates shown in
Table 3-11 which will meet specification requirements using the
trial-and-error method.
Examination of the grading indicates that it is possible to find a blend
that falls within the specification limits, possibly a 50-50 blend because of
the relative distance of the curves to the center of the band (desired grada-
----=t1:on)-:-Tlie1irstffial blend can be determmed more mtelhgently 1f certam
critical sieve sizes are selected. By inspecting the gradations, it is seen that
all material retained on the 3/8-inch or 9.5 mm sieve (100-80 = 20 per-
cent desired) has to come from aggregate A, and all material finer than the
No. 30 (0.6 mm) sieve must be furnished by aggregate B. With regard to
~ Q
~
~ ~
w
'f,_. i
,_.
Sieve Size
r$" I "'
('Jq
Aggregate Percent 3/4 inch 1/2 inch 3/8 inch No.4 No.8 No.30 No.50 No.100 No. 200 0
'"1-'j
Used
Aggregate A 100 100 90 59 16 3 0 0 0 0 ~
0
Aggregate B
50 percent A
50 percent B
100
50
50
100
50
50
100
45
50
100
29.5
50
96
8
48
82
1.5
41
51
0
25.5
36
0
18
21
0
10.5
9
0
4.5
f~
~
Blend 100 100 95 79.5 56 42.5 25.5 18 10.5 4.5
Desired 100 100 90 80 60 43 24 18 12 7
Specification 100 80-100 70-90 . 50-70 35-50 18-29 13-23 8-16 4-10
,_.
8:
166 Hot Mix Asphalt Materials, Mixture Design and Construction
Now looking at the No. 30 (0.6 mm) sieve it is apparent that 75 per-
cent of the material larger than a No. 30 (0.6 mm) sieve must come from
materials A and B. Since 75 percent is the total desired to be retained on
the No. 30 (0.6 mm) and 0.64(95) is that portion provided by aggregate A.
Chapter 3-Aggregates 167
Based on these calculations, it appears that the best first estimate would be
64 percent of aggregate A, 14 percent of aggregate B, and 22 percent of
aggregate C. It is obvious, however, that using 22 percent of aggregate C
would produce too much material passing the No. 200 (75 µm) sieve. The
percent of aggregate C should probably be no more than 8 percent, since
that aggregate alone will produce approximately 7 percent passing the
Aggregate
Median of
Sieve Size A B C Specification Specifications
1 inch 100 100 100 94 to 100 97
½ inch 63 100 100 70 to 85 78
No: 4 19 100 100 40 to 55 48
'No. 8 8 93 JOO 30 to 42 36
No. 30 5 55 100 20 to 30 25
No. 100 3 36 97 12 to 22 17
No. 200 2 3 88 5 to 11 8
No. 200 (75 µm) sieve. Based on this analysis, the first estimate should be
71 percent of aggregate A, 21 percent of aggregate B, and 8 percent of
aggregate C. The calculations are shown in Table 3-13.
The estimated percentages of 71 percent A, 21 percent B, and 8 per-
cent C produce a blend that meets the specification requirements.
However, the blend can be improved slightly to get it closer to the desired
gradation. To improve the blend, use 2 percent less of aggregate C and 5
percent less of aggregate A and, hence, 7 percent more of aggregate B. For
each 1 percent that two aggregates are changed the resulting blend will
change by the difference between the percent passing for any given sieve
size for the two aggregates times 1 percent. For instance, increasing aggre-
-----g~a=te-B-oy:, percen:-r-.rn-d~d-ec::re"lfsin:g-.rggreg:i:te-Aby 5 percent will result in
a change of (.81) 5 or 4.0 on the No. 4 sieve.
After making the calculations (Table 3-13) it can be seen that the
amount of each aggregate to use is 66 percent A, 28 percent B, and 6 per-
cent C.
168 Hot Mix Asphalt Materials, Mixture Design and Construction
Sieve Size
Aggregate Percent 1 Inch 1/2 No.4 No.8 No.30 No.100 No.200
Used Inch
C 8 8 8 8 8 8 7.8 7.0
Desired 97 78 48 36 25 17 8.0
C 6 6 6 6 6 6 5.8 5.3
'~
Blend 100 100 75.6 46.5 37.3 24.6 17.9 7.4
There are several graphical methods that have been used to determine
the percent of each aggregate to use in a blend. These methods are gener-
ally not used since the trial-and-error method is quicker and easier in most
cases. These methods are especially difficult to use when more than two
aggregates are being blended, which is normally the case for HMA. For
more information on graphical methods, additional references should be
reviewed (§, 12.,_ 21, 22). ·
Computer programs have been developed to determine the optimum
blend for a specified number of aggregates. None of these programs have
been very successful since an experienced engineer can determine the
desired proportion of up to four aggregates by the trial-and-error method
- - - - 5 y tne time ne could mput the data mto tlie computer. Another probli-::ec:::m=------
with a computer method is that it is difficult to quantitively describe the
optimum blend. For example, the significance of a 2 percent deviation
from the desired gradation is not as critical for a No. 4 (4.75 mm) sieve as
for a number 200 (75 µm) sieve. Many factors must be considered when
Chapter 3-Aggregates 169
Surface Area
One important property that can be computed from the aggregate gra-
dation is the surface area. The aggregate surface area is important since it
affects 'the amount of asphalt needed to coat the aggregate. Dense-graded
asphalt mixtures are usually designed to contain a desired amount of air
voids; hence, the aggregate surface area is not a design factor. It is possible
to increase the surface area of an aggregate and at the same time reduce the
optimum asphalt content. One way to do this is by increasing the dust
content of a mixture. Asphalt mixes that have high surface area and low
optimum asphalt content are undesirable because these mixes will have a
thin asphalt film on the aggregate and will probably not have adequate
durability. A further discussion of asphalt film thickness is provided in
Chapter 4.
The most common method. used for estimating the aggregate surface
area involves calculations made from the aggregate gradation. The surface
area is determined by multiplying surface area factors by the percentage
passing the various sieve sizes. These factors can be calculated by assuming
an aggregate specific gravity and assuming all particles are rounded or cube
shaped. The surface area per pound of aggregate can then be determined
for each aggregate size.
The surface area factors for various sieve sizes are shown in Table 3-14.
The surface area is determined by multiplying the factor by the percentage
passing for each sieve size and totalling for all sieve sizes. The units for these
calculations are square feet per pound. All surface area factors for all sieve
sizes shown in Table 3-14 must be used. The surface area is only slightly
------affectecl: by the percent passmg the larger sieve sizes-oiifis significantly
affected by the percent passing the smaller sieve sizes.
One of the primary reasons for estimating the surface area is to deter-
mine the asphalt film thickness. This is an estimated value, but it does
allow comparisons to be made for various mixtures.
170 Hot Mix Asphalt Materials, Mixture Design and Construction
Sieve Percent
Size Passing
3/4 inch 100
1/2 inch 88
3/8 inch 72
No.4" 60
No.8 47
No. 16 35
No. 30 25
No. 50 20
No. 100 10
No. 200 5
Using the surface area factors and the percent passing each sieve shown
above, the surface area (SA) is calculated below:
. . I
SA= 2(100) + 2(.60) + 4(.47) + 8(.35) + 14(.25) + 30(.20) + 60(.10)
+ 160(.05)
REFERENCES
INTRODUCTION
In 1995, a few states began to use the Superpave design procedures for
a small number of projects. It is anticipated that within the next 2-5 years
most HMA projects will be designed using the Superpave concept.
Most HMA produced during the 50 years between the 1940s and mid
1990s were designed using the Marshall or Hveem methods. According to
a survey in 1984; approximately 75 percent of the state highway depart-
ments used some variation of the Marshall method while the remaining 25
percent used some variation of the Hveem method (1). Some states used
both methods ori the same materials to get a better understanding of the
characteristics of the mixtures. It is important to recognize that all three
methods are designed to help the asphalt technologist choose an appropri-
ate asphalt content with which to begin field construction. It is important
to understand that if the service conditions for which a mixture is being
designed are different from those for which the design method was devel-
oped, the mixture design may not be adequate for service even though it is
designed according to the method. It is incumbent upon the user of each
of these methods to understand the limitations of that design method to
ensure that the method is used appropriately. To enhance that understand-
ing, these three mixture design methods are discussed in detail in this chap-
ter. The design of only dense-graded HMA mixtures is discussed.
Crawford has traced the history of asphalt mix design dating back to
the 1860s and has given a very interesting account of mix design develop-
ment in Reference 2_.
174 Hot Mix Asphalt Materials, Mixture Design and Construction
Dense Gradation
Hubbard-Field Method
dard orifice. The load was reported as the stability value. Stability numbers
corresponding to various asphalt contents were plotted and the optimum
binder content determined. This was probably the first attempt to quanti-
fy empirical mix stability values at various asphalt contents. The method
was widely accepted and survived in some states for a long time. The test
was modified in the mid 1950s to test six-inch (150 mm) diameter speci-
mens to accommodate mixes containing coarse aggregate up to 3/4 inch
(19 mm) maximum size. However, the modified version was used for a
very short time because the Marshall test had started to gain popularity
during that period.
ated tamper foot. The foot is raised after a specified pressure is sustained
by the specimen, the base rotates 1/6 of a revolution, the tamper foot is
automatically lowered, and the specified pressure is applied again around
the perimeter of the specimen. The area of the tamper foot is one-fourth
the cross sectional area of the specimen.
A second mechanical test device called a cohesiometer was developed
along with the Hveem stabilometer. It was designed to measure the cohe-
sive strength across the diameter of a compacted specimen on which the
stability test had already been conducted. The specimen is placed in the
cohesiometer, the specimen is secured, and the load is applied by lead shots
flowing from a reservoir into a bucket at the end of the moment arm
(details are given in ASTM D1560). Load is applied until the specimen
fails, and at that time the shot supply is automatically shut off. The total
applied force is determined and the cohesion calculated by a formula. This
test was aimed at measuring a tensile property of the oil mixes that could
be related to a minimum level to preclude ravelling of surface mixes under
tractive forces. This test proved to be of little value in characterizing HMA
surfacings. HMA surfacings were made with asphalt cement and always
had cohesion values large enough to prevent ravelling. Therefore, when oil
mixes were replaced by HMA surfacings after World War II, the cohe-
siometer test served no real purpose and it gradually fell out of favor.
By 1959 the Hveem stabilometer procedure for design ofHMA mixes
had evolved into its final form. Several states, primarily in the West, adopt-
ed the Hveem procedure. However, many states modified various elements
in the procedure.
Asphalt Content
Selection Limiting Criteria
Test Property
HMA Sand Asphalt HMA Sand Asphalt
Stability, lbs. (N) Peak . Peak 500 (2220) min 500 (2220) min
Flow, 0.01 inch (0.25 mm) - 20 max 20 max
Unit Weight Total
Mix, pcf (g/ml) Peak Peak
Voids Total Mix, % 4 6 3-5 5-7
Voids Filled with
Asphalt,% 80 70 75-85 65-75
During the late 1940s and early 1950s, aircraft size and weight con-
!-----tinued to increase-with-the-resul:t--that-rire-pressures-were-i.ncreased-ro-2-0 - · ~ - - - - -
psi (1380 kPa) in order to keep aircraft gear and tire sizes about the same.
This increase in tire pressure significantly increased the near surface com-
pactive effort of traffic. The WES began a study to determine what labo-
ratory compactive effort was required to correlate with traffic-induced
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 179
densities and to modify the low pressure mix design and quality control
criteria. The original full-scale test sections were trafficked by 30,000 lb
(133.3 kN) wheel loads on 200 psi (1380 kPa) tires and the pavement per-
formance monitored.
Even though the surface density continued to increase with the appli-
cation of the high pressure tire loads, the surface rutting measurements
were confounded as a result of a failure in the asphalt binder course under-
neath the surface. Using the data from both the field study and the labo-
ratory compaction study, the following criteria were adopted for the high
pressure tires (200 psi or 1380 kPa) which included adjustments in stabili-
ty and flow based on test data from the trafficked surface as well as a mod-
ification to the laboratory compaction effort:
Selection of
Test Property Asphalt Content Limiting Criteria
and performance of HMA: asphalt content and density. In the field it is the
highest satisfactory asphalt content at a density achieved under traffic that
is significant. In the laboratory the important feature is selecting a com-
paction procedure that represents traffic-induced density and then select-
180 Hot Mix Asphalt Materials, Mixture Design and Construction
ND
IIT
SD
'If'(
~HVEEM
C]MARSHALL
~ TEXAS
-GRADATION
Objectives
Fatigue Resistance. The mix should not crack when subjected to repeated
loads over a period of-time. Complex repeated load (either constant stress
or constant strain controlled) tests discussed in Chapter 5 can be run to
estimate the number of cycles to cause failure (cracking) in HMA mixtures.
11-----S'kid- Resistance;-'Fhis-requiremencis'-orrly-a:ppticable-to-stirface-mixe-s
1
PERCENT PASSING
8
- .
0 0
co g i [j! 0
\
\
\
f\ \
I\ I\
i\
\ \
I\
I\ f\
I\ I\ I\
ffi
3
0
a. \ I\
~~ \ \
0
~ ~ I\ i\
0
l- - ;;;~ [3
N
a I'\. '
[", I'\.
\ i\ \ cij
~ I\ [\ w
<
0::
~ ~
\ \ >
w
en
~ I'\ I\
I'\ I\ [\ I\
[3
N ~
' f" I"\ I'\ \ I\' I\I\
(/)
w
Gi "' r---,.
1'-,,
f" I'\
cij 1'.
1'-,, I'\ I'\. I\f\ i\
I"\~ \
" t---- r-.... t---- '
~
r---,. ' ~I'\ I\ [\\
--t--..
r--.. r--..
t----
r--..
'r---.
I'---
~
I"--
'~ "' "~
r---,. l\L'\. ~
\
0
"l
-
0
r--.. ~
r-- 'r---. ', ~ ~ i
0
~
r--.. ............ I"-..
....... i::::: '"~ ~~
~
~\ !
0
.,0
"' 0
0
·-------+--------9NISS~d-J.N3::>!:13d- - -
0
0 0 0 g 0 0 0 0
N
0 0
Ol CD CD
"' "' N
~
-
~ z ,,,,
~
0
\ I\ /<f.
i==
\ l;- <f.
I'. Cl
<f.
I'\ f\ ' a::
~~ V
(!)
~
:i-
;;;
~
' )\
~
I
......... ......... a) 17I/ ~ f\.. ~
•W
~N
Cf)
u " .\ k' .w
~>
ci5
"[\_~ ~ ~
"\ ' "'w
ci5
'- st
~~
' I\.
'
CD
"~ ~ I',.,_
"'
I'\ ~ 0
p
' \.
, .~
.,0
1::-.;
' .~ 0
~
-·- -~--
0
' 0 0 0 0 0
0
r-
Ol
"' "' "'
E>NISS'vd 1N3:::>H3d
Figure 4-3. Three Common Methods of Plotting Maximum Density
Lines on 0.45 Power Gradation Chart
Chapter 4-Hot Mix Asphalt Mixture Design Methodology ' 185
too much asphalt cement binder to cause the HMA mix to flush and cre-
ate a slippery surface. Skid resistance is discussed in detail in Chapter 8. It
is also discussed briefly later in this chapter.
The solid line in Figure 4-3 shows the gradation of a binder course
mix (as an example):
These gradation data are used three different ways to produce the maxi-
mum density lines labeled A, B, and C in figure 4-3. The construction of
each line is described below.
Tp =
vase
SA x W (1,000
)
where,
TF = Average film thickness, microns;
V asp = Effective volume of asphalt cement (liters);
SA = Surface area of the aggregate (m2 per kg of
aggregate); and
W = Weight of aggregate (kg).
Recall that
vase
SA X w (1,000)
0.2847%
0.002847
Weight ofabsorbed asphalt 0.002847 (weight ofaggregate)
vasp
SA X w (1,000)
107.4 (1,000)
7.23 (2340 X 0.95)
Tp = 6.7 microns
Laboratory Compaction
Based on the discussion of the evolution of both the Marshall and the
Hveem methods of mixture design, it is apparent that the laboratory com-
paction methods included in the initial methods were referenced to field
conditions. The reference field conditions were not based on the as con-
structed densities but on the density achieved after the pavements had sus-
----~t.ai.n~_d_a_.r_e_a_sonable level of traffic;, Th_e_S_up_e.rp_ave methQdis also designed
to provide a density equal to that after some amount of traffic.
Therefore, it is clear that if the engineer intends to design a mixture in
the laboratory that will sustain the loads imposed on it by traffic, that lab-
oratory mixture should be compacted to the same general density (void
190 Hot Mix Asphalt Materials, Mixture Design and Construction
content) to which the traffic will finally compact it under service condi-
tions. Otherwise, the traffic levels may be so high in service that the mix-
ture may be compacted to such a dense state that the void content is too
.low and instability results. Alternatively, the traffic levels may be so light
that the compaction energy used in the laboratory produces a density too
high for the actual traffic level. The result is too little asphalt cement and
a mixture that has low durability in the field. This second type of mix does
not exhibit instability in the field, but the film thicknesses are too thin for
good durability under light traffic. It is apparent then that the selection of
the laboratory compaction level is critical in designing a mixture that will
provide adequate field performance.
Figure 4-4 illustrates the type of relationship that exists between air
void content and level of compaction in th:-e laboratory and the time the
same mix is subjected to field traffic as reported by Arena (11). Notice in
the plot that as the number of 'revolutions of the gyratory compactor
increases, the air voids go down. In the gyratory compaction device, a com-
pressive load is applied to a specimen while the mold containing the spec-
imen is tilted from side to side in a gyrating action by a rotating flanged
9
MIX NO. 2
ASPHALT 5.8%
8
100 PSI
•
Cl)
0
0
>
I- 6
z
w
7
\-L •
~ 5 BLOWS, PLANT CMECHJ
u
c::
w •
a.. 5
4,
0
·--.
10 20 30 40 50 60 70
NUMBER OF GYRATIONS
mold chuck (see Figure 4-5). Notice also in Figure 4-4 that several refer-
ence points are noted on the curve. The first point, "ORIGINAL," corre-
sponds to the original as constructed density in the field which is normal-
ly specified as some percentage oflaboratory density; the second point, "75
BLOWS PLANT (MECH)," corresponds to the air void level achieved
from the field mixed material that was compacted in the field laboratory
using 75 blows from a mechanical Marshall compactor; and the next 3
points, "6 MONTHS," "15 MONTHS," and "36 MONTHS," represent
air void contents determined from cores taken from the wheel paths after
the mixture was subjected to traffic for those respective periods. While the
Marshall compaction effort in the field laboratory corresponded to an air
void content of slightly less than 6 percent, the mixture appeared to stabi-
lize under traffic at an air void content between 3 and 4 percent. Notice
also that the gyratory compactor is capable of achieving air void contents
much lower than that achieved by mechanical Marshall hammer compac-
tion. The compaction curve shown in Figure 4-4 applies only to the spe-
cific mix investigated and cannot be considered as a typical plot for all
mixes.
If surveys show that the laboratory compaction method consistently
produces densities substantially lower or higher than those achieved in the
field under traffic, the specifying agency should consider changing the lab-
oratory compaction methodology or blows (compaction effort) in order to
achieve laboratory densities which are consistent with those being pro-
duced by traffic.
Fixed Roller
Gyrolory Angle
Adjustment Knob
VT COMPACTED WT
SPECIMEN
Vy Air Voids 0
V,
Ji'7M = __! X 100
VT
W.4gg
VAa = ----=---
Sp Gr.4gg x Yw
' V,
YFA = .&tC V x 100
V.&(C + y
where,
VT = Total volume of compacted specimen;
Vv = Volume of air voids;
VEAC = Volume of effective asphalt cement;
VMC = Volume of absorbed asphalt cement;
VAss = Volume of aggregate;
WT = Total weight of compacted specimen;
WEAc = Weight of effective asphalt cement;
WMc = Weight of absorbed asphalt cement;
WAss = Weight of aggregate;
VTM = Voids in total mix, percent;
VMA = Voids in mineral aggregate, percent;
VFA = Voids filled with asphalt, percent;
AC = Asphalt content by weight of mix, percent;
'Yw = Unit weight of water; and
Sp Gr = Specific gravity.
Figure 4-6. Weight-Volume Relationships for Hot Mix Asphalt
absorbed asphalts
Impermeable
voids
effective asphalt
water
permeable voids
voids
impermeable voids
voids
impermeable voids
impermeable voids
permeable voids
Bulle Specific Gravity, Dry (ASTM C127 and C128)-The ratio of the
oven dry weight in air of a unit volume of a permeable material (including
both permeable and impermeable voids normal for the material) at a stat-
ed temperature to the weight of an equal volume of gas-free distilled water
at a stated temperature (see Figure 4-9).
WssD - Wsub
where,
WO = Dry weight, grams;
W 550 ,,; Saturated surface dry weight, grams; and
Wsub = Saturated surface dry weight submerged m water,
grams.
impermeable voids
volume of aggregate
impermeable
voids volume of voids not filled with
asphalt
simplifying,
V (bulk)
VMA = 100 -' ..tu (100)
VT
W. IG
VMA = 100 - AU sb 100
WJGm1,
substituting,
and simplifying
(1)
where,
VAgg = Volume of aggregate;
VT= Total volume of compacted mixture;
WAgg = Weight of aggregate;
WT= Total weight of mixture;
G.b = Bulk specific gravity- of aggregate;
Gmb = Bulk specific gravity of compacted mixture; and
Pb= Asphalt content.
By definition,
Vy
Jl7M = - X 100
VT
substituting,
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 201
(2)
where,
VTM = Voids in total mix;
Vv = Volume of air voids;
Vy = Total volume of compacted specimen;
VEAc = Volume of effective asphalt cement;
VAgg = Volume of aggregate (bulk);
Wy = Total weight of compacted specimen;
Gmb = Bulk specific gravity of compacted specimen; and
Gmm = Maximum theoretical specific gravity of mixture.
G,, =
since the effective specific gravity (G,J is equal to the weight of aggregate
-----Jivtded-by-rhreffective-voforne··ohhe-a:ggrega:te-(V:e.f!}.The effective vol-
ume is the volume of the aggregate plus the external voids minus the exter-
nal voids filled with asphalt cement. In the test for maximum specific g~av-
ity (Gmm), the total volume of the voidless mix is measured and the volume
of asphalt cement is determined from the weight of asphalt cement and the
202 Hot Mix Asphalt Materials, Mixture Design and C(i)nstruction
Hence,
G,, =
substituting,
G = WT - Pb WT
se WT
-- ---
w;..c
G_ Gb
simplifying,
Gu =
where,
w.gg = Weight of aggregate;
VEff = Effective volume of aggregate;
VAC = Volume of total asphalt cement;
Wr = Total weight of mixture;
Vrv = Total volume of voidless mixture;
Pb = Asphalt content;
WAc = Weight of total asphalt cement;
Gb = Specific gravity of asphalt cement; and
Gmm = Maximum specific gravity of mixture.
Substituting,
simplifying,
where,
Wagg = Weight of aggregate;
WAc = Weight of total asphalt cement;
VEff = Effective volume of aggregate;
VAc = Volume of asphalt cement;
WT= Total weight of mix;
Gse = Effective specific gravity of aggregate;
Gb = Specific gravity of asphalt cement; and
Pb= Asphalt content, percent by weight of mix.
pba = ( ; : ) 100
The volume of absorbed asphalt is the difference between the bulk volume
- - - - - - t i :£.-ihe-aggregate-and-the-effecti:v:e-volume.-Hence----- -.--·
204 Hot Mix Asphalt Materials, Mixture Design and Construction
simplifying,
where,
Pba = Absorbed asphalt by weight of aggregate;
W MC = Weight of absorbed asphalt;
W Agg = Weight of aggregate;
VMC = Volume of absorbed asphalt;
Vbulk = Bulk volume of aggregate;
VEff = Effective volume of aggregate;
Gb = Specific gravity of asphalt cement;
G,e = Effective specific gravity of aggregate; and
G,b = Bulk specific gravity of aggregate.
B-1. Determine appropriate asphalt cement grade for type and geograph-
ical location of mixture being designed.
B-3. Determine asphalt cement specific gravity and plot viscosity data on
a temperature-viscosity plot.
206 Hot Mix Asphalt Materials, Mixture Design and Construction
Prepare Marshall Specimens using the procedures given in Steps C-1 thru
C-7 for individual specimens. As an alternate, HMA for a desired asphalt
content can be prepared in bulk to produce the required number of com-
pacted Marshall specimens and to provide a sample for determination of
maximum (Rice) specific gravity.
C-1. Dry and sieve aggregates into sizes (preferably individual sizes) and
store in dean sealable containers. Separate enough material to make 18
specimens of approximately 1150 g each. Minimum aggregate and asphalt
cement requirements to prepare one series of test specimens of a given gra-
dation are 50 pounds (25 kg) and one gallon (4 liters), respectively.
C-2. Weigh out aggregate for 18 specimens placing each in a separate c;:on-
tainer and heat to mixing temperature determined in Step B-4. However,
the total aggregate weight should be determined as discussed in Step C-3.
where,
C-6. Remove the hot aggregate, place it on a scale, and add the proper
weight of asphalt cement to obtain the desired asphalt content.
C-7. Mix asphalt cement and aggregate until all the aggregate is coated. It
is helpful to work on a heated table. Mixing can be by hand, but a mechan-
ical mixer is preferred.
C-9. Place a paper disc into an assembled, preheated Marshall mold and
pour in loose HMA. Check the temperature. Spade the mixture with a
heated spatula or trowel 15 times around the perimeter and 10 times over
the interior. Remove the collar and mound materials inside the mold so
that the middle is slightly higher than the edges. Attach the mold and base
plate to the pedestal. Place the preheated hammer into the mold, ~d apply
the appropriate number of blows to the top side of the specimen.
C-10. Remove the mold from the base plate. Place a paper disc on top of
the specimen and rotate the mold 180° so that the top surface is ori bot-
tom. Replace the mold collar and attach the mold and base plate to the
pedestal. Place the hammer in the mold _and apply the same number of
blows to the opposite side of the specimen~ .
C-11. Remove the paper filters from the top and bottom of specimens.
Cool the specimens and extrude from the mold using a hydraulic jack.
Place identification marks on each specimen with an alphanumeric code
using a grease pencil. Allow specimens to sit at room temperature
overnight before further testing.
C-12. Determine the bulk specific gravity for each specimen by weighing
in air. Submerge the samples in water and ailow to saturate prior to getting
submerged weight in SSD condition. Remove the sample and weigh in air
in saturated surface dry condition. This test is conducted in accordance
with AASHTO T166.
C-13. Measure the Rice specific gravity on the loose HMA mi?c samples in
accordance with AASHTO T209 (ASTM D2041). This specific gravity is
required for voids analysis. If the mix contains absorptive aggregate it is
recommended to place the loose mix in an oven (maintained at the mix
temperature) for 4 hours so that the asphalt cement binder is completely
absorbed by the aggregate prior to testing. Keep the mix in a covered con-
tainer while in the oven. If the test is run in triplicate on the mix contain-
ing near optimum asphalt content, average the three results, calculate the
!: effective specific gravity of the aggregate, and then calculate the Rice spe-
.~___cific gravity _or__.TMD_for__.the-remaining-m.ixes-with-d-iffa.r-e-m-asphal:t-ron-- - - - - -
f tents. If one TMD test is conducted on each mix containing a different
I asphalt content, then calculate the effective specific gravity of aggregate in
1
each case. Calculate the average effective specific gravity, and then calculate
the TMD values using the· average for all five mixes.
r
I Chapter 4-Hot Mix Asphalt Mixture Design Methodology 209
D-1. For each specimen, use the bulk specific gravity (Gmb) from Step
C-12 and the Rice Specific gravity (Gmm) from C-13 to calculate the per-
cent voids or VTM.
D-3. Calculate the VMA for each Marshall specimen using the bulk spe-
cific gravity of the aggregate (G, 6) from Step A-2, the bulk specific gravity
of the compacted mix (Gm6) from Step C-12, and the asphalt content by
weight of total mix (Pb):
D-4. Calculate the VFA (voids filled with asphalt) for each Marshall spec-
imen using the VTM and VMA as follows:
E-2. After heating for the required amount of time, remove a specimen
from the bath, pat with towel to remove excess water, and _quickly place in ',"-"
the Marshall testing head.
E-3. Bring the loading ram into contact with the testing head. Zero the
pens if using a load-deformation recorder or zero flow gauge, and place the
gauge on the rod of the testing head~ Apply the load at 2 inches/minute
(50.8 mm/minute) until maximum load is reached. When load just begins
to decrease, remove the flow meter, stop ram movement, and record the
stability (maximum load) in lbs (Newtons) and flow in 0.01 inches
(0.25 mm). Testing should be completed within 1 minute from the time
the specimen is removed from the bath.
E-4. Repeat Steps D-2 and D-3 until all specimens have been tested being
careful to check that:
F-1. Tabulate the results from testing, correct the stability values for spec-
imen height (ASTM D1559), and calculate the average of each set of 3
specimens.
G-1. The following two methods are commonly used to determine the
optimum asphalt content from the plots:
G-2. Failure to meet the criteria for any property requires that special
action be taken to determine if the mix can be used in construction. Local
experience with a mix which fails to meet'particular criteria may verify that
the mix performs satisfactorily. If that is the case, the criteria ~ay be mod- .
ified only for those mixes that serve satisfactorily under specified traffic and
environments. If the VMA. criteria are not met, the aggregate gradation
must be modified and the mixture redesigned.
G-3. The laboratory developed mixture design forms the basis for the ini-
tial job mix formula QMF). The initial JMF must often be adjusted early
I, in the construction project because of slight differences between the labo-
ratory supplied aggregates and those being used in the field. Often, the
asphalt content developed in the laboratory is too high for field construc-
tion and must be reduced a few tenths of a percent, especially if the VTM
of the produced mix is low. Construction experience during the first few
days should be carefully reviewed and the JMF modified to provide a mix
that can be satisfactorily mixed, placed, and compacted. Ensure that· the
produced mix satisfies all Marshall parameters.
An example of Marshall mixture design is given later in this chapter.
11:
! :
Marshall Criteria
As mentioned earlier, the criteria used by the various states that design
I
mixes using the Marshall method vary considerably. However, in order to
I:
include all the criteria in this discussion, the reader must understand that
....jl.,._':---=n=ot~al=l these criteria ar.e us.e_d_b.y_alLsp.ecify.ing-ageR-G-ies.-GeI1~-al-ly-m-0si:--------
~
()
::,-
~
-t
CT'
n i...,
t
~ ::i:::
~
Marshall Traffic
~
t,:j ~ ,
Method
Light Medium Heavy ~ ~ >
Mix
"' =-
n "' .r,
::,-
Criteria Minimum Maximum Minimum Maximum Minimum Maximum ~~ ~
Compaction,
No. of
blows/side
35 50 75
~
~-~ Iti
l"'I
n ~-"'
:::s
Stability, lb (N) 750 (3333) --- 1200 (5333) --- 1800 (8000) --- t::J ~
n
"'.... 9-
0
Flow ~
0.
g_
0.01 inches 0
(IQ
....
n
Air Voids,% 3 5 3 5 3 5 ....
l"'I
~
Voids in 8"'
l"'I
Mineral (Graphically shown in Figure 4-13)
Aggregate Vl
=l"'I
~
n
n
"'
~
....,_.
N
i:i..
214 Hot Mix Asphalt Materials, Mixture Design and Construction
Mllllmetres1,2
co
(/J
"l
rn
,.._ U'l U'l
ru
0
oi
0
ui
U'l
.,:
,., 0 ,.,
(\J .,f oi (\J U'l (/J
4◊- 40
30 30
Normally
~i1t 0
l
'\-
Acceptable
VMA
~j~arp°:c1flc
Gravity ot Aggregate
<i.
:i 20
>
E
E
~
Deficient In
Either Asphalt
or Air Voids
10 10
No.16 No.8 No.4 318In. ll2in. 314in. lln. I ½Jn. 2In. 2½Jn.
Sieve Slze 1r2
'standard Speclflcallon lor Wire Clolh Sieves /or Testing Purposes, ASTM Designation Ell
IAASHT0 Deslgnallon M92J.
2For processed aggregate, the nominal maximum particle size Is the largest sieve size listed In. the
applicable speclflcallon upon which any material Is retained.
3Mixture in the I% tolerance band shall be permitted only when experience lndlcales that lhe mix-
ture wlll perform satisfactorily and when all other criteria are met.
the percent of voids filled with asphalt (VFA) be within a specified range.
Each of these criteria is discussed below.
thin asphalt cement film, and strip the asphalt cement off the aggregates.
It is very important that the HMA be compacted to a laboratory density
that approximates the ultimate density achieved under traffic and at the
same time have an air void content in the 3 to 5 percent range. The
in-place air void content should initially be slightly higher than 3 to 5 per-
cent to allow for some additional compaction.
I
r
I
Chapter 4--Hot Mix Asphalt Mixture Design Methodology 217
effort to force them closer together. The density reaches a peak and then
begins to decrease because the additional asphalt cement produces thicker
films around the individual aggregates, thereby pushing the aggregate par-
ticles further apart and resulting in lower density.
Flow. The flow is measured at the same time as the Marshall stability.
The flow is equal to the vertical deformation of the sample (measured from
start of loading to the point at which stabil1ty begins to decrease) in hun-
dredths of an inch. High flow values generally indicate a glastic ~ix that
will experience permanent deformation under traffic, whereas low flow val-
~
ues may indicate a mix with higher than normal voids and insufficient
asphalt for_cl!:!I~bility and one that may experience pre~ure crac}ung due
to mixture b r i ~ d g g ~ life of tb.~,~~nt.
For 75-blow Marshall designs that are used on interstates or high vol-
ume roads or airfields, the flow value is usually specified to be in the range
of 8-16. For 50-blow Marshall designs, used on lower traffic volume
roads:the flow is sometimes allowed to be as high as 20.
Percent Voids filled with Asphalt. There are a number of states that
include percent voids filled with asphalt cement (percent VFA) in addition
to the criteria listed above. If a specifying agency includes a VMA require-
ment and exercises air void control during construction, percent VFA is a
redundant requirement for dense-graded HMA. Most states that include
percent VFA requirements generally specify that the VFA range from 7(}JQ_
85 percent.
The basic philosophy of the Hveem method of mix design was sum-
marized by Vallerga and Lovering (1) as containing the following elements:
These elements were combined into a summary statement which forms the
basis for selecting the optimum asphalt content in the Hveem method: Use
a dense, well-graded aggregate with high internal friction without an
excess of fines and add as much asphalt cement as the mixture will tolerate
without losing stability. At least 3 percent air void content is desired in the
Hveem mix design.
Just as with the Marshall Method discussed previously, the procedure
is presented in steps to show how the various elements of the design pro-
cedure fit together. Steps A and B are the same as in the Marshall method
and will not be repeated here; the Centrifuge Kerosene Equivalent test for
the fine aggregate is outlined in Step C; the surface capacity test for the
coarse aggregate is outlined in Step D; the procedure for estimating the
optimum asphalt content is outlined in Step E; stabilometer specimen
preparation is outlined in Step F; calibration of the displacement of the
Hveem stabilometer is contained in Step G; the procedure for the sta-
bilometer test is outlined in Step H; procedures for voids and density
analysis are outlined in Step I; and Step J includes procedures for analyz-
ing Hveem mix design test results to determine the optimum asphalt con-
tent. An example of Hveem mix design is given later in the chapter.
Details of specimen preparation with the California Kneading
Compactor and testing by the Hveem apparatus are given in ASTM
Dl561 and ASTM D1560, respectively.
The purpose of the tests and procedures in Steps C, D, and Eis to esti-
mate the optimum asphalt content at which to compact Hveem test spec-
imens. If previous information on aggregates is available to provide guid-
ance on asphalt content, go directly to step F.
Seive Size Percent Passing* Surface Area Factor Surface Area ft2/Jb (m2/kg)
(mm) (SAF) Percent Passing x SAF/ 100
maximum 100 2 (0.4) 2.00
size
C-5. Stop the centrifuge, remove the cup and weigh, and record the
weight of the aggregate and absorbed kerosene as Ww. Calculate CKE as
percentage of kerosene retained by aggregate:
C-6. Obtain specific gravity (SG) of fine aggregate from Step A. If the SG
is greater than 2.70 or less than 2.60, correct the CKE using the following
relationship:
' SG
CKECORRECTED = CKE X --
2.6S
D-3. Drain at room temperature for 2 minutes. Place the funnel and sam-
ple in an oven and drain at 140°F (60°C) for 15 minutes.
,...--
i
p--4. Remove from the oven, cool, and weigh (Ww). Calculate the amount
of oil retained as a percent of dry aggregate:
Ww- WD
Percent Oil Retained = - - - - x 100
WD
D-5. Obtain coarse aggregate specific gravity (SGcoursJ from Step A. If the
SG is greater than 2.70 or less than 2.60, correct percent oil retained using
the following relationship:
E-1. Using the CKE from Step C-6, enter the chart in Figure 4-14 to
determine the surface constant, Kr, for fine material.
E-2. Using the percentage oil retained from Step D-5, enter the chart in
Figure 4-15 to determine the surface constant, Kc, for coarse material.
E-3. Using the surface area of aggregate (Step C-:-1), the percent coarse
aggregate (percent retained above No. 4 sieve), and the absolute value of
K:-Kf (Steps E-2 & E-1), determine the correction to Kr from the chart
in Figure 4-16.
K.n = Kr + Correction to Kr
Correction to Kr is:
negative if Kc-Kr is negative
positive if Kc-Kr is positive
E-5. Calculate the Approximate BituI?en Ratio (ABR) for a mix that uses
cutbacks, i.e., an oil mix (cutbacks· RC-250, MC-250 and SC-250).
~-----Note:..A-cor:rection-wi1Lbe...made.Jor_the_viscosicy-_o£paving grade asphalt
cements. Using SA from Step C-1, Average Aggregate SG from Step A, Km
from Step E-4, and the chart in Figure 4-17, determine oil ratio if oil was
a cutback. If oil is a paving grade asphalt, go to Step E-6 for a viscosity cor-
rection, otherwise stop.
222 Hot Mix Asphalt Materials, Mixture Design and Construction
~,,, 3,0
'~ll'>_, ,....,~
~<i> .
u
u
~~~ ' ~--:-.,.-.
~,~~ l'>I'"
1.0
.9 ~
cil
.8
I' "~'!, N. ,~r-.~~
r,.. ~ .7
r-.,••'l ,, r-;
r,<"o ~,,,r,.. I:: ' .6
.ll
20
1B '-
16 '
tj 14
, 12
a.
