NCHRP RPT 531

NATIONAL
NCHRP REPORT 531

COOPERATIVE
HIGHWAY
RESEARCH
PROGRAM
Relationship of Air Voids,

Lift Thickness, and
Permeability in Hot Mix
Asphalt Pavements
TRANSPORTATION RESEARCH BOARD EXECUTIVE COMMITTEE 2004 (Membership as of July 2004)
OFFICERS
Chair: Michael S. Townes, President and CEO, Hampton Roads Transit, Hampton, VA
Vice Chair: Joseph H. Boardman, Commissioner, New York State DOT
Executive Director: Robert E. Skinner, Jr., Transportation Research Board
MEMBERS
MICHAEL W. BEHRENS, Executive Director, Texas DOT
SARAH C. CAMPBELL, President, TransManagement, Inc., Washington, DC
E. DEAN CARLSON, Director, Carlson Associates, Topeka, KS
JOHN L. CRAIG, Director, Nebraska Department of Roads
DOUGLAS G. DUNCAN, President and CEO, FedEx Freight, Memphis, TN
GENEVIEVE GIULIANO, Director, Metrans Transportation Center and Professor, School of Policy, Planning, and Development, USC,
Los Angeles
BERNARD S. GROSECLOSE, JR., President and CEO, South Carolina State Ports Authority
SUSAN HANSON, Landry University Professor of Geography, Graduate School of Geography, Clark University
JAMES R. HERTWIG, President, CSX Intermodal, Jacksonville, FL
GLORIA J. JEFF, Director, Michigan DOT
ADIB K. KANAFANI, Cahill Professor of Civil Engineering, University of California, Berkeley
RONALD F. KIRBY, Director of Transportation Planning, Metropolitan Washington Council of Governments
HERBERT S. LEVINSON, Principal, Herbert S. Levinson Transportation Consultant, New Haven, CT
SUE MCNEIL, Director, Urban Transportation Center and Professor, College of Urban Planning and Public Affairs and Department of
Civil and Material Engineering, University of Illinois, Chicago
MICHAEL D. MEYER, Professor, School of Civil and Environmental Engineering, Georgia Institute of Technology
CAROL A. MURRAY, Commissioner, New Hampshire DOT
JOHN E. NJORD, Executive Director, Utah DOT
DAVID PLAVIN, President, Airports Council International, Washington, DC
JOHN H. REBENSDORF, Vice President, Network Planning and Operations, Union Pacific Railroad Co., Omaha, NE
PHILIP A. SHUCET, Commissioner, Virginia DOT
C. MICHAEL WALTON, Ernest H. Cockrell Centennial Chair in Engineering, University of Texas, Austin
LINDA S. WATSON, Executive Director, LYNX—Central Florida Regional Transportation Authority, Orlando, FL
MARION C. BLAKEY, Federal Aviation Administrator, U.S.DOT (ex officio)

SAMUEL G. BONASSO, Acting Administrator, Research and Special Programs Administration, U.S.DOT (ex officio)
REBECCA M. BREWSTER, President and COO, American Transportation Research Institute, Smyrna, GA (ex officio)
GEORGE BUGLIARELLO, Chancellor, Polytechnic University and Foreign Secretary, National Academy of Engineering (ex officio)
THOMAS H. COLLINS (Adm., U.S. Coast Guard), Commandant, U.S. Coast Guard (ex officio)
JENNIFER L. DORN, Federal Transit Administrator, U.S.DOT (ex officio)
EDWARD R. HAMBERGER, President and CEO, Association of American Railroads (ex officio)
JOHN C. HORSLEY, Executive Director, American Association of State Highway and Transportation Officials (ex officio)
RICK KOWALEWSKI, Deputy Director, Bureau of Transportation Statistics, U.S.DOT (ex officio)
WILLIAM W. MILLAR, President, American Public Transportation Association (ex officio)
BETTY MONRO, Acting Administrator, Federal Railroad Administration, U.S.DOT (ex officio)
MARY E. PETERS, Federal Highway Administrator, U.S.DOT (ex officio)
SUZANNE RUDZINSKI, Director, Transportation and Regional Programs, U.S. Environmental Protection Agency (ex officio)
JEFFREY W. RUNGE, National Highway Traffic Safety Administrator, U.S.DOT (ex officio)
ANNETTE M. SANDBERG, Federal Motor Carrier Safety Administrator, U.S.DOT (ex officio)
WILLIAM G. SCHUBERT, Maritime Administrator, U.S.DOT (ex officio)
JEFFREY N. SHANE, Under Secretary for Policy, U.S.DOT (ex officio)
CARL A. STROCK (Maj. Gen., U.S. Army), Chief of Engineers and Commanding General, U.S. Army Corps of Engineers (ex officio)
ROBERT A. VENEZIA, Program Manager of Public Health Applications, National Aeronautics and Space Administration (ex officio)
NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM

Transportation Research Board Executive Committee Subcommittee for NCHRP
MICHAEL S. TOWNES, Hampton Roads Transit, Hampton, VA JOHN C. HORSLEY, American Association of State Highway
(Chair) and Transportation Officials
JOSEPH H. BOARDMAN, New York State DOT MARY E. PETERS, Federal Highway Administration
GENEVIEVE GIULIANO, University of Southern California, ROBERT E. SKINNER, JR., Transportation Research Board
Los Angeles C. MICHAEL WALTON, University of Texas, Austin
NCHRP REPORT 531

Relationship of Air Voids,
Lift Thickness, and
Permeability in Hot Mix
Asphalt Pavements
E. RAY BROWN
M. ROSLI HAININ
ALLEN COOLEY
GRAHAM HURLEY
National Center for Asphalt Technology—Auburn University
Auburn, AL
S UBJECT A REAS
Materials and Construction
Research Sponsored by the American Association of State Highway and Transportation Officials
in Cooperation with the Federal Highway Administration
TRANSPORTATION RESEARCH BOARD

WASHINGTON, D.C.
2004
www.TRB.org
NATIONAL COOPERATIVE HIGHWAY RESEARCH NCHRP REPORT 531
PROGRAM
Systematic, well-designed research provides the most effective Project 9-27 FY’01
approach to the solution of many problems facing highway
administrators and engineers. Often, highway problems are of local ISSN 0077-5614
interest and can best be studied by highway departments ISBN 0-309-088070
individually or in cooperation with their state universities and
Library of Congress Control Number 2004111830
others. However, the accelerating growth of highway transportation
develops increasingly complex problems of wide interest to © 2004 Transportation Research Board
highway authorities. These problems are best studied through a
coordinated program of cooperative research. Price $19.00
In recognition of these needs, the highway administrators of the
American Association of State Highway and Transportation
Officials initiated in 1962 an objective national highway research
program employing modern scientific techniques. This program is
supported on a continuing basis by funds from participating
member states of the Association and it receives the full cooperation
and support of the Federal Highway Administration, United States NOTICE
Department of Transportation.
The project that is the subject of this report was a part of the National Cooperative
The Transportation Research Board of the National Academies
Highway Research Program conducted by the Transportation Research Board with the
was requested by the Association to administer the research
approval of the Governing Board of the National Research Council. Such approval
program because of the Board’s recognized objectivity and reflects the Governing Board’s judgment that the program concerned is of national
understanding of modern research practices. The Board is uniquely importance and appropriate with respect to both the purposes and resources of the
suited for this purpose as it maintains an extensive committee National Research Council.
structure from which authorities on any highway transportation
The members of the technical committee selected to monitor this project and to review
subject may be drawn; it possesses avenues of communications and
this report were chosen for recognized scholarly competence and with due
cooperation with federal, state and local governmental agencies, consideration for the balance of disciplines appropriate to the project. The opinions and
universities, and industry; its relationship to the National Research conclusions expressed or implied are those of the research agency that performed the
Council is an insurance of objectivity; it maintains a full-time research, and, while they have been accepted as appropriate by the technical committee,
research correlation staff of specialists in highway transportation they are not necessarily those of the Transportation Research Board, the National
matters to bring the findings of research directly to those who are in Research Council, the American Association of State Highway and Transportation
a position to use them. Officials, or the Federal Highway Administration, U.S. Department of Transportation.
The program is developed on the basis of research needs Each report is reviewed and accepted for publication by the technical committee
identified by chief administrators of the highway and transportation according to procedures established and monitored by the Transportation Research
departments and by committees of AASHTO. Each year, specific Board Executive Committee and the Governing Board of the National Research
areas of research needs to be included in the program are proposed Council.
to the National Research Council and the Board by the American
Association of State Highway and Transportation Officials.
Research projects to fulfill these needs are defined by the Board, and
qualified research agencies are selected from those that have
submitted proposals. Administration and surveillance of research
contracts are the responsibilities of the National Research Council
and the Transportation Research Board. Published reports of the
The needs for highway research are many, and the National
Cooperative Highway