~ 10
,2 9
.., B
7
~
s
6
ll
4
~ 3
~-----m•'------ff-&---------------------7-------
•SUrface area, kg • 0.204816 T Courfuy of Calfornla Deparfmenl of Tr-portallon.
3.0
/
2.8
2.6
/
/
2.4
2.2 /
,:22.0
/
1.8
/
I-
z<( /,,
1.6
I-
en
z
0 1.4
u V
/
w
u 1.2 /
/
<(
LL.
a::
::>
en 1.0
V
.8
/
/
/
1.5 2 3 4 5 6 7 8 9
PER CENT OIL RETAINED - CORRECTED FOR SP. GR. OF AGGREGATE
Material Used• .Aggregate - Passing 9.5 mm (3/8•>, Rel 4.75mm (#4) Sieve
Oil - SAE 10
% Oil Rel. Corrected .. % Oil· Rel. " sp. gr. of Coarse Aggregate
2.65
0.4
~
0.3
.2
C
(.0 ,g
1i5 0.2 f"
st. a
-u
0
C\J
0
..E I"'
.,., OJ
300 ...+-+--.l--!-,-.+-.+-,,-+,-+--.-1,
200 f-+-+--"tr-l--\-+-'~...+--1-1-\....JI-I
i0.
;;: 100~-1--1--14-+--Y--li
,.. 90·1-'--~l-l-++-4--l
3 eol-'---1-1-H+-:-!\
! 70•1----+......l--l+-
f eo,1-----1-1--1---+-+-'l-+-++-.,....+--r-++-1
"'
8' 50 l-'----'\-1--l't-+-.li+-'r-\l--\-'f-l-'H
: 401----+-I---H--l-_....._ _ _1-+l+•U
.
ls
'° 30
]
~
-Cour-tesy-ot-Calltomla-Departmenl-oHranspo)'latto
"'!j
Q
1't :i
l 100
D
I
.g
!'i
....
:-,l
.: 90
~
.S
i
BO
70
60 /LI
,
, ,, ,,
I
/
CHART FOR COMPUTING APPROXIMATE
BITUMEN RA TIO CABRI FOR DENSE
GRADED BITUMINOUS MIXTURES
t
~
• ,.:J~LY
I' I'/ ~
~Q ; 50 'l,!6
v ><
gs //"'?
~/'II~
140 _.,..,l,,,j
.i-"17:~ ►
~
::r'
§ 30
~
g· 8"> j
r/ ~ ~ ,I V
7, V //
;~
i:: "'I
"'s= vv
C':) ~ ~ [/ 1/V
a·! 15
~~ [/ V
I/
v v 8
c-:.l· p«~ ~ '.t-rf.i.':',1!1!".. ti
~
~-
a a.=
20
10
I/ ~-:,,~~~.~":.-~
I/[/ [/ I5
~:!:=t =~:.1~r:1,t;1 ~
w
::c: (Jq I/[/ ,_
lfnewllh1calo8•ollrcllo. e-
0
~ 0
l'I)
9 -
....
I/
) .
'M('_,,,
~p
~
-
,,. ~-
V..,.. -::;; ;.,
....,..,
_....._
10
9
8
Can 2.. OIYen aurfaoe area,
of aggrogala.
Flnd
1p.
;.~~s·
~f,:;: ~
.,,,,.
V ...... ::::::,,I.,' :,,:v V / .,,,,. 6,g
5
f:~::L"'tl:, loolKfdtil.hen horlzonlallr
60
!
~
.,,,,.
f:::C: ~;:::: t:-t:::: ~v vi:,,,,-
0:
Is
0Q ratio • lbs. of oll 1,er I00 lb.- 40
!
0 4 ~
8"> :::::~ [::: ~~ ~,.,,,.v- Vv
::;;-i;.,
ri :l,9f!l !1 ~~2~g '/.\'c~;~lyandt
0
,:l..
~--i.,,vv
0 "'I
~
t:::- t:= V t::J-6 V
3
:~-!~ ro~ct~·:-1t~~r.r•• 30
f.
~~
'"">
povlng osphalls. Fig. IV 6
t::, t::, _v V vv
fl l'I)
.... = 2 V v
l•J 1-J 8 2
20
!
~ 7 1.5 1.0 0.5 0.5
I
OJ AOJ 0.5 0.5
m2 112
~
~
I
"'ib'"
l
I O........., •St&rface Area, 0.204816
1.8 1,8 2.0 2.2 2.4 2.8 2.8 3.0 3J!
c 10
Sp, Gr, of Aggrogqfa
••Kg. al oil' per 45.36 kg, N
,:l..
~
CHART FOR CORRECTING BITUMEN
Jorj
I N
N
1'
n
REQUIREMENT DUE TO INCREASING VISCOSITY
t (I) 121
In Tables I and 2 of AASH O Sfceclflca-
lion M 226 and ASTM 8/eclficat on
"""'
po Llqllld Paving Grades CAR)
D 3381 ore used llnsteo of AR grades
In Tobie 3l, substitutions on Scale A may
D Asphalts
be made as follows• rl
o~n 2 .----:::i
8 ~ 8
0 4
~ 2~ ~
0
n =r
'"I') 0 A AC 2.5 for AR 1000
t, ~ e; (\I m !!! AC 5 for AR 2000
10 for AR 4000 t-13
;, el ... 3
Olis and Asphalts•
AC
AC 20 for AR 8000
~
=0
... !:!l'I
Arr~1e• according lo Absolute Viscosity
AC 40 for AR 16000
aq '"'
Sn l
~
4 0) S MC or RC
121 A • Asphalt Residue
12
:I:
"' 0 ::,
er ~
n '"' ILi 5 II
g-;
0 '"'
"' n
... .,
~ e.
C
lo .f
;g 8
;,.,
►
o· =
'"l')Cl(l
e~
~l,
8 10
l~
"'
'Cl
::,'
g, PROCEDURE
~
'O ... ::, 8
CJ)
7 9 i~ .,~
E-.o
[i
"' txi
,.. n
:61
~! 8
BY MEANS OF A STRAIGHTEDGE
CONNECT THE POINT ON SCALE AF
WHICH REPRESENTS THE GRADE 0
!/
8
:E-!
l!ID
t
F
ct
::i.
>
n= ... BITUMEN TO BE USED WITH THE
POINT ON SCALE B REPRESENTING
§
n
a ,.c~ ""
:§ 9
THE SURFACE AREA OF THE AGGRE-
GA TE. THROUGH THE POINT OF
7 ~2
g 0:: INTERSECTION ON LINE C PLACE A ...
C>
~ ij, 5 10
STRAIGHTEDGE CONNECTING WITH
THE PREVJOUSLY DETERMINED OIL B C1
C>
::C: f!l.
;a an RATIO VALUE ON SCALED, THE
INTERSECTION OF THE STRAIGHT
(IQ
.,::,::,
aa II EDGE WITH SCALE E REPRESENTS
THE BITUMEN RA TIO CORRECTED
FOR VISCOSITY OF THE BITUMEN.
~ B
5
p..
(')
~ t, 4 0
12 ::,
E "'
n i::
S- n m2 112
..Surface area,-• 0,204816b
a'
0
0 ... kg I ,:t.
i:i..o 0
::,
r
i
E-6. Correct for viscosity of paving grade asphalt cement. Using asphalt
cement viscosity, surface area of aggregate, oil ratio from E-5, and chart in
Figure 4-18, determine the viscosity corrected Bitumen Ratio.
E-7. Bitumen Ratio is lbs (kg) of oil per 100 lbs (kg) of aggregate.
Therefore, asphalt content (AC) by weight of total mix is calculated as fol-
lows:
0
cerce
nt AC
=-------
Bitumen Ratio
100 + Bitumen Ratio
X 100
F-3. Heat the aggregates and the asphalt cement to the required tempera-
ture, take the aggregate out of the oven, form a crater in the aggregates, and
pour in the required amount of asphalt cement. Note: prepare specimens
at asphalt contents corresponding to:
At least one set of specimens should have enough asphalt cement to flush
after compaction. Flushing occurs when asphalt cement is forced to the
outside resulting in a shiny surface.
228 Hot Mix Asphalt Materials, Mixture Design and Construction
F-6. Place the mold in a mold holder, place a steel shim under the edge of
the mold, tighten the positioning screw, and insert a paper disc to cover the
base plate.
F-7. Spread the preheated mix in the feeder trough, transfer halfof the mix
into a compaction mold and rod the mix 20 times in center of the mold
and 20 -times around the edge using a bullet nose steel rod. Add remainder
of the mix and rod as before.
F-8. Place the mold assembly into position in a mechanical compactor and
apply semi-compaction tamping blows as follows:
F-9. Remove the shim, release the mold tightening screw, increase the foot
pressure to 500 psi (3450 kPa), and apply 150 tamping blows. Observe-the
surface of each specimen and note each specimen that flushes. (Continue
adding asphalt cement until at least 1 specimen has a moderate or heavily
flushed surface.) Flushing can be ranked as follows:
~-------1-;-Slighra-if-surface-has-a-s-I-ighrsheen.
2. Moderate-if paper sticks to the surface but no surface distortion is
observed.
3. Heavy-if there is free asphalt cement puddled on the surface or if
surface distortion is observed.
Chapter 4-Hot Mix Asphalt Mixture. Design Methodology 229
F-10. Place the mold and specimen in an oven at 140°F (60°C) for 1½
hours; then apply a levelling load of 1000 psi (6900 kPa) at a rate of 0.25
inches/minute (6.4 mm/minute), releasing immediately once 1000 psi
(6900 kPa) is reached. Allow the specimen to cool in the mold.
G-2. Place the stabilometer on a compression machine and apply 1000 lbs
(4.45 kN) to the top of a dummy specimen.
G-3. Open the fluid valve and rotate the pump handle clockwise until the
pressure is 5 psi (34.5 kPa) (tap gauge gently to ensure that the needle valve
is operating properly). NOTE: If the pressure exceeds 5 psi (34.5 kPa),
reduce the pressure by turning counterclockwise to below 5 psi (34.5 kPa)
and then turning clockwise to achieve 5 psi (34.5 kPa). Adjust the dis-
placement dial gauge to zero.
G-4. Rotate the pump handle clockwise in a continuous motion until the
pressure increases from 5 to 100 psi (34.5 to 690 kPa). Read the dial indi-
cator and record the initial displacement (0.1 inch (2.5 mm) dial reading
= 1 turn displacement). ·
G-6. Repeat Steps G-2 through G-5 until the displacement is within tol-
erances.
H-3. Set the compression machine head speed at 0.05 inch/minute (1.3
mm/minute).
H-4. Place the stabilometer on a base and adjust the stabilometer base so
that the distance from the bottom of the upper tapered ring on the sta-
bilometer to the top of the base is 3.5 inches (88.9 mm).
H-5. Remove the mold and specimen from the oven and place over the
top of the stabilometer. Using the plunger, hand lever, and fulcrum, force
the specimen from the mold into the stabilometer. Be sure the specimen is
in straight and firmly seated. Place the follower on top of the specimen and
position the entire assembly in a compression machine for testing.
H-6. Using the displacement pump, raise the lateral pressure on the spec-
imen to 5 psi (34.5 kPa) initial pressure, tap the needle valve to ensure an
accurate reading, and then close the displacement pump valve.
H-7. Apply the test load at a head speed of 0.05 inch/minute (1.3
mm/minute). Record the test gauge readings at 500, 1000, and at 1000 lb
(2.22, 4.45, and at 4.45 kN) increments up to 6000 lbs (26.7 kN).
Immediately reduce the load to 1000 lbs (4.45 kN) as soon as 6000 lbs
(26. 7 kN) is reached.
H-8. Open the displacement pump valve and reset the pressure to 5 psi
- - - -·(ignore-the readmg om~he compression tester). Set ilie Ames dial gauge on ~----=c-------
the pump to zero using'the thumbscrew.
H-9. Turn the displacement pump handle clockwise rapidly (at 2 revolu-
tions/second) until the pressure test gauge reads 100 psi (690 kPa). Read
r
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 231
the Ames dial, record in 0.01 inch (0.25 mm) units (0.1 inch (2.5 mm)
equals 1 revolution or turn of the crank handle) as the displacement of the
specimen, D 2 •
H-10. Remove the compressive load, reduce the pressure on the test gauge
to zero, back the displacement pump past zero 3 turns counterclockw-ise,
remove the specimen from the stabilometer, and measure the height< in
inches.
22.2
where,
If the specimen height is between 2.4 and 2.59 inches (61.0 and 65.8 mm),
no height correction is necessary. Otherwise, correct S for specimen height
using the chart in Figure 4-19.
H-12. Repeat Steps H-5 through H-11 until all specimens have been
tested.
1-2. Calculate the percent air voids (or VTM) and density of the compact-
ed specimens as per Step E of the Marshall mix design method.
232 Hot Mix Asphalt Materials, Mixture Design and Construction
--g
0
CD
L
40
L
0 II
U 30
CD
:,
0
>
L
CD 20
"i
E
..2
:a
0
ci5 10
-0 - -- - - --
0 IO 20 30 40 50 60
Stabilometer Value Before Height Correction
Figure 4-19. Chart for Correcting Stabilometer Values to Effective
Specimen Height of 2.5 inches (64 mm)
Chapter 4--Hot Mix Asphalt Mixture Design Methodology 233
1. Write the design series asphalt contents in each box with the lowest
on the left.
2. Referring to observations made in Step F-9, mark out the asphalt
contents in which moderate to heave surface flushing was observed.
3. Write the 3 highest remaining asphalt contents in the second level
of blocks in the pyramid.
4. For these 3 asphalt contents, check the Hveem stability and mark
out any which do not meet minimum stability.
------~S-:-Seleet-the-2-highest-asph-alt-content"S-that-meet-minimum stability
and record in the third level of the pyramid.
6. Referring to the plots prepared in Step J-1, select the highest
asphalt content that has at least 4 percent air voids and enter at the
top of the pyramid. This is the optimum asphalt content.
234 Hot Mix Asphalt Materials, Mixture Design and Construction
J-4. The laboratory developed mixture design forms the basis for the ini-
tial job mix formula QMF). The initial JMF must often be adjusted early
in the construction project because of slight differences between the labo-
ratory supplied aggregates and those being used in the field. The construc-
tion experience during the first few days should be carefully reviewed and
the JMF modified to provide a mixture that can be adequately mixed,
placed, and compacted.
Background
The Hveem and Marshall methods of mix design have been used since
the 1940s and 1950s. These mix design methods have performed well for
many years but with more traffic and heavier loads it was determined in .
the early 1980s that an improved method of mix design was needed. A
method was needed that could be used to rationally design mixtures for
various traffic volumes, axle loads and environments. It was hoped that this
new mix design procedure would provide tests and models that could use
these test results to predict rutting, thermal cracking, and fatigue cracking.
In 1988, the Strategic Highway Research Program (SHRP) was initi-
ated with a primary goal of developing an improved mix design procedure.
(Z, 16, lZ ,18,) The SHRP program was completed in 1993. The effort
consisted of hundreds of researchers working on various parts of the
research program and it took a great deal of effort to bring all of the results
to a cohesive solution in five years. Even though the volumetric portion of
the mix design procedure is essentially complete, a lot of additional work
is on-going to verify portions of the procedure and to make modifications
where needed.
As directed the study primarily focused .on asphalt binder evaluation,
mix design, and mixture analysis. However, no part of the research effort
looked specifically at aggregates. Certainly, the mixture part of the study
included aggregates but there was no effort to evaluate aggregate separate
from the other mixture components.
Since aggregates are a very important part of the mixture the SHRP
----±•1esearc:hers-r:eali.zed-t:hat-·S0me-g1:1idanee-m-t1st-be-provided-:-'Fhe-g1:1:ida:neP------
that was eventually provided was based on a consensus approach by a
group of experts on aggregates. Recommendations were made for desired
aggregate properties for various traffic levels and for various depths below
the surface of the pavement.
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 235
Materials
The numbers shown in Table 4-2 are numbers that were developed
through concensus testing with a number of experts. These values are
believed to be acceptable for most applications of HMA. However, local
experience may indicate that the specifications should be different than
those values shown in Table 4-2. Local experience and materials should
always be considered when preparing project specifications.
The fine aggregate angularity is measured by determining the amount
of voids when the fine aggregate is poured into a cylinder as shown in
Figure 4-20. The fine aggregate is poured into the top end of the cylinder
and the amount of voids in the loose aggregate is determined. The higher
the amount of voids the more angular the aggregate. The test is often
referred to as the National Aggregate Association (NAA) flow test. The test
should be conducted in accordance with AASHTO TP 33 Method A "Test
Method for Uncompacted Void Content of Fine Aggregate (as Influenced
by Particle Shape, Surface Texture, & Grading)". The amount of uncom-
pacted voids in the cylinder can be calculated from the following equation:
I
uncompacte
TT. vot"ds -- V-VW/G,b
dTl'. X 100
~------where-----------
I Maximum Percentage of
I
TRAFFIC particles with length/thickness ·
I.
!
ESALs (millions) greater than 5
I
< 0.3 -
<1 -
<3 - - ···-~ ~-· -- 10
< 10 10
< 30 10
< 100 10
i
2 100 10
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 239
The applied loads must ultimately be carried by the aggregate. For sat-
isfactory performance an aggregate meeting the qualityrequirements dis-
-----,crrs~-e<::I-eadier with the proper grading must be used. The gradation of the
aggregate is important to ensure 1) the maximum aggregate size is not too
large or too small, 2) VMA requirements are met, and 3) a satisfactory
aggregate skeleton is obtained.
According to Superpave, the nominal maximum aggregate size is one
240 Hot Mix Asphalt Materials, Mixture Design and Construction
IO
N
Cl)
II-- ' ......
' 1·
r.q
N E
..- E
r.q cl
w
Cl)
s0
~
0\ a..
IO
-
II '<I"
~ ci
-
N
CZ)
IO
I'-;
'<I"
w
N
en
-~
w
~0 -
,• !---11-··!
(0
~
>
w
w
z N
(X)
"-:
..-
.,_~
0
0
' " " • , - - -.. MOMH-• (")
0 IO
..-
·•--• IO
r---
'-=!
0
0
0 0 0 0 0 0 0 0 0 0 0
0 Cl) GO r--- (0 IO '<I" (") N ..-
..-
DNISSVd .LN3::J1I3:d
sieve size larger than the ~ieve tC> __!e~ain.~(),re than 1_Q__pe~,:L.QLt~e
material. The maximum aggregate size is defined as one sieve larger than
the nominal maximum s:ze (Figure 4-21). If a 19 mm (3/4 inch) maxi-
mum aggregate size is specified, then 100 percent of the aggregate must
pass the 25 mm (1 inch) sieve size. At least 90-100 percent of the aggre-
gate must be finer than the nominal maximum aggregate size (19 mm or
3/4 inch). Less than 90% of the aggregate must pass the 12.5 mm (1/2
inch) sieve. Control points are also set on the 2.36 mm (No. 8) and the
0.075 mm (No. 200) sieve sizes. The aggregate gradation must be within
the control limits to meet the Superpave requirements.
/.
Another part of the grading specification is the restricted zone. This is
the most misunderstood portion of the grading requirements. The restrict-
ed zone is meant to be a guide to help ensure that too much natural sand
is not used in the__mixt.uE._and to help ensure that minimum V1v1A require-·
ments are met. However, there are aggregates that can pass through the
restricted zone that are well crushed and meet the minimum V1v1A require-
ments. Since the restricted zone is only a guide, aggregate blends that pass
through the restricted zone that do not use excessive amounts of rounded
aggregates and that meet the minimum V1v1A requirements should per-
form satisfactorily.
There is evidence to support the belief that mixtures closer to the low
end of the control limits (those going underneath the restricted zone) pro-
vide a better stone structure. Most Superpave designs are performed below
the restricted zone to provide a mixture that is believed to be more resis-
tant to rutting. Typical specifications for 9.5 mm nominal size aggregates
to 37.5 mm nominal size are shown in Tables (4-6 thru 4-10).
Ni (NJ0.45
Nm (Nd)l.lO
Compaction of Samples
All samples for mix design and. volumetric determination are required
to be aged at 135°C (275°F) in an oven for 2 hours after mixing and prior
to compaction. This aging is needed to allow absorption of the asphalt
cement into the aggregate pores prior to compaction and cooling. This
absorption produces a mixture that is more representative of the mix in the
field. The sample is spread in a thin layer for the aging process. If the mix
is not allowed to age a lower optimum asphalt content will be selected
which could result in some loss in durability during the life of the pave-
ment.
During the design process all samples are compacted to Nm and the
voids are determined by machine calculations at Ni, Nd and Nm. The sam-
ple height is measured and recorded during the compaction process which
allows determination of mixture volume. 'Fhe sample weight is input ·into
the computer and the sample volume is calculated as it is compacted by
knowing the diameter (150 mm or 6 inches) and the sample height. The
estimated bulk density can be determined from these values. This estimat-
ed density is slightly incorrect, however, due to the surface voids on the
top, bottom, and sides of each specimen being considered as part of the
specimen volume. After each sample has been compacted to Nm it is
removed from the compactor and the actual bulk density is determined by
weighing in air and water. The actual density at Nm is compared to the cal-
culated density at Nm to get a correction factor. The calculated densities at
Ni and Nd are corrected based on this correction factor. For example,
the correction factor (CF) is calculated on samples compacted to Nm
gyrations as:
Actual density
CF=
Calculated density
After the aggregates and asphalt binder have been determined to meet
the quality requirements and after the aggregate proportioning necessary to
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 247
92 ~ 4.S%AC
-6---- 5.0 %AC
-6- 5.S¾AC
84
IO 100 1000
Number of Gyrations
After all of the data has been plotted as shown in Figure 4-23, pick the
asphalt content that will provide 4 percent voids at Nd. This is done by
drawing a vertical line at Nd and a horizontal line at 96 percentTMD. The
point at which these lines cross is the optimum asphalt content. This will
likely require some interpolation between the actual data points. It is satis-
factory to interpolate but one should not extrapolate. The asphalt content
selected to produce 96 percent ofTMD is the optimum asphalt content.
Now at Ni draw a vertical line up to the optimum asphalt content and then
horizontally over to determine percent of TMD. The percent of TMD
should be less than 89 percent. Now do the same thing at Nm. The num-
ber selected should be less than 98 percent. The requirements for Ni and
Nm are met for this example (Figure 4-23). Remember that the TMD is
different for different asphalt contents. \X&en the TMD is known at one
asphalt content, it can be calculated at, another asphalt content by deter-
mining the effective specific gravity of the aggregate and by making a foy
calculations as discussed elsewhere in this chapter. .
! If the air voids. are 4 percent at ~d, _greater ~an 11 percent at Ni and
I greater than 2 percent at Nm, then this is the optimum asphalt content. If
) the mixture fails to meet Ni or Nm then the aggregate grading or possibly
(' aggregate source will need to be changed. Two other criteria that must be
) satisfied are VMA and voids filled with asphalt (VFA). The VMA criteria
\ are shown in Table 4-12 and the VFA criteria are shown in Table 4-13.
If the mix design meets all of the requirements except VFA, the asphalt
content may be able to be adjusted slightly to keep the air voids close to 4
percent and to bring the VFA within limits. The VFA is a somewhat redun-.
dant term since it is a function of air voids and VMA. From a practical
point of view, the volumetric quality of a HMA can be controlled by air
voids and VMA.
Two items need to be mentioned about Superpave mix designs.
Individuals that have experience designing mixtures with 100 mm (4 inch)
samples will find that much more material is needed for a 150 mm (6 inch)
design. It is estimated that approximately 4 times the material is needed for
a 150 mm Superpave design as that for a 100 mm (4 inch) Marshall or
Hveem design. Since the samples must be aged in an overi, one will also
realize that a substantial amount of oven space is needed. Depending on
the number of samples to prepare and compact, this may require that more
than one oven be available during the mix design process.
The Superpave mix design system is certainly a step in the right direc-
tion to improve the quality of HMA. This is a new process that will see
substantial changes within the next few years. Changes are expected with
any new technology and the Superpave system will almost certainly under-
go some modifications during the first few years of its existence.
OTHER CONSIDERATIONS
There are other considerations that should be discussed relative to var-
ious features of mixture design, including moisture susceptibility, recycled
mixture design, and skid resistance for surface mixtures. Each of these con-
siderations is discussed briefly below.
Moisture Susceptibility
Hot Recycling
More and more mixtures being produced around the USA contain
some amount of reclaimed asphalt pavement (RAP) simply because this
material usually contains valuable aggregates and asphalt cement that can
be reused. Field Investigations (21) have shown no significant difference
between the recycled and virgin HMA pavements in terms of in-situ mate-
rial properties and pavement performance. If RAP material is not used in
a recycled mix, it is typically used.on shoulders, as bases, or is disposed of
in landfills. None of these uses take advantage of the monetary value of this
product. Additionally, with the growing use of milling machines to remove
unwanted pavement layers and badly distressed materials, there is a grow-
ing supply of RAP which in most states belongs to the contractor. The con-
tractor wants to use this material in the most cost-effective way and that is
in recycled HMA mixtures. However, any mixture containing RAP mate-
rials should be required to meet the same specifications as a virgin mixture.
Therefore, the contractor must carefully consider the quality and unifor-
mity of the RAP stockpiles to ensure that the HMA produced meets the
project specifications. Additionally, the HMA facility must be modified to
handle RAP materials and often special stockpiling procedures must be fol-
lowed to ensure that the material can be easily handled during production.
In order to incorporate a RAP material in a mix design, the gradation
of the RAP and the properties of the asphalt cement must be determined.
Reference 14 suggests that three samples be taken from the RAP stockpile
and an asphalt cement extraction process be performed on each sample.
The extraction process involves dissolving the asphalt cement off the aggre-
- - . gates·.-Once performed;-rhe-aggregate-gra:ctation andasphaltcontentcanb~e------
determined for each of the three samples. The RAP stockpile gradation is
the average of the three sample gradations, and the RAP asphalt content is
the average of the three sample asphalt contents. The sqlvent containing
the dissolved asphalt cement should be collected and an asphalt cement
Chapter 4--Hot Mix Asphalt Mixture Design Methodol?gy 251
Once the grade of the virgin asphalt to be added to the mixture has
been determined, the amount of RAP and virgin asphalt can be deter-
mined for each asphalt content (total) to be used in the mixture design.
The following formulas are given in Reference 14 to calculate these quan-
tities:
RAP - ( B X C ) (
100 - D
WT)
100
where,
RAP = weight of reclaimed pavement, g;
B =_ percent of all aggregate in total mix;
C = percent reclaimed aggregate in the aggregate blend;
D = percent asphalt cement in RAP;
Wr = total weight of test specimen, g;
Av = weight of virgin asphalt cement, g; and
A= percent asphalt cement in total mix.
lar tires for the life of the project after construction is complete. These sur-
faces must not only be stable under acceleration, deceleration, and vertical
loads, but they must also provide sufficient resistance to sliding to permit
normal turning and braking movements to occur. This means that it is the
responsibility of the paving engineer to design a surface mixture so that the
necessary tire-pavement friction is attained even under the variable envi-
ronmental and vehicle operating conditions occurring in the field. The mix
designer must choose aggregates that will provide adequate frictional resis-
tance for all surface mixtures and ensure that the asphalt content of the mix
is not so high that bleeding occurs with subsequent loss in frictional re-
sistance.
Four aggregate characteristics should be evaluated in order to select a
material that provides desirable frictional resistance: texture, shape, size,
and resistance to wear and polish. Selecting the source of the aggregates
largely determines such factors as texture, mineralogy, and chemical com-
position. Factors like particle shape and size are affected by mineralogy but
are also significantly affected by the type of crushing process selected for
manufacture of the aggregate. Resistance to wear and polish is affected by
all the above plus the hardness of the various mineral constituents of the
rock, the size of the individual minerals in the rock, and various traffic fac-
tors.
While it is necessary for the designer to consider the texture charac-
teristics of the aggregates available for use in a mixture design, some agen-
cies do not permit the use of carbonate stone such as limestone which gen-
erally tends to polish under traffic. The quality of limestone aggregate
varies considerably; hence, some limestone aggregates do not tend to pol-
ish while others do.
Numerous states and the National Stone Association have shown that
it is possible through proper combinations of aggregates of different wear
rates to use most sources of limestone aggregates successfully in surface
mixes. Laboratory studies indicate that blending aggregates with different
polishing rates is effective because the frictional resistance of the blend is
proportional to the weighted average based on the percentage of the coarse
aggregate in the blend. However, blending requires careful testing and eval-
uations during preliminary studies leading to mix design. Blending is cer-
tainly worth considering because in the USA there is an enormous amount
of aggregates that tend to polish. Therefore, considerable effort should be
~-----x-pe-aa-©a-i-ri.-d~t~r-mining-h-ow-much-of-these--aggregates can be used in
combination with other aggregates to produce surfaces with adequate fric-
tional resistance. Consideration should also be made for tailoring the
blends of materials to meet the demands of traffic both in terms of volume
and speed of traffic.
254 Hot MixAsphalt Materials, Mixture Design and Construction
Sieve Size
(1/2 Inch 3/8 Inch No.4 No.8 No.16 No. 30 No. 50 No:100 No.200
(12.5mm) (9.5mm) (4.75mm) (2.36mm) (1.18mm) (0.600mm) (0.300mm) (0.150mm) (0.075mm)
Trial No. 1
No.1
(25%) 25 22 1 0 0 0 0 . 0 0.0
No.2
(25%) 25 22 14 2 2 1 1 1 0.6
No.3
(25%) 25 25 25 20 18 12 6 4 2.5
No.4 .,,.,..
(25%) 25 25 25 25 24 22 17 4 0.6
Total
100 97 65 47 44 35 24 9 3.6
Specs 95·.
100 100 47 - 77 · 52- 70 46 - 63 37- 57 24 - 39 8 - 25 2-8
Trial No. 2
No.1
(25%) 25 22 1 0 0 0 0 0 0.0
No.2
(15%) 15 15 9 2 1 1 1 1 03
No.3
(35%) 35 35 35 28 25 16 9 5 3.5
No.4
(25%) 25 25 25 25 24 22 17 4 0.6
Total
100 97 70 55 50 39 27 10 4.4
Specs 95 -
100 100 47- 77 52- 70 46 - 63 37 - 57 24 - 39 8 - 25 2-8
The bulk specific gravity of the sample Gmb is then determined by:
~
Gmb= - - - - -
WssD - Wsub
The volume of the sample in ml is equal to W 550-W,ub when weighed in
grams. The bulk density (y) of the sample in pcf is
Y= Gmb X Ow
~
()
::,'
-t
Welghtln lbeordl-
Sample A,phaO Weight la Water Welghtlll Bulk cal Mu. Vokk M...wed Coneded er .g
No. Content Ah-(D,Y) (SSD) Air(SSD) Volume Density Density AlrVoida VMA FIJlcd Slabllily st.,bUky flow
0 !:f
518.3 140.6
ta
1 5.0 1167.8 650.7 1169.0 2400 2400 11
Average
1169.4 650.9 1171.0 520.1 140.3
2620
13
12 t f
!ti
1 6.0 1170.4 656.7 1171.0 514.3 142.0 2620 2620 13 ~
,::i.. (l)
2 1181.l 664.7 1181.9 517.2 142.5 2710 2710 13
~ ~-
3
Average
1187.3 670.9 1189.0 518.l 143.0
2770
12
13
~- ::;
~
1 65 1174.2 661.6 1174.7 513.l 142.8 2BOO 2800 12 a° I
"0 0.
%
2
3
1185.3
1182.3
661.1
667.7
U86.0
1182.9
518.J
5152
142.7
1432
2730
2900
2730
2900
13
14 a.
0
0{I)
°"'<
Average 142.9 149.9 4.7 18.3 74.3 2810 13
Avenge
1192.8 675.7 1193.3 517.6 143.8
2780
15
14
5i:r
1 15 1111.9 663.3 1182.3 519.0 142.1 2650 2650 16 ~
2380 16
~
2 1173.0 6602 1173.5 513.3 142.6 2380
The following calculations are macle for sample 1 at 5.0 percent asphalt
content.
1167.8
Gmb= - - - - - - = 2.253
1169.0 - 650.7
The voids in total mix (VIM) are determined for each sample by com-
paring the average bulk density for each asphalt content to the theoretical
maximum density (TMD) for that asphalt content. The most common
method to determine the TMD is the Rice-method as specified in ASTM
D2041. The VTM is determined by the following equation.
or
For this problem, the average of the three samples at an asphalt content of
5.0 percent had a bulk density of 2.252 g/ml and the TMD for this same
mixture was determined to be 2.454 g/ml. The VTM is then
2.252 )
V.MA. = (1 - - - x 100
2.454
= 8.2%
2.252
= 100 1 - _ (1 _
_....._ - 0.05)]
_ ~
[ 2.620
= 18.3
The voids filled with asphalt (VFA) are determined for the 5 percent
asphalt mixture as shown below.
= 55.2 perce_nt
After the samples were weighed in air and water and all calculations
made, the samples were then tested for stability and flow. The load to fail-
ure (pounds) when tested in the standard Marshall equipment is the
Marshall stability. The flow is the deformation in 0.01 inches (0.25 mm)
from the start of loading to the point where the load begins to decrease.
The measured stability is the actual load measured, and the corrected sta-
bility includes a correction for sample volume. For this design all volumes
were in the required range (ASTM D1559); hence, no correction was
needed. The stability and flow results are shown in Table 4-16.
After all the data was collected, plots were developed to show the rela-
tionship between the various properties and asphalt content as shown in
Figure 4-24. Properties that are usually plotted include: unit weight,
Marshall stability, flow, voids in total mix, voids filled with asphalt cement,
and voids in mineral aggregate.
Now to select the optimum asphalt content the project specifications
must be known. For this mixture design the specifications are shown
below:
There are a number of methods used to select the optimum asphalt con-
tent once the data is developed. Each agency that is involved with pave-
ment construction has 1ts own method for selecting optimum.
I
Ii
, I
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 261 ! !
i !
'I
I
8 •
b
a..
146 (/'J
7
'" "
~ 145 96
0 l
:z: >
~144 a: 5
"1
3
1-143 . :.t4 I
~
I,' ~ r,,_
~-142 17 3
141 Ii' 2
5.0 5.5, 6.0 6.5 7.0 7.5 5.0 5.5 6.0 6.5 7.0 7.5
% AC BY WGT. OF MIX % AC BY WGT. OF MIX
~
..J
3050
90
>-
!::
2950 • II
82
~ 2850 ,_
..... ~74
~
(/'J
2750 , > I'
V
15 f]
I ""'
I,
~13 19.6
" <(
812 ~ 19.2
-
T II
3
--.;, >
~ 18.8 ,
I
310
LI..
18.4 ,....
i--..., v
9 18.0
5.0 5.5 6.0 6.5 7.0 7.5 5.0 5.5 6.0 6.5 7.0 7.5
% AC BY WGT. OF MIX % AC BY WGT. qF MIX
·O
60
>-
I-
d
al
5o
<( ...........
I- V .......... ~
U) /
::::t / ....... r--..
w •
~ 40
:::c
0
'
w
I-
u
~ 30
a::
0
u
20
4.0 5.0 6.0 7.0
Binder Evaluation J
Determine that the binder to be used meets the requirements for PG
64...:.22. For this example it is assumed that these tests meet the require-
ments discussed in Chapter 2.
Aggregate Evaluation
Property Specification
1,,
Test Results
Aggregate Aggregate Aggregate Aggregate
Pro12e!];}'. 1 2 4
Coarse Aggregate Angularity 100/100 100/100
I. Fine Aggregate Angularity 48 47 42
Flat & Elongated Coarse Aggregate Particles 3 5
Fine Aggregate Sand Equivalen~ 52 58
After reviewing the test results for each aggregate shown in Table 4-18
and the requirements shown in Table 4-17, it is seen that all individual
aggregates meet the requirements for coarse aggregate angularity, flat and
elongated count, and sand equivalent test. The only question is whether or
... -~-- not the fine aggregate-meets-the-speGifiecl-requirement-for-fi.ne-aggregat·----
11 angularity. Aggregate 3 meets the minimum requirement of 45 while
I I: adggregathte 4 dhoesbnot meet the requirement. If aggregate 4 is used in thedmfcix
· !i esign en t e lend of aggregate 4 and aggregate 3 must be evaluate or
fine aggregate angularity.
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 265
Trial blend No. 1 goes above the restricted zone while blends No. 2
and No. 3 go below the restricted zone. There is not a large difference
----between-the-gradati0n~0f-aggregate-fer-0lend-2--and blend 3. With the
aggregates available, this is about the limit of variation that can be done
and still stay within the control limits and below the restricted zone. Even
this small variation may result in a significantly different aggregate struc-
.ture. A plot of the three trial gradations is shown in Figure 4-26.
266 Hot Mix Asphalt Materials, Mixture Design and Construction
The fine aggregates are blended for each of the three trial gradations to
evaluate the fine aggregate angularity. The values are determined to be 46
for gradation 1 and 47 for gradations 2 and 3.
Trial Blends
After estimating the optimum asphalt content for these blends, speci-
mens are prepared to evaluate the aggregate structure. In this case the opti-
mum asphalt content was estimated from past experience with similar
materials. The optimum asphalt content for all three aggregate blends was
estimated to be 5.0 percent.
The number of gyrations required for this design is 128 as shown in
Table 4-11 (based on 15,000,000 ESALs and 41-42°C average design high
air temperature). N and Nm are shown to be 9 and 208 respectively. So, all
mixtures are compacted up to 208 gyrations. The densities at Nd and Ni
are determined by calculating the volume from height measurements dur-
ing compaction and making corrections based on a comparison of mea-
sured density and actual density at Nm.
100.00
~ Blend 1
-0- Blend2
80.00 --A- Blend 3
bO
.E 60.00
"'
"'
"'
i:i...
1:l
Cl)
~
Cl)
i:i... 40.00
,,]
I
20.00
0.00 ........---::~,...,..-,,.....,..-----------------
0.075 0.3 2.36 9.5 12.5
Sieve Size (mm) Raised to 0.45 Power
Figure 4-26. Plot of three trial blend gradations
Chapter 4-Hot Mix Asphalt Mixture Design Methodology 267
Sample
Sample No. of Weight Ht Volume Uncorrected Uncorrected Corrected
Number Gyrations (grams) {mm) ctn3 Density TMD Density Density
Trial Blend
No.1
Sample 1 9 4693.6 123.6 2184.3 2.149 2.453 87.6 88.4
128 114.3 2020.5 2.323 94.7 95.6
208 JJ2.2 1982.8 2.367 96.5 97.4
Sample2 9 4684.0 123.3 2179.7 2.149 2.453 87.6 88.5
128 114.5 2022.7 2.316 94.4 95.4
208 112.2 1982.8 2.362 96.3 97.3
Trial Blend
No.2
Sample 1 9 4750.5 123.3 2262.4 2.100 2.447 85.8 86.5
128 114.5 2036.9 2.322 95.3 96.1
208 112.2 2005.3 2.369 96.8 97.6
Samele2 9 4744.3 128.0 2254.7 2.104 2.447 86.0 86.9
128 115.3 2034.7 2.332 95.3 96.3
208 113.5 2005.2 2.366 96.7 97.7
Trial Blend
No.3
Samele 1 9 4729.8 127.8 2259.1 2.094 2.449 85.5 86.6
128 115.4 2039.6 2.319 94.7 95.9
208 JJ3.7 2009.9 2.353 96.J 97.3
Samele 2 9 4721.1 127.6 2254.5 2.094 2.449 85.5 86.5
128 115.3 2037.6 2.317 94.6 95.7
208 113.7 2010.0 2.349 95.9 97.0
The trial blend samples are compacted and the volumetric results are
shown in Table 4-20. The test results are collected ori intervals of 10 gyra-
tions up to 100 and then by intervals of 25 through Nd and Nm. The
results are presented here for 9, 128, and 208 gyrations only since these
gyrations are used to evaluate the quality of the HMA. The height of these
samples is determined as the samples are compacted. The volume is then
calculated from the diameter (150 mm) and the measured height. The
·uncorrected density is then calculated from the weight of the sample and
the measured volume. The actual density is determined for each specimen
compacted to Nm. A correction factor is determined from the actual den-
sity at Nm and the measured density at Nm. This factor is applied to the
uncorrected density at other gyration levels to obtain the corrected densi-
ty.
:A-summary-of-the-compaction-data-for-the-trial-blends is shown in
Table 4-21. This data needs to be converted to that expected for optimum
asphalt content (that asphalt content that provides 4 percent air voids) to
fully evaluate the expected mix quality at optimum asphalt content. First
correct the asphalt content used (5.0 percent) to the estimated optimum.
268 Hot Mix Asphalt Materials, Mixture Design and Construction
where,
Aggregate Voids in
Bulk Total Voids o/oGmm o/oGmm
Identification Gravity %AC Mix VMA Filled at Ni at Nm
Blend No.I 2.631 5.0 4.5 15.4 70.8 88.4 97.4
Blend No. 2 2.625 5.0 3.8 14.8 74.3 86.7 97.6
Blend No. 3 2.622 5.0 4.2 15.0 72.0 86.6 97.2
So the estimated optimum asphalt contents for trial blends 1, 2, and 3 are
5.2, 4.9, and 5.1 respectively. Now estimate VMA at the estimated opti-
mum asphalt content using the following equation:
where:
The estimated VMA at optimum asphalt content is 15.3, 14.8, and 15.0
;-- respectively-forolenas~1-;-2-;-ana~-:-Esfimate the percent ofvoiasfilTe-ar-w-It_,_h_ _ _ __
asphalt at optimum asphalt content using the equation below:
For the three trial blends then the VFA values for blends 1, 2, and 3 are
73.8%, 73.0%, and 73.3% respectively. Now estimate the percent of Gnim
at Ni for optimum asphalt content. The equation to be used is:
Estimated percent of Gmm @Nm= percent of Gmm for trial @Nm - (4.0- VJ
Substituting.into the equation gives 97.9, 97.4, and 97.4 for blends 1, 2,
and 3 respectively.
Now summarize the estimated volumetric and compaction data for the
optimum asphalt content as shown in Table 4-22.
After reviewing the ·data, it appears that blend No. 3 is the best choice.
Blend 1 barely meets the Ni and Nm requirements for percentage of Gmm·
Blend 2 does not meet the VMA requirements for 9.5 mm nominal size
aggregate.
~ 4.6%AC
92.00
J
~
88.00
84.00
voids, number of gyrations, and asphalt content. It can be seen from this
plot that the optimum asphalt content is 5.2 percent and that the require-
ments for Ni and Nm are met.
Now verify that the compaction and volumetric properties ai:e satis-
factory at 5.2 percent asphalt content. The properties include percent of
Gmm at Ni, percent ofGmm at Nm, VMA, and voids filled. This can be done
best by plotting each of these values versus asphalt content as shown in
Figure 4-28. The data is calculated as done earlier for the trial blends. The
data is summarized in Table 4-24.
6.00 88.00
ii
2.00 ~ 85.00
""
1.00 84.00
4.40 4.80 5.20 5.60 6.00 6.40 4.40 4.80 5.20 5.60 6.00 6.40
Asphalt Content, % Asphalt Content, %
15.80 100.00
~ E 98.00
i 15.40
....t.5
> 0
97.00
15.20
J 96.00
15.00 95.00
4.40 4.80 5.20 5.60 6.00 6.40 4.40 4.80 5.20 5.60 6.00 6.40
Asphalt Content, % Asphalt Content, %
90.00 \
80.00
~
t 70.00
-- - -----i;o:oo-
4.40 4.80 5.20 5.60 6.00 6.40
Asphalt Content, %
An inspection of Table 4-24 and Figure 4-28 shows that the mix at
5.2 percent asphalt content should have 4.0 percent air voids and should
meet the requirements for V1v1.A, voids filled, percent Gmm at Ni and per-
cent Gmm at Nm. Interpolating between the values indicates that at opti-
mum asphalt content the VMA is 15.2, the voids filled is 74.2, the percent
Gmm at Ni is 86.4 and the percent Gmm at Nm is 97.6. All of these values
meet the specified requirements. Therefore, the optimum asphalt content
is selected to be 5.2 percent.
After the mix design is completed water susceptibility tests should be
conducted. If the mixture fails to meet the specified requirements for water
susceptibility the mix design will have to be adjusted.
REFERENCES
INTRODUCTION
Although the need for materials characterization has been around for
many years, and a great deal of work has been done, there ~!!!e agree-
ment throughout the industry with regard to the types of tests and test pro-
cedures required to predict mixture performance Tne approa2;. to solving
the problem has, at best, '""5een piecemeal involving many different
researchers, from many different agencies working relatively independent-
ly yet striving for an optimal solution to a singular problem without con-
sideration for all of the ramifications. For the past 50 years or more, pave-
ment design agencies have always pressed for an immediate answer to their
needs and problems. Long term, well thought out, sequential efforts have
usually been rejected because of the time arid expense involved. The effect
... ---has been that structural-and-mixture-design--procedures-have-suffered-a:nA------
an integrated design and analysis system has not evolved.
The Strategic Highway Research Program (SHRP) was undertaken in
an attempt to remedy this situation. SHRP involved extensive research to
develop performance-based asphalt cement specifications, fundamental
Chapter 5-Characterization of Asphalt Mixtures 277
tests, and an improved structural design and analysis system. The SHRP
work was completed in 1993. Additional' work is being performed to
develop a model that will predict ~an.ce. This effort should be com-
pleted prior to tlieyear 2000.
-Materfal-Variability-and-Projecr-Siz~~--
Project Size. Project size and cost are also important in establishing a
testing program to characterize a HMA mixture. As the size of a project
increases, the variability of material properties may also increase since
material properties may change with time and as equipment wears. When
·-·---ffro]ecf s1ze of cost-increases, the JUSt1fication for a more extensive testing
program increases. Thus, the extent and nature of the testing program ulti-
mately relates to the variability of the material and test method, the cost of
the project, and the consequences of failure. More test results allow the
engineer to more accurately measure the true quality of the mixture but
Chapter 5-Characterizatio.n o.f Asphalt Mixtures 279
excessive testing can be expensive and may not be justified when the test-
ing cost exceeds some established percentag~of the overall project cost.
en
'iii Q.
Q.
0
0 0
0 t\l
t\l II 'iii
11 "iij Q.
.c.
.c.
Q.
-o
-o
g'(.\J
g'I")
G)
G) _:II
.._ II 0
(X)
C/)C/) I 0
(X)
U) Cl) G) I
(\J C\J
O'I
G)
O'I
...
C
...
C
G)
I
I
G) >
> 0 <( I 0
<( I
,;;t" V
(\J I C\J
I
0 0
0 I 0
(\J ·u; I C\J en
a. a.
\
OU) \ OU)
(0 U) \ U)
U)
-w \ -w
er. \ er.
1- \ I-
0 U) OU)
(\J C\J
C C
_g _g
0 0 0
(X) :§ (X)
>
G)
>
G)
C C
0
...
-c
C 0 0
-c
~
V
-c m V
-0
C
n C
C II 0
en
en-
G) Cl)
'en Cl)
en- 'en
... Q. G> en n
cno
-ct\l
II
Cl)
... Q.
'en 0 Cl)
-ct\l
~n ~ II
'5CI) '5CI)
.5 .5
G)
G)
O'I
C O'I
0
G)
G)
~ ~
Figure 5-1. Graphical Presentation of Failure for Two Conditions with
the Same Mean Value But Different Variability (ll
,,
':1
Another criterion relates to the ability of the test to measure the fun-
damental properties previously mentioned. Empirical test results are only
of value to those empirical design procedures from which they were devel-
oped. Attempts at using empirical tests to estimate the fundamental prop-
erties through correlations are poor at best and should be used with cau-
tion only when better information cannot be obtained. Most correlations
between actual performance and test properties have shown that the phys-
ical properties of the asphalt mixture have produced the best correlation
with performance. Substantial research is still needed to develop practical
fundamental tests that can be used in mix design and quality control test-
ing to evaluate mixture performance at a high level of reliability.
Ease of Testing
CURRENT TESTS
Physical Properties
Air Voids. The physical property most often correlated .to per-
formance is air voids in the mixture (see Chapter 4 for calculation of air
voids). Re~shown that the amount of air voids in a HMA mix-
ture affects the ~ t y -and durability. When the in-place air voids
decrease to less than 3 percent, rutting of the asphalt mixture is likely to
occur (2, 1, ~ due to plastic fl.ow. However, when the air voids are above
approximately 8 percent, the mix is permeable to air and water, and the
rate of oxidation
__, ...__,.- . ~-
of the asphalt binder is significantly
-·-··---···-··~··incre;:i.sed
....
(i, g, 2)
resulting in premature cradang and/or ravefling. The rate of oxidation is
a function of the air voids as vidl as th~--aiph.tlt film thickness. The labo-
ratory compaction effort (50-75 blows/side for Marshall Method) is select-
ed to provide a density in the HMA that provides air voids approximately
equal to that expected in-place after 2-3 years of traffic. Air voids in lab-
oratory compacted samples then provide an estimate of the minimum air
voids expected after traffic.
Voids Filled with Asphalt Cement. The voids filled with asphalt
cement are inyersely related to the air voids. As the percentage of air voids
approaches _{ero:the-p~rceritage ofvoias filled with asphalt cement approach-
es I 00. The asphalt mixture is initially constructed to some percentage of
voids filled (usually 50-70 percent) with asphalt cement. This perce'ntage-of
voids filled increases as the asphalt mixture continues to densify under traf-
fic. When the voids filled exceed approximately 80%, to ~5%,. the asphalt
mbcture typically becomes unstable and rutting is fik:elyt~ occur.
,__ .,--···•,-.v•·•·=··- ,p•--..,. . --- • ·- -~ •
Mechanical Tests
The two most common tests used to evaluate or design HMA mixtures
are the MarJ~-~!.~!~£ig t:y test and the Hveem ~!"~2H}]:'.;..,,,test. Both of these
0
tests and mix design procedures have served the industry well and will con-
tinue to be used until better test methods are developed and accepted._
Since both of these tests are empirical, and not highly correlated to perfor-
mance, improved test methods are needed.
Resilient-Modulus-(-Repeated-b0ad-Indirect-Tensile}fMR7:-B-e.:li:nse-o-f
its simplicity and applicability to test field cores, the MR test has become
popular with many laboratories. This is the most common method of
measuring stiffness modulus for HMA. A test setup somewhat similar to
that for the indirect tensile test is used to measure MR. The primary dif-
Chapter 5-Characterization of Asphalt Mixtures 287
ference is that the test equipment must be able to apply repeated loads.
The procedures for conducting this test are described in ASTM D4123
and the data is typically plotted as shown in Figure 5-2. The equations for
computing the tensile str,ength and strain to failure for the indirect tensile
test can be used for computing stress and strain in the resilient modulus
test. The applied stress is calculated the same way as that for tensile
strength; however, the specimen is not loaded to failure when measuring
resilient modulus. For the resilient modulus test, the specimen is normal-
ly loaded to a stress level between 5 and 20 percent of indirect tensile
strength. Therefore, prior to conducting the resilient. modulus test, the
tensile strength must be measured or estimated. The load is typically
applied for 0.1 seconds along with a rest period (no load) of 0.9 seconds.
Hence, the HMA sample receives one load cycle per second. Other com-
binations of loading times and rest periods have also been used.
The equations that have been derived to estimate Poisson's ratio differ .
slightly for 4-inch and 6-inch samples. The two equations are shown
below:
µ =
H - 0.27
3.59 V for 4 inch (IO 1.6 mm) samples
C
a b
(\ '\ (', r
I I \ I '
I \ / l
Time
(al Load-Time Pulse
a = load duration
b = recovery time
c = cycle time
where,
µ = Poisson's ratio;
H = horizontal deformation, inches; and
V = vertical deformation, inches.
where,
MR = resilient modulus, psi;
P = applied load, pounds;
H = horizontal deformation, inches;
t = sample thickness, inches; and
µ = Poisson's ratio.
Assuming that Poisson's ratio is 0.35, the above equation can be further
reduced to:
M R-_ 0.62P
Ht
The resilient modulus test is becoming more routine within many labora-
tories because of the emphasis being placed on the measurement of funda-
mental engineering properties. Thus far, there is no good correlation
between MR and rutting, but MR at low temperatures is somewhat related
to cracking. It has been shown that stiffer mixtures .at low temperatures
tend to crack earlier than more flexible mixtures.
Remember that
1
MR-_ 0.62P
Ht
Chapter 5-Characterization of Asphalt Mixtures 289
substituting,
M _ 0.62(100)
R - 0.00015(2.5)
solving,
MR= 165,000 psi (1138.5 MPa)
where,
E = dynamic modulus;
cr = applied stress; and
- £ = measured vertical strain.
Load Cell
Loading Head
Specimen
Strain Gages
--- --- - - - - - - - - - - - - - - H - - - - - - - - - - - - - - - - -
Poisson's ratio can also be determined from the test set-up by measuring
the horizontal deformation during the test. Attempts to measure Poisson's
ratio of HMA mixtures have usually been unsuccessful.
The primary purpose for measuring the dynamic modulus is to deter-
mine the stress-strain relationships in. a pavement structure under an
--
appliedJoad. Because the dynamic modulus~ 1s not a measure of strength,
a~alt mixture with a higher dynamic modulus does not necessarily
have higher strength. A higher dynamic modulus indicates that a given
applied stress results in lower strai1:;jp. the mixture.
This test h~ limited use because of the time involved, complex equip-
ment required to conduct the test, and height requirement of test speci-
mens. This test cannot be used to test cores taken from pavements, since
the pavement layers are normally no more than two inches thick and the
cores must be at least 8 inches (203.2 mm) tall for a 4-inch diameter spec-
imen. In some cases, cores have been stacked on top of each other to pro-
vide the required 8 .inches (203.2 mm), but this practice is questionable.
LOAD
I\ ,,
I I I
I '
I
I I
---
J I \
I I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I
O,l sec.
STRAIN
,_ I"'-- -- -- __ -- _, _ _ ,_ -- - ---
,
i,.-i- 1,i-
· - - - - - - - - ·--
,.,i- 1 ......
I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
Figure 5-4. Recorded Trace from a Dynamic Complex Modulus Test (1)
292 Hot Mix Asphalt Materials, Mixture Design and Construction
KEY
I. Reaction Clamp
2. Restrainer
3, End Plates
4. Specimen
5. Loading Rod
6. Stop Nuts
7. Load Bar
8. Piston Rod
9. Thompson Ball Bushing
10. LVDT Holder
I I. LVDT
The stress and strain at the outer fibers, and the stiffness modulus after
about 200 load applications, are calculated using basic relationships for
stresses and strains in beams. Simplifying this information provides the
following equations:
E = Pa(3l 2-4a2)
s (48Id)
Chapter 5-Characterization of Asphalt Mixtures. 293
where,
cr tensile stress in the outer fibers, psi;
=
tensile strain in the outer fibers, inches/inch;
£ =
E5 = flexural stiffness modulus, psi;
a = distance between support and first applied load,
inches; ·
P = total dynamic load with 1/2 P applied at third
points, lbs;
b = specimen width, inches;
h = specimen height, inches;
1= reaction span length, inches;
I = moment of inertia of specimen, in4; and
d = dynamic deflection of beam at the center, inches.
1/)
0..
1/)
1/)
-
....
Cl)
Cl)
"C
-~
0..
0..
<(
0
C"I
0
_j
ness values generally perform better than materials with lower stiffness val-
ues. Another plot that is often made for a constant stress mode ofloading
is log of initial strain versus log of cycles to failure.
In the constant strai~ mode of loading, the deflection is maintained
constant and the applied load is allowed to decrease with increasing cycles.
Failure is more difficult to define in the constant strain test since the sam-
ple usually does not break. Failure is normally taken as the point where the
applied load reaches some preselected percent of the original applied load.
For instance, 25 and 50 percent of the original load are typical values that
are often selected as failure. Two plots are normally prepared from the data
obtained in this test. The first is the log of applied strain versus the log of
cycles to failure and the other is the log of initial stress versus the log of
cycles to failur~. Materials that are more flexible (lower stiffness) perform
better in the constant strain test when the log of applied strain is plotted
against log of cycles to failure.
Example 5-2. A beam is being tested for flexural fatigue. The reac-
tions are 12 inches (305 mm) apart and a total load of 100 pounds (444 N)
is being applied at third points (50 pounds (222 N) at each 1/3 point).
After 1000 stress applications, the measured deflection is 0.0011 inches
(0.28 mm) per load application. Determine the flexural stress, flexural
strain, and stiffness modulus. The cross sectional dimensions of the beam
are 3 inches by 3 inches (76.2 mm X 76.2 mm).
The flexural stress has been determined to be
3aP
a=-
bh2
hence,
= 3(4)(100)
0
- 3(3)2
Simplifying the equation, the flexural stress for the load conditions in
this example is
12hd
e
312 - 4a 2
hence,
e= 12(3X0.0011)
3(12)2 - 4(4)2
E = Pa(3l2 - 4a 2)
s (481d)
Recall that the moment of inertia (I) for a rectangular cross section is:
bh 3
l=-
12
Substituting,
I= 3(3)3
12
solving,
solving,
Es= 413,000
296 Hot Mix Asphalt Materials, Mixture Design and Construction
It can also be shown that the stiffness modulus is equal to stress divid-
ed by strain. For this example:
E = .!!_ = 44.4
• e 108 X 10-6
solving,
N = K (..!..)111
1 1 C1
where,
Nf = number of cycles to failure;
--flexural-stress1.n-extreme-fiber;-psi;-arr,1---------
K1,n2 = regression constants.
where
Nf = number of cycles to failure;
£ = flexural strain in extreme fiber, inches/inch; and
K2,n2 = regression constants.
The same two equations can be used when analyzing initial stress or initial
strain. For example, the initial stress, determined in the constant strain
mode ofloading, can be plotted versus cycles to failure (1.1). The constants
KI> K2 , np and n 2 are determined from a regression analysis of the labora-
tory data.
Rallhby 8.
Sterling
n,-3.87
K~-3.65xl 0 11
T=77°
Pell et al
n2 -5.3
K~•l.lxl0 19
T=32°F
Pell et al
n2 -3.9
K~-3.0xl0 12
T=50°F
Kennedy et al
n2 -3.88
K~-3.0xl0 11
T=75°F
0
10 ,02
Stress, psi
Figure 5-7. Typical Stress-Fatigue Life Relationships (1.i)
298 Hot Mix Asphalt Materials, Mixture Design and Construction
This flexural fatigue test is used to estimate the fatigue life of a ,pave-
ment. For a given pavement and loading condition the maximum flexur-
al stresses and strains can be computed. By knowing the stresses or strains
that the pavement experiences and by knowing the relationship between
stress or strain and cycles to failure, the number of traffic passes to failure
can be estimated. Examples of test results are shown in Figure 5-7. Since
it is impossible to duplicate field conditions in the laboratory, the predict-
ed traffic applications to failure is not a good estimate of actual applications
to failure. Even in a highly controlled laboratory environment, the results
of flexural fatigue tests are highly variable. Thus, this test is normally used
to compare various mixtures and rank them for relative performance and
is not used to provide accurate estimates of the 'amount of traffic to failure
for specific projects.
(1
2P
·=-
z 1tdt
6P~-------------------
y
1tdt
Chapter 5-Characterization of Asphalt Mixtures 299
where,
c;x = horizontal tensile stress at center of specimen, psi;
cry = vertical compressive stress at center of specimen,
psi;
Ef = tensile strain at failure, inches/inch;
P = applied load, lbs.;
d = diameter of specimen, inches;
t = thickness of specimen, inches; and
x, = horizontal deformation across specimen, inches.
The above equation applies for 4-inch (101.6 mm) diameter samples hav-
ing a 0.5 inch (12.2 mm) curved loading strip and for 6-inch (152.4 mm)
diameter samples having a 0.75 inch (19.0 mm) curved loading strip.
that can tolerate high strains prior to failure are more likely to resist crack-
ing than mixtures that cannot tolerate high strains.
The indirect tensile test is normally conducted at a deformation rate of
2 inches/minute (50.8 mm/minute) and ata temperature of77°F (25°C).
A deformation rate of 3 inches/minute (76.2 mm/minute) should be used
for 6-inch (152.4 mm) samples if a strain rate similar to that for 4-inch
(101.6 mm) diameter samples is desirable. In practice, the same machine
speed is normally used for 4-inch (101.6 mm) and 6-inch (152.4 mm)
samples; however, the measured tensile strength for 6-inch (152.4 mm)
samples is expected to be smaller since the strain rate is lower. Tensile tests
can also be conducted at other temperatures (especially low temperatures)
to evaluate mix properties closer to the anticipated performance tempera-
ture. When tested _at lower temperatures, the data can be used in thermal
crack prediction models. '-
Recall that the tensile strength for a 4-inch (101.6 mm) diameter
specimen is equal to
cr = 2P
rcdt
substituting into the equation,
cr = 2(835)
3.14(4)(2)
solving,
cr = 66 psi (455.4 kPa)
The tensile strain at failure for a 4-inch (25.4 mm) diameter sample
(assuming 0.5 inch loading strip is used) is
e = 0.52x,
1
The horizontal deformation was measured to be 0.0023, hence,
solving
e
1
= 0.0012 inches/inch
Chapter 5----'Characterization of Asphalt Mixtures 301
Loading - - ---Unloading
Strain
Time
Figure 5-10. Typical Creep Relationship
There are 3 general ways that the creep test can be performed:
sample fails), and the test temperature cannot usually exceed 104°F
(40°C). In-place asphalt mixtures are typically exposed to truck tire pres-
sures of approximately 120 psi (828 kPa) and maximum temperatures of
140°F (60°C) or higher. Therefore, this test does not closely simulate
in-place conditions. ,
One type of creep test that has been used that is relatively simple and
easy to perform is the static confined test,. which more closely relates to
field conditions. By applying a confining pressure (usually approximately )
20 psi (138 kPa)), the sample can be tested at a vertical pressure up to 1:20
psi (828 kPa) (or higher) and at a temperiture up to 140°F (60°C). These
test conditions are more closely related to actual field conditions than those
for static unconfined. Research suggests that the static confined creep test
does a better job of predicting field performance than the static unconfined
creep test.
An integral part of the creep test that may be utilized is the rebound
portion of the curve obtained after the load is removed. Creep in pave-
a
ments is not necessarily problem if significant rebound occurs after the
load is removed. In the laboratory, a static load is typically applied for one
hour; then the load is removed and the sample is allowed to rebound for
another hour.. The permanent deformation is the maximum deformation
minus the amount of rebound. When comparing various mixtures i~ is this
permanent deformation that is important for estimating performance.
Creep compliance is often calculated from the creep test results. The
creep compliance, J,, is. calculated by dividing the strain by the applied
stress as follows:
-0
Cl)
c
::i C:
...
E"a
:::,
u-
0 U)
<-c:
-CD
0 C:
t:I
EE
..c: ...
-CD
15 a..
t:l'I
0
_J Arithmetic Value I
where,
Ea = accumulated permanent strain;
I = intercept with permanent strain axis (arithmetic ,
strain value, not log value) (Figure 5-11);
N = number of load applications; and
S = slope of the linear portion of the logarithmic
relationship.
where,
£r = resilient strain which is considered to become
constant after a few load applications (Figure 5-12).
he-origina:l-corrcept:-crti:-lized,epeated-load7:estnoproduce accumu-
lated permanent strain relationships. Subsequently, sequential creep tests
were utilized. An evaluation of different tests indicates that the permanent
strain relationships for tests, which involve tensile stresses, are similar but
different from those for the triaxial compression tests.
304 Hot Mix Asphalt Materials, Mixture Design and Construction
t
w CNl = resilient
strain
V__-1'_ _.,__
€a = accumulated
permanent strain
the action of rollers during construction. The GTM has the flexibility of
·varying the vertical pressure, gyration angle, and number of gyrations to
simulate field compaction equip_lll~f!~ and subsequent traffic.Typically, the
vertical pressure applied is 12(Ypsi (828 k:Pa), whichis approximately equal
to truck tire inflation pressures. The s;ttings for the gyration angle and the
number of revolutions vary between laboratories but typical values are 1
degree gyration angle and 300 revolutions. ·
During compaction of a specimen in the GTM, several mixture prop-
erties are determined. The gyratory shear index (GS!) is a measure of mix-
ture stability and is related to permanent deformation. The GSI is deter-
mined by dividing the intermediate gyration angle by the initial angle.
The gyration angle increases during compaction for unstable mixtures due
to plastic Bow of the asphalt mixture. The gyration angle does not increase
significantly for stable mixtures. GSI values close to 1.0 have been shown
to be typical for stable mixtures and values significantly above 1. 1 usually
indicate unstable mixtures.
The GTM also has the capability of measuring the shear strength of
the mixt_~r~ _during__compaction The pressure required to producetfie
desired gyration angle is measured and can be converted to shear strength.
Strength, which is measured during compaction at high temperatures, is
primarily a measure of aggregate properties, since the viscosity of the
asphalt is low and the mixture has little cohesion. It is these aggregate
properties that must provide the support to resist permanent deformation
caused by traffic.
The GTM can be used for mix design or quality control of HMA.
This equipment does a good job of achieving the ultimate density that is
obtained in the field. It also has the flexibility of being adjusted to simu-
late the tire pressures of any traffic type including cars, trucks, and aircraft,
and is the only device which can monitor the change in mixture response
with densification. This becomes critically important for traffic densifica-
tion studies or effects.
i
,:I
I
Chapter 5-Characterization of Asphalt Mixtures 309
REFERENCES
f .••. •ii, 33. AAl-ShwalailCmi, S. anMd _R.L. TeUrr~l. thEvalEuati?n of WaalteCr Damd_ ~ge, of
sp t oncrete 1xtures smg e nvironment on monmg
' System (ECS). Journal of the Association of Asphalt Paving
I Technologists, Vol. 61, 1992.
CHAPTER 6. EQUIPMENT AND
CONSTRUCTION
INTRODUCTION
This chapter includes a discussion of equipment and procedures
involved in the construction of HMA. The equipment consists of HMA
facilities, trucks, placement equipment, and compaction equipment.
Construction specifications and procedures are also discussed along with
the most common methods of selecting contractors. This chapter provides
the guidance needed, after a HMA mixture is designed, to manufacture,
place, and compact the HMA.
Typical Layout
Drum Mix Facility. A schematic diagram of the components in a typ-
ical drum mix facility is shown in Figure 6-1. The :flow of materials is basi-
314 Hot Mix Asphalt Materials, Mixture Design and Construction
EB
..0
® -uB>"
c
>
C
0
@~
:iI
0
:c
'0
II>
:,1 II>
LL
11
'0
I 0
I, u
'·
@
II
cally from left to right. The aggregates are typically moved from the stock-
pile area to the cold feed bins via a front end loader which charges the cold
feed bins. The minimum required number of bins depends on the num-
ber of different aggregates to be used in the HMA. Most facilities have 4-5
cold feed bins. The metal cold feed bins have slanted sides with a vibrator
attached (if gravity feed is inadequate) to keep a constant supply of mate-
rial fl.owing out of the bottom of the cold feed bin. Typically an adjustable
gate and a variable speed feeder are located at the bottom of each cold feed
bin to proportion the material from each bin to meet the job mix formula
QMF) gradation. A gathering conveyor brings the material to the cold
feed elevator which moves the material to the dryer. An automatic weigh-
ing system on this elevator continuously weighs the amount of aggregate
plus moisture going into the drum mixer. In the control room, a correc-
tion to the total weight is made for the ·moisture content of the aggregates
so the asphalt proportioning system can pump the proper amount of
asphalt cement into the drum mixer. The moisture correction is made by
physically determining the moisture content of the aggregate stockpile or
aggregate from the gathering conveyor and providing this information to
the computer.
A typical drum mixer is of a parallel fl.ow design in which the aggre-
gates move in the same direction as (parallel to) the exhaust gases, although
a number of counter fl.ow units are also being built. Therefore, the aggre-
gate is fed into the drum at the burner end and then it is dried and heated
as it moves down the drum toward the discharge end. The asphalt cement
is pumped from the asphalt cement storage tank and enters the drum at a
point about 1/3 the drum length from the discharge end. Some plants use
a coater which is similar to a pugmill. The asphalt cement is added to the
aggregate in the coater after the aggregate exits the drum. When the
asphalt cement is added into the drum, it is pumped into the bottom of
the drum at about the same location that the mineral filler and/ or bag-
house fines are reintroduced. Adding asphalt cement and dust in close
proximity allows the asphalt cement to trap the fines and to coat them
before they are picked up by the high velocity exhaust gas stream. The
exhaust gases are passed through a dust collection system where enough of
the dust is trapped and removed to meet emission requirements.
--------1-he-asph-alt-e:ement-e0ats-1:he-aggregates-as-the-materials move down
the lower 1/3 of the drum. The HMA exits the drum through a discharge
chute into a conveyor system which transports the HMA into a surge silo
(or other type of storage bin) from which it is discharged into the trucks
via an automatic scale system. It is then transported to the laydown site.
316 Hot Mix Asphalt Materiais, Mixture Design and Construction
All material movements from the cold feed to the surge silo are mon-
itored from the control room of the I-IMA facility. Sensors at various loca-
tions in the facility monitor conveyor speeds, aggregate weights, tempera-
ture, and other critical functions which affect efficient operation.
Drum Mixers
The purpose of the drum mixer is to dry and heat the aggregate eco-
nomically and to mix the aggregate with the asphalt cement, so that the
HMA produced at the discharge end is uniformly coated and is at the
desired temperature. In order to accomplish this purpose, the drum mixer
incorporates a variety of elements which are described below.
Aggregate Feed. The virgin aggregates from the cold feed bins are
delivered to the drum through a series of conveyors and introduced into
the drum either through an inclined chute or on a slinger conveyor. The
inclined chute feeds the aggregate by gravity into the drum at an angle that
directs the material toward the discharge end and toward the bottom of the
drum and away from the flame. The slinger conveyor, located below the
burner, is operated at a speed high enough to throw (sling) the aggregate
away from the burner toward the discharge end so that it does not inter-
.i
··' fere with the flame.
·•1:!
·'·i i
1
' Drum. The aggregate enters the drum and begins to move down the
1llii
'11i! inclined drum by gravity action. The time the aggregates spend in the
1
!t drum (dwell time) depends on the length of the drum, rotational speed of
l'!i! • the drum, angle of the drum, design and m.imber of flights, and the aggre-
!i j gate size. Typical dwell times vary from 3 to 4 minutes.
... Early drum mixers used the same drum length-to-diameter ratio of
\I 4: 1 as used in batch facility dryers, even though the drum design was par-
::.1.:.11;,1 allel fl.ow while the batch design was counter fl.ow. RAP materials are
: 11
I added about the midpoint of the drum in some systems, necessitating
:: .1 additional length to dry and heat this material properly. Therefore, drum
I' ! length to diameter ratios have increased to 5. I or 6: 1 for drums of more
lil recent design.
~~I---th_e_d-±;::!i;::::7!:::;~u;:t;;f~~-r:~::7~t::;:~;::l~~:~:~;~s-----
: jil of drum facilities operating at high efficiency. As the drum diameter
increases, the production rate increases nominally as the square of the
drum diameter for aggregate with l~w moisture content. As moisture con-
Chapter 6-Equipment and Construction 319
different p atterns, shapes, numbers, and locatfons for the flights in their
drums, all have been designed in an attempt to get the aggregates heated
and dried, using the minimum amount of fuel, and to efficiently coat the
-----.a-g:greg.a.re.....w.ith..as.phalLc.eme=n-t.~---------
The first set of flights (kicker flights) are configured at a steep angle rel-
ative to the length of the drum in order to push the aggregates down the
drum and away from the flame. If the aggregates penetrate the flame near
the entry p9int, flame quenching can occur, causing inefficient combustion
320 Hot Mix Asphalt Materials, Mixture Design and Construction
and allowing unburned fud droplets to adsorb onto the aggregate surfaces.
Thin films of unburned fuel on aggregates can significantly .reduce the
adhesion of the thin film of asphalt cement to the aggregate, reducing
durability and strength of the HMA. Pushing the aggregate away from the
flame allows the flame to expand, burn efficiently, and radiate heat down
the drum. ·
The next set of flights (cup, notched, and tapered flights) lift part of ,
the aggregate from the bottom of the drum and, as the drum rotates, aggre-
gates fall from thf flights through the hot exhaust gases. More of the aggre-
gate is lifted and cascaded through the exhaust gas flow as the aggregates
proceed down the drum. As the.material moves down the drum, the flights
are designed to pickup larger and larger amounts of the drying aggregate.
The flight designs change from tapered to notched to cup flights, each of
which picks up more material than the previous one. Once past the cup
flights, the cascading aggregate produces a curtain across the drum, increas-
ing heat transfer efficiency and preventing the flame and very hot gases
from reaching the entry points of either the RAP or the asphalt cement.
Near the middle of the drum length, immediately before the RAP
material inlet, some manufacturers have installed devices to retard the vir-
gin aggregate flow. These aggregate retention rings, also called donuts or
dams, are attached to the circumference of the drum and act as an obsta-
cle to flow, causing aggregates to build up in front of the ring. When this
larger quantity of aggregate is picked up and cascaded across the drum, a
denser veil of aggregate occurs immediately in front of the RAP inlet,
which helps to improve drying and to prevent the flame from penetrating
to the RAP, which would produce blue smoke and associated environmen-
11;1, tal problems. Other manufacturers use kicker flights in this region of the
ii'' drum to kick the aggregate back upstream toward the front of the drum.
'ri1I Either arrangement produces a concentration of aggregate just before the
:·i
,,I RAP inlet.
,IiI; The mixing flights are located just past the middle of the drum. These
1
!. flights, often called J-flights because of their shape, are designed to pick up
;···:il.·1
the aggregate and asphalt cement, mix them by producing a rolling action,
1
11
l'i and then throw the combined material through the exhaust gases. As the
l!lf asphalt cement encounters the moisture laden exhaust gas, it begins to
J,]1--i\--~£=oam,.in_c.r_e_as_ing_its_v:.ohim.e_.__The_c.aating_o£.the,_ag.gi:egates-with-asphal-t-----
lI
1
1 cement occurs very quickly and the continued cascading of coated aggre-
1
'!.'I
'i'I gate through the exhaust gas stream completes the mixing phase of the
, :I drum mix process.
j i A set of discharge flights are located at the end of the drum. These
ti
1,,,1
,!
1111
Chapter 6-Equipment and Construction 321
flights occupy avery short segment and are designed to move the material
out of the drum at either the side or rear through a discharge chute.
EXHAUST
FAN
"'
the air m.ust be provided by the exhaust fan. Therefore, sufficient air
flow must be provided by the exhaust fan to handle the exhaust gas com-
ponents made up of steam and water vapor, combustion products, excess
air, and leakage. Reference 3. provides an indication of the relationship
between these exhaust gas components for a drum mix facility operating
at 250 tons per hour (TPH) (227 Mg/Hr) production rate (see Table
6-2). The volume of steam and water vapor increases more significant-
::i:: ly than any other component of the exhaust gas stream as material mois-
:,iii
1
,','1':':
[!,1!
ture content increases.
'i:j1 i
1
Table 6-2. Exhaust Gas Components and Their Volume for Different
Aggregate Moisture Contents (J)
Data Based on 250 TPH (227 Mg/Hr) production, 6 percent asphalt, 310°F (154°C) exhaust tem-
perature, 5 percent leakage, 25 percent excess air, operation at sea level. 1 cubic foot = 28.3 liters.
Drying and Heating Process. At the tip of the burner flame the tem-
peratures exceed 2,500°F (1371 °C) and as the induced air begins to fl.ow
around the flame its temperature begins to rise rapidly to about 1400°F
324 Hot Mix Asphalt Materials, Mixture Design and Construction
O . ·:,.···:·,,.;-,•,,•:,,.,,. . Exhaust
/ • : ·,.,.........,....Recycle Asphalt
, .....~~-~ ···•.. t
Liquid
Asphalt
lnilial _J
Adjustment I Il£onstant I Superheated
Drying f"Virgin Aggregate
-+ Dry & Heat Rap
t- .. _JI
Mean Aqgreqate Temoeralure
Mixing
However, the aggregate temperature profile is quite different with the con-
stant drying zone much shorter. The virgin aggregate is kept in the hot
zone of the drum and superheated to about 500°F (260°C). Kicker flights
or dams are included in the middle of the drum to increase the aggregate
dwell time in the first half of the drum. The superheated aggregate is used
to heat the 50 percent RAP material and develop the temperature profile
in Figure 6-5. Notice that the mean aggregate temperature drops as the
RAP material is heated and its moisture evaporated in the "dry and heat
RAP zone" of Figure 6-5. The mix is brought up to discharge temperature
in the last portion of the drum.
The heat transfer process in a drum occurs by three distinct mecha-
nisms: (1) by convection-heat from the exhaust gases, (2) by conduc-
tion-heat from hotter aggregate particles transfers to cooler ones as a
result of physical contact between the two surfaces, and (3) by radiation-
the hot me~al flights and drum shell radiate thermal energy to the cooler
aggregates in the immediate vicinity of these hotter metal parts which have
been heated by the burner gases.
cement discharge varies widely but generally tends to be beyond the 2/3
point down the drum toward the discharge point. For most facilities, the
asphalt cement discharge point can be adjusted to match drying conditions
in the drum.
One manufacturer has eliminated the asphalt feed line from some
models and has added a shaft mixer, called a coater, at the discharge end of
the drum which is similar to the design used in a continuous mixer. A sub-
sequent variation by the same manufacturer is the double drum mixer , ·
which has concentric drums, with the mixing taking place in the annular
space between the inner and outer drum walls.
charged from the drum, all or part of these baghouse fines are delivered to
a surge silo and are metered through a vane feeder into the supply pipe. If
mineral filler and baghouse fines are both fed back into the drum, a com-
mon supply pipe or separate pipes may be used. Some facilities use screw
conveyors rather than vane feeders to return baghouse fines to the drum.
60°F (28-33°C) cooler than that of a parallel flow dryer of a drum facili-
ty. The increased thermal energy transfer efficiency of the counter flow
designs not only produces aggregate at the same temperature for lower cost
but lower exhaust temperatures are easier on the baghouse.
The dryer is also equipped with longitudinal flights which lift the
aggregates and cascade them through the hot burner gases. The tempera-
ture of the aggregates exiting the dryer is the primary factor that deter-
mines the temperature of the final mixture. If the aggregates are too hot,
mixing causes the asphalt cement to oxidize excessively, producing prema-
ture hardening and possibly premature cracking of the HMA mixture. If
the aggregates are not hot enough, coating the aggregates is difficult.
Once the aggregates leave the dryer, they are transported by bucket ele-
vator to the top of the tower unit containing tlie screening unit, hot bins,
and pugmill. Each of these units is described in subsequent sections.
Burners. Burners for a batch facility are the same as those used and
discussed for drum mix facilities.
Screens and Hot Bins. The hot screening unit is shown schematical-
ly in Figure 6-6. The hot aggregates are fed into the hot, vibrating screens
from the left. On the first screen, the oversized material is scalped and
moved to the right over the slanted vibrating screen. The oversize materi-
al is fed into a discharge chute and usually dropped to the ground below.
The coarse aggregate is retained between the top and second screen and
moves into Bin No. 4 on the far right. The material retained between
screen numbers 2 and 3 is the medium coarse aggregate which vibrates
through the screen into Bin No. 3. Bin No. 2 contains the intermediate
fine material that is retained between screen numbers 3 and 4. Bin No. 1
contains the finest aggregate particles. The material in these four bins must
be combined in proportions that will produce the job mix formula QMF)
developed in the mixture design. In addition to the aggregates in these
bins, two additional aggregate components that may be added to produce
the JMF are mineral filler and RAP material.
The surface area of the hot screens must be balanced with the output
from the dryer, the aggregate proportions from the aggregate cold feed sys-
tem, and the capacity of the pugmill. If the materials from the cold feed
----,1·-re-out-of-proportion-wi-rh-rhe-quanti--ry-ofm--a1:erial.1equired-fr"'""o=m~e~ac=1chc----=of-----
the bins, then some bins will be overcharged and others will be under-
charged. Such proportioning problems could result in the mix being
rejected because it does not meet the specifications or the output of the
facility being reduced because of delays in waiting for the undercharged
Chapter 6-Equipment and Construction 329
0 Rejected
oversize
aggregates
The amount of material to be used in a batch from each bin can be cal-
culated by multiplying the percentage of the batch to be taken from that
bin times the aggregate batch weight.
~~/4~~~~~/4~
PPPPP~P~~--~..XEA
I
LIVE
ZONE
drum. Since the generally accepted industry standard for velocity of air
through a drum is about 1,000 feet (305 m) per minute, the exhaust fan
must be sized to ensure that adequat1:: air flow is provided. When drum
facilities were first introduced, exhaust air velocities were between 600 and
800 feet (183-244 m) per minute, but were found to be inadequate in
areas where high production rates were required for aggregates with mod-
erate to high moisture contents. The maximum exhaust gas volumes con-
tained in Tables 6-1 and 6-2 are based on an exhaust gas velocity of 1,000
feet (305 m) per minute.
The velocity of the exhaust gas is a primary factor affecting the poten-
tial amount of airborne material for an aggregate with a given size distrib-
ution. The amount of airborne dust increases with the square of the gas
velocity.
One other factor that affects the amount of dust entering the pollution
control system in a drum mixer is the location of the asphalt cement inlet.
The closer the asphalt cement inlet is to the flame, the larger the quantity
of airborne dust that is captured by the asphalt cement and prevented from
being picked up by the dust collector system. Moving the asphalt cement
inlet closer to the flame produces two potential problems. First, the
asphalt cement may be oxidized or stripped of its low volatile components
which produce blue smoke in the stack. Second, the aggregate is not dried
as well before being coated. The facility operator must then optimize the
location of the asphalt cement inlet to minimize the amount of dust and
smoke in the pollution control system while at the same time meeting the
HMA moisture content requirements of the specifying agency.
The air pollution control system i.q. most HMA facilities includes one
or more of the following types of dust collectors (1) primary dry collectors,
(2) wet collectors, or (3) fabric filters (baghouses). Each type of collector
is discussed in the following sections.
The centrifugal collector has a tangential inlet which forces the heavier dust
to the outside wall where it slows down and falls to the bottom and is either
wasted or fed back into the HMA The cleaned exhaust air goes up through the
inner cylinder enroute to the remaining pollution control equipment. These
primary collectors are between 60 and 85 percent efficient and when used with
;i. wet scrubber, reduce the volume of sludge that must be handled.
EXHAUST GAS
WITH WETTED OUST
PARTICLES
---
OUST LADEN
EXHAUST GAS
FROM THERMOORUM
--------~-----------~---
' SLUDGE
' - TO SETTLING PONO
As the name implies, settling ponds are used to settle (separate) the
dust from the water and reduce the total volume of water required for the
operation by recirculation. Water, laden with dust, enters the settling pond
and the heaviest dust settles quickly to the bottom of the pond. Typically,
---+---_ _th_e_s_e_s_e_tt_li_n--o_f1onds are divided into sec_tions_with_the.._w.ate.r_b.ecoming----
cleaner as it moves from one section to another. Water in the last section
of the pond is drawn off the top and circulated back to the scrubber unit.
The collected fines must be removed from the ponds and disposed of since
excess buildup adversely affects the efficiency of the ponds.
Chapter 6-Equipment and Construction 335
When using a scrubber unit, the dust particles in the sludge are lost to
the HMA, thereby making the gradation of the produced materials differ-
ent from that of the incoming aggregate. If a large volume of dust is cap-
tured in the exhaust gas, a significant change can occur in the mixture gra-
dations particularly in the fine aggregate fraction. If the materials in a par-
ticular area of the country are so low in fines that mineral filler is required,
the use of a wet dust collection is not recommended since the dust cannot
be fed back to the plant.
Fabric Filter (Baghouse). Baghouses are very efficient and can remove
over 99 percent of the dust from the exhaust gases. The operational prin-
ciple for the fabric filters is very simple. The dust-laden exhaust gases are
pulled through a filter cloth that traps the dust on one side but allows the
clean air to pass through the filter cloth. The operational principle is sim-
ilar to that of a common home vacuum cleaner with the primary difference
between the two units being one of size. The size (total cloth area) of a
baghouse is determined by the volume of air it must filter. Experience has
shown that an air volume to cloth area ratio (volume in cubic feet per
minute/square feet) between 5: I and 6: I is very effective in cleaning the air
without reducing plant production due to a lack of dryer air volume.
Typically a baghouse contains between 300 and 800 separate bags.
These bags are sewn into the shape of a cylinder closed at one end and
mounted over a metal cage to prevent them from collapsing during the fil-
tering process. They are constructed from a felted, nylon fabric that can
withstand temperatures up to 450°F (232°C) and repeated flexing cycles
during the cleaning process.
The dust-laden exhaust gases are brought through the ductwork from
the primary collector to the front of the first set of filter bags (Figure 6-10).
As the dirty exhaust gases encounter the filter fabric, the dust particles are
trapped on the outside, and the clean air continues to the exhaust fan and
out the stack (Figure 6-1 I). The dust particles on the outside of the filter
bag build up with time and form a dust cake which is very important for
proper functioning of the baghouse. As a dust coating builds up on the
bag, the filtering efficiency of the bag increases and finer and finer dust par-
ticles are trapped. However, if the coating on the bags becomes too thick,
the bags are blinded and air is unable to penetrate the cake, causing the
baghouse to stop functioning. Because of this the dust cake must period-
ically be removed from the bag. This is accomplished by the cleaning
----m-ecnams-rn:--'TliI':-d:uscfa:tlroff-ene-outstcl.eohne-bags-aha-falls to the bot-
tom of the baghouse and is either fed back into the HMA or wasted. Since
there are so many bags in a collector, the cleaning is usually performed in
groups so that only a small percentage of bags will be off-:line at any given
time in order to maintain efficiency.
336 Hot Mix Asphalt Materials, Mixture Design and Construction
_J __
Transportation Operations
The transportation portion of the project involves the trucks used to
haul the HMA from the HMA facility to the paver, and the equipment and
procedures used. These procedures include loading the HMA into the
truck, weighing and ticketing, protecting the HMA from cooling, trans-
porting to the job site, dumping into the paver hopper, and returning the
truck to the 'HMA facility. This whole transportation operation is often
referred to as the truck cycle and must be analyzed carefully when deter-
mining how many trucks of a specific size are required to balance the out-
put from the HMA facility wipi the compaction and laydown equipment
at the road. The number of trucks needed for a project is affected by the
length of waiting times for loading and unloading, the loading time, time
to ticket and cover the HMA, the distance between the production site and
the laydown site, the average truck speed, the output of the HMA facility,
and the availability of surge or storage facilities. In this section, the dis-
cussion begins with a description of the types of trucks used to haul HMA
from the production site to the laydown site. The primary types of vehi-
cles, categorized by the method of discharge ofHMA, are end dump, bot-
tom dump, and live bottom (conveyor) vehicles. Regardless of the dis-
charge method, all v.ehicles should have metal bodies which are clean,
smooth, and free of holes. Insulation extends the haul distance (time) that
is acceptable for a project.
End Dump Vehicles. These vehicles are typically one of three AASH-
TO classes: a single unit vehicle with either a tandem rear axle (3D) or
triple rear axle (4T); or a tractor-semitrailer with tandems on the rear of
both the tractor and the semitrailer (3-S2). Each of these vehicles has a
maximum gross weight which is controlled by the weight restrictions
imposed by either state or federal statutes. The federal bridge formula
restricts the gross vehicle weights (GVW) and payload of vehicles, primar-
ily as a function of axle spacing.
The end dump vehicle must be capable of lifting and rotating the
- - - - r r t ruck-bed-ro-discharge-aH--H-MA-wi:thour-rhe-rea:rof-rh-e---rraiterpressirr·o-----
down on the paver hopper. 'When discharging HMA from the truck to the
paver hopper, the truck bed should be elevated slightly before the tailgate
is opened to allow HMA to flood the hopper with a surge of asphalt mix-
ture thereby preventing a trickle of coarse material into the hopper which
Chapter 6-Equipment and Construction 339
produces segregated spots. However, filling the hopper should not cause
material to spill in front of the paver.
Bottom Dump Vehicles. The bottom dump vehicles are gondola type
3-S2 vehicles which unload by dumping directly onto the roadway or into
a spreader box beneath the vehicle which then discharges onto the road.
For vehicles that dump directly onto the road, the discharge rate is con-
trolled by the width of the opening of the clam shell jaws that are in the
bottom center of the trailer bed. The bed has slanted sides to facilitate
smooth flow of material from the bed into the discharge area. Control of
the width, and therefore the volume, of the material is critical since the
material is often supplied to the paver by a windrow pickup machine. This
machine has no significant storage capacity to store excess material or to
supply extra material if there is a deficiency of material in the windrow in
front of the paver. Therefore the quantity of material discharged by the
bottom dump vehicles must be accurately controlled if a windrow pickup
machine/paver unit is used to spread the HMA.
'!
Live Bottom Vehicles. There has been an increase in the use of live
bottom trailers which contain a conveyor at the bottom of the trailer.
These vehicles have been used successfully on some projects to help mini-
mize segregation of materials, especially at end of truck loads. The slat
conveyor pulls material from the truck bed and discharges it into the hop-
per of the paving machine without elevating the bed. Use of this type of
vehicle increases capital investment, because of the additional cost of the
trailer, and increases operating costs as a result of additional maintenance
on the conveyor system. Some advantages are increased payload, and lower
cost per ton mile compared with 3D and 4T vehicles, and increased safety
when compared with rear dump 3-S2 vehicles.
the sides of the cone of dumped material. The typical pattern of segrega-
tion of these coarse materials in a truck bed filled with a single dump is
shown in Figure 6-I-3a. The segregation in the truck bed shown in Figure
6-13a shows up on the road behind the paving machine as a distinct pat-
tern since the segregated areas may be scattered across the entire paving
lane. The segregation is increased when the paver wings are lifted and the
material is dumped into the paving hopper. This segregation is highlight-
i ed at the end of truck loads.
~,'+--------1-he-slight-segregati0fl-i-n-the-tmek-bed-produced-by-multip-le-dt1mp-----
as shown in Figure 6-13b may not show up at all on the road, especially if
dumps 1 and 2 are made close to the ends of the truck bed thereby pre-
venting the segregation that occurs at the front and rear when single dumps
are used. With the segregation concentrated in the center of the truck bed,
Chapter 6-Equipment and Construction 341
the paver augers have an opportunity to slightly remix and move the large
aggregates around to minimize the segregation on the roadway.
The gate configuration on the silos or surge bins should be large and
elongated to drop a large mass of HMA material quickly and thereby pre-
vent significant coning of the drop which accentuates segregation. The
worst drop situation is a slow trickle of material into the center of the bed
which allows the coarse aggregate to run down the pile producing segrega-
tion at the edges. Studies have shown that one of the biggest causes of seg-
regation on the roadway is improper use of the storage silo.
Protection During Haul. When the weather is cool or the haul time
is long, the HMA should have a protective covering to prevent excessive
cooling and the formation of a crust on the surface. It is important, how-
ever, that protective covers be securely tied down to prevent air from get-
ting between the cover and the asphalt mixture. Therefore all covers
should lap over the sides of the truck bed and be snugly tied down all along
the sides and rear of the truck bed.
Figure 6-14. Storage Silo Showing Bucket Elevator and Weigh Batcher
many practical advantages for special applications such as paving test tracks
and race tracks where high superelevation, stringent grade control, and
unusual construction requirements are included in the construction con-
tract. Experience has shown that the use of a MTV results in improved
surface smoothne_ss.
Asphalt Pavers
Principles of the Asphalt Paver. As seen in Figure 6-15, the paver is
the prime mover which receives material from a truck, stores it on board
and has a conveying system that takes the material from the hopper to the
rear of the machine and dumps it on the prepared surface. After being
dumped in front of the screed, a pair of laterJ augers distribute the mate-
rial transversely across the width of the screed. The paver is comprised of
two basic machines: the first unit is the tractor, which can be mounted on
rubber tires or tracks that propel the paver and which provides power for
running the mechanical systems; the second unit is the screed unit at the
rear of the paver with its pivot arms that are mounted on the tractor. The
screed unit is free to rise and fall since. no torque is exerted at the pinned
connection point on the tractor. The plan view of the paver shown in the
bottom of Figure 6-15 shows the two material handling systems that (i)
move the material from the hopper at the front of the paver to the rear of
the paver and (2) distribute the material transversely across the paving lane.
The most critical feature of the paver is the self-levelling screed unit
(Figure 6-16) which determines the profile of the material being placed.
As the screed is towed down the road there are a number of forces acting
on the unit. First, there is a towing force which is provided by the tractor
unit and which is exerted at the tow point. Second, there is a force by the
asphalt mixture resisting the towing force; this material is shown in Figure
6-17 immediately in front of the screed unit. Third, the weight of the
screed unit presses down vertically on the material unµerneath the screed.
Fourth, there is additional compactive force applied by the screed either
through tamping bars, vibrators, or both. Fifth, there is a frictional force
between the bottom of the screed and the material underneath it. Sixth,
there is a resistive vertical force against the bottom of the screed from the
material being compacted. These forces are seeking an equilibrium posi-
-4'11-----t-1o_n_a_t__,.l times, therefore a free body diagram can be drawn to demon-
strate the relationship between these forces as shown in Figure 6-18.
The HMA in front of the screed is loose but the material discharged
behind the screed is slightly compacted. The screed weight plus the vibrat-
ing and tamping energy it imparts produces the increase in density. If the
Chapter 6-Equipment and Construction 345
PUSH ROLLERS
AUGERS
''
)
. , , .. __ ;r ------ ' ___ , .
MATERIAL
F\.OW
MATERIAi.
FLOW
' ,
,~.----'--------'- -- ....,
( ~
'·------ ----- -----
LEVELLING ARM
~~:g;-~~. '
~~~~t~,I~N SCREED
t F2~f~~~
Fl (~!'.:~~i::;;;~~:.~_:_;,►;";:).•:. ;-,,·.,. :--.-· •:'t :- .. .•.• , ..--t-.:• .. ,-~~~ • ..!-~•.••·• :
F 1 = Towing Force
I
F2 = Weight of F = Closing force required by equilibrium which
1 5
Screed is due to both the resistive force of the HMA
~under the screed and the compaction effort
1
1
supplied to the screed.
F, = Force of Material in
front of screed
-----~-----
I --'-----.!. ____ i _____,__ 'i,,._ _____ _
-
T
.....: .....: .....: ~
:
~~ ~ ~ ~ :
,, 633/o I ; : I ;
Tl
STEP AMPLITUDE= 100%
I 7' I
TOW POINT PATH
..I.-
I
1L
I
2L
I
3L 4L SL 6L
ing continuously; in these cases the paver should be stopped quickly but
smoothly, loaded as quickly as possible, and accelerated to paving speed
again quickly and smoothly.
A second job-related factor that affects the screed is a change in paving
speed. During the paving operation, the material passing under the screed
is subjected to certain weight and compactive forces that cause the density
of the HMA to increase as it Bows from the front to the back of the screed.
If the paver speed is increased, the mix Bowing under the screed is exposed
to the compactive forces for a shorter period of time, thereby providing a
lower density material. This lower density material does not provide as
high a resistive force as the higher density material produced at the lower
speed, with the result that force F4 (Figure 6-18) is lower. The resistive
force increases to achieve equilibrium and the- screed falls. Therefore, if a
smooth pavement is desired, it is important that the paver maintain a con-
stant speed to avoid the screed undulations caused by speed changes.
Control of the head of material in front of the screed is maintained by
the proper setting of the Bow gates and augers on the paver. An adjustable
Bow gate is located above each of the slat conveyors immediately in front
of the augers. The purpose of the Bow gates is to control the amount of
material that is supplied by the slat conveyor to the augers. It is important
that the Bow gates be set at a height that delivers sufficient material to the
auger chamber to have only about 50 percent of the auger sticking out of
the mix and to cause the augers to turn when the paver is moving forward.
The auger motion is controlled by sensor elements that activate the augers
when the depth of material in the auger chamber falls below a preset value.
Each of the augers operates independently and distributes the materi-
al transversely across the paving lane. At the junction of the two augers, in
the center of the paver, is the auger gear box. In order to force HMA under
the gear box, there is typically a reverse auger adjacent to the gear box, to
ensure that sufficient material is forced into that location.
Historically, screeds produced compaction by the action of a tamping
bar which was mounted in front of the screed plate (Figure 6-20). A typi-
cal tamping bar moved up arid down 5/32 inch (4.0 mm) at a rate of 1500
tamps per minute. The compaction actually occurred just in front of the
beveled face of the tamping bar, which confined the material when coupled
..------~-_,ci.th._the_fo..cw..ard.moYement-o£the_p.av.:er__The..m.ateriaLcame-in-contact-w.ith.__ _ __
the face of the bar and was forced down when the bar moved down. This
tamper bar compaction system had a large number of moving parts, required
substantial regular maintenance for efficient operation, and was very heavy.
The compaction produced by the bottom 1/4 inch (6.4 mm) of the
Chapter 6-Equipment and Construction 349
24"
Tamping bar
of Motion
Plate
60"
-i·lr"t-
5132 in. eccentricity
1500 lamps/minute
I I
Direction
of Motion
\A
t.h I -
~
Figure 6-21. Schematic of a Vibrating Screed
350 Hot Mix Asphalt Materials, Mixture Design and Construction
Strikeoff
Direction Plate
screed. When variations in paving width occur often, it may be more eco-
nomical to have the flexibility of the hydraulically extendible screed as
compared to the down-time required to add and adjust the rigid screed
extensions. One problem with the hydraulically extendible screed is that
the surface texture is different and the HMA density is lower under the
screed extension than that under the basic screed unit.
suming and expensive. In addition, any errors in the stringline grade are
reproduced by the paver.
Another disadvantage of the erected stringline is that all equipment
and personnel must stay away from the line so that the line is not dis-
turbed. Therefore, truck drivers must be very careful when entering and
leaving the area in front of the paver. Unless grade is an important crite-
ria on a paving project, the added cost and care associated with the use of
an erected stringline is seldom warranted. However, the use of stringlines
is essential for many airfield paving projects since this method controls the
actual grade. On airfields, for instance, where the final pavement must tie
in with taxiways, concrete slabs, and other existing structures, and where
close controls are specified for grade changes, the stringline method is a
necessity. On these large paved surfaces, such '2.S airfields and parking areas,
the use of stringlines also prevents bird baths.
Laser technology has recently been applied to the paving operation,
offering an alternative to the stringline grade control system. A heli-
um-neon laser transmitter is used to establish a horizontal reference plane
over the job site. Current laser technology permits the establishment of a
horizontal plane, a single grade plane or a dual grade plane for both grade
and slope control. These systems can be coupled with existing paver con-
trols for automatic or manual operation and have been demonstrated to
work in both dust-laden and in high-electronic and light-noise areas like
airports. Advantages claimed for these systems include more uniform
depth control and smoother riding quality than for conventional automat-
ic control systems. The laser method is difficult to use when the pavement
surface is designed to have frequent grade changes.
The purpose of the mobile reference is to average the effect of devi-
ations in the existing pavement surface over a distance greater than the
wheelbase of the tractor unit. Therefore, the mobile reference system
expands the base length of the paver in an attempt to smooth out the
profile of the surface on which the paver is moving (Figure 6-23).
Manufacturers have used several types of mobile reference systems
including:
This third type of reference system includes a short shoe, or ski, which
slides on an adjacent pavement or curb to which the mat being placed is to
be matched. This type of mobile' reference is only used when the grade
Chapter 6-Equipment and Construction 353
being sensed is smooth. Because of its short length, this reference system
does not remove any variations occurring in the adjacent pavement surface
but rather duplicates that grade in the mat being placed.
COMPACTION
Compaction is the process by which the volume of air in a HMA mix-
ture is reduced through the application of external forces. The expulsion
of air enables the mix to occupy a smaller space thereby increasing the unit
weight or density of the mass.
The compacted mixture should have sufficient voids to allow the
asphalt cement to expand and contract as temperature changes without :fill-
ing the voids resulting in flushing. The voids should be high enough to
allow for some subsequent traf:fic~induced densi:fication during the first
few years of service without the void content falling below about 3-4 per-
cent for dense-graded mixtures. If the void content in a dense-graded
asphalt mixture drops below about 3 percent, significant permanent defor-
mation can occur. However, for dense-graded mixtures, the pavement life
is reduced about 10 percent for each percent increase in voids above 7 per-
cent, according to Linden, et al. (i).
The compaction process is affected by the condition of confinement of
the HMA being compacted. In the laboratory, confinement is provided by
the sides and bottom of the mold and the hammer. Compaction energy in
the laboratory process can be calculated. However, in the field, confine-
ment is provided by the surrounding HMA material, the underlying layer
and the compactor contact area in its zone of influence. The amount of
compaction energy put into field material cannot be determined.
However, its effectiveness can be determined by measuring the increase in
density produced by the compaction energy.
;[i
1.:1.
Mixture Factors Affecting Compaction
For effective compaction t<;> occur, the compactive forces exerted by the
roller must exceed the forces resisting compaction within the mixture. The
mixture resistance is a result of the _combined effect of the aggregate phase
and the asphalt cement phase which fills the voids in the aggregates (.g). An
Chapter 6-Equipment and Construction 355
S=c+crtan<p
where,
S = shearing resistance of the soil;
c = cohesion;
cr = applied vertical stress;
<p = angle of internal friction due to the interlocking
resistance of the particles.
1. The voids factor should have a minor effect, since most mixtures
are designed for approximately 4 percent voids;
2:-Tne fitler-15itumen factor is quite-imp-ortant~ as~ir-affects the mass
viscosity of the matrix that surrounds the coarse aggregate particles;
3. The effectiv;e particle size of the filler is important since generally
the smaller the effective diameter of the filler, the greater its mass
viscosity; and
356 Hot Mix Asphalt Materials, Mixture Design and Construction
The shear rate is also very important and it is one factor over which
the contractor has absolute control. Therefore, if the mass viscosity for a
particular mixture is higher than normal, the contractor can reduce the
speed of the rollers in order to decrease the shear rate to produce shearing
stresses high enough to move the aggregate into a denser position. Of
course, the contractor might be able to increase the weight of the roller to
produce the same effect.
Compaction Equipment
Self-propelled compactors are used to provide the compaction energy
required to reduce the volume of freshly laid HMA sufficiently to produce
the specified density. This compaction train usually consists of two or
more rollers whose objectives are:
: :.'i!: one providing the drive ·energy to propel the roller. This arrangement
allows the loose HMA to be tucked under the drive drum. If a non-dri-
en-drnm·encounters--freshlyfaid-HMA-;-drere-f·satena.eficyfofl:ne Mrn-----
1 :,I! to shove horizontally and produce a ridge of material immediately in front
'11.11 ! of the drum which remains after the roller reverses directions and produces
,I a bump. Hairline cracks (checking) also develop when an asphalt mixture
Ji is rolled in this way. A PTR can also be used as a breakdown roller but it
l!I
i II
II!
Chapter 6-Equipment and Construction 357
I
is sometimes difficult to remove roller depressions with the finish rollers.
The hot asphalt mixture also tends to stick to the rubber tires causing sur-
face deficiencies. Therefore, a PTR is most often used as an intermediate
roller. Finish rollers are steel-wheeled rollers.
Rolle;s operate in a rolling zone behind the paver whose length is
determined primarily by the number of compaction passes required to
achieve density. The number of passes are in turn determined by the effec-
tiveness of the roller in producing the required density, the compactability
of the HMA being placed, the width of the roller, the width of the paving
lane, the speed of the roller, and the cooling rate of the mat. Compaction
must be accomplished before the mat cools to a temperature below approx-
imately 175°F (80°C), called the cessation temperature. The cooling rate
of the mat is affected by the mat thickness, the temperature of the mat
when placed, the ambient temperature, the temperature of the base on
which the HMA is placed, and the wind conditions. A sample of the types
of data available showing cooling time versus factors affecting cooling rate
is shown in Figure 6-24.
Compaction always results from the application of pressure over a con-
tact area (SJ. The number of contact points in the aggregate mass increas-
es as the particles are moved into more intimate contact. Thus compaction
occurs only when the HMA is confined as a result of the roller, the base
support, and the surrounding mix particles. If sufficient confinement pres-
sure is not present, the HMA will move laterally producing decompaction
rather than moving vertically producing compaction. This possibility for
a decrease in density must be considered when developing a roller pattern
(Figure 6-25). As the temperature of the mix drops and the number of
contact points increases, there is an increased resistance to further com-
paction. This process continues until no additional densification occurs.
At this refusal density point the internal resistance of the mix equals or
exceeds the compaction contact pressure. If the refusal density does not
equal or exceed the specification density, the operator must increase the
compaction pressure. Modification of the mixture design to improve
in-place density is a common mistake made on many paving projects. As
a general rule, the mix design should not be modified for the purpose of
increasing in-place density.
The ability of the roller operator to change the contact pressure
improves the operator's chances of achieving the specification density. The
options available to the operator for increasing the contact pressure
include:
01 lO
02
0£ Cl)
Ov w
m; IC) 5
09 z
(./)
Cl)
01
~~
~
u
09 J:
I-
t() I-
<(
~
C\J
0
-----,---.....-------....-----+-
0 0 0 0
0
IC) I'() C\J
LOW 0 2 3 4 5 6 PASSES
COMPACTION CURVE
Figure 6-25. Establishing a Roller Pattern Using a Test Strip
(140-175 cm) in diameter and 18-20 inches (46-51 cm) wide. This roller
is typically used as a breakdown roller during the initial part of the rolling
operation which occurs immediately behind the paver. These rollers are
not widely used anymore. Their biggest advantage is their ability to apply
large pressure, due to the narrow back wheels, an:d thus improve density in
areas such as longitudinal joints.
Tandem steel-wheeled rollers are two axle rollers with both drums
having the same width. They typically weigh from 5 to 14 tons (4.5 to 13
Mg) (12). These rollers can be ballasted with water or sand for extra
weight. Typically, the,drums vary from 48 to 60 inches (122-152 cm) in
diameter and are 42 to 54 inches (107-137 cm) wide. Many of these
rollers have a single drive drum to propel the machine. The drive drum is·
often larger in diameter than the steering (tiller) drum.
Since hot asphalt tends to stick to the steel drums, all rollers have a
series of mats, mounted on a transverse bar, which are kept wet and in con-
tact with the drum surface The mats are kept wet to help minimize stick-
ing, especially when the drum is cold. The roller also has a scraper across
the steel wheels to remove any asphalt mixture that sticks to the wheels.
The water system for wetting the mat is used during breakdown rolling but
is not used for finish rolling.
The actual compactive effort supplied by steel-wheeled rollers is deter-
mined by the contact pressure between the steel drum and the HMA mix-
ture being compacted. The vertical load is determined by the gross weight
of the roller and ballast. The contact area is determined by the diameter of
the drum and the depth of penetration of the drum into the HMA mate-
rial, which is a function of the compaction characteristics of the HMA
(Figure 6-26). Table 6-3 contains contact pressure information for sever-
al rollers of different drum diameters and static linear loads for different
depths of penetration of the drum into the freshly laid mat (h 1-h2 from
Figure 6-26). & the penetration depth increases for a given roller, the
contact pressure decreases since the contact area is larger. For a given con-
tact pressure, larger drums have lower angles of contact than smaller drums
since the angle is a function of the arc length, AIA2, divided by the radius
of the drum (Figure 6-26). Since larger drums have smaller contact angles,
they also have a lower component of horizontal force that tends to push
against the mat being compacted.
Figure 6-26. Contact Angle of a Steel Drum with a HMA Mat Being
Compacted (11)
tact pressure on the mat. The design of the pneumatic-tired rollers provides
a kneading action by the tires, which are individually mounted to the roller
frame. Typical configurations of pneumatic-tired rollers provide 4, 5, 6 or
7 tires on the front of the roller and 3, 4, 5 or 6 tires on the rear with the
rear tires aligned to provide compaction for the material passing between
the front tires. These rollers typically vary in wheel weights from 1,500 to
10,000 pounds (0.7-4.5 Mg) as reported by Geller (JQ). The tires are
smooth, i.e. they have no tread, and must be able to withstand high infla-
tion pressures for long periods of time. The size and ply rating of the tires
can vary significantly. Inflation pressure is controlled by tire ply rating but
the combination of contact pressure (controlled by tire inflation pressure)
and area (controlled by weight of roller) are important in producing densi-
fication of the HMA. High contact pressures are only attainable with high
tire ply ratings, whether for the 7.50 or 20-inch (19 or 51 cm) rim sizes. If
----~•-P_'.f.R-i-s-t0-be-1:1sed-fe.r-G0-mpaGti0-B.-pur:poses,-then-the-tire.inflation pres-
sures should be as high as the behavior of the HMA will permit without
severe rutting which may be difficult to remove by the finish roller.
PTRs can be used in the breakdown or intermediate roller positions.
In the breakdown mode, PTRs usually cause deformation that may or may
.,,
Si ~
i .
flass Drum Static Lineal Penetration Depth, in.
-
c:r
n,
0\
Diameter Load ~
12-15\Ton Static 69" Conditions 1/16 1/8 3/16 1/4 3/8 1/2 5/8 3/4 f(l n0
3°Wheel
Contact Pressure
n,
t!,. a
Without
Ballast
341 pli psf 174 112 92 83 , 66 58 53 48 ~a
n, ~
t!,. ..,
n, n,
With 429 pli psf 220 142 116 104 83 73 66 60 ~~
I Ballast ~·=
@: a
::i::
$a
~
53 45 38 ~-
rem
8-12 /ron Static 60" Without 180 pli psf 96 64 33 29 26
A
Ballast
"' ~ ►
With
Ballast
250 pli psf 132 88 74 63 53 46 41 36
~~... 1
~
i • 0 ~
Over 14 fon Vibratory
T~ndem
60" 190 pli psf 101 68 56 47
'
1
40 35 31 27
=
"'
;:!.
e.
Y'
I g
Chapter 6-Equipment and Construction 363
not be removed with finish rolling. Therefore, the mix behavior deter-
mines whether a PTR can be effectively used as a breakdown roller (12).
Geller (1Q) points out that the effectiveness of a PTR as a breakdown roller
can only be determined after the next roller levels the surface, so that a
nuclear gauge can sit fl.at on the surface and density measurements can be
performed.
Traditionally, PTRs have been used in the intermediate roller position.
Geller (1Q) recommends ground contact pressures between 75 and 90 psi
(517 and 621 kPa) which are most easily met by the 20-inch rim with 16
or 18 ply tires. In addition to producing densification, the PTR in the
intermediate position can help to remove roller checking produced by a
steel-wheeled breakdown roller. Checking is the term used to describe
fine, hairline, transverse cracks which occur in the surface of a HMA mat
typically under a steel-wheeled roller. Checking is normally caused by
horizontal movement of the mixture during compaction with the
steel-wheeled rollers. This can be caused by mixture problems (tender
mixture) and/or improper rolling techniques.
Since asphalt cement sticks to cold rubber tires, wetting mats, like
those described for steel-wheeled rollers, are used with individual mats
contacting each tire. Once the tires become warm, the asphalt cement does
not stick to the tires and the water is no longer needed. Rollers should not
be parked on the hot mat since this causes depressions that cannot be
removed with additional rolling. If the tires cool, they have to be brought
back to operating temperature in the same way as for initial cold starts.
Pneumatic-tired rollers are reported to offer several advantages when
compacting dense-graded HMA (12):
------¥ibrat-o-ry-Steel-"Wh.eeled-R-0llen.-T'.h@-m0st-recent-development in
rollers for compacting HMA is the vibratory steel-wheeled roller or vibra-
tory roller. Vibratory rollers require more operator discipline than nonvi-
bratory rollers in carrying out the roller pattern. Selecting the wrong force
level, rolling too fast, and making too many passes with the vibrator
364 Hot Mix Asphalt Materials, Mixture Design and Construction
engaged can cause probleml!. However, current roller designs have numer-
ous features to assist the operator in maintaining roller pattern discipline
and the features being added are improving UJ!).
Five types of vibratory rollers which may be seen. in use for HMA com-
paction include:
Vibratory rollers are the only type of HMA compactors that have a
dynamic load component and are typically lighter than non-vibratory
steel-wheeled rollers because the total, compactive effort is the sum of the
static and dynamic loads. Typical drum widths for vibratory rollers are 58 (
to 84 inches (147 to 213 cm) with typical drum diameters of 40 to 60
inches (102 to 152 cm). The dynamic load applied by the vibratory roller
is the major force component producing densification of the HMA.
The dynamic load is produced by an eccentric weight attached to a
rotating shaft in the center of the drum. When the eccentric weight is
rotating toward the surface being compacted, the total applied load is the
sum of the static and centrifugal loads. When the eccentric weight is rotat-
ing away from the surface being compacted, the total applied load is the
static load minus the centrifugal load. Geller (lff) reports that vibration
reduces the mechanical friction during compaction but results in an
increased mechanical interlock afterwards. The vibration is especially
effective on harsh mixtures which have high volume concentrations of
coarse aggregate.
In order to compact a particular HMA layer effectively, the frequency
and amplitude must be selected to satisfy the conditions on the job so that
the total force applied is adequate to produce the density desired but not
so large that decompaction occurs because of the large vibrating force.
Generally thicker lifts require a higher amplitude and .a lower frequency
compared to thinner lifts. Monitoring changes in field density with each
roller _pass using a nuclear density gauge is probably the simplest way to
establish the proper settings for amplitude and vibration for a particular
project.
Several organizations have evaluated the relationship between impact
spacing and marking of the pavement surface (13.). For surface mixes, a
smooth ride is one of the primary criteria for acceptance; therefore, the
impact spacing should be short enough so that no vibration marks are
observed in the finished surface. To accomplish this, it is recommended
that the impact spacing be 1.5 inches (38 mm) or less (8 or more impacts
per foot).
the cooling rate of the HMA mat and the length of time the material is hot
enough to be compacted. Information on cooling rates and time available
for compaction is typically presented in either tabular or figure form. Table
6-4 contains one set of recommended minimum laydown temperatures for
a range of base temperatures and mat thicknesses. These compaction tem-
peratures are estimates and will vary with different asphalt cements and
aggregates. This table was developed for typical asphalt mixes used in the
United States (ld). Notice that for thin mats, the time available for rolling
is short. For example, a 3/4-inch (19 mm) mat placed at the recom-
mended minimum laydown temperature has only 6 minutes to be com-
pacted to achieve the target density. The roller speeds cannot be increased
significantly without adversely affecting density; hence, additional rollers
may be required when paving at low temperatures.
Rolling Patterns
---+------tiffer.ent.rolling_patterns..ace-used-on-mat.s-with-G0-n-sta1H-c-cF0ss-sl0-p@--atr------
com pared to those with crowns or super elevation. In rolling unsupport-
ed mats with constant cross slope, the first pass of the roller should be on
the outside (low) edge of the mat? with each successive forward pass mov-
ing toward the high edge of the mat as shown in Figure 6-28. The next
Chapter 6-Equipment and Construction 367
CD
forward pass should overlap the first forward pass by at least 6 inches (150
mm). The roller passes should end at different points to prevent the devel-
opment of a bump. The last forward pass of the roller; at the high edge,
should overlap the upper edge by at least 6 inches (150 mm).
Rolling of transverse joints is important to ensure that a smooth dense
joint is obtained. If possible, transverse joints should be rolled in a trans-
verse direction. This requires that runoff boards be used adjacent to the
HMA mixture to prevent damage to the edge. The runoff boards should
be the same thickness as the asphalt mixture being placed. Many times,
due to traffic considerations and working room, the transverse joint must
be rolled in a longitudinal direction and, thus, runoff boards are not need-
ed. When rolling in a longitudinal direction, the hot mat tends to push
away from the cold asphalt mixture resulting in a depression.
Longitudinal joints are important because of their extent on most
HMA paving projects. It is important that longitudinal joints be com-
pacted properly to ensure a smooth joint that is barely noticeable when
_ _ _ __.,_rossed.--Iflongitudinal-jointsare_noLcompacted properly, the joint may
be rough and noisy when crossed, it may be a future location for water
damage or raveling, and it may be a potential driving hazard. Longitudinal
joints are compacted using the rollers available for compacting other por-
tions of the HMA mixture.
368 Hot Mix Asphalt Materials, Mixture Design and Construction
that perform satisfactorily; however, all these techniques have also been
misuse1d, producing pavements that do not perform satisfactorily.
One difficulty in preparing density specifications is describing the
method of measurement. The two primary methods for measuring densi-.
ty are (1) taking cores from the in-place pavement and then measuring
bulk density in the laboratory, and (2) using nuclear gauges to measure the
in-place density. Most engineers agree that nuclear gauges do not measure
density as accurately as the core density procedures. Many agencies use the
nuclear gauge to develop roller patterns but specify that cores be taken and
tested for acceptance or rejection of the in-place mixture. However, some
agencies believe that a properly calibrated and. properly operated nuclear
gauge provides results accurate enough for acceptance testing. Some cores
should routinely be taken throughout the paving project to verify that the
1
nuclear gauge is in calibration.
The voids in an asphalt mixture are directly related to density; thus,
density must be closely controlled to ensure that the initial in-place voids
are within an acceptable range. There is considerable evidence to show
that the initial in-place voids for dense-graded mixtures should be no
higher than approximately 8 percent and never fall below approximately 3
percent during the life of the pavement Cl§,_ 12. l1J.,. 12,_ 2Q,_ 21). High
voids allow water and air to penetrate into the structure resulting in water
damage, oxidation, raveling, and cracking. Low voids lead to rutting and
shoving of asphalt mixtures. This recommended range of in-place voids
applies only to dense-graded HMA mixtures. Other mixtures, which are
specifically designed to contain higher or lower air voids, can perform
properly when designed using proper techniques.
Ford ( 16) showed in a study for the state of Arkansas that HMA mix-
tures should be designed and constructed so that the in-place air voids stay
above 2.5 percent. At void levels above 2.5 percent, Ford's data showed
that the expected rut depth would be less than 10/32 inch (7.9 mm)
(Figure 6-29) based on air voids determined from cores and rut depth
measurements of in-s_ervice pavements.
Brown and Cross (12) showed that significant rutting would likely
occur in mixes that had laboratory compacted voids of 3 percent or less.
When a suitable aggregate was used and the voids stayed above 3 percent,
little rutting was observed. Huber (11f), in a study of HMA pavements in
----\Gafl.-a@,e0n-si-derecl-a-ft1:l:l"l'l:0er-0-fea1:tSes-0f-rt1:tting--and-determined-that·on·-----
of the primary causes was air void content below 3 percent.
Zube (12) showed that dense-graded HMA pavements become high-
ly permeable to water at approximately 8 percent air voids. Figure 6-30
shows that as long as the voids are below 8 percent permeability should not
Chapter 6-Equipment and Construction 371
C\I 30
-
I'()
Best Filled Equation
Log RUT = 1.188 - 0.6947 Log AIR VOIDS
R2 = 0.456, RMSE = 0.186
:C: 20
I--
CL
w
Cl
I- 10
:::)
0::
15-----,----,----,----r----ir-----,
----- --
--·
V
0
0 i---
0 ... i.--
101----c~>+-I/__.~-L'~_YO"'l------+-=0---+----+-~->----1
V O I) 0
~
Cb ..,.~ > o
I
CJ)
o,/o 0
a
0
>
o--_.......___....____._______._____._____,
0 I 00 200 300 400 500 600
FIELD PERMEABILITY ML/MIN
Figure 6-30. Relationship Between Air Voids and Permeability for 10
Projects in California (12)
0.7
0.6
-.E
Q)
: :,
C:
-en
Q)
.c:
o.5
(.)
C:
-;_0.4
.J:l
C
Q)
E
a5 0.3
Cl.
Q)
-
.::::
C
a5 0.2
a:::
_0.1
the air void content of the laboratory compacted material is not satisfacto-
ry, the mix must be adjusted so that an acceptable air void level is obtained.
Often the only adjustment required is a slight modification of the asphalt
------ccomem-:--T--:he-clensi-ty-p·1:00:1:1:eea-d1¾r-ing-mixtui:e-design-should not be used
as the reference density since the laboratory produced material is usually
somewhat different from the material produced in the HMA facility.
Aggregates tend to break down during mix production, creating an
increase in dust, thus altering the properties of the compacted asphalt mix-
374 Hot Mix Asphalt Materials, Mixture Design and Construction
1,I ·100
90
:i:I sci
i ,70 10
I
60
i :I ~ 50
11'
-0 9
] 40
0 0
i ~-
C 30
15
13
I
.2
"§
ai
~ 20
CL
104'------'6e-----!:a,------+.1o:c----:'::12,---,---=1'-=-4---t,5·
ture. The material produced in the field may be different from that pro-
vided for mix design due to normal sampling and testing errors prior to
and during mix design.
In the development of the Marshall method, the density produced
with the Marshall manual hammer was correlated with density in the field
after traffic (22). Hence any other method of compaction (mechanical or
otherwise) must be calibrated to produce a density equal to that obtained
with the hand hammer or, better yet, should be calibrated to produce a
density equal to that obtained in the field after several years of traffic. The
procedures specified in ASTM Dl559 and MSHTO T245 for the
Marshall test require that the manual hammer be used or that the chosen
compaction method be calibrated to produce the same density as that
obtained with the manual hammer.
Suppose a mix is designed to have 4 percent voids and the field com-
paction is required to be at least 95 percent of laboratory density. This
specification results in up to 9 percent voids (4 percent at 100 percent lab-
oratory-density-plus-s-pel'tenn:noreac95 percent of laboratory density)
immediately after compaction and should result in approximately 4 per-
cent voids after several years of traffic. The initial voids (9 percent) may be
a little high, but the final voids (4 percent) should be acceptable. The high
initial voids may result in rapid oxidation resulting in an increase in crack-
Chapter 6-Equipment and Construction 375
ing and raveling.· If this mix is subjected to a high volume of traffic, then
a small rut (5 percent oflayer thickness, 0.10 inch (2.5 mm) for 2-inch
(51 mm) layer) results in the wheel paths. This 0.10-inch (2.5 mm) rut,
which is insignificant, is caused by densification under traffic.
·If a mixture is designed to have four percent air voids and is compact-
ed to a density greater than 100 percent, rutting may develop early in the
life of the pavement since the initial in-place voids are less than four per-
cent. If the laboratory compactive effort is sufficient to produce a density
similar to that obtained in the field after a few years of traffic, then past
experience has shown that the contractor cannot reasonably compact the
mix to a density higher than the laboratory density. Thus, if the density
on any project continually approaches or exceeds 100 percent of the labo-
ratory density, it is likely that the reference laboratory density is low for
various reasons and not likely that the contractor is applying excessive
compactive effort in the field.
This technique of evaluating field compaction results in good perfor-
mance for properly designed mixes if 1) laboratory samples are compacted
during construction to establish a reference density, 2) correct laboratory
compaction techniques are used for compacting laboratory samples, and 3)
a minimum compaction requirement is set to insure that in-place air voids
after field compaction are approximately 8 percent or less.
strip method require that the control strip be compacted to some mini-
mum percentage of the design laboratory density or to some minimum
percentage ofTMD. If the specifications do not require some minimum
density for the control strip, then the inspector must closely evaluate the
compaction equipment and rolling procedures to ensure that reasonable
compactive effort is applied to the asphalt mixture. Any significant
changes in the mixture during construction require that a new control strip
be constructed and evaluated.
This method of density control allows the compactibility of a mixture
to be evaluated and included as part of the specification requirements.
Monitoring the density with a nuclear gauge provides the information nec-
essary to know when maximum density for the rollers being used has been
obtained. If this maximum density is lower than desired, the size of rollers
should probably be increased or the rolling techniques improved. A large
number of factors affect density and a change in any one of these factors
can make the test strip invalid. Common factors affecting density include
aggregate gradation (especially percent passing the Number 200 sieve),
asphalt content, moisture content, mixture temperature, air temperature,
layer thickness, firmness of the underlying layer, roller weight, roller pat-
tern, roller speed, etc.
A typical control strip specification requires some minimum density
for the control strip. The remaining construction must achieve some min-
imum percent of the control strip density. This specification can be made
simpler by requiring that the compacted mix meet some percentage of lab-
oratory density or TMD. A control strip could be constructed to ensure
that the specified density is achievable and an appropriate rolling pattern is
developed. For example, assume that a specification requires that a control
strip have a density of at least 94 percent ofTMD and that all HMA placed
after the control strip have a density of at least 98 percent of the control
strip. This specification could be made simpler by requiring that the
in-place mixture be compacted to a minimum density of 92 percent of
TMD since both of these examples result in similar compaction require-
ments (0.94 X 0.98 = 0.92).
The control strip method of specifying density can be used to obtain
satisfactory results. However, the specifications should be written so that
------1c.-he-initiaLin=place-v:oids...in._th.e..HMA...do__no_t_exceed_approximately 8 per-
cent, and the in-place voids after traffic do not fall below approximately 3
percent. To ensure that the voids never fall below 3 percent, laboratory
specimens must be compacted during construction to estimate the
in-place air voids after traffic.
378 Hot Mix Asphalt Materials, Mixture Design and Construction
Measurement of Density
Core Method. The core method of measuring density is the referee
procedure, which is the standard to which the nuclear gauge is compared.
Requiring that density be r measured with cores does impose a delay
between the completion of compaction and a determination of the densi-
ty results, since the pavement has to cool before cores can be taken and
then the cores must be dried to obtain dry weight. In most cases the den-
sity results using the core method are obtained the day after construction
is completed.
After cutting the core from the pavement layer, the freshly placed
material must be separated from the underlying material attached to the
core. In some cases paper or other bond breakip.g material is placed on the
existing surface prior to placing the HMA. When this is done, the core can
be easily separated; however, the location of the paper must be carefully
marked so that the core can be taken in the proper place. There are some
problems in using paper. or other material as a bond breaker, as there is
some concern that a lower density in the asphalt mixture may be obtained
over the bond breaker.
The method most often used to obtain core samples is to identify ran-
dom sample locations, to cut a full depth core through the asphalt pave-
ment, and to saw or otherwise separate the layer being tested from the
underlying material. This method is believed to be the most suitable
method of evaluating the overall density of the pavement.
Burati and Elzoghbi (2.1) showed that the variability of density test
results was lower with core measurements than with nuclear density mea-
surements. Their study included nuclear gauges made by three different
manufacturers which were used on two constrµ_ction projects. They found
that there was a statistically significant difference in the average density
measured from cores and with the nuclear gauges.
Nuclear Gauge Method. Nuclear gauges have been used for a num-
ber of years to measure the bulk density of HMA mixtures. This technique
has the advantage that results can be determined rapidly and nondestruc-
tively.
Most nuclear density gauges measure density of HMA in the back-
scatter mode. A gauge is placed on the selected spot on top of the pave-
---ment and a reading is made:The density determmeclrrom the readmg rep-
resents the average density for the top several inches of the pavement. For
instance, the average density may be representative of the top 6 inches
(150 mm) of the pavement, howeve::r, the thickness of the layer of interest
may only be 2 inches (51 mm). One way to eliminate the effect of the
Chapter 6-Equipment and Construction 379
underlying layers is to calibrate the nuclear gauge to produce the same den-
sity as that obtained from cores cut at the same spot. This calibration cor-
rects for the density of the underlying material. Even with this calibration,
errors still exist due to variations in layer thickness and variations in densi-
ty of the underlying layers along the length of the project.
In recent years, a nuclear. gauge has been developed that measures the
density of thin lifts. This new gauge should provide greater accuracy in
measuring density as compared to the previous gauges; however, the over-
all accuracy of these gauges is not known.
The best use of nuclear gauges is in the development of rolling patterns
by quickly monitoring changes in densification of HMA materials. Some
nuclear gauges have been attached to rollers to continually measure densi-
ty during the rolling operation. Used regularly, the nuclear gauge can be a
valuable quality control tool. Because of the possibility of errors with
nuclear gauges, it is not recommended that these gauges be used alone for
acceptance testing, even with daily calibration of the gauge. Some cores
should routinely be taken to verify the accuracy of the nuclear gauge and
to ensure that an acceptable density is obtained.
MIXTURE SEGREGATION
Segregation is another significant mat deficiency problem which can
cause performance problems. Segregation occurs when the HMA materi-
al behind the laydown machine does not have a consistent gradation across
the mat, usually as a result of the coarse aggregate separating from the fine
aggregate. The surface texture of a segregated mixture is very coarse when
compared to a normal mat surface. Segregation can be caused at any point
in the construction operation between the stockpiles and the laydown
operation (Figure 6-33). Because there are so many potential sources of
segregation, it is often very difficult to identify the cause or causes of a par-
ticular case of segregation. Based on the type and location of the segrega-
tion, the potential causes can be identified and checked to determine how
to correct the problem. An excellent segregation troubleshooting guide
was prepared by Brock (25). Segregation is a significant construction prob-
lem in much of the United States. The types of segregation discussed by
Brock are:
-----------+1.----1,Random----segregati-0n;-----------
2. Each side segregation which is evenly spaced between trucks;
3. Center line segregation;
4. Each side continuous segregation; and
5. One side segregation evenly spaced between trucks.
380 Hot Mix Asphalt Materials, Mixture Design and Construction
General
Plans and specifications are used to describe to the contractor specifiy
cally how a project is to be built. The plans provide the dimensions and
layout of all aspects of the project. The specifications describe the materi-
als, workmanship, and other general requirements for the project. It is
- - - - · es~ential _that the specificatiJ::ms_ are_clear to .both the contractor..and..th<e-------
:
1
Bidding Methods
There are several approaches that have been used by owners to select a
contractor for a particular project. The three most common approaches
are low bid, best bid, and negotiated contract. Each of these methods have
some advantages and disadvantages that are discussed below.
Low Bid. The low bid method is the most common approach used by
public agencies for selecting a contractor. This method involves supplying
all the information about a project to a number of contractors and allow-
ing each contractor to submit a bid providing a detailed cost breakdown
for performing the work. It is assumed that each contractor will perform
satisfactory work since the plans and specifications describe the minimum
requirements necessary for a satisfactory project. The contractor with the
lowest bid is awarded the contract. This method encourages competition
among the contractors and results in a contract with lower initial cost. A
major problem with this method, however, is that competition is often so
keen among contractors that, in order to secure the contract, the winning
contractor may cut corners in his bid, making it difficult to perform the
construction for the contract costs. When this happens, construction qual-
ity may suffer.
Specification Development
The first step in ensuring satisfactory construction of a HMA is prepa-
ration of adequate specifications for the project. The second step is to
ensure the specification limits are met during mix design and construction.
Generally the two types of specifications used to describe asphalt pavement
construction are method and performance specifications.
ties to specify and what limits to set for these properties. For this type of
specification to be acceptable, these specified properties should be directly
or indirectly related to performance.
An end-result specification for density may state for example; "The
asphalt mixture will be divided into lot sizes of 1,000 tons (909 Mg) each
for evaluation. Each lot will be divided into 4 sublots and one random
sample will be taken from each sublot. The asphalt mixture will be cored
after compaction and the bulk density will be determined in accordance
with ASTM D2726. The average density of the four samples shall be at
least 98% of laboratory density. If the average density is less than 98%
then the mix will be accepted at a reduced price." A table for this reduced
. price is provided in the specifications. This specification does not require
that the inspector be on the project at all times; the tests can be conduct-
ed sometime after construction.
An end-result specification should clearly identify several items to be
effective (2.6). These items include:
l. Lot size;
2. Mixture properties to be evaluated;
3. Number of tests per lot;
4. The point of random sampling;
5. The sampling method;
6. Number of measurements per test result;
7. Test or measurement method;
8. Target value of measured characteristic;
9. Realistic tolerance; and
10. Action to be taken in case of noncompliance with specifications.
-----.Sampling-of-Aggi:eg.ate.-and-Calibration-of_Eeeders.
R= 1.8 WS
B
where,
= rate otfeea, tons1hour;
W = weight of sample from B feet of belt, lbs;
B = length of belt sampled, feet; and
S = speed of belt, feet/second.
388 Hot Mix Asphalt Materials, Mixture Design and Construction
The speed of the belt on which aggregate samples are taken can be easily
determined, by measuring the belt length and determining the time it takes
to make one revolution.
After the rate of feed is determined for several feeder belt speeds, the
data is plotted on a graph similar to that shown in Figure 6-34 as X, Y, Z
or Q. After calibration of the feeders, the belt speed for any aggregate
being used can be selected to feed any desired amount.
Some HMA facilities (usually drum mix) have weight sensors on the
belt between the cold feeders and the dryer. These weight sensors allow the
operator to calibrate the cold feeders from the control room. In this case,
belt samples do not have to be taken so the entire process is faster. Care
must be taken, however, to ensure that the weight sensors are correctly cal-
ibrated. This can be verified by feeding the- aggregate into a truck for a
given amount of time and determining the weight of the material collected.
Feeder Number I
"-
::,
0
-
I
( /)
C:
0 Feeder Number 2
I-
"O
..
(I)
(I)
-
LL
0 Feeder Number 3
-(I)
0
c:::
Feeder Number 4
Remember that
R= 1.8 WS
B
The rate of feed for the 20% belt speed is,
These three measured tonnage rates should be plotted along with the
corresponding belt speed. The relationship between belt speed and tons/
hour are approximately straight lines. After being plotted, these graphs can
be used to select the belt speed for any desired tons/hour of aggregate.
390 Hot Mix Asphalt Materials, Mixture Design and Construction
Sampling HMA
Samples of HMA are normally taken either from the loaded truck or
the roadway behind the paver. Because some segregation of the asphalt
mixture occurs when the truck is loaded, care must be exercised when col-
lecting samples. The coarse aggregate coated with asphalt cement tends to
• roll down the side of the mound of HMA mixture and congregate next to
the sides and the ends of the truck bed. Though this problem always
occurs to some extent, as long as it is minimized, problems during laydown
are not noticeable. This type of segregation affects test results if the sam-
ple is not properly taken. A segregated sample containing a high percent-
age of coarse aggregate has a low percentage of asphalt cement due to the
low surface area. Normally several small samples are taken at various
points along a truck load of HMA. These samples are usually taken just
beneath the surface at the top and on the sides of the truck load of HMA.
These small samples are combined to produced one large sample used for
testing.
When HMA is sampled on the roadway, care must be taken to ensure
that contamination with tack coat material is minimized. The roadway
sample can be representative (meaning several small samples) or random
(sample from one location) depending on what is being evaluated by the
test result. In most cases, random samples are required so that variability
of the construction process can be evaluated.
STATISTICAL .CONCEPTS
Variations in test results on paving materials and mixtures depend on
variability of materials, equipment, and testing procedures. Well con-
trolled projects provide mixture properties with low variability while poor-
ly controlled projects result in higher variability in mixture properties.
Chapter 6-Equipment and Construction 391
•1
E (X; - 'i)2
Variance= s2 = -'=-1 - - - -
•1
E <xi - x)2
i=1
Standard Deviation = a = {s2 =
where,
result for the ith sample;
x 1 = test
x = average of measured test results;
n1 = number of test results;
n,_= number of samples that are averaged to y1eld-a~single test
result.
>..
u
C:
Q)
:::,
0-
Q)
:....
LL
Asphalt Content
Figure 6-35. Histogram of Distribution of Asphalt Content
a = .[s2 = 1.13
a 1.13
CV = -= x 100 = - - x 100 = 1.21 percent
X 93.0
Now, what if the specification requires that all density test results exceed 91
percent ofTMD? Using the estimated average density of the lot as 93.0
and the estimated standard deviation as 1.13, what is the probability that
a test result from a random sample is outside the specification require-
ments?
To analyze the problem, the data is reduced to a standard normal devi-
ate in order to use the standard probability distribution tables for a normal
distribution:
where,
Ka = standard normal deviate;
y = random variable;
x = estimate of the population average; and
cr = estimate of the population standard deviation.
So for the problem being evaluated, x = 93.0 and cr = 1.13. For this prob-
lem the lower specification limit of 91 is the random variable y. Then sub-
stituting x and cr into the equation for Ka,
Ka and x are identified for the standard distribution shown in Figure 6-36.
394 Hot Mix Asphalt Materials, Mixture Design and Construction
From the statistical Table 6-5 it can be seen that for Ka= I.77 the nor-
mal distribution function has .9616 (96.2 percent) of the area below x +
1.77 and 3.8 percent of the area above x + 1.77 since the total area under
the normal distribution function is 1.00. Due to symmetry then x - 1.77
p.as 3.8 percent of the data below that point. Given the estimated average
and standard deviation of the population, it would be expected that
approximately 4 percent of the individual test results would fall below 91
percent or fall outside the lower specification limit.
x= 93 percent of TMD
Ka = -1.77 Ka = 1.77
91 percent of TMD 95 percent of TMD
CT=cr/Tu
cr = standard error of the mean;
cr = population standard deviation; and
n = number of individual samples which are included in
each average.
Assume for instance that four test results are averaged to produce one
reported mean test result. Then
Ka
Ka .00 .01 .02 .03 .04 .05 .06 .07 .08 .09
o.o .5000 .4960 .4920 .4880 .4840 .4801 A761 .4721 .4681 A641
0.1 .4602 .4562 .4522 .4483 .4443 .4404 .4364 .4325 .4286 A247
0.2 .4207 .4168 .4129 .4090 .4052 .4013 .3974 .3936 .3897 .3859
0.3 .3821 .3783 .3745 .3707 .3669 .3632 .3594 .3557 .3520. .3483
0.4 .3446 .3409 .3372 .3336 .3300 .3264 .3228 .3192 .3156 .3121
0.5 .3085 .3050 .3015 .2981 .2946 .2912 ..2an .2843 .2810 .2776
0.6 .2743 .2709 .2676 .2643 .2611 .2578 .2546 .2514 .2483 .2451
0.7 .2420 .2389 .2358 .2327 .2296 .2266 .2236 .2266 .2177 .2148
0.8 .2119 ;2090 .2061 .2033 .2005 .19n .1949 .1922 .1894 .1867
0.9 .1841 .1814 .1788 .1762 .1736 .1711 .1685 .1660 .1635 .1611
1.0 .1587 .1562 .1539 .1515 .1492 .1469 .1446 .1423 .1401 .1379
1.1 .1357 .1335 .1314 .1292 .1271 .1251 .1230 .1210 .1190 .1170
1.2 .1151 .1131 .1112 .1093 .1075 .1056 .1038 .1020 .1003 .0985.
1.3 .0968 .0951 .0934 .0918 .0901 .0885 .0869 .0853 .0838 .0823
1.4 .0808 .0793 .ona .0764 .0749 .0735 .0721 .0708 .0694 ~1
1.5 .0668 .0655 .0643 .0630 .0618 .0606 .0594 .0582 .0571 .0559
1.6 .0548 .0537 .0526 .0516 .0505 .0495 .0485 .0475 .0465 ~
1.7 .0446 .0436 .0427 .0418 .0409 .0401 .0392 ~ .0375 ~7
1.8 .0359 .0351 .0344 .0336 .0329 .0322 ~14 .0307 .0301 .0294
1.9 .0287 .0281 .0274 .0268 .0262 .0256 .0250 .0244 m39 .0233
Ka .o .1 .2 .3 .4 .5 .6 .7 .8 .9
· ♦ From Toblu tfArtlM ilt Two Toil.r/J/ld ilt OM Toil tftM Nonnol.C,,,.,., by Ftoderict B. CroxlOD. O>pyrighl 1949 by Prmlice-Hall,
Inc.
396 Hot Mix Asphalt Materials, Mixture Design and Construction
From the statistical table it can be seen that for Ka= 3.52 the area of
the curve below x + Ka is 0.9998, this means that 99.98% of the data falls
below x + 3.52 and only 0.02% is above x + 3.52 or below x - 3.52. By
averaging 4 samples for each data point, the percent of material expected
to be .below the specification requirements is reduced from approximately
3.8 percent to approximately 0.02 percent.
Past experience often provides a good indication of the expected value of
cr for various mixture properties for a typical project. However, cr is unique
for each project and cannot be accurately assumed for any project. After suf-
ficient data is generated for a project, cr can be computed, the variability of
test results can be determined, and a judgement can be made as to whether
good control occurs (cr is low) or poor control occurs (cr is high).
A large amount of data has been collected to evaluate the typical vari-
ability of pavement properties (26). Typical values for asphalt content and
density are shown below:
Other work has shown that the standard deviation for asphalt content can
be as high as 0.30.
Standard deviation values for aggregate gradation have been shown to
be related to percent passing a particular sieve size. Typical results are
shown in Figure 6-37. Now consider a practical problem encountered on
many paving projects. HMA is being produced and tested by both the
owner and the contractor. The asphalt content is being measured using the
solvent extraction test. .Because of high aggregate absorption, some of the
asphalt cement is trapped in the pores of the aggregates and cannot be dis-
solved by the solvent, therefore it appears that the asphalt content is low.
The contractor's tests show more asphalt cement removed by the solvent
than does the owner's tests; hence, t:F-o_p_o_p_ulations....oLdata...exisL(.Eigun,___ _ _ __
6-38). Neither the owner's nor the contractor's data represents the actual
asphalt content since both sets of tests leave some amount of asphalt
cement in the aggregate. The bias in the test results is defined as the dif-
ference between the true value of the property and the actual test result
Chapter 6-Equipment and Construction 397
5
z
0
i= 4
<(
>
W3
---r--..--
0
~- -- -7-
EXTRACTION-
_--:;:
~2
<(
0
z1
<(
~~
.,,,-r- HOT BINS-
-
r- ...._
....
I-
C/) 0 1/
0 10 20 30 40 50 60 70 80 90 100
PERCENT AGGREGATE PASSING
Figure 6-37. Typical Standard Deviation Values for Aggregate
Gradation (26)
>,,
(..)
C:
Q)
::,
er
Q)
:,...
LL
% Asphalt
Figure 6-38. Comparison of Owner's Test Results, Contractors Test
Results, and True Value
398 Hot Mix Asphalt Materials, Mixture Design and Construction
observed. The bias from the owner's tests is different from the bias from
the contractor's tests. The primary objective in testing is to measure the
true property by reducing the bias to approximately zero and to minimize
the variability associated with the testing process. If both the owner and
contractor average a number of samples for each data point, the variability
of the data points is reduced; however, the bias is not affected. The bias
can only be reduced by removing more of the asphalt cement from the
aggregate during· the extraction test.
The third location for sampling aggregate is in the HMA batch facili-
ty hot bins, (drum mix facilities do not have aggregate hot bins). Causes
for variability here include improper gradation fed from cold feeder, errat-
ic feed from dust collector system, changing production rate (screening
efficiency changes with production rate), blinding screens, holes in screens.
or bin walls, and sampling and testing errors. The hot bins, if operated
correctly, will partially correct for gradation fluctuations coming into the
plant.
The fourth location for determining gradation is from the asphalt mix-
ture. The sample is normally taken from loaded· trucks but can be taken
' of the asphalt paver. This test, which is performed on the
from the back
finished product, must be controlled because it is the one on which accep-
tance of the mixture is normally based. Variability of gradation at this
point (for a batch plant) could involve incorrect hot bin gradations, incor-
rect percentage of material from each hot bin, segregation of mixture (in
storage silo or in truck), and sampling and testing errors. For a drum mix
facility, the causes of variability at this point include improper cold feed
gradation, erratic feed from the dust collector, segregation of aggregate
travelling through the plant or in the storage silo, and sampling and test-
ing errors.
Evaluation of the gradation at several locations allows the engineer to
troubleshoot the gradation problem and quickly identify the location
where it is occurring. For instance, if the stockpile gradation is satisfacto-
ry but the cold feed gradation changes, then the problem areas are likely to
be segregation of mixture, improper loading of cold feed bins, or sampling
and testing errors. These items can be quickly checked and modifications
made to correct the problem.
to the asphalt mixture to dissolve the asphalt cement. The asphalt cement
and solvent are then passed through a piece of filter paper, but the aggre-
gate is not allowed to pass. This is not a highly accurate test but it is wide-
ly used for measuring asphalt content. One advantage of the extraction
test is that it allows determination of the aggregate gradation of the mix-
ture. A disadvantage of the extraction test is that the solvent used is haz-
ardous and is difficult to dispose. Biodegradable solvents are being used by
some states but additional work is needed prior to their widespread accep-
tance. The current asphalt cement metering devices, such as variable speed
pumps, give accurate control of asphalt content. These devices should be
calibrated by the HMA producer on a regular basis.
The National Center for Asphalt Technology (NCAT) has developed
a test method to determine the asphalt content of the HMA mixtures by
ignition. The test method is based on research started in 1990 at NCAT
(22). In the NCAT ignition method, a sample of HMA mixture is sub-
jected to an elevated temperature of 538°C (1000°F) in a furnace to ignite
and burn i:he asphalt content from the aggregate. NCAT's work has result-
ed in a test procedure and equipment that automatically measures the
asphalt content in 30-40 minutes. The grading of the aggregate can then
be determined using standard sieve analysis. Based on round robin stud-
ies conducted by NCAT in which 12 laboratories participated, the accura-
cy and precision of the NCAT ignition test was found to be better than
those of the solvent extraction method. Therefore, this test method is
increasingly replacing solvent extraction methods, which are being elimi-
nated due to growing health and environmental concerns associated with
the use of chlorinated solvents.
Improper asphalt content can be caused by several factors in a batch
HMA facility. These causes include inaccurate aggregate scales, inaccurate
asphalt cement scales, leaking valve in asphalt cement pot, segregation, and
sampling and testing errors. Causes of incorrect asphalt content in a drum
mix facility include inaccurate aggregate belt scales, improperly calibrated
asphalt cement meter, incorrect moisture content correction for aggregates,
segregation, and sampling and testing errors. Properly sampling the
asphalt mixture from a truck load of material is difficult. When a sample
------of'mixture is obtained wiililiigh coarse aggregate content aue to segrega-
tion after mixing, the measured asphalt content will be lower than speci-
fied since the coarse aggregate has a lower surface area than the fine aggre-
gate.
402 Hot Mix Asphalt Materials, Mixture Design and Construction
Test Frequency
Stockpile gradation 1 per day
Cold feed gradation 1 per day
Hot bin gradation (if applicable) 1 per day
Extracted asphalt content and gradation 2 per day
Laboratory compacted samples
voids, stability, flow 2 sets per day
Theoretical maximum density 2 per day
Temperature Regularly throughout day
Visual inspection Regularly throughout day
In-place density 6-10 per day
Smoothness Regularly throughout day
Elevations After completion of lot
with control charts. Control charts are simple methods of graphically dis-
playing the QC data as it is developed. The two most common types of
control charts are plots of individual test results and plots of running aver-
ages of the test results.
Example 6-5. Assume that the results in Table 6-6 are obtained for
the density of a HMA mixture. The specifications indicate that the densi-
ty must be at least 92 percent and no more than 96 percent of theoretical
maximum density for the running average of four samples. Plot control
charts showing the data and specification limits as shown in Figure 6-39.
These control charts show no obvious trends nor data outside the spec-
ification limits on the running average of four plots. For proper evaluation
of the process being used to produce the mixture, the engineer must look
for trends moving upward or downward and for test results outside the
specification limits. Identification of these trends during construction
allows the problem to be solved before it becomes significant. Figure 6-39
does show some slight trends which are likely due to normal material vari-
ation. Test numbers 10 and 19 show that the running average is approach-
ing the lower limit. When this happens the contractor should probably
-----t·ak.e some action to increase: dre density.
Density,
Density, Percent of 4-point
Test No. Theoretical Maximum Running
Density Average
1 93.0
2 94.5
3 92.1
4 95.1 93.J
5 93.3 93.8
~-
6 94.2 93.7
7 93.5 94.0
8 91.6 93.2
9 92.5 93.0
10 93.7 92.8
11 94.2 93.0
12 92.2 93.2
13 91.5 92.9
14 95.6 93.4
15 93.6 93.2
16 94.1 93.7
17 92.3 93.9
18 91.7 92.8
19 92.1 92.4
20 94.3 92.5
21 96.0 93.4
22 92.2 93.6
"
23, 91.3 93.4
24 93.5 93.2
Chapter 6-Equipment and Construction 407
~
...
0
0:::
...
Q)
">C
Q)
-( /)
C
(l)
I
I
Test No.
Figure 6-40. Effect of Process Changes on Density Test Results
which that change occurred allows the engineer to determine if the change
affects process control as shown on Figure 6-40. The increase in roller
weight appeared to increase the density of the mixture. Changing roller
operators did not appear to affect the average density, but the variability
between test results appears to be higher for the new operator.
Another plot sometimes used to analyze variability of test results is the
range plot shown in Figure 6-41. As the range values increase or decrease,
the variability changes in a similar way.
-
( /)
C
--
..J
Q)
O'\ Cl)
(/)
C :::,
Cc
0::>
~¢
(/)
C
Cl)
-------,O--t:======::::::::::::::::::::::::::::::::::::::::::::::::::::::;;;;;;;;:;;;;;;;;:~~-
T est No.
Figure 6-41. Density Range Plot Sh~wing Range of Last Four Data
Points
408 Hot Mix Asphalt Materials, Mixture Design and Construction
The properties that are typically plotted on control charts for HMA
construction are: asphalt content, aggregate gradation (for selected sieve
sizes such as 3/s-inch, No. 8, No. 50, and No. 200), in-place density, voids
in laboratory compacted samples, theoretical maximum density, laborato-
ry density, stability, and Bow Other properties not mentioned above may
be plotted as needed.
REFERENCES
1. Brock, J .D. Dryer Drum Mixer. Astec Industries, Chattanooga, TN,
undated.
2. Scherocman, J.A., T.W _Kennedy, M. Tahmorelsi, R. Holmgreen.
Construction of Asphalt Concrete,-- Pavement. Center for
Transportation Research, The University of Texas at Austin,
September 1986.
3. Brock, J.D. and E.G. Mize. The Drum Mix Process. Astec Industries,
Technical Bulletin NumberT-110, Chattanooga, TN, 1984.
4. Fugitive Dust Control and the Hot Mix Plant. National Asphalt
Pavement Association, NAPA Report IS-73, 1980.
5. Linden, et al. Effect of Compaction on Asphalt Concrete Perfor-
mance. Transportation Research Board, Transportation Research
Record 1217, 1989.
6. Nijboer, L.W Plastici-ty as a Factor in the Design ofDense Bituminous
Road Carpets. Elsevier Publishing Co., Inc., New York, N.Y., 1948.
7. Principles of Construction of Hot Mix Asphalt Pavements. The
Asphalt Institute, Manual Series No. 22 (MS-22), January 1983.
8. Compaction Handbook. 6th Edition, Hyster Company,
Construction Equipment Division, Kewanee, IL, October 1986.
9. Tunnicliff, D.G., R. W Beatty, and E.H. Holt. A History of Plants,
Equipment and Methods in Bituminous Paving. Proceedings,
Association of Asphalt Paving Technologists, Volume 43A, 1974.
10. Geller, Myron. Compaction Equipment for Asphalt Mixes. Placement
and Compaction ofAsphalt Mixtures, American Society for Testing and
1
1
Materials, Special Technical Publication 829, ET. Wagner, Ed., 1984.
:1,li, 11. McLaughlin, A.L. Field Compaction of Bituminous Mixes for Airport
Pav~ments. Federal Aviation Administration, Report No .
..i,1.----..L.o.=IOUJ.=7-7-=42.,..ApriLL9-7-t-.------------------
1
'
able and provide rapid drainage for subsurface water. However, to be effec-
tive, the layer must be either daylighted or connected to edge or tile drains.
Their cost is significantly lower than dense-graded HMA mixes primarily
because of the lower asphalt content. Two or three roller passes are enough
to seat the aggregate and compact the mix. Excessive rolling tends to break
the aggregate. "When this type of mix is placed between a concrete pave-
ment and a dense-graded HMA overlay, it minimizes reflection cracking.
The use of "big rock mixes" has been reported in Tennessee and
Indiana (1). These mixes have supported heavy truck traffic without dis-
tortion. The resistance to deformation comes from relatively large aggre-
gate size and good interlock of aggregate particles.
ATPM has been used by several states to improve the subsurface
drainage of both flexible and concrete pavements. Cedergren and Lovering
(3) recommended a two-layer drainage blanket for flexible pavement sys-
tems consisting of subbase and ATPM. In the case of a two-layer system,
the water being drained can pass upward through the underlying'-'p_e_r_m_e_-_ _ _ _ __
---a-,ble subbase layer mto the ATPM ana be quickly drained from any spot
underneath the surface of the roadway. However, in a single layer system
consisting of sub base material only, any water entering the sub base must be
carried all the way to the drainage outlet in that material. Thus, the effec-
tive drainage path for the two-layer ·system is the thickness of the subbase,
Chapter 7-Special Mixtures, Recycling, and Additives 413
and the drainage path for the one-layer system is half the width of the
roadway, assuming the water is removed from the roadway at the two
shoulders by daylighting or edge drains. The problem is compounded
when the roadway has multiple lanes and is very wide.
Permeability of ATPM generally exceeds 10,000 feet (3,000 m) per
day (.1). It has been used successfully to drain water from badly pumping
concrete pavements when used as a drainage layer between the concrete
pavement and dense-graded HMA overlay. However, the concrete pave-
ment needs to be cracked and seated prior to placing the ATPM if water is
to be drained from underneath the concrete.
Asphalt content is estimated from the surface capacity CK) -of the
aggregate retained on a No. 4 (4.75 mm) sieve. Kc is determined from the
percent of SAE No. IO oil retained by the aggregate which represents the
total effect of superficial area, the aggregate's absorptive properties, and sur-
face roughness. Void capacity of the coarse aggregate is determined by
using a vibratory compaction apparatus. The optimum amount of fine
aggregate is that which fills the voids between the coarse aggregate particles
without bulking the coarse aggregate. This aggregate mixture must provide
enough room for the required asphalt content and interconnected voids
needed for drainage. The FHWA procedure establishes optimum mixing
temperature by running asphalt drainage tests at different temperatures.
The use of rubberized asphalt as the binder results in a mixing tem-
perature that is generally higher than that for conventional asphalt mix-
tures. A mixing temperature that is too high allows the asphalt binder to
drain through the aggregate, resulting in a high variation of asphalt con-
tent throughout the mix and bleeding spots in the pavement. If the mix-
ing temperature is too low, the aggregate will not be completely coated.
The asphalt drainage problem limits the amount of storage time of the
OGFC in a storage silo. A long storage time results in excessive drainage
of asphalt through the mixture. When the OGFC is stored in a silo even
for short periods of time, it is essential that the mix be observed to ensure
that noticeable drainage does not occur.
The asphalt may tend to drain through the aggregate while being
hauled from the mixing facility to the laydown site. If this occurs, asphalt
may leak around the tailgate while the truck is waiting or may form a pud-
dle in the bottom of the truck and hence cause a rich spot on the pavement
underneath the truck when unloading. This asphalt on the surface of the
OGFC results in a slick spot.
When the OGFC arrives at the laydown site, it tends to look rich due
to the greater film thickness and the lack of fine materials. This rich look
is normal. Typically the OGFC is placed appro~imately 3/4 inch (19 mm)
thick using an aggregate of 1/2 inch (12.5 mm) maximum size.
An OGFC should not be used over an existing pavement surface that
is uneven. An existing pavement that is uneven prior to overlay with an
OGFC should be levelled with a dense-graded binder course prior to over-
laying with the OGFC. Another problem that occurs when overlaying
___uneven-pa.v.em~-nts-is-puU-i-ng-aflcl-t:ea:ri-ng-wherrdre-a-sph:alt paver places ffi'1:-"""e- - - - - -
OGFC thinner than approximately 3/4 inch (1,9 mm).
The OGFC should not be laid in cold weather. Several failures have
been attributed to loss of bond between the OGFC and an existing surface
because of cold-weather laydown. Most states specify a minimum air tern-
Chapter ?-Special Mixtures, Recycling, and Additives 415
perature in the range of 60 to 70°F (16 to 21 °C) when OGFC mixes are
placed.
There is no density specification for OGFC mixtures. The normal
procedure for rolling is to make 2-3 passes over the surface with a steel
wheel roller. Vibratory and rubber-tired rollers should not be used. The
vibratory roller degrades the aggregate and the rubber-tired roller tends to
pick up the asphalt, especially when a rubberized asphalt is used. Another
problem that occurs with the rubber-tired roller is excessive closing of the
voids due to the kneading action of the rubber tires. Low voids, if present,
do not allow good drainage of water .across and through the surface.
The quality control procedures for an OGFC pavement are essential-
ly the same as that for dense-graded HMA. The primary difference is that
there are no requirements for laboratory or field compaction of the mix-
ture. The required tests generally include asphalt content, aggregate gra-
dation, thickness, and smoothness. These tests are conducted routinely to
ensure that a satisfactory mix is produced.
Proper placement of an OGFC is an art and hence is difficult ro spec-
ify. The quality control process must include close inspection of the mix-
ture by a qualified inspector during production and placement of the mix-
ture. Deficiencies to look for include drainage of the asphalt, pulling and
tearing behind the paver, and improper rolling procedures.
An OGFC surface has been shown to improve skid resistance, espe-
cially in wet weather. However, there has been one problem with perfor-
mance that has reduced its use in some states. A number of states have
reported that stripping occurs in the layer directly underneath the OGFC.
Many states believe that they have solved this problem by using antistrip-
ping agents in the underlying layers, and they continue to use OGFC sur-
faces on a regular basis. Some states (.2) have reported that the use of sand
or cinder during snow-and-ice-control operations tends to fill the surface
voids of the OGFC, preventing proper water drainage.
tant to studded tire wear than any other type of hot mix asphalt. Since that
time most European countries have disallowed the use of studded tires but
SMA has continued to be used because of other desired properties. The
primary reasons for using SMA mixtures in Europe today are its resistance
to rutting and its increased durability.
Just as in Europe, the primary reasons for using SMA in the United
States have been its improved resistance to rutting and its improved dura-
bility. The macrotexture of the mixture has also been shown to improve
surface drainage over that of dense-graded mixtures and to increase fric-
tion during wet weather.
Today SMA is a standard mixture type in several states. As its use has
increased, the price per ton of mix has decreased substantially to a point
slightly above (typically 20-25%) that for conventional HMA.
Description
Stone Matrix Asphalt is a mixture of crushed coarse aggregate, crushed
fine aggregate, mineral filler, asphalt cement, and stabilizing agent. The
stabilizing agent is used to prevent draindown of the asphalt cement and
typically consists of fibers and/ or polymers.
The SMA mixtures are designed to have a high coarse aggregate con-
tent (typically 70-80%) a high asphalt content (typically over 6%) and
high filler content (approximately 10% by weight). Using high coarse
aggregate content results in stone on stone contact that produces a mixture
that is highly resistant to rutting. A view of a typical SMA mixture and for
comparison a typical dense-graded mix is shown in Figure 7-1. Notice the
STONE-ON-STONE CONTACT
high stone content for the SMA mixture. Also notice that the coarse aggre-
gate in the dense-graded mixture appears to be floating in the fine aggre-
gate matrix. Typically the coarse aggregate in the SMA mixture carries the
load while the fine aggregate in a dense-graded mixture must carry the
load. The SMA mixture is more resistant to rutting since coarse aggregate
can develop more shear strength than fine aggregate.
The SMA mixture is designed, produced, and placed using methods
similar to those for dense mixtures. There are some differences in SMA
and dense-graded mixtures that are discussed in the following paragraphs.
Materials
As stated earlier, the materials used to produce SMA include aggregate,
filler, asphalt cement, and stabilizer. Most SMA projects that have been
constructed have used all crushed aggregate to provide the maximum resis-
tance to rutting. Typical specifications (2) require 100% of the coarse
aggregate to have at least one fractured face and at least 90% with 2 or
more fractured faces. This is a premium asphalt mixture and therefore
high quality materials must be used for best performance.
The Los Angeles abrasion value required by some states for SMA is
lower than that required for dense-graded mixtures. Since the SMA mix-
ture has stone on stone contact there is more stress under traffic loads on
the coarse aggregate particles and many 'engineers believe that because of
this higher stress the L.A. Abrasion requirements should be lower. A typi-
cal L.A. abrasion requirement that has been used for aggregate for SMA is
30. However, due to a lack of unlimited quantities of high quality aggre-
gates, many states allow the L. A. Abrasion requirement to be higher.
The typical gradation requirement for SMA mixture is shown in Table
7...:2 (2). Notice that this mixture has a relative low percentage passing the
4.75 mm sieves (20-28%) and a relative high percentage passing the 0.075
mm sieve (8-10%). The low percentage passing the 4.75 mm sieve is nec-
essary to provide a mixture with stone on stone contact and to 'Provide a
mixture meeting the minimum VMA requirements (typically 17%). The
high percentage passing the 0.075 mm sieve is necessary to adequately stiff-
en the binder so that the mixture is rut resistant and to ensure that drain-
down of the asphalt cement does not occur during construction.
- - - - - ~ r h : e h'i:gh-p-ercentagro·f-rrrare·fra:l-p·assirrg-the-Et0'15-mrn:·sieve requires
that a commercial filler be added to the mixture. In dense-graded mix-
tures it is common to have up to 6 percent material passing the 0.075 mm
sieve coming from the fine aggregate source. The SMA mixtures however
have a lower percentage of fine aggregate resulting in a lower percentage of
418 Hot Mix Asphalt Materials, Mixture Design and Construction
'
Table 7-2. Typical Gradation Requi,rements for SMA Mixture (.2)
19.0 mm 100
12.5 mm 85-95
9.5mm 75 maximum
4:75 mm 20-28
2.36mm 16-24
0.60 mm ""·-
12-16
0.30 mm 12-15
0.075 mm 8-10
Mix Design
The mix design procedure for SMA mixtures is similar to that for
dense-graded mixtures but there are some significant differences (1.Q). The
Marshall and Superpave design methods have been used to design SMA
mixtures. The criteria for both methods have been set to ensure coarse
stone on coarse stone contact and to ensure that high asphalt contents are
selected.
One step in the mix design process is to ensure that stone on stone
contact exists in the SMA mixture. One way to measure this is to place
only the coarse aggregate fraction into a mold and rod it to determine the
dry rodded unit weight. This represents the density of the coarse aggregate
for stone on stone contact. Certainly, if the density of the coarse aggregate
portion of the SMA is compacted to a higher density than the dry rodded
density, stone on stone contact would be expected to exist in the SMA.
This 1s evaluated by measuring the voicl.sinthecoarse aggregate (VCA) in
the SMA mixture by the following formula: ·
VCA = (100 (1
420 Hot Mix Asphalt Materials, Mixture Design and Construction
where
Gmb = Bulk specific gravity of coarse aggregate
Pca = Percentage of coarse aggregate by weight in total
mix
G,6 = Bulk specific gravity of total aggregate
Since VCA is inversely proportional to density, a lower VCA is need-
ed in the S~ mixture than in the compacted coarse aggregate only for
stone on stone contact.
When the VCA in the mixture is greater than that in the coarse aggre-
gate only compacted sample, the VCA can be decreased by reducing the
amount of fine aggregate. As the amount of fine aggregate is decreased, the
VMA will normally increase for SMA mixtures and the VCA will decrease
(Figures 7-2 and 7-3).
20
19
18
17
- 16
~
< 15
::.
> 14
13
12
11
IO
15 20 25 30 40 so
Percent Passing 4.75 mm Sieve(%)
60
55
50
45
- 40
~
~ 35
u
> 30
25
20
15
IO
15 20 25 30 40 50
Percent Passing 4.7S mm Sieve (o/o)
For the SMA mix design, the fine aggregate content is varied to deter-
mine which aggregate structure will provide adequate VMA and stone. on
stone contact. In other words, the percent passing the 4.75 mm sieve will
be reduced (beginning on the high side of the specification band) within
the allowable specification range until the minimum VMA and maximum
VCA requirements are met.
Work with SMA mixtures has.shown that for most aggregates stone on
stone contact begins to occur at approximately 30 percent passing the 4.75
mm sieve. This is the primary reason that the high side of the specifica-
tions for aggregate gradation for SMA mixtures are usually set at approxi-
mately 30 percent passing the 4. 75 mm sieve.
To ensure high optimum asphalt content in an SMA mixture the min-
imum VMA is set relatively high. A typical minimum requirement for
VMA is 17. The VMA typically begins to increase rapidly as the percent-
age passing the 4.75 mm sieve falls below approximately 30 percent. If a
VMA problem exists, it can generally be solved by reducing the percentage
passing the 4.75 mm sieve a few percentage points. Some aggregates tend
to breakdown excessively resulting in low VMA values.
Aftenhe gradation has been selected to provide stone on stone contact
and minimum VMA requirements, the asphalt content is adjusted to pro-
vide the desired air void level. SMA mixtures have been designed to have
air voids somewhere between 3 and 4 percent with most mixtures designed
closer to 4 percent air voids. Close control must be exercised when pro-
ducing SMA mixtures at lower air void content to ensure that bleeding
does not occur after the mix is in-place. The air void content of laborato-
ry compacted samples must be controlled during construction to ensure
that the air voids stay within the desired range.
SMA mixtures have been designed using the Marshall and Superpave
compaction procedures. When the Marshall compaction is used, 50-blow
compactive effort has been utilized. Additional compaction does not
increase density significantly but does tend to break down the aggregate
excessively. Some SMA mixtures have been designed with the Superpave
gyratory compactor. Although the number of gyratory revolutions have
not been established for SMA, mixes have been designed at 100 revolutions
for Ndesign·
After the mixture has been designed a draindown test should be con-
------.-1---U€t-ed--1:0-ens1:1Je-~ha-t-El:Fa:i-a-a-0-wn-is-n0-t-a-prebl~m-(-H).~A-simple test has
been developed to measure the draindown potential. The test requires that
a sample of SMA mixture be placed into a wire basket and placed into an
oven at a specified temperature (typically the mixing temperature) for a
specified amount of time. The asphalt binder that drains through the bas-
422 Hot Mix Asphalt Materials, Mixture Design and Construction
ket is caught and weighed. The amount of drainage from the loose SMA
mixture should not exceed some specified requirement typically 0.3 per-
cent for I hour. A photo of the basket is shown in Figure 7-4. If exces-
sive draindown occurs, the mix should be modified, usually by increasing
the amount of polymer or fiber.
Construction
The use of SMA mixtures requires some changes at the HMA plant
(12). A high amount of commercial filler, usually 3-5 percent, is fed into
the plant. Most plants are not set up to handle this high amount of filler,
and this can be a bottleneck in production capacity. The preferred method
of addition is with a silo where the filler is accurately metered into the mix-
ture. Some projects have fed the filler though one of the aggregate feeders,
but this approach is undesirable.
When a polymer is used, it is normally blended with the asphalt
cement prior to delivery to the HMA plant. However, in some cases the
polymer has been added at the plant. In this case, the polymers may be
blended with the asphalt cement and aggregate at the same time. One
potential problem with pre-blending a polymer is the potential for separa-
tion during handling and storage. The primary problems often encoun-
tered when blending at the plant are lack of proper facilities, inadequate
1 - - - - ~ lenciiiig;-aii.a-ina6ificy-to test mod1fielasphafr prior to iii:troduction into
_ _ _ __
the mixture.
When fibers are used, their introduction into the mixture must be
closely controlled. Even though the fiber content is small, it has a signifi-
Chapter 7-Special Mixtures, Recycling, and Additives 423
cant effect on the mixture properties. The fibers can be weighed and added
in batches in a batch plant. The fibers are normally blown into the drum
at the desired rate in a drum mix plant.
The SMA mixtures are typically stiffer than conventional dense-grad-
ed mixtures and may require higher temperatures and more mixing time at
the plant. The primary reasons for the increased stiffness are the high filler
content, modified asphalt, and/or fibers.
Placement and compaction procedures are similar to that for
dense-graded mixtures. SMA mixtures are typically stiff and more difficult
to work. Because of this stiffness and large coarse aggregate content, con-
struction of good longitudinal joints is more difficult. However, experience
has shown that a good joint can be built. Draindown of the asphalt cement
may occur if the mixture is not properly designed or if the mixture is not
properly controlled, resulting in fat spots in the compacted pavement.
Rollers must be kept immediately behind the paver for good com-
paction. As the SMA mixture cools, it tends to set up quickly and experi-
ence has shown that compaction becomes very difficult once the mix
begins to cool. Rubber tire rollers are not used on SMA primarily due to
the mix sticking to the tires. Static and vibratory steel wheel rollers have
been successfully used. Under some situations, the vibratory rollers may
tend to excessively break down the aggregate. If this is a problem, only sta-
tic rolling should be used. SMA mixtures are typically compacted in-place
to an air void content of 5-6 percent.
Performance
Since the first SMA projects were constructed in the U.S. in 1991, the
performance history is insufficient to make final conclusions. However,
the Europeans have used SMA mixtures since the 1960's and have found
them to provide excellent performance.
SMA mixtures have been shown to provide excellent resistance to rut-
ting on high traffic volume roads. SMA mixture has also been shown to
be less susceptible to cracking. It appears that SMA mixtures have the
potential to improve stability and at the same time increase the durability
of HMA mixtures.
-----------------------~------
DENSE-GRADED LARGE STONE MIXES
Premature rutting of heavy duty asphalt pavements has been a signifi-
cant problem in recent years. This rutting problem has primarily resulted
from higher pressure truck tires and increased wheel loads. The design of
424 Hot Mix Asphalt Materials, Mixture Design and Construction
HMA mixtures, which served reasonably well in the past, needs to be re-
examined to withstand the increased stresses. Various asphalt additives are
being promoted to. i~trease the stability of HMA pavements at high tem-
peratures. However, most asphalt technologists believe that fundamental
properties of the aggregate component of the HMA (such as size, shape,
texture, and gradation) ~e inost important to ensure that a rut resistant
mixture is obtained. There is a general agreement that the use of large size
stone in the binder and base courses will minimize or eliminate the rutting
of heavy duty asphalt pavements.
The use of large stone mixes is not new. As reported by Davis (13),
Warren Brothers Company had a patent issued in 1903 which specified a
top size aggregate of three inches (76 mm). Most paving companies start-
ed to use smaller stone mixes at that time 'i:o avoid infringement of the
patent. Although the patent expired, such use is still prevalent today.
Marshall mix design procedures are used by 76 percent of the states in
the United States according to a survey conducted in 1984 (1§). The
equipment specified in the Marshall procedure (ASTM Dl559) consists of
a 4-inch (100 mm) diameter compaction mold, which is intended for mix-
tures containing aggregate up to I-inch (25 mm) maximum size only.
This has also inhibited the use of HMA containing aggregate larger than
one inch because this mixture cannot be tested by the standard Marshall
mix design procedures. There are other test procedures, such as gyratory
compaction, TRRL refusal test, and the Minnesota DOT vibrating ham-
mer, which use 6-inch (150 mm) diameter molds accommodating I½ to
2-inch (38 to 51 mm) maximum aggregate size (15). However, most
,agencies are reluctant to buy new equipment because of cost and/or com-
plexity. They prefer to utilize the existing equipment and/ or methodology
(such as the Marshall test) with some modifications. NCHRP's AAMAS
(Asphalt-Aggregate Mix Analysis System) research study showed that a
laboratory gyratory compactor better simulates the aggregate particle ori-
entation obtained in the field compared to that obtained with an impact
compactor used in the Marshall procedure (lfi). However, it was realized
in 1989 that it would be a few years before many agencies start to imple-
ment the AAMAS study's recommendations and use gyratory compactors
or other proposed compactors. In the meantime there was an urgent need
to start designing large stone HMA using modified Marshall design proce-
----dures-6-a:s·ed-cm-catrenrktrnwhxtge and experience.
The term "large stone" is a relative one. For the purpose of this sec-
tion, large stone is defined as an aggregate with a maximum size of more
than one inch (25 mm) which cannot be used in preparing standard
4-inch (100 mm) diameter Marshall specimens.
Chapter ?-Special Mixtures, Recycling, and Additives 425
of blows needed for 4-,---inch (100 mm) diameter specimens that are
21/z'inches (63,5 mm) high in order to obtain an equivalent com-
paction level fof a given volume of material.
Since the hammer 'Yeighs 22.5 pounds (10.2 kg) and the number of
blows on each side is 75 or 112 depending on the anticipated traffic, some
crushing of the aggregate at the surface has been observed. However, it is
believed that the effect of this minor amount of crushing on Marshall
properties is minimal.
-
::,-
O"' .g
n Ef
~
--i
Agency (Year data obtained) No. of Blows Ratio I
~ Cl)
~ra
'Cl
4 inch 6 inch Stability Flow
I
"
D.
e.
PennDO1f (1969) 50 75 2.12 · 1.62
~ ~
X
PennDojr (1970)
I
PennDO1f (1988)
50
50
75
75
2.81
1.95
1.15
1.39
.s
0
...8
"
Y'
::0
PennDojr (1988) 50 75 2.17 1.58
l-+-)
ti)
"
C)
'<
I
PennDO1f (1989) 50 75 1.68 1.40 g. [-
<J<>
*Note: The average stability and flow ratio for these five mixes compacted with 75/112 blows are 2.28 and 1.49, respectively.
~
...
Vl
0 .I>,
1:1 N
n --l
428 Hot Mix Asphalt Materials, Mixture Design and Construction
should be 2.25 times the load applied to a 4-inch (100 mm) specimen of
the same mix. This means the stability ratio should be 2.25.
Flow units measi,ired by the testing machine are the values for the
total movement of the breaking heads to the point of maximum stabili-
ty. When flow is ~nsidered on a unit basis (inches per inch of diame-
ter), the flow value for a 6-inch (150 mm) diameter specimen is 1.5
times that of a 4-incp. (100 mm) diameter specimen. This means the
flow ratio should be 1.5.
The average stability and flow ratio of specimens compacted with 75
and 112 blows (4-inch (100 mm) and 6-inch (150 mm) mold, respec-
tively) are 2 . 28 and 1.49 which are very close to the theoretically derived
values of 2.25 and 1.50, respectively.
It is recommended that the minimum Marshall stability requirement
for 6-inch (150 mm) diameter specimens be 2.25 times the requirement
for 4-in.ch (100 mm) diameter specimens. For example, if 1,000 pounds
(4,500 N) minimum stability is currently being specified using ASTM
D1559 ·(4-inch (100 mm) specimen), then 2,250 pounds (10,125 N)
minimum stability should be specified for large stone mixes using the
6-inch (150 mm) Marshall testing equipment. Similarly, the range of flow
values for 6-inch (150 mm) specimens should be adjusted to 11/z times the
values required for 4-inch (100 mm) specimens. For example, if the spec-
ified range for 4-inch (100 mm) specimens is 8-18, it should be adjusted
to 12-27 for 6-inch (150 mm) specimens.
inch to No. 4) or #4 (1 ½ inch to 3/4 inch) was used for coarse aggregate
to incorporate + 1 inch (25 mm) material in the mix. The data from field
trials in Kentucky indicate that achieving the desired density (compaction)
in the field does not appear to be a problem if the compaction process is
optimized (12.). The average in-place void content of three reported pro-
jects was approximately 6.5 percent.
Careful attention to details is needed to assure uniform delivery and
laydown of large stone mix without any significant segregation. The fol-
lowing factors are considered important (22):
plates. If extensions are not used, coarse particles tend to roll to the
out~r edge of the spread, creating a porous area with low density.
8. Paver speed is very important. The lowest rate of travel that accom-
modates production should be used. Slower rate of movement per-
m.its 111:rre uniform feeding of mixture under the screed and sup-
plies more vibrating compaction by the screed. Both permit better
positioning of coarse particles. Avoiding "stop and go" paving
reduces segregation, improves the texture of the spread, and elimi-
nates any tendency for screed settlement.
Only the asphalt patching materials which can be stockpiled (Types 2 and
3) will be discussed in more detail in this section.
Desirable Properties
In a majority of cases the failures of cold mixes may be traced to
improper construction practices (such as "dump and run" where the mate-
rial is dumped into a pothole on the run with little or no compaction).
Failures are also often related to the lack of desirable properties in a stock-
pile patching mixture. The following properties are desirable in such mix-
tures (.21):
I.Stability
A patching mixture should be stable after placement and com-
paction to resist vertical and horizontal displacement due to
imposed traffic loads. Lack of stability causes dishing and shoving
of the mixture.
2. Stickiness
Stickiness causes the mixture to adhere to the underlying pave-
rrreurarrd-srde·s-ofth:-qmth:-o-I-e~an-d-is-esp·ecially-desirable when the
mixture must be feathered to thin edges. Quite often mainte-
nance personnel do not take time to clean and dry the hole thor-
oughly so that proper tack coat can be applied. In such cases a
stickier patching mix is especially helpful.
432 Hot Mix Asphalt Materials, Mixture Design and Construction
1. Aggregate Gradation
For good mixture workability, it is desirable to have an open gra-
dation, but after the mix is placed, the gradation needs to be
denser to improve stability and durability.
2. Aggregate Shape
To obtain good workability, angular aggregate should be avoided.
However, once the mix is in place, high angularity is desirable for
better stability.
3. Binder Viscosicy:
Lower binder viscosity is desired for storageability and workabili-
ty, but after placement and compaction, higher viscosity is desir-
able for better cohesion and stability of the mixture.
4. Binder Content
It is desirable to have higher residual asphalt content in the mix-
ture to obtain thicker films on the aggregate for stickiness and
durability, but there are limits due to the binder drainage problem
in the stockpile just after stockpiling while the mix is hot. A high
binder content may also produce an unstable mixture.
New Concepts
In the past, use oflarger aggregate size (1/2-3/4 inch or 12-19 mm)
in the stockpile mixture has been promoted to obtain higher stability.
Such a mixture can be successful if the patching technique is ideal (such as
434 Hot Mix Asphalt Materials, Mixture Design and Construction
making edges vertical, cleaning, applying tack coat, and compacting ade-
quately). However, ideal patching techniques an: not always used, and
mixtures with larger aggregates tend to ravel prematurely under traffic,
resulting in failure of the patch. Another concept is to disregard the sta-
bility and make the mixture finer and more pliable so that it is more toler-
ant of abuse during the placement and performance under traffic. This
finer mix, if placed less than 3 inches (76 mm) deep in one lift in a con-
fined area, shoufd cbe stable. For deeper and/ or larger holes, the mixture
has to be compacted in layers.
The cohesive and adhesive qualities of a mix are mainly dependent
upon the composition of the mortar (asphalt binder plus fines). The pres-
ence of excessive fines or dust (material passing No. 200 (0.075 mm) sieve)
in the mixture results in a mortar that is lean, less tacky, and friable. It is
no coincidence that most of the expensive commercial. patching products
are made. from clean stone. Several extraction tests performed on such
products have revealed that the fines (minus 200 fraction) are usually less
than one percent. The absence of excessive fines causes mixtures to be very
tacky and sticky; therefore, tack-coating of the pothole is not required.
Many conventional stockpile patching mixtures do not perform satisfacto-
rily because of excessive fines. Such mixes are dull and friable, and lack the
cohesive and adhesive qualities.
Improved Formulation
In view of the challenges of mix design and the new concepts, the fol-
lowing characteristics appear desirable for a satisfactory and economical
stockpile patching mixture (25.):
1. Finer and Predominantly One-Size Gradation
Atypical gradation used very successfully by a state DOT is given
in Table 7-4. Ideally, two sizes of patching mixtures should be
Some states require the asphalt supplier to conduct the wet coating
Chapter 7-Special Mixtures, Recycling, and Additives 437
test, the static immersion test, and the stripping test using the job
aggregate.
Stockpile patching mix made from the improved formulation gave
excellent performance in the study conducted by the Strategic Highway
Research Program.
I. Mix Composition
Extraction tests are performed or an automated HMA facility's
printed tickets are examined to verify the conformance of the mix
composition to the job-mix formula.
2. Stripping Test
Different forms of stripping tests are employed by different agen-
cies. Some use. the ASTM Standard Designation D 1664
"Coating and Stripping of Bitumen Aggregate Mixtures." Some
states use a boiling water test in which the mixture is placed in
boiling water for 3 minutes and the retained coating is evaluated
(at least 90 percent coating is often required). There is no satis-
factory test which has been related to performance.
3. Workability Test
A scientific test for quantifying the workability is not available at
the present time. Most specifications require that the "mix shall
remain workable and capable of being handled and placed easily."
Some states use a subjective test. "-For example, in Pennsylvania
approximately 5 lbs (2.3 kg) of the mixture is cooled to 20°F
(-7°C) in the laboratory. The mixture should be capable of being
broken up "readily'' with a spatula having a blade length of
approximately 8 inches.
be stockpiled higher than 4 feet for the first 48 hours to prevent drainage
of the asphalt binder and consolidation. This allows the stockpile to cool
before it is stacked too high. Some cases of heat build-up and burning of
high stockpiles prompted this suggested requirement.
of asphalt made any reclaimed asphalt material a valuable asset. Before the
oil embargo, the price of asphalt cement was so low that the cost of remov-
ing, stockpiling, and recycling old asphalt pavements was more than that
for purchasing, mixing, and placing new material.
A second item that has had a great impact on recycling is the milling
machine. Prior to the development of the milling machine, old asphalt
pavement had to be ripped from the roadway, then crushed prior to using.
This process was awkward, required full depth removal of the HMA, and
usually required major rework of the roadway prior to overlaying. This
process also often required that the roadway be shut down for extended
periods of time. Another process that was used prior to the milling
machine was the heater planer. This process required a large amount of
fuel to heat th~ pavement, resulted in pollution, and damaged the removed
HMA ·by overheating. The milling machine has solved most all of these
problems observed with heater planers. If can remove any amount of
material desired and produce any desired grade. It does not prod,uce appre-
ciable pollution since heat is not required. The material removed with a
milling machine does not have to be crushed since it is fine enough, imme-
diately after being removed, to recycle. A milled surface can be opened to
traffic temporarily until the overlay has been completed.
A third reason that recycling has become popular is the development
of the drum mix facility. Before the development of the drum mix facili-
ty, engineers did not know how to recycle in a batch plant. Reclaimed
material could not be fed through the dryer since this would overheat the
asphalt and cause a major pollution problem. Since the development of
the drum mix facility, the Minnesota process has been developed for the
batch facility which superheats the new aggregate in the dryer. The super-
heated aggregate mixes with the reclaimed asphalt pavement (RAP) in the
weigh box and pugmill, thus transmitting the heat of the aggregate to the
RAP. In one of the drum mix processes the RAP is added about halfway
down the drum where it is mixed with the new aggregate and new asphalt
cement. There are other processes which are described in Chapter 6.
ture. The removal of chunks makes it difficult to maintain grade and may
require a major rework of the base course. In this case it may be simpler
and cheaper to apply an overlay. A major item that must be considered
prior to repair is the overall condition of the existing pavement. A pave-
ment .exhibiting alligator cracks has a structural problem; hence, enough
HMA must be placed to provide adequate structural properties. The need
to add 2-3 layers of asphalt mixture may indicate that overlays would be
quicker and cheaper. If it is determined that only one layer is needed to
provide adequate support, then recycling should be considered so that the
damaged layer is removed and replaced during recycling. The cracks are
likely to reflect through the overlay causing poor performance if the exist-
ing mix is not removed.
Recycling may not be desirable if the aggregate in• the existing mixture
does not meet the specified requirements. For example, if the existing
aggregate is uncrushed, then recycling produces another mixture with a sig-
nificant amount of uncrushed material. The amount of RAP used in the
recycled mixture in this case has to be minimized. A high percentage of
material passing the No. 200 (0.075 mm) sieve in the existing mix may be
difficult to recycle. The milling operation creates additional material pass-
ing the No. 200 (0.075 mm) sieve, and it may be difficult to keep the per-
cent of aggregate passing this critical sieve size within the specification
requirements. If the aggregate in the existing mix has a tendency to pol-
ish, then care must be used in trying to recycle this mixture if it is used for
the surface course. Most of these problems can be overcome if small
amounts of RAP are used.
Many additives are being used to improve the performance of asphalt
mixtures. Some of these additives when used in an asphalt mixture may be
difficult or impossible to recycle. Some additives that have been used
include latex rubber, crumb rubber, sulfur, polypropylene, and carbon
black. .No additive should be used in an asphalt mixture if this additive
prevents the mixture from being recycled at a later date. Reheating mixes
that contain sulfur can be dangerous since sulfur combined with asphalt
gives off toxic fumes when heated to high temperatures. There has been
some indication that milling mixes that contain latex rubber may be diffi-
cult. On some projects, the .teeth on the milling machine have been
gummed up with this rubberized asphalt. There is no known way to deter-
------m1·nnh-e-aspfra:l:rcon-rent-rrora:sphait-binder-proper-ties-of old asphalt rub-
ber mixes. The rubber may also result in more pollution during produc-
tion of recycled mixtures. Several years ago asbestos was used as an addi-
tive in some asphalt mixtures to improve the durability. Obviously, the
asbestos has made.it difficult or impossible to recycle those mixtures.
442 Hot Mix Asphalt Materials, Mixture Design and Construction
There are many candidate pavements for recycling, but each project
has to be evaluated to determine if recycling is desirable. The recycling
process can be used to reduce reflective cracking, to match curbs and gut-
ters, to reduce construction cost, to maintain better traffic flow during con-
struction, to maintain overhead clearance, and to expose an underlying
deteriorated layer that should be removed or modified._ The existing pave-
ments that obviously should be considered for recycling are those with sig-
nificant rutting, cracking, or on city streets or other places where existing
structures such as curbs and gutters must be matched.
I cP=I0"3 Pa • s
l c = {I.F - 32) /1,8
0
The next step in the design process is to vary the amount of new
asphalt cement added to the selected combination of RAP and aggregate
blend. The amount of asphalt added is less than that for new (virgin) mix-
tures since the RAP contains a significant amount of asphalt cement.
The prepared samples of recycled mixture are compacted and tested in
the same way as that required for a new mixture as discussed in Chapter 4.
Most agencies require that the recycled mixtures have the same properties
as that required for new mixtures.
The final test that must be conducted on recycled mixtures, but which
is not performed on new mixtures, is a test on the recovered asphalt binder.
The asphalt cement is extracted from the mix by the extraction test (ASTM
D2172) and recovered from solvent by the Abson method (ASTM
D1856). The properties of the asphalt binder from this recycled mixture
should be approximately the same as that for asphalt cement recovered
from new mixtures. This test should be conducted on laboratory prepared
samples as well as samples during production to get an estimate of the
asphalt cement properties. To be accurate this test should be conducted on
asphalt mixtures actually produced in a HMA facility.
The properties of the recovered binder such as penetration and viscosity
must be within the specification limits. The grade of new asphalt cement added
to the recycled mixture must be changed or the percent of RAP used in a mix
should be changed to modify the resulting binder properties.
The mix design on a recycled mix is a preliminary design that must be
finalized during production of recycled mixtures at the HMA facility (211).
Modifications to the mix design ·will likely be required once the plant oper-
ation begins.
PPER
NEW
BITUMEN
RECYCLING AGENT I
~
OF NEEDED>
\U0;)
problems when recycled include sulfur, rubber, and carbon black. Some of
these create problems when milling, some cause more pollution during mix
production, and some have mix and material properties that are difficult to
measure and control.
Recycled mixtures can be designed and constructed to perform as well
as new mixtures. The necessary tests ahd evaluations to ensure quality con-
trol must be performed on a regular b~sis throughout the construction pro-
ject; otherwise, poor performance1is likely the outcome.
All asphalt binders and mixtures do not require modification. The fol-
lowing are some specific technical reasons for using additives and modifiers
inHMA:
------=•_Qb_tain._s_tiffe.unixtur.e.s_at_bigh_ser:v.ice-temper.atur,ewo-minimiz,c------
rutting.
2. Obtain softer mixtures at low service temperatures to minimize
non-load associated thermal cracking.
3. Improve fatigue resistance of HMA mixtures.
Chapter ?-Special Mixtures, Recycling, and Additives 449
-
Cl)
C Asphalt-aggregate · temperatures
mix
!-
-8
>,
"in
en
>... Conventional_/
Cl)
"C
C
asphalt-aggregate
mix
ixi •••
, Practical •
limits of
modification
developed at this time to check the compatibility in all instances. (e) How
should the binder be stored to maintain its properties over a length of time?
(f) How should the binder be specified? (g) Does the modifier affect rou-
tine test results? For example, an accurate determination of asphalt con-
tent in HMA may not be possible when certain modifiers are incorporat-
ed in HMA. (h) Can the HMA be recycled? For example, this question
has been raised about the HMA containing crumb rubber modifiers. (i)
What is the effect on life cycle cost? Are the initial increased costs justified
based on life cycle cost analysis? (j) What are the effects of the modifier on
worker health and safety?
Classification
It is possible to classify additives and modifiers in different ways.
However, an excellent generic dassification system was first suggested by
Terrel and Walter (32). A modified version of this system is shown in Table
7-6. Several additives exist un.der each category; however, this is a dynam-
ic list that is rapidly changing. A discussion of each additive or modifier
type follows.
Fillersj
Mineral fillers include mineral dust from the crushing and screening of
aggregates (including baghouse fines), lime, portland cement, and fly ash.
Other materials such as carbon black and sulfur have been used primarily
to modify asphalt binder properties but they do have a role of filler. Fillers
may be used to (1) fill voids and hence reduce optimum asphalt content,
(2) meet specifications for aggregate gradation, (3) increase stability, and
(4) improve bond between asphalt cement and aggregate. Typically, an
increase in filler lowers the optimum asphalt content, increases the densi-
ty, and increases the stability (33).
Mineral fillers have been used primarily to fill in the voids between the
aggregate particles and meet specified gradations for HMA. However, it has
been documented extensively in the literature that mineral fillers such as bag-
------,h01:1Se-fu:i©s-<ie-affe-Gt--the-h~ha¥i.or-0£asphalt-binders-and-HMA-mixtures C31.,.
.35). Mineral fillers cause a general stiffening or reinforcing effect in HMA mix-
tures. However, the extent of stiffening effect is generally dependent upon
Rigden void content (voids between the fine particles in dry compacted state)
which in turn is affected by mineralogic type, size distribution, particle shape,
452 Hot Mix Asphalt Materials, Mixture Design and Construction
2. Extender • Sulfur
• Lignin
POLYMERS
and surface texture of the mineral fillers (31:). Some baghouse fines, primarily
those containing clay, can increase the potential for stripping in Hl\1A mixtures.
All fillers must be fed to the asphalt mixture consistently and in cor-
rect proportions; otherwise, the mix properties are adversely affected.
Excessive amounts of filler usually reduce the VMA to a point that suffi-
cient asphalt content for a durable mix cannot be added. High filler con-
tent also increases the aggregate surface area and thus greatly reduces the
asphalt film thickness. Some specifications place limits on the dust to
asphalt cement ratio. A typical specification requires the dust to asphalt
cement ratio to be benveen 0.6 and 1.2 by weight. An ideal specification
should require such a ratio based on volume and should take Rigden voids
(which generally control the stiffening effect) into account.
Although a mineral filler, lime is primarily added as an antistripping
agent to Hl\1A to minimize stripping problems. Lime will be discussed
later as an antistripping agent.
Carbon black can be considered as a "microfiller" because. it is very
fine. Different grades and types are available because carbon black is wide-
ly used in other industries, such as reinforcement for rubber tires. The
grade which has been used in Hl\1A consists of sub-micron size particles
(about 10 14 particles per gram). Since this material is very fine it could read-
ily become airborne and cause air quality problems.· Therefore, carbon
black is usually combined with a high boiling point maltene oil to form
pellets; which are easier to handle in Hl\1A facilities. Recommended con-
centrations of carbon black pellets range from 10 to 15 percent by weight
of asphalt cement. To obtain desirable low temperature properties in cold
climates while maintaining desirable high temperature properties, some
highway agencies have used a softer grade of asphalt cement compared to
the common grade generally used. The carbon black can be supplied in
polyethylene bags or in a bulk container. It can be preblended with asphalt
cement either at the refinery or at the Hl\1A facility. A 10 to 20°F (6° to
11 °C) increase in mixing temperature may be needed to compensate for
the stiffening effect of the carbon black. It has been claimed that the car-
- - - - - - v..on..black=.eeinfo.rced...as.phalLcemen.Lb.indecimp.r.oyesJfMA.Jesistance to
rutting, improves Hl\1A resistance to wear and abrasion, and reduces
oxidative hardening and embrittlement of binder with exposure to weath-
ering. The use of .carbon black in Hl\1A has been evaluated by various
researchers (3..6, 32).
454 Hot. Mix Asphalt Materials, Mixture Design and Construction
Extenders
The concept of extending asphalt binder became prominent after the
1973 oil embargo. It appeared at that time that the price of asphalt cement
would continue to increase and that asphalt cement might become scarce.
The Federal Highway Administration (FHWA) promoted the investiga-
tion of extending asphalt binders by partially substituting with sulfur and
lignin. Complete substitution of asphalt cement with sulfur (sulphlex) was
also evaluated (18). Both sulfur and lignin are largely byproducts of other
industries: sulfur from natural gas desulfurization and lignin from pulp
and paper production. The supply and price of sulfur have fluctuated
widely since the 1973 oil embargo, so the use of sulfur has been sporadic.
At the present time (1996), because of lowensphalt cement prices and ris-
ing sulfur prices, sulfur can not compete with asphalt cement.
Commercial use of lignin in HMA has also been negligible. Much of the
lignin is burned in-plant for fuel.
Sulfur is an abundant, non-metallic chemical element. It is a yellow,
crystalline solid at room temperature. It melts into a free-flowing,
straw-colored liquid at about 240°F (116°C). Above 310°F (154°C), liq-
uid sulfur becomes highly viscous. Within the range of 250°F to 290°F
(121 °C to 143°C), usually employed for HMA production and construc-
tion, the viscosity of sulfur extended asphalt (SEA) binder is lower than
that of asphalt cement alone. Sulfur has been used to modify HMA base,
binder, and surface courses.
Since the specific weight of sulfur (specific gravity is about 2.07 at
68°F or 20°C) is about twice that of asphalt cement, the specific weight of
SEA binder increases as the sulfur content increases. Sulfur contents have
ranged between 10 and 50 percent by weight of the binder; however, the
most frequent substitution rate has been between 20 and 35 percent by
weight (approximately equal to 10 and 17.5 percent by volume) of the
binder. Sulfur may be delivered in either solid or liquid form, usually in
bulk by rail car or truck. Sulfur has to be added to the asphalt cement just
prior to the HMA mixing operation. For a"liquid sulfur system, it is nec-
essary to have a separate liquid sulfur storage tank, sulfur pump, metering
system, and in some cases a preblending unit. For a solid sulfur system, a
storage pad area, a conveyor system, a small crusher, and an auger or a van
1-----eecl.erancl. 15lower need to be arranged. Vanous methods and eqmpments
have been developed for combining sulfur and asphalt cement prior to
their addition to the aggregate at the HMA facility (22). Good tempera-
ture control combined with adequate ventilation are very important to pre-
vent any build-up of sulfur dioxide (SO 2) or hydrogen sulfide (H 2 S) gases.
Chapter ?-Special Mixtures, Recycling, and Additives 455
Polymers
It can be seen in Table 7-6 that both rubber and plastic materials are
referred to as polymers. Simply stated, a "polymer" is a very large molecule
made by chemically reacting many (poly) smaller molecules (monomers)
to one another in long chains or clusters. The sequence and chemical
structure of the monomers determine the physical properties of the result-
ing polymer. Copolymers consist of the combination of two different
monomers which can be either in a· random or block arrangement. For
example, polystyrene is a hard, brittle plastic whereas polybutadiene is soft
and rubbery. If these two distinctly different monomers are randomly
mixed and reacted together, a new polymer called copolymer with very dif-
ferent properties is created. Polymers can be engineered to obtain a broad
range of physical properties. However, they can be divided into two gen-
eral categories: elastomers (rubber) and plastomers (plastic).
Elastomers or rubbers used as asphalt modifiers include natural rubber,
styrene-butadiene rubber (SBR) latexes or SBR, styrene-butadiene-
styrene (SBS) block copolymers, styrene-isoprene-styrene (SIS) block
----·copolymers,polychforoprene-1-a:texes,and-crumb-rubber-modifiers (from
ground tires). Elastomers resist deformation from applied stress by stretch-
ing and recovering their shape quickly when stress is removed. These poly-
mers add very little strength to the asphalt cement until they are stretched.
However, their tensile strength increases with elongations (~.
456 Hot Mix Asphalt Materials, Mixture Design and Construction
desired results. The aggregate is fully coated with neat asphalt cement
prior to introducing the latex. For batch facilities, latex is normally intro-
duced 8-10 seconds after addition of neat asphalt cement. The total wet
mix time is about 10 seconds longer than the conventional wet mix time.
In case of drum mix facilities, latex should be introduced into the drum at
a point between the introduction of neat asphalt cement and the discharge
end of the drum. No significant changes in HMA laydown and com-
paction are normally required when SBR latexes are used.
SBS block copolymers are available in solid form as pellets, crumb, or
ground material in bags or bulk. The usual dosage rate is about 5 percent
by weight of asphalt cement. High shear mixing equipment is used for
blending this modifier with hot asphalt cement maintained at 350-380°F
(177-193°C). The modified asphalt binder can be prepared at the refin-
ery and delivered to the HMA facility like a conventional asphalt cement.
No agitation is required during storage. Since most mix design procedures
require the asphalt binder to be within specified viscosity ranges during
mixing and compaction, it will probably be necessary to increase the mix-
ing and compaction temperatures while conducting laboratory work.
However, field experience has shown no unusual difficulties at conven-
tional mix temperatures when processing and compacting the HMA pro-
duced with this high viscosity modified binder. It is believed that the ease
of handling is due to the relatively higher shear forces experienced during
manufacturing in the field compared to the lower shear forces used in lab-
oratory viscosity measurements. Therefore, supplier's recommendations
for mixing and compaction temperatures should be followed in the labo-
ratory and field.
Reclaimed rubber is primarily obtained from tires. About285 million
tires are discarded every year in the United States. Of these, about 55 mil-
lion are retreaded or reused (resold), and about 42 million are diverted to
various alternative uses such as combustion for generating power and addi-
tives for HMA mixes. The remaining 188 million tires are added to stock-
piles, landfills or illegal dumps. According to EPA estimates, 2 to 3 billion
discarded tires were available in 1992 (di'). Several states have enacted leg-
islation to regulate the scrap tire problem.
Crumb rubber obtained from tires can either be ambient ground
(grinding at room temperature or above) or cryogenically ground (grind-
i.--ng-b-efow-embrittlemenrremp-era:mre-;-liqui_--d-nitrogen--is -often - used).
Ambient ground crumb rubber has a sponge-like surface. Due to very
high surface area this rubber reacts with asphalt cement reasonably fast.
Some cryogenically ground rubbers may have undesirable particle mor-
phology (structure). This process produces clean flat surfaces which, in
458 Hot Mix Asphalt Materials, Mixture Design and Construction
turn, may reduce the reaction rate with hot asphalt cement. In one study,
.cryogenically ground rubber gave lower elastic recovery compared to the
ambient ground rubber (1§;.
The use of crumb rubber to modify asphalt cement has been devel-
oped over the past 25 years. Crumb rubber is primarily used in HMA
mixes by two processes; generally referred to as the wet process and the dry
process:
recycled mix be designed? If these concerns are not addressed it may not
be possible to recycle the HMA mixes containing crumb rubber. In that
case, the disposal of RAP (reclaimed asphalt pavement) may become a
much greater problem than disposing the tires.
The "McDonald process" (wet process) and "PlusRide" (dry process)
as discussed earlier are patented processes. However, the original patent on
the "McDonald process" expired in 1992. Recently some newer concepts
which include the generic dry process, chunk rubber asphalt concrete and
continuously blended asphalt rubber, have been introduced (15).
The generic dry process allows incorporation of crumb rubber into
available "generic" aggregate gradations rather than gap-graded aggregates
required in the PlusRide system. Experimental field applications of the
generic dry process have been made in NewYork, Florida, Iowa, Kansas,
Oregon, and Illinois (15).
The continuous blending asphalt rubber concept has been used by
Florida DOT. It is a wet process which uses 5 to 15 percent of fine crumb
rubber (180 µ or No. 80 mesh) to avoid reaction tanks, and to facilitate
continuous blending. The mixing of crumb rubber and the asphalt cement
is accomplished in a self-contained portable blending/metering unit.
range. Only moderate agitation (such as low shear, stirred pot) or circula-
tion is necessary for proper blending. Blends can be stored for several
weeks without any separation. The EVA must be compatible with the neat
asphalt cement to obtain the desired properties.
Fibers
The use of natural fibers such as asbestos has been well known for
many years. However, since asbestos was identified as a health hazard, it is
no longer in use in the U.S. Manufactured or synthetic fibers have been
promoted to replace asbestos. Fibers are most commonly used as rein-
forcement or stiffener in the manufacture of roofing and industrial water-
proofing membrane systems. They have the potential of reinforcing and
improving the tensile strength and cohesion ofHMA mixtures. Fibers also
permit higher asphalt contents than conventional HMA mixtures without
any significant draindown problems. This is especially important for
open-graded friction courses and stone matrix asphalt mixtures. The
resulting thick binder films from high asphalt contents increase the dura-
bility of paving mixtures. Fibers are generally added to the mixture at the
HMA facility. Most manufacturers supply fibers in easily melted plastic
bags because, being small and finely divided, they cannot easily be handled
in bulk. For drum mixers, pneumatic systems are available for air blowing
the fibers into the drum. The fiber manufacturers usually supply or oth-
erwise assist the HMA contractors with special equipments to handle their
------peeifie-fi.ber-s.-'J'.B.e-H-lvlA-m-Bf:fl:l-Fe-Gen-1,ainiag-filiC:lrs-may-be-difficult to lute
and place by hand.
Polypropylene fibers about 10 mm in length have been used as an
asphalt reinforcement. Typically, 0.3 percent fibers by total weight of
HMA are added. The specific gravity of polypropylene fibers is about
462 Hot Mix Asphalt Materials, Mixture Design and Construction
0.91. The HMA mix temperature should not exceed 300°F (149°C) and
the aggregate temperature should not exceed 290°C (143°C) to prevent
polypropylene fibers from softening or melting. In a batch facility, good
dispersion of the fibers is usually obtained by a 10-second dry mixing cycle
followed by the usual 35-second wet mixing cycle. The finished HMA
mixture should be examined for presence of clumps of fiber. If clumps are
present, the wet mixing cycle should be increased in 5-second increments
until a uniform .niix is obtained. For drum mix facilities, feed systems for
delivering fiber at a metered rate are available from the fiber suppliers.
Polyester fibers of different lengths have been used to reinforce HMA
mixtures. The length varies from 1/2 inch (13 mm) for base course mix-
tures to 1/50 inch (1/2 mm) for fine-graded mixtures. The specific grav-
ity of these fibers varies from 1.32 to 1.40.-The minimum melt tempera-
ture and the minimum ignition temperatures are usually 480°F (249°C)
and 1000°F (538°C), respectively. Typical dosage rate varies from 3 to 7.5
pounds per ton (1.5 to 3.8 kg per Mg) of HMA mixture (about 0.15 to
0.40 percent by weight of HMA mix) depending upon the application.
Additional asphalt cement binder ranging from 0.2 to 0.3 percent by
weight of the HMA mix needs to be incorporated in fiber-modified mix.
Polyester fibers are handled and mixed similar to polypropylene fibers.
Mineral fibers are used in Europe in dense-graded, open-graded, and
stone matrix asphalt (SMA) mixtures. Typically, mineral fibers are manu-
factured at high temperatures from diabase as a raw material. The usual
dosage rate of mineral fibers in SMA is 0.4 percent by weight of total mix.
A maximum fiber length of 6 mm is generally used in SMA.
Recommended specifications for mineral fibers for SMA and how they are
introduced in HMA facilities, are given in Reference£ Mineral fibers are
supplied in prepackaged condition or in bulk. Methodology and equip-
ment for metering bulk (loose) fibers into batch and drum mix facilities
have been developed and are available.
Cellulose fibers are extensively used in SMA in Europe with the pri-
mary purpose of increasing asphalt content without any draindown prob-
lem. These fibers have been used in many SMA field trials in the United
States. Specifications for cellulose fibers and guidelines on handling are
given in Reference ...iZ Typically a maximum fiber length of 6 mm is used.
---~A-bout..20....p_ercenr_o£fihers_p.ass__thr.ough-N.o.-I-O.O-(0..-l.;;-m-m....si<W.e)-.-Th.ec..------
dosage rate for cellulose fibers in SMA is usually 0.3 percent by weight of
the total mix. Cellulose fibers are supplied in prepackaged condition or in
bulk. The methodology and equipment for metering bulk (loose) fibers
into batch and drum mix facilities have been developed and are available.
Chapter 7-Special Mixtures, Recycling, and Additives 463
Oxidants
Sometimes it is desirable to increase the stiffness ofHMA mixtures for
improved strength. This may be appropriate for low stability mixes such
as sand-asphalt. One approach is to use a harder asphalt cement at the
begin~ing. However, this approach may reduce the long term perfor-
mance. The other approach is to use oxidation catalysts which stiffen the
asphalt cement binder after the HMA is placed on the pavement. An
oil-like liquid containing a manganese compound was used in field trials
throughout the U.S. in the 1980s. Typical concentrations range from 1 to
4 percent by weight of binder. Some agencies chose to use a softer grade
asphalt cement than normally required (for example, AC-10 in lieu of
AC-20). Initially, this modifier lowers the viscosity of asphalt cement at
77°, 140°, and 275°F (25°, 60°, and 135°C). In some cases, this softening
may prevent the early placement of heavy traffic on the pavement. The
viscosity of the modified binder continues to increase on exposure to air for
some weeks. The HMA becomes significantly stiffer after this "cure" peri-
od. Early field trials in the U.S. showed excessive stiffening that resulted
in premature thermal cracking. Later formulations and use of soft asphalt
cement reduced this problem. The use of this modifier is not prevalent at
the present time (1996). It appears that this modifier may be more appro-
priate for use in base courses rather than surface courses.
Antioxidants
Asphalt cements undergo oxidative hardening during HMA produc-
tion and construction, and subsequently during service. Antioxidant addi-
tives consisting of lead or other compounds are available to minimize
oxidative hardening of asphalt cements, thereby increasing-the durability of
HMA mixtures. The experience with the antioxidant additives in HMA is
rather limited, therefore, no guidelines are available. Most of the research
work has been done in the laboratory and there is a lack of adequate field
experience and validation. Some field work has been donein ~11stralia to
retard oxidative hardening of binders used in chip seals. Because the
short-term aging of the asphalt cement during HMA production is mini-
mized due to the presence of antioxidant, there is a potential for tender mix
problems during laydown and compaction operations.
I
Ii,
464 Hot Mix Asphalt Materials, Mixture Design and Construction
Hydrocarbons
It may not always be poss~ble to obtain an asphalt cement of desired
properties (such as viscosity, penetration, and ductility) for a particular
project. However, modification of the available asphalt cement is possible
by adding either softer or harder hydrocarbon materials. For example,
aged asphalt binder in the recycled asphalt pavement (RAP) needs to be
softened or rejuvenated by the use of recycling agents. On the other hand,
some asphalt cements need to be hardened with hardening agents when
used in HMA for high stress areas such as intersections and access to toll
booths. A discussion of recycling agents and hardening agents consisting
of hydrocarbon materials follows.
2. Rejuvenating agent:
(a) Lube extracts (highly naphthenic or aromatic fractions
removed from lube stock by solvent extraction)
(b) Extender oils (aromatic oils from lube stock, mostly used for
--------<e•Xtmding-asphal.t=rubber-blendSci-----------------·--,
flash point of 425°F (218°C) for all grades to ensure safety. Maximum vis-
cosity ratio of 3 and maximum weight change of 3 or 4 percent, after the
neat asphalt cement is subjected to thin film oven test or rolling thin film
oven test, are intended to preclude light materials that would evaporate
quickly in the HMA facility or from the HMA pavement.
Most recycling agents are viscous liquids at room temperature. For hot
mix recycling, a storage tank, pump, and accurate metering system is
required at the HMA facility. A separate injection pipe is needed in the
event only a recycling agent is used without asphalt cement. Careful con-
trol of the quantity of the recycling agent is essential because small devia-
tions in dosage rates can cause large deviations in HMA behavior and per-
formance (22). The structural properties of recycled HMA are generally
satisfactory when the proper recycling agent is selected (1.2,_ .5.Q).
Trinidad lake asphalt (TLA) is one of the natural asphalts which were
discussed in Chapter 2. It contains 50 to 57 percent asphalt. The remain-
der is colloidal clay of volcanic origin, plus some inert organic matter.
Trinidad lake asphalt has a specific gravity of about 1.41, a softening
point in 200-207°F (93-97°C) range, and a penetration in 3-10 range.
It has been used in HMA for high stress areas such as intersections and
access to toll booths. It has also been used in Gussasphalt (51) which is a
pourable mastic asphalt concrete. A blend of 25% TLA and 75% AC-20
asphalt cement is recommended for heavy duty pavements. TLA is usu-
ally supplied in 520-pound (235-kg) drums for crushing and blending
' in suitable tanks. TLA can also be supplied in powdered form, which can
be added through filler lines. It can be preblended with asphalt cement;
however, the blend should be stored in suitaBle stirred tanks. Pumps with
bearings exposed to the blend, which contains fine silica based filler,
should not be used.
Antistripping Agents
Antistripping (AS) agents are used to minimize or eliminate stripping
of asphalt cement from aggregate in HMA mixtures. Stripping is discussed
in Chapters 4 and 8. Both liquid AS additives and lime additives are used
to resist stripping.
J
asphalt cement is stored at high temperatures for a prolonged period of
time.
The simplest and most economical way is to mix the AS agent with the
asphalt cement in a liquid state prior to mixing the asphalt cement with the
aggregate. Although this method is most commonly used, it is inefficient
because only a portion of the AS agent reaches the aggregate-asphalt
---fl·ement-i-nt-erf-aee:-Bi-reet--a:p-p-lieati-on-of-the-A-S agent to-rhe-aggregate-sur-~-----
face is undoubtedly the most efficient and possibly the most effective.
However, a uniform dispersion is not possible because very small amounts
of AS agents (for example 0.5 percent by weight of asphalt cement) are
normally used, and the HMA mix contains a substantial amount of fines.
Chapter 7-Special Mixtures, Recycling, and Additives 467
mixers by injecting lime into the drum just ahead of the asphalt
cement. The pick up of lime by the gas stream is prevented by
modifications of the flights and providing suitable baffles inside the
drum. (5§). Some asphalt paving technologists believe that the use
of dry lime is not consistently effective, although many agencies
including Georgia DOT have had very satisfactory results with dry
lime.
2. Hydrated lime slurry: This method requires additional water to be
added to the aggregates which results in increased fuel costs and
reduced HMA production rates.
3. Dry hydrated lime to wet aggregate: In this method dry hydrated
lime is added to wet aggregate, us~ally containing 3-5 percent
water, and then mixed in a pugmill or tumble mixer to obtain a
homogeneous mix.
4. Hot (Quicklime) slurry: The use of quicklime (CaO) slurry has at
least two advantages: (a) its cost is equal to that of hydrated lime
but when slaked the hydrated lime yield is 25 percent greater, and
(b) the heat from slaking results in an elevated temperature which
helps in the evaporation of the added moisture. It should be han-
dled with caution because it can cause skin burns.
Waste Materials
Numerous waste materials result from manufacturing operations, ser-
j
vice industries, sewage treatment plants, households, and mining CiZ).
Legislation has been enacted by several states in recent years to either man-
date the use of some waste materials or to examine the feasibility of s u c h -
usage. The hot mix asphalt (HMA) industry has been pressured in recent
years to incorporate a wide. variety of waste materials into HMA pave-
___m.ents_.Jhis_has_raisecLthe...follo:.w.ing.legitlmate-GGn.c;er,1=1s:-~a}-~agine~i:ing
concerns such as effect on the engineering properties (for example, strength
and durability of HMA), impact on its production, and its future recycla-
bility: (b) environmental concerns such as emissions, fumes, odor, leach-
ing, and handling and processing procedures; and (c) economic concerns
Chapter ?-Special Mixtures, Recycling, and Additives 469
such as first cost, life cycle costs, salvage value, and lack of monetary incen-
tives.
The waste materials can broadly be categorized as follows: (a) indus-
trial wastes such as cellulose wastes, wood lignins, bottom ash, and fly ash;
(b) municipal/domestic wastes such as incinerator residue, scrap rubber,
waste glass, and roofing shingles; and (c) mining wastes such as coal mine
refuse. A discussion of some waste materials which have been used in
HMA on an experimental basis follows.
Miscellaneous
There are some miscellaneous materials that are not classified but have
been used as additives and modifiers in HMA mixes. Two materials: sili-
cone and deicing calcium chloride granules, are discussed.
(25°C) is most widely used. The silicone fluid is added to asphalt cement
binder in very small amounts, usually two to three parts per million (ppm)
to realize the following benefits:
REFERENCES
1. Fehsenfeld, Fred M. Performance of Open-Graded Big Rock Mixes
in Tennessee and Indiana. National Asphalt Pavement Association,
Information Series 104, 1988.
2. Cedergren, H.R., J.A. Arman and K.H. O'Brien. Development of
Guidelines for the Design of Subsurface Drainage Systems. Federal
Highway Administration Report RD-73-14, February 1973.
3. Cedergren, H.R. and WR. Lovering. The Economics and
Practicability of Layered Drains for Roadbeds. Highway Research
Board, Highway Research Record 215, 1968.
4. Kandhal, P.S. and M.E. Wenger. Experimental Asphalt Stabilized
Base under Portland Cement Concrete. Pennsylvania Department of
Transportation Research Project 71-5, Final Report, February 1974.
5. Halstead, WJ. Open-Graded Friction Courses for Highways
Transportation Research Board, NCHRP Synthesis 49, 1978.
6. Halstead, WJ. Criteria for Use of Asphalt Friction Surfaces.
Tr~nsportation Research Board, NCHRP Synthesis 104, 1983.
7. Kandhal, P.S., R.J. Brunner and T.H. Nichols. Design, Construction,
and Performance of Asphalt Friction Courses in Pennsylvania.
Transportation Research Board, Transportation Research Record 659,
1977.
8. Bukowski,] .R., SMA in America. Paper presented at the 1993 Annual
Meeting of the Transportation Research Board, Washington, DC,
1993.
9. Guidelines for Materials, Production, and Placement of Stone Matrix
Asphalt (SMA). National Asphalt Pavement Association, IS 118,
August 1994.
10. Brown, E.R. et al. Designing Stone Matrix Asphalt Mixtures.
National Academy of Sciences, Transportation Research Board,
National Cooperative Highway Research Program, Report No. 9-8/1,
1995.
11. Brown, E.R. and RB. Mallick, A Laboratory Study on Draindown of
Asphalt Cement in Stone Matrix Asphalt (SMA). Transportation
Research Record 1513, 1995.
12. Scherocman, J.A., "The Design, Construction and Performance of
----£1:0-FJ:e-Mas-t-ie-Pa:v-e·mef.l·t-ba-yers:-1:he-Gontinuing-Sto~Proree-ding·,--,- - - - - ~
Canadian Technical Asphalt Association, Vol. XXXVII, November
1992.
13. Davis, Richard L. Large Stone Mixes: A Historical Insight. National
Asphalt Pavement Association: Report IS 103/88, 1988.
Chapter ?-Special Mixtures, Recycling, and Additives 473
SERVICEABILITY/PERFORMANCE CONCEPTS
1. Highways are designed and built for the comfort and convenience
of the travelling public. Therefore, a good highway is one that is
smooth and safe.
2. One user's opinion about how well a highway is serving its function
is subjective. ·
3. There are objective characteristics of the road which, when mea-
sured and properly combined, can be correlated to the average 1?Ub-
jective evaluation of the highway.
4. The serviceability of a road can be adequately represented by the
arithmetic average of opinions of a group of highway users. Single
rating values should not be used since honest differences in opin-
i:0n are too divergent to produce a reliable rating.
5. Performance is defined as the area under a serviceability-time curve
from the time of construction to the time performance is being
evaluated.
where,
period is represented by the area under the PSI versus cumulative traffic
curve. Comparisons between competing thickness design combinations
can be made using the estimated or observed historical performance.
The performance area under the curve during a particular period is sig-
nificantly affected by the initial PSI for that period. Therefore, consider-
able emphasis has been placed on the initial PSI by many state agencies. In
fact, many states have adopted roughness or smoothness specifications that
specify the maximum allowable roughness that a roadway can have for the
contractor to receive 100 percent pay. Many of these same agencies have a
maximum permissible roughness level. If this level is exceeded, the con-
tractor may be required to repair the surface to meet specification require-
ments. Many of these same specifications pay the contractor a bonus if the
smoothness significantly exceeds the requirements.
Pavement Roughness
C:
C I,
(!)
----~0-=l'--+----+--+--+---t---11---+--+---+---+--+----+---------
o 10 20
Frequency (Hz)
Figure 8-2. Frequency Response Curve for a Mays Road Meter
Chapter 8-Perfonnance / Distress of HMA 481
ning a profile measuring device over the test section (1). Using one of these
techniques, the profile elevations can be defined, and using standard math-
ematical techniques such as Fourier analysis, the wave length of the profile
can be determined. This wave length is the input that the road provides to
a vehicle travelling over it. Different vehicles have significant differences in
wheel base suspension characteristics as well as different tire and wheel
response characteristics. Also, the way each of these vehicles responds to a
unit amplitude of road output varies with the frequency of the input. For
example, Figure 8-2 shows a response curve for a vehicle containing a
Mays Road Meter, a type of device that measures roughness by determin-
ing the amount of relative vertical movement between the rear axle of the
vehicle and the rear deck of the body of the vehicle. Notice in Figure 8-2
that the ordinate is the relative amplitude which represents the vertical
response of the vehicle to a unit of vertical input. In some frequency ranges
a unit of input produces a unit of output. Typically however, the vehicle
attenuates any vertical input by 20 to 80 percent. However, in other fre-
quency ranges, a unit of input produces significantly more than a unit of
output. These regions of response are near the natural frequency of some
component of the vehicle, either the whole vehicle or the tire wheel assem-
bly or the spring-shock absorber-axle assembly.
Since the road profile is static, it should be recognized that as a vehi-
cle travels over a fixed profile the frequency of a particular wave length
input decreases as speed increases. This relationship is demonstrated in
Figure 8-3 which shows that the typical resonant frequency range of most
automobiles lies between 1 and 10 cps. This plot shows that at any speed
there is a road profile wave length that can induce the natural frequency of
the vehicle. If the amplitude of that frequency is large, the quality of the
ride can be significantly affected.
1. Rolling straightedge;
482 Hot Mix Asphalt Materials, Mixture Design and Construction
60
50
e,~
30
I CPS Resonant Frequency
0::
l3 20
/0
0
0 /0 20 30 40 50 60 70 80
WAVE LENGTH (FeefJ
Most state agencies utilize these devices because they believe the advantages
outweigh the disadvantages.
The Surface Dynamics Profilometer (SDP) was originally developed
by General Motors and described by Spangler and Kelley in 1962 W).
484 Hot Mix Asphalt Materials, Mixture Design and Construction
The original device measured road profile in only one wheelpath; how-
ever, it has since bee::n modified to measure the profile in both wheelpaths
simultaneously.
Figure 8-4 shows a schematic diagram of how the SDP works. In the
early devices, a set of wheels were used to provide contact with the road
surface. These small wheels were held in contact with the road surface by
a set of torsion bars which applied a 300 lb. (1350 N) force to each wheel.
A potentiometer was mounted between a yoke above the road follower
wheel and the body of the panel truck or van in which the device was locat-
ed. The potentiometer output measured the relative displacement between
the road follower wheel and the body of the vehicle. An accelerometer was
mounted directly above the potentiometer base plate inside the vehicle.
The accelerometer senses the vertical acceleration of the body of the vehi-
cle. This vertical acceleration signal is integrated twice by an onboard com-
puter to produce the vertical movement of the vehicle body. Because of the
sensitivity of the accelerometer to long wave length inputs which would
saturate or overcharge the accelerometer, a set of signal conditioning filters
were built into the computer to attenuate the amplitude of the long wave
length signals which correspond to grade changes or topography.
Therefore, the computer double integrates the filtered accelerometer signal
and adds the potentiometer signal to it to produce a representation of the
road profile.
The profile data is in either analog or digital form and can be
processed using sophisticated techniques such as power spectral density
techniques. Exce::pt for onboard processing techniques, the summarization
of the road profile data is complicated and time consuming and requires
considerable development work to analyze and select the types of profile
summaries that work best for the user. Tl-iere are significant advantages to
using the SDP; however, the high capital and operating personnel costs,
the complexity of the system, and the high cost of data analysis have severe-
ly limited the use of this device within highway agencies.
Several significant advantages of the SDP are listed below:
0::
w 0::
I- w
w I-
:::E w
0 :::E
0::
w 0
_J I-
E w
u
z
3: w
u I-
n <(
- 0
a.
-
-0
-0
N
\
t:::
+
E E
NI !N NI
~ ~
z
0
I-
<(
(!) I-
a:::,
_JQ.
<( :::E
zo
<(U
Since the mid-70s, several other devices have been developed and mar-
keted that collect other roadway information in addition to roughness;
among these devices are the ARAN (ldJ, the APL (15) and GERPH O (15)
units. These devices typically collect condition survey data in addition to
roughness. Some have special capabilities for taking pictures of the road
surface or have options to mark in the computer records the occurrence of
bridges, culverts, intersections, etc. A discussion of these devices is outside
the scope of this book.
Rideability Specifications
Survey
Year Using . Considering
1981 17 13
1984 24 12
1987 32 5
'--
The equipment used to measure rideability varies substantially from
state to state. In the 1987 survey, 3 different types of equipment were used:
l. Profilographs
California-21 states
Rainhart--4 states
2. SDP profilometers-2 states
3. Mays Road Meters (MRM)-2 states
state or contractor supplied equipment, but in all five cases, the state per-
sonnel reduced the data.
Reference 16 notes that no state indicated that a study had been con-
ducted to quantify benefits of a smoother pavement as a basis for estab-
lishing bonus thresholds: However, the results of a study conducted by
Janoff (12) indicated that annual savings of up to $600 per lane mile could
result if the initial pavement smoothness was reduced from 35 to 5 inches
per mile (555mm to 80mm per km) as measured with the Mays Road
Meter. The data set used by Janoff was obtained from the states of Arizona
and Pennsylvania and contained data on pavement condition, environ-
ment, traffic, pavement cross-section, and maintenance cost. Janoff con-
ducted a statistical analysis of the data and reported that the pavements
with lower initial roughness (1) had lower roughness after 10 years of ser-
vice than'did pavements with higher initial roughness, (2) had lower crack-
ing levels after 10 years of traffic than did pavements with higher initial
roughness, and (3) required lower average annual maintenance costs after
10 years of traffic than did pavements with higher initial roughness.
--------------------,~
0 -E
I")
--
C:
IO U)
C\J U)
w
z
I
0
C\J
r-
0
0
~
U)
IO
_J
-<i
r-
0 z
0 IO O 10 0 IO O IO O IO
---'~~ st___,q:__~_!D--~-~'---------------
(!W/U!) SS3NH9n0H ~H3.l 9N01
Figure 8-5. Initial Pavement Roughness Versus Roughness after
Serving Traffic for 8 Years (after Janoff, 12)
Chapter 8-Performance / Distress of HMA 491
--------------------~
-... _ IO
rt)
o-
--
1") E
C
IO Cf)
C\J Cf)
w
z
I
0 1-
C\J 0
0
::E
IO Cf)
_J
<{
1-
0 z
IO
o...____..___o""!!-------------------.......-----...i..-~o
0
rt)
~ ¢
0
0
m
~
~
~
W
2 g g
IO rt) C\J
O
rt)
0
I
494 Hot Mix Asphalt Materials, Mixture Design and Construction
Cracking
able to the engineer to control the occurrence of thermal cracking are pen-
etration at 77°F (25°C), viscosity at 140°F (60°C), and temperature sus-
ceptibility. Increasing the asphalt penetration is the most often suggested
designer contr9lled factor for minimizing this type of cracking. However,
as pointed out in Chapter 2, asphalts with high temperature susceptibili-
ties should be avoided as these asphalts become extremely stiff at low tem-
peratures regardless of the penetration at 77°F (25°C) or viscosity at I 40°F
(60°C) The asphalt must be soft enough at the lowest anticipated service
temperature to provide some flexibility; otherwise, cracking will occur.
Low temperature cracks are typically repaired by sealing with a liquid
asphalt or other type of sealing material (ASTM D5078 and D3405).
Sealing the cracks prevents moisture from entering the base course and
subgrade and helps to minimize ravelling adj-acent to the cracks. HMA can
provide years of satisfactory service after developing transverse cracks if the
cracks are kept sealed. When the sides of the crack are at different eleva-
tions, grinding has been performed to minimize the bump at the crack,
thereby improving the ride quality. If cracking becomes too severe, the
HMA layer may have to be removed and replaced with an overlay. If all the
cracked pavement layer is not removed before the overlay, it is likely that
the underlying cracks will reflect though the overlay during the first or sec-
ond winter.
8-9. They most often occur at the joint between adjacent lanes of asphalt
mixture or at the edges of the wheel paths in a rutted pavement. These
cracks allow water to penetrate into the underlying layers, possibly soften-
ing nonstabilized layers and accelerating the development of fatigue cracks
radiating outward from the longitudinal crack. Ravelling of HMA materi-
al adjacent to the longitudinal crack often widens the crack and accelerates
the overall damage to the pavement structure.
The longitudinal cracks between adjacent lanes can be induced by low
temperature, since the density at the joint between paving lanes is lowest,
resulting in low tensile strength. Typically the density of the HMA near a
longitudinal joint is at least 2 to 3 percent lower than the remaining pave-
ment. Low tensile strength, coupled with the possibility of oxidative hard-
ening of the asphalt due to high voids in the mixture in the joint area,
makes that material particularly susceptible to cracking and subsequent
ravelling. Poor adhesion between the asphalt and the aggregate can aggra-
vate the problem adjacent to longitudinal cracking and cause more rapid
deterioration.
Longitudinal cracks at the edges of the wheel path of a tutted pave-
ment are most often caused by heavy loads or high tire pressures applied in
the wheel path. As the HMA in the wheel path is densified by traffic, resid-
ual stresses develop in the HMA adjacent to the wheel path. "When these
stresses exceed the tensile strength of the HMA, cracking occurs. This
problem is aggravated in colder climates when heavy loads are applied to a
coid asphalt mixture over a soft subgrade. Under such circumstances, a rel-
atively high deflection can occur, causing a longitudinal crack to develop
in the brittle asphalt mixture.
During early stages, these cracks are most often repaired by sealing.
After ravelling begins to develop, the repair is more extensive, since the
materials adja~ent to the crack must be removed and replaced. By the time
a HMA pavement gets into this condition, an overlay is usually the best
solution.
-·- J
Figure 8-10. Typical Case of Block Cracking in a HMA Pavement
Chapter 8-Performance / Distress of HMA 499
The name reflection cracking indicates that the crack is initiated by action
in the underlying layers that produce stresses in the HMA surface which
-----·XGeed-the-str-e.ng-th-0£-the-m-a-ter.iaL-Ther.efore,-if-the-new-surface .is bond-
ed to the old surface using a standard tack coat, cracks in the underlying
layer almost always propagate through the new surface within 1-2 years.
Some engineers believe that placement of a thin stress or strain absorb-
ing layer over the old surface absorbs much of the horizontal movement in
500 Hot Mix Asphalt Materials,'Mixture Design and Construction
the underlying layer, thereby attenuating the crack before it penetrates the
top layer of HMA. Asphalt rubber and geotextiles have been used as this
intermediate layer. Another repair technique for old HMA surfaces
involves reworking a portion, the top 2 to 4 inches (51 to 102mm), of the
old surface using recycling techniques. This reworking of the old surface
eliminates the cracking pattern in at least a portion of the old pavement
and prevents, or at least retards, the development of reflection cracking.
If the reflection cracking is caused by horizontal and vertical move-
ment of two adjacent portions of a cracked concrete pavement, reflection
cracks through the HMA surface can hardly be prevented. One technique
that has been successfully used to prevent the recurrence of these cracks is
breaking up the underlying slabs into small pieces and then seating them
into the subbase layer using heavy rollers. This cracking and seating process
is followed by a thick (usually equal to or greater than 4 inches or 102mm)
HMA overlay. Another technique that has been successfully used to treat a
concrete pavement prior to overlay is to rubblize the existing concrete. This
method requires a thicker overlay than the crack and seat method but is
not as likely to develop reflective cracks. A third process that has been used
----ro treat reflective··cracks overconcrete pavemertts·ts-to saw cut fnrotignfn 6
'e~------
HMA overlay directly over the existing joints in the concrete. These cuts
are made shortly after construction and before the reflection cracks devel-
op. This produces a straight cut in the HMA that can be easily sealed with
a material such as that specified in ASTM D3405 instead of a wandering
Chapter 8-Performance / Distress of HMA 501
Distortion
tives which can provide positive drainage of water from within the pave-
ment structure. However, the surface cracks can be kept sealed and the
drainage ditches adjacent to the shoulders can be kept free flowing in order
to quickly remove surface water to prevent intrusion of excess water into
the foundation of the pavement structure. Since most soils and many gran-
ular materials lose strength at high water contents, it is good practice to
keep moisture levels as low as practical in order to maintain pavement
strength. ·
Rutting can be caused by densification in these same pavement mate-
rials because of insufficient initial compaction. The pavement layers may
be compacted to meet specifications; however, the specified density may be
too low. An NCHRP study on granular bases conducted by Rostron, et al.,
(22) showed that many states used AASHTO"T99 compaction efforts on
granular bases even as late as 1971. Base materials compacted to this refer-
ence density are susceptible to significant additional densification under
traffic. Therefore, it is imperative that materials of all types be compacted
during construction to levels very near those densities which will be
reachea after serving traffic for several years.
Perhaps the most common recent cause of rutting is associated with
the HMA layer, especially on routes with heavy loads and high tire pres-
sures. Much of this rutting can be attributed to improper mix design. Some
of the most common mistakes when designing heavy duty HMA mixtures
are:
Most results have shown that for high volume roads the laboratory
compactive effort should be at least equivalent to 75 blows per side with
the Marshall manual hammer. Some field studies indicate that even 75
blows with a manual hammer do not produce the density ultimately
achieved in the field under traffic. The obvious result is overfilling the
voids of the in-place HMA mixture with asphalt cement which leads to
plastic flow of the HMA and rutting.
Another problem that has resulted in rutting on a number of projects
has been the use of an excessive amount of material passing the No. 200
sieve (filler). Typically, filler material fills the voids in an asphalt mixture
and lowers the optimum asphalt content. Some amount of filler is neces-
sary to obtain the desired stability, but excess filler results in a mixture at
optimum asphalt content that is brittle and which tends to crack. The
asphalt content must be adjusted for higher filler contents; otherwise, rut-
ting will likely occur.
Mineral filler characteristics vary with the gradation of the filler. If the
size of the mineral filler particles is smaller than about 10 microns, the filler
acts as an extender of the asphalt cement since the thickness of most
asphalt films in dense-graded HMA is less than 10 microns. If the miner-
al filler size is larger than 10 microns, it acts more like an aggregate. If an
excessive amount of this larger sized mineral filler is present, the asphalt
demand may increase because of increased VMA. Therefore, care must be
taken to consider not only the amount of mineral filler, but also its size in
evaluating whether an excessive amount of fine material is present in a mix
being designed or controlled in the field. There is an interaction between
the mineralogy of the fines and the asphalt cement which can cause the
behavior of some material combinations to vary. Discussion of these fac-
tors can be found in numerous reports in the literature by Anderson (25,
26, 22), Craus, J., et al. (2.i), Eick and Shook (2.2), and from several high-
way departments including California (312), West Virginia (3.1),
Washington (32), and Pennsylvania (33).
Mixtures designed with high amounts of natural sands and uncrushed
gravels have low stabilities and are prone to excessive rutting. Many states
require that the coarse aggregates be sufficiently crushed to incxease
Marshall stability above a minimum level. Natural sand is often rounded
or subrounded, which leads to low mixture stability and aggravates the rut-
----~ri~o-g. p.r.o.blem--..Na:t:.ural sands sometimes contain cfay _balls and organic
material which lead to an overall reduction in performance. Many states
limit the amount of natural sand in heavy duty HMA mixes to 10 to 25
percent. The project specifications should specify the percent of fractured
faces for both the fine and coarse aggregate.· Typical requirements for
506 Hot Mix Asphalt Materials, Mixture Design and Construction
crushed coarse aggregate call for at least 75 percent of the particles to have
two or more fractured faces and that at least 90 percent of particles to have
at least one fractured face. Typical requirements for crushed fine aggregate
call for at least 90 percent of the particles to have two or more fractured•
faces not including the natural sand.
eration. Bus stops are often the first locations to exhibit premature distor-
tion in the form of shoving. Shoving can be easily detected at edge lane
markings where these lines bend outward in the vicinity of the shoving.
Corrugations typically show up in intersection areas on the side where
vehicles stop, but can occur on the acceleration side if high acceleration
forces occur. This unstable HMA forms ridges of material which are per-
pendicular to the direction of the applied shear forces which cause the flow.
A typical corrugation pattern is shown in Figure 8-15.
Shoving and corrugations occur when the applied forces exceed the
shear strength of the HMA or underlying layers. Shoving can occur when
thin HMA surfaces are placed over granular bases and subbases. Shoving
and corrugations are primarily due to unstable HMA layers. Most of these
problems are similar to the same problems discussed under rutting.
Disintegration
by a loss of bond between the aggregate particle and the asphalt coating. A
number of conditions can lead to ravelling:
Wear Loss. Wear can either be gradual or it can occur very quickly in
some instances. After asphalt cement has oxidized and is eroded, the shear
stresses between tire and surface may be sufficient to remove aggregate par-
ticles. This type of wear is accentuated by studded tires and can become
catastrophic on surface treatments. Brittle materials are particularly sus-
ceptible to this type of wear.
Generally, wear is of the gradual, continuous type, at least as far as
individual aggregate particles are concerned. This wear is greatly accelerat-
ed by loose particles on the surface which serve as an abrasive. Studies of
wear of various minerals were made (~ by subjecting ten mineral sam-
ples, predominantly oxides, to abrasion from dry fine abrasives.
Microscopic photographs of worn surfaces revealed two phenomena,
- - - - scratching and pitting~ Rapidly wearirii m1nerals suffered both types of
damage while slow wearing minerals displayed no scratching and only a
small amount of pitting. Rapid wear occurred when the abrasive was hard-
er than the mineral and was found to be proportional to load and inverse-
510 Hot Mix Asphalt Materials, Mixture Design and Construction
Air voids in the HMA pavement may become saturated with water
even from vapor condensation due to water in the subgrade or subbase. A
temperature rise after this saturation can cause expansion of the water
trapped in the mixture voids resulting in significant void pressure when the
voids are saturated. The pore pressure from stresses induced by traffic can
also cause the failure of the binder-aggregate bond. Initially, the traffic
stresses may further compact the mixture and trap or greatly reduce the
internal water drainage. Therefore, the internal water is in frequent motion
(cyclic) and considerable pore pressure is built up under the traffic action.
Tell-tale signs of water damage to HMA overlays (over concrete pave-
m.ents) have been described by Kandhal et al (35). They observed wet spots
on the HMA overlay surface scattered throughout the project. Usually at
these wet spots water oozed out during hot ah@rnoons. Some of the wet
spots contained fines suspended in the water which were tracked on the
pavement by the traffic and appeared as white spots. Most white spots
turned into fatty areas (resulting from asphalt stripping and migrating to
the surface) which usually preceded the formation of potholes. Figure 8-18
shows all three stages: white spots, fatty areas, and potholes on a four-lane
highway. Figure 8-19 shows severely stripped aggregate in a pothole.
2. Inadequate Compaction:
The presence of dust and clay coatings on the aggregate can inhibit an
intimate contact between the asphalt cement and aggregate, and provide
channels for penetrating water. The asphalt cement coats the dust coating
and is not in contact with the aggregate surface. It has also been hypothe-
sized that some very fine clayey material may cause stripping by emulsify-
ing the· asphalt cement binder in presence of water.
-------0-ev-er-aLst-ates-i.n-the..south~tem-Uni-ted-States-experienced stripping
in the HMA course underlying open-graded asphalt friction course
(OGFC) during the late 1970s. It has been hypothesized that the OGFC
retains moisture for a longer time and does not dry out after rain as fast as
a conventional dense-graded HMA surface. The water in OGFC is also
514 Hot Mix Asphalt Materials, Mixture Design and Construction
pressed into the underlying course by the truck tires initiating the stripping
action which can cause flushing, rutting or shoving at the surface. Several
states suspended the use of OGFC in early 1980s. Some studies have also
shown that the stripping in the layers underlying OGFC resulted from
their high air void content (lack of adequate compaction). Evidently, it is
all the more desirable to have an impervious HMA course below the
OGFC to minimize stripping. If the OGFC must be used, the underlying
HMA course should be tested for moisture susceptibility.
Potholes. Potholes are relatively small holes that penetrate all the way
through the surface of HMA and down into the base course. Potholes
occur most frequently on roadways with relatively thin HMA surfaces, typ-
ically 1 to 2 inches (25 to 50mm), and seldom occur on roadways with
HMA surfaces more than 4 inches (100mm).
The occurrence of potholes often coincides with a period of heavy
rainfall during which water penetrates the HMA surface through cracks,
usually closely spaced alligator cracks, and softens the granular base course.
In northern climates, potholes generally occur during the spring thaw. If
the cracked pieces of the pavement lose their bond to the surrounding
HMA material, they are removed by the action of the tire as it rolls over
the material. Fine material is pumped through the cracks so that the under-
lying base support is weakened, resulting in removal of the adjacent mate-
rial by traffic. Once the first piece is dislodged, the pothole grows rapidly
since all other pieces are more easily dislodged than the first piece.
Potholes can develop along any crack that occurs in a pavement. For
example, either longitudinal or transverse cracks can deteriorate at loca-
tions where the wheel continually crosses the crack. Since all unsealed
cracks allow water to penetrate, base softening occurs which leads to either
ravelling adjacent to the crack or fatigue cracking on the edges of the crack.
Once ravelling or fatigue cracks begin to occur, pothole development is a
real possibility.
Pothole development can be minimized if good surface and subsurface
drainage is provided. Potholes are normally repaired by patching with
either cold mix or HMA. In either case, proper patching techniques are
absolutely essential if the integrity of the pavement is to be restored in the
pothole area. Discussion of proper maintenance techniques is contained in
Chapter 9.
------~~--------------------- ----- ------ -
FN = 100 FIL
where,
FN = friction number;
F = frictional resistance to motion at the pavement
surface; and
L = load normal to the pavement surface.
On the road, most of the polishing probably takes place while the
pavement is dry because most pavements are dry 96-97 percent of the time
_ _ _ ______,.4@__The-number-0£passes,_w.hethe.c...0£.a_pacLor-a...tire,-go:v:ernuhe. polish-
ing rate, but typically it takes two or more years of service before the sur-
face has been polished to its final state. The frictional resistance may, how-
ever, continue to drop because wear continues to reduce the macrotexture,
decreasing the channels available for clearing the water from under the tire.
522 Hot Mix Asphalt Materials, Mixture Design and Construction
REFERENCES
1. The WASHTO Road Test, Part 2: Test Data Analyses and
Findings. Highway Research Board Special Report 22, 1955.
2. Carey, W.N. and P.E. Irick. The Pavement Serviceability -
Performance Concept, Highway Research Board Bulletin 250,
1960.
3. Haas, R. and WR. Hudson. Pavement Management Systems.
McGraw-Hill Book Company, New York, NY, 1978.
4. Spangler, E.B. and WJ. Kelley. GMR Road Profilometer: A
Method for Measuring Road Profile. General Motors Corporation,
December 1964.
5. Hveem, EN. Devices for Recording and Evaluating Pavement
Roughness. Highway Research Board, Bulletin 264, 1960.
6. Hudson, WR., WE. Teske, K.H. ·Dunn, and E.B. Spangler. State
of the Art of Pavement Condition Evaluation. Highway Research
Board Special Report 95, 1968.
7. Roberts, EL. State of the Art of Estimating Pavement Serviceability
Using Roughness Measurements. Proceedings, American Society of
Civil Engineers, Specialty Conference in Pavement Design for
Practicing Engineers, Atlanta, GA, June 1975.
8. Wambolt, J.D. State of the Art of Measurement and Analysis of
Road Roughness. Transportation Research Board, Transportation
Research Record 836, 1981.
9. Hudson, WR., G.E. Elkins, W Uddin, and K.T. Reilly. Improved
Methods and Equipment to Conduct Pavement Distress Surveys.
Federal Highway Administration Report FHWA-TS-87-213,
1987.
10. Brokaw, M.P. Development of the PCA Road Meter: A Rapid
Method for Measuring Slope Variance. Highway Research Board,
Highway Research Record 189, 1967.
11. Phillips, M.B. and G. Swift. A Comparison of Four Roughness
Measuring Sy$tems. Highway Research Board, Highway Research
Record 291, 1969.
12. Rainhart Co. Mays Ride Meter Booklet, Austin, Texas, 1972.
---~U._S_p_angler,_E...R_and__W...J_.__Kelle.y_Semo.:S.eis.m-i-C-Method-0£_ _ _ __
Measuring the Road Profile. Highway Research Board, Bulletin
328, 1962.
14. Automatic Road Analyzer Mobile Data Acquisition Vehicle.
Product Bulletin, Highway Products International, Inc., 1985.
Chapter 8-Perforrnance / Distress of HMAs 523
INTRODUCTION
1310
:r:
<!lu
z~
u J 8
<{ Cl)
LJ...C/)
a::w
~~ 6
Cl)!:!
:::,:C
o1-
z1- 4
:EZ
~'.j
iii~ 2
5
0
w 0
T5-3/4"
1
1,000,000 2,000,000 3,000,000
ACCUMULATED 18,000 LB. EQUIVALENT AXLES
,..,.,.,..,=== $ 56,000
ili\~1111
Before Failure After Failure
Figure 9-3. Cost Comparison for Overlay Strategies Shown in Figures
. 9-1 and 9-2 (1)
VERYriiiiiiiiiiiiiiiiiiii7~5.;%;T:l:ME~--~:t
GOOD
40 % EACH $1.00 OF
QUALITY DROP RENOVATION
GOOD
!~ COST HERE •••
FAIR
40 % ~ILL COST $4 TO
·- -;JIii-
POOR QUALITY DROP $5 IF DELAYED
TO HERE
VERY
POOR
4 8 12 16
YEARS
by the same group of conditions such as roadway type, traffic volume and
makeup, federal aid available, and expectations of the public.
The importance of effective maintenance is reflected in the fact that
while, nationwide, the number of miles of the roads in poor condition sta,-
bilized by 1977, the number of miles deteriorating from good to fair con-
dition increased. The FHWA estimated that the amount of total highway
expenditure used nationally for physical maintenance in 1975 conJinued
to rise at a rate of about $300 million per year and amounted to almost
one-third of total highway expenditures. By the late 1980s the rate of
increase in maintenance funds grew to almost $600 million per year and
amounted to over half of total highway expenditures. This change in
expenditure category from a capital intensive program to a maintenance
intensive program reflects the maturing of tli:e highway system, especially
as the construction o(the interstate system has been completed and the
emphasis changed to rehabilitation and reconstruction of the oldest sec-
tions of the interstate system.
The proper identification of pavement defects and the determination
of their causes is crucial to the selection of proper maintenance treatments
and repair strategies. In addition, a recognition that different pavements
deteriorate at different rates is important in selecting the proper time for
maintenance. Factors which affect pavement deterioration rates are traffic
loads, weather, materials, thickness, construction quality, and effectiveness
of previous maintenance. Generally, the rate of deterioration increases with
use and age.
The primary reasons for maintenance are to correct deficiencies pro-
duced by distress, to protect the investment, and to provide a smooth, safe
road for the user. Therefore, it is important to determine which distresses
relate most to these factors. Darter and Shahin (3) identified what they
believed to be the principal factors which initiated maintenance and the
types of distresses that were related to those factors (shown in Figure 9-5).
Only alligator cracking, ravelling/weathering, and rutting show up under
all three of the principal factors which initiate maintenance.
MSHTO (.1'J has identified 17 types of HMA pavement distresses,
which are listed in Table 9-1. Table 9-2 relating distresses and mainte-
nance/ rehabilitation methods prepared by Hudson et al. (i) should be use-
ful in familiarizing the reader with the elements of typical maintenance and
---;ehabil-i1:a1:i-0-R-1:eehn-i-q_-1:1-@s.-'I'al,le~-9-gi:ves-t-he-p0ssie-le-eauses--of--l-7--El.-ist.r~s,,,_____
types and rehabilitation alternatives recommended by the FHWA (.6).
Various maintenance, rehabilitation, and reconstruction techniques gener-
ally used by the highway agencies are discussed in the following sections.
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 531
• Pothole Potential
• Cracking
Weathering/
Ravelling
• Traffic Volume
MAINTENANCE
Generally, maintenance activities are divided into two categories, pre-
ventive and corrective maintenance. Preventive maintenance is that group
of activities performed to protect the pavement and decrease the rate of
deterioration of the pavement quality. Corrective maintenance is that
group of activities performed to correct a specific pavement failure or area
of distress. Some maintenance methods serve both functions. The follow-
ing sections include descriptions of the most common maintenance meth-
ods, distresses which are repaired by each method, and typical life
expectancies for each maintenance method.
Preventive Maintenance
Surface seals and crack sealing are discussed in this section. Surface
seals are those maintenance activities consisting of applications of asphalt
alone or asphalt and aggregates which are applied continuously to the
whole surface of a traffic lane. These surface seals are most often applied to
532 Hot Mix Asphalt Materials, Mixture Design and Construction
Note: It is difficult to attribute all of a particular distress to only one cause. It is recognized that
in the above list, some types of distress might arguably have more than one cause.
(1) rejuvenate or retard the oxidation of asphalt at the surface; (2) restore
skid resistance of the surface; (3) seal fine cracks which have appeared at
the surface; (4) prevent the intrusion of water into the pavement structure
through cracks which penetrate through the HMA layer; and (5) retard the
ravelling of aggregates from a weathered, disintegrating surface. The most
common types of seals applied to HMA surfaces are (1) fog seals, (2) reju-
venators, (3) chip seals or surface treatments, and (4) slurry seals (.2). Each
type of surface seal is discussed separately.
-------1Fog-Seals.-fug-seal-s-inv:0-l¥e-a-light-ap-pliGat:i0:F1-0£-a-slew-se1:1ing--eFnul-------
sion to an oxidized pavement surface. The primary objective is to restore
or rejuvenate the asphalt cement at the surface of the roadway. Often the
aged surface has a light gray rather than dark gray or black appearance with
fine cracks apparent and, in many cases, there is evidence of ravelling. The
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 533
-- _--...
n
PoulbleCau11 RehabUltalloa
\0
Con-- , &
Dlllnn Type Stntdural
Fallun
Mb
....
Compod·
Temp. or
MolUun
Chana;e1
lloa
Suda«
Reqdlna;
Thlft
Ottrlay
. .,....
Op,,,-
Gnd<d
..,_uni
"-lay 11,q<llq
~~
a~g an
1. AJligalor or Fatigue Cracking X X X .x
2. Blttding X X X X
3. Block Cracking X X X X X
4. Com.1gation X X X X x' X X X
~~
....
5. Deprn1ion X X X X n
6. Joint ReOed:ion Cracking X X X ~9 ::.:
from PCC Slab
;. ~
S;::
~
7. Lane/Shoulder Dropoff or Heave X X X
;;;·
8. Lane/Shoulder Separation X X X
"' ►
9. longiludlna.1 and Transverse Cracking X X X .... 'g.
JO. Palcb Decerioralioa X X X X = ~
S;::
11. Polished Aurcgale X X X X X
12. Polho~ X X X
i
X ~ ~
8.
i:r
~
,,,
e?..
-........
X X
13. Pumping and Waler BJc.cding
X
X
X X X
.... S;::
;;;·
X X X X xi X X X
.... ij
~
15. Rulling
(I)
16. Slippage Cracking X
0 ti
17. Swell X X X = (I)
[!l.
{IQ
:::,
....,
0
8.
Notes: I. If problem is extensive enough (')
2. When rutting is minor
~
0
:::,
3. Over milled surface
'"g.. g
e.
....
~-
0
0
:::,
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 535
public will complain about loose aggregates on the surface which are
picked up by vehicle wheels and can chip paint and break windshields.
Surface treatments add little load-carrying strength to the pavement struc-
ture. However, when properly designed and constructed, they are eco-
nomical, easy to place, and reasonably durable. Not only do they provide
a waterproof cover over the existing pavement structure, they resist surface
abrasion by traffic. Surface treatment is also used in skin patches which are
placed directly on localized distressed areas.
Cover aggregate should be clean; otherwise, it does not adhere to the
asphalt material. The ideal shape of the aggregate is· cubical. Flat or elon-
gated particles are not desirable because they lie flat on the surface and nor-
mal amounts of asphalt binder covers them, resulting in bleeding or flush-
ing surface conditions. A one-size cover aggregate is preferred for surface
treatment. However, many states use a graded cover aggregate because
one-size aggregate is not economical.
Asphalt binder can be asphalt cement, cutback asphalt, or emulsified
asphalt. Most states prefer emulsified asphalts (rapid setting grades)
because they can be used with damp aggregate, set up more rapidly than
cutback asphalts, and eliminate the fire hazard that is associated with the
use of cutback asphalt. Asphalt cement is used in hot climate areas.
However, the aggregate must be dry and very clean, and be applied before
the asphalt cement cools down to ambient temperature. Cutback asphalts
Quantities
Nominal AASHTO
Aggregate Aggregate Aggregate 1.l Asphalt 1,3
Size Range, in. Designation lb/sy gal/sy
3/4 to 3/8 6 40-50 0.40-0.50
1/2 to No. 4 7 25-30 0.30-0.45
3/8 to No. 8 8 20-25 0.20-0.35
No. 4 to No. 16 9 15-20 0.15-0.25
Sand M-6 10-15 0.15-0.20
1Toe lower application rates of asphalt shown are for aggregates on the fine side of the spec limits.
----lli+iligher-application-rates-are-for-aggregates-on-the-coarse-side·ofthe,pec-limi,-.,.~_- - - - - - - - - -
2Weights are based on an aggregate specific gravity of 2.65. If the specific gravity is greater than
2.75 or less than 2.55,' multiply the table value by the aggregate specific gravity divided by 2.65.
3
Asphalt is an emulsion of acceptable grade and type. Consult Reference 2 or state specifications
for recommendations.
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 537
are normally not used due to environmental problems, cost, and safety
problems (fire hazard).
Although Table 9-4 gives a range of application rates for asphalt
binder and cover aggregate for specific aggregate sizes, it is recommended
to use a design method to compute the application rates. A vast majority
of highway agencies still use the quantities of binder and cover aggregate
determined by experience and/or precedent. This may result in surface
treatments that have poor performance characteristics, but past experience
with particular aggregates and asphalts must be used to ensure best perfor-
mance.
An excellent review of surface treatment design procedures existing
prior to 1968 was made by Herrin, et al. (JJ.). These include the following
design methods: (a) Hanson, (b) California, (c) Nevitt, (d) modified
Kearby, (e) Lovering spread modules, (f) European, (g) McLeod, (h)
Mackintosh, (i) American Bitumel, and (j) Asphalt Institute.
Design formulas for simple surface treatments using one-size and
graded aggregate are given in Reference 2- Reference 10 gives McLeod's
method in detail. His method is a modification of the method developed
by the County Roads Board of Victoria, Australia. The method is based on
the following principles:
While some aggregate loss always occurs, most of the typical problems
' can be alleviated by using well maintained equipment and knowledgeable
construction crews. Typical construction problems and their solutions
include:
HMA facility. Using 0.6 to 1.2 percent asphalt cement by weight of the
aggregate is sufficient for precoating. The precoated aggregate should be
free flowing so that it can be. applied with an aggregate spreader.
EMULSION
CONVEYOR
-------------,; ~:~>----------~
!' SLURRY ~;X/ .
Figure 9-8. Flow Diagram of a Continuous Slurry Mixer (1f!)
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 541
5. After allowing the sealant to cool, broom off the excess sand,
remove traffic control devices, and open to traffic.
A wide variety of crack filler materials are in use today. However, rub-
berized asphalt crack sealers have been found quite effective (.1.§J.
Specifications for such crack sealers are contained in ASTM D5078. The
prepackaged material is heated and applied using a special piece of equip-
ment that heats and dispenses the material. Since the heating kettles used
by most manufacturers involve flame heating, all appropriate safety proce-
dures should be followed, including the use of goggles and heavy protec-
tive clothing by the sealing crew. Special precautions are required to keep
water away from hot asphalt since hot asphalt will foam dramatically
(approximately 15 times its volume) when it comes in contact with water.
Serious burns and fires can result when handling and safety procedures are
ignored.
Corrective Maintenance
Corrective maintenance generally consists of patches, chip seals, and
thin HMA overlays. Since chip seals have already been discussed, only
patching and thin overlays are included in this section.
The patching procedure described above achieves the best results when
repairs are made during warm, dry weather. Unfortunately, not all repair
can be postponed until after inclement weather. Therefore, it is possible
that some patching must be done during cold, wet weather. In such cases,
temporary repairs are made using cold mix asphalt patching materials and
emergency, temporary patching procedures.,_In such cases, the following
patching procedure is often followed until permanent repairs can be made:
4 ,,~··:;,·,·:=:.:~,:-.; :·!'.=~·~·-,·-o'.:,
1
inu ~o1:pih b~~ ~i~~f!~i~d
%,·~-:~:~·o:; "::"!···; :"-:·:6. ::~. :•..:·,f;·_ ~~o.··~.;#q
Table 9-6 shows that the pothole procedures can be divided into three
different groups as a function of the types of materials and equipment used
and the intensity of the effort applied to make the patch permanent. The
groups are as follows:
548 Hot Mix Asphalt Materials, Mixture Design and Construction
Group I, which involves the use of cold mix material, utilizes no effort
to remove loose debris from the hole and little effort to compact the cold
mix placed in the hole. For method No. 1, compaction is simply whacking
the loose material twice with the backside of a shovel; for method No. 2,
compaction is with the truck wheels. The life of these patches is short, I to
2 months, and the cost is very high at about $300/ton and $200/ton annu-
ally for methods I and 2 respectively.
Groups 2 and 3 of Table 9-6 include the use of HMA, tacking the
sides of the hole with asphalt, and compaction with power equipment.
Methods 3 and 4 include shaping the hole with hand equipment, remov-
ing loose material from the hole, and compacting with a wacker tamper for
Method 3, and with a pup roller for Method 4. These improvements in
procedure produce significant increases in life expectancy of the patch
from 1 or 2 months (for methods 1 and 2) up to 12 months at about 1/3
the annual cost per ton of material. Group 3 (Method 5) is similar to
Group 2 except that the patched area is shaped with power equipment
which slightly increases the cost per ton of material, but extends the aver-
age life of the patch.
The use of Method 1 is warranted only in the most extreme circum-
stances because the costs per ton of mix are so high. All maintenance forces
should be trained in the procedures described in Methods 3, 4 and 5 and
be told the cost consequences of using Methods 1 and 2.
ply does not remove that roughness. Therefore, the total overlay thickness
should vary depending on the type of roughness components present in the
road profile in order to allow greater overlay depths when warranted by a
roughness survey. Many DOTs utilize levelling courses in an attempt to
~
(')
-
::r
O"' ~
n ....ft
i
\0
Cost per ton
Tons In place
per shift
or material
in place
ure or
patch
Annual cost
per ton
~
ei.
No, .. I l Description or Method >X Equipment Required (7.5 hrs) ($) (months) ($)
~n a
(>
::,
~ ~ §
Fm hole in one lift with mixture, and
corpact by hitting the patch with the back
Dump truck
shovels
18.0 25.64 1 307.68
e.,... 0
'
2 I Sate as method no. 1, except compaction Dump truck 12.0 31.80 2 190.80 ~
....
fl>
is erformed with the tire of the dump shovels 0 F
truck.
=
fl> §
p.
3 I Sh~pe the area to be patched with an axe Dump truck, pickup, 6.0 63.29 1• 63.29 ~
anr sledge, remove loose asphalt with heating kettle, wacker, . 8"'
>'t
(>
0
0
mhtock, sweep area clean, tack the exposed axe, sledge, brooms, ::,
.,,
su aces of patch area, shovel in material,
anlt level with lute. Compact with wacker
mattocks
....t:, g
~
0
i:t.
(vi~rat_ory compactor) and seal edges with 0
ta~k 011 and # 1B stone. g
....
::,
....,
0
~
4 I Sa~e as method no. 3, except a pup roller Dump truck, pickup, 7.0 61.41 7• 61.41
is rsed for compaction. heating kettle, pup roller,
5
I
I Saitle as method no. 4, except the area-to
be: patched is shaped with a pavement
axe, sledge, brooms, mattocks
e-~
•For methods 3, 4 and 5 it was assumed that failure would occur after 12 months, though in ail probability patch life would extend beyond one year.
-
0
n "'
.ia,.
\C
550 Hot Mix Asphalt Materials, Mixture Design and Construction
smooth out a profile that exhibits severe distortion in either the horizontal
or transverse profile. In many cases it is more satisfactory and economical
to use a milling machine to mill the old pavement to a plane surface rather
than to use levelling courses.
Where strengthening the pavement structure is the primary objective
for the overlay, thin overlays are seldom cost effective, even if several are
placed over a period of years, primarily because thin overlays cool quickly
and therefore are difficult to compact to the required air void content
which yields optimum strength and durability. Also, thin HMA overlays
are generally prone to debonding from the existing surface. Therefore, it is
imperative, when pavement strengthening is needed, that the engineer per-
form an adequate analysis to establish the cg_ndition of the existing pave-
ment; determine the strength properties of the in situ material; determine
the length of the performance period required for service and estimate the
traffic to be served; and determine the required overlay thickness using one
of the overlay design methods included in References 1., 12, 20, or the
appropriate procedure of the agency. Only by using good engineering
design procedures and sound life-cycle economic cost analyses can the
most cost effective combination of overlay design thickness and future
maintenance activities be determined.
If thin HMA overlays must be used, it should be ensured that (a) the
maximum aggregate size in HMA mix is appropriate for the overlay thick-
ness, (b) proper tack coat is applied, (c) work is carried out in warm weath-
er so that desired compaction level is obtained, and (d) good HMA con-
struction quality control is maintained.
REHABILITATION
Placing the HMA Overlay. New HMA is placed over the scarified
(and rejuvenated if necessary) mix. If the purpose of surface recycling is to
remove only surface deficiencies in a structurally adequate pavement, a
thin HMA overlay will suffice. On the other hand, properly designed
552 Hot Mix Asphalt Materials, Mixture Design and Construction
Cold Milling
Milling machines have become more efficient during the last few years
______a_s_m_an-ufacturers have mcreased the size of the power umts, simplified the
hydraulic systems used to control the machines, made many parts inter-
changeable hetween model lines, and made improvements in tooth design,
including tooth shape and materials to allow more tonnage removed per
554 Hot Mix Asphalt Materials, Mixture Design and Construction
'21,11t111!ll(1,1YP
(a) Wedge sections used to correct a depression.
~ 2 n d layer
I st layer
Structural Recycling
Unlike hot surface recycling, which is usually limited to the top 1-inch
of HMA pavement, structural recycling can be accomplished to any
r-
Structural Overlays
I . Bending stresses increase as the load gets closer to the crack and
reach a peak when the load is directly over the crack, point B, and
then reduce as the load moves away.
2. Shear forces build up in the HMA overlay as the load approaches
the crack.
(a) The shear force reaches a maximum just before the load reach-
es the crack, point A, and then reduces to zero immediately as
--------he-lo-ad-crosses-over-tlre-crac1~ . - - - - - - - - - - - - - - - - - -
(6) The shear force then builds up rapidly in the opposite direc-
tion as the load approaches point C and then gradually
reduces toward zero as the vehicle approaches the center of the
concrete slab.
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 557
OVERLAY
~en
a. Zen
j:: 15~
w:.:: QI-
:cu men
I-<
1-5
<w A
en:c
WI-
en ~en
en LL ~en
WO
a:::
a:::w
I-
<a:::
w1-
en V5en
The time period over which this loading sequence occurs is fairly
short, approaching 0.05 seconds at highway speeds. At these loading rates,
the stiffness of both the concrete and the HMA material is very high.
------~~ohably_the__m_Qs_t_s.ignific.an.t.foadmgJ.sJl.o_uhaLd_u_e_tQ_whed loads,
but those induced by temperature cycling. Figure 9-14 shows schemati-
cally the effects of temperature on the thermal contraction and curling of
a concrete slab and the effect of that movement on the deflected shape of
the BMA overlay.
558 Hot Mix Asphalt Materials, Mixture Design and Construction
TEMPERATURE DECREASE
OVERLAY
be divided into 4 groups: (1) increasing HMA overlay thickness, (2) per-
forming special treatments on the existing pavement, (3) performing spe-
cial treatments only on the cracks and/or joints in the existing pavement,
and (4) special consideration of the HMA overlay design.
roller. The Arkansas DOT has experimented with the use of an asphalt sta-
bilized, thick (up to 6 inches) (150 mm) interlayer oflarge, single-sized (2
to 3-inch) (50-75 mm) aggregate before overlaying with HMA. This mix,
called the Arkansas mix, has been tried with mixed results (2!SJ in that the
Arkansas mix has delayed reflection cracking but not prevented it.
One other technique which has been used to retard reflection cracking
is to coat the existing surface with a thin layer of asphalt, asphalt rubber,
asphalt and fabric, or a seal coat of asphalt and aggregate. This layer is often
called a stress-absorbing membrane interlayer (SAM!) since the objective
of the treatment is to provide a very flexible, soft layer of asphalt or asphalt
rubber which will not readily transmit the stress produced by horizontal
movement of the underlying layer to the HMA overlay. This soft, strain:..
absorbing layer prevents propogation of the ;rack tip from the old surface
to the HMA overlay because it is expected to absorb high strains witho~t
cracking. Field experience with these layers has also been varied with vari-
able results reported from different full scale trial sections.· Perhaps one of
the basic problems with many of these techniques is that there is no mate-
rial property measurable in a laboratory which can be related to a theoret-
ical model which can predict the rate of crack propogation from an under-
lying layer. As a result, it is difficult to design or select a material with par-
ticular properties that can resist propogation of reflection cracks.
15 in.
r-----------------==-Fabric
Moveable Plate
Cracking. Tight cracks that permit load transfer across the crack through
aggregate interlock with little loss of structural value, generally applicable
to plain jointed concrete pavement CTCP).
Saw and Sealing. Involves sawing joints in the HMA overlay directly
above joints and working cracks in the existing concrete pavement and
sealing the joints, generally applicable to joint reinforced concrete. pave-
ments with long joint spacings in good condition. ·
564 Hot Mix Asphalt Materials, Mixture Design and Construction
Using these definitions and the description (in Reference 2.Z) of the
pavement conditions for which these techniques have been applied, a guide
for selection of an appropriate treatment technique for concrete pavements
before overlaying has been prepared and shown in Table 9-7. Notice that
crack and seat is appropriate only for JCP, while break and seat is appro-
priate only for JRCP. Saw and seal could be used for both JCP and JRCP,
but the joint spacing on most JCP is so short that it makes the cost of saw
and seal impractical for JCP. Rubblizing and compacting is appropriate for
any type of severely deteriorated concrete pavement.
S3 ~
f
Energy Range Production Rate
"ti
i
V)
EqulpmJ.nt Type Applications (ft. lbs.) (lane-miles/day) Advantages I
I
-Umitatlons
aa
ball l
Crane and Jrecking small areas where
specialized equipment not
.25 10 .75 Equipment generally
available
Requires skilled crane
operator ta
r, Ill
justified
JCP I 3:~
Whiphammef 1.00 High productivity covers Not effective on JRCP
4o
full lane width
~,.,..
Drop ~amm¢r JCP and wire 'mesh 4,800 • 15,000 .15 to .75 Effective with JCP I Developed unusual
&~ ~
~
reinforced JRCP crack pattern with mesh
reinforced JRCP.
Several passes required
=
.;·
Guillotine JCP and JRCP 12,000 - 120,000 .75 to 4.0 Versatile - effective with
JCP and JRCP;
a
n,
=
.... i
preferred by several
states; high;productivity;
~ f
2,000
covers full lane width
Rubblized portions of
'"O
~-
J
~
Impact hamter
concrete below cracks
Pile-driving I JRCP 8,000 • 115,000 .27 Covers full lane width I Low productivity
(j
=i-
8
@
~
Resonant
breaker
paljement JCP, JRCP, CRCP <2,000 .so Excellent for rubblizing I Not effective for
making transverse
cracks i........
.,
....fl>
~-
8.
()
I
(')
fl>
s
i::i...
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 567
cps, and the input amplitude can be varied from 1.25 to 1.5 inches at an
impact force of about 2000 pounds. This force system breaks the concrete
slab apart and destroys the bond between any concrete and steel, enhanc-
ing the removal of the reinforcing steel or wire fabric if required. The steel
is often left in place, but protruding bars are cut. Joint sealers are easily
removed manually.
Saw and Seal. The basic concept behind saw and seal is that, since
reflective cracks occur in an HMA overlay on concrete pavement, it is
easier to seal and maintain a sawed straight groove than a narrow, mean-
dering crack. Kilareski and Bionda (3.1) report that saw and seal was first
suggested in 1954 but was not tried in the field for several years. The pro-
cedures first used and those in use today appear to be very similar with the
Figure 9-21. Heavy Rubber Tire Roller Used to Seat Broken Concrete
Slabs
Chapter 9-Maintenance, Rehabilitation, and Reconstruction of HMA 573
standard procedure for saw and seal shown in Figure 9-23. A review of the
literature by Kilareski and Bionda and a summary of the results from their
study indicate that:
1. The bulk of the field experience with saw and seal is in the north-
eastern U.S.; however, in recent years several other states have elect-
ed to try this method. States with some experience include Arizona
(G), Connecticut (G), Louisiana (M), Maine (G), Massachusetts
(G), Michigan (M), New Jersey (G), New York (G), North Carolina
(G), Ohio (G), Pennsylvania (G), and Rhode Island (G). (G means
good effectiveness while M means marginal effectiveness of the
method) ,
2. It is absolutely crucial that the joint or crack be accurately located
before the saw cut is made; if not correct to within 1 inch, sec-
ondary reflection cracking occurs and causes serious roughness and
maintenance problems.
3. Thick overlays (5 inches or more) perform better than thin overlays
(2.5 inches or less).
4. Compared to the control sections, properly constructed saw and
seal joints reduce roughness by 20 percent and transverse reflection
cracking by 64 percent. The saw and seal technique cannot be
applied to random cracks in concrete pavements.
RECONSTRUCTION
OLD PCC
PAVEMENT
---t - - - OPENING
i
VERTICAL
DIFFERENTIAL
DEFLECTION
Since cold recycled mix does not have adequate resistance to abrasion
by traffic and moisture induced damage, it must be capped with a wearing
course ofHMA or a single or double surface treatment. In-place cold mix
recycling is also suitable for roads which have been built over the years with
layers of surface treatments and/or cold mixes. Recycled cold mix in such
cases provides a reasonably good, uniform base course which can be over-
laid with HMA to provide a durable riding surface.
Drainage
Inadequate subsurface drainage is frequently a primary cause of the
pavement distresses which necessitate reconstruction. Therefore, when
diagnosing the cause(s) of pavement distresses, drainage must be investi-
gated thoroughly. If drainage is found to be inadequate, it must be
improved as an integral part of the reconstruction project, including
structural overlays.
Aggregate base courses or subbases are most often not free-draining as
intended. Some dense-graded granular bases may actually become saturat-
ed with water that cannot drain from the base. Subbases under some old
concrete pavements get saturated with subsurface water, and concrete slabs
start to pump water through the joints and cracks. If the concrete pave-
ment has a HMA overlay, the water under pressure can cause stripping and
potholing of the overlay (33). Asphalt treated permeable material (ATPM),
as discussed in Chapter 7, can be used as a drainage layer between the con-
crete pavement and HMA overlays.
Longitudinal drains are required only along the lower edge of travelled
lanes in flat terrain. Lateral drains must be provided in rolling or hilly ter-
rain to intercept water in the drainage layer because water is likely to flow
longitudinally on steep grades.
Reconstruction gives a unique opportunity to improve the existing
drairrage-sysrem;-irrdrrdirrg-additiona:l-rrreasures-such-as transverse drains
cut into the pavement structure and changes in the drainage pattern.
Reference 3d gives details on the hydrology and hydraulics of both surface
and subsurface drainage and describes methods for the design and con-
struction of drainage systems.
576 Hot Mix Asphalt Materials, Mixture Design and Construction
REFERENCES
INDEX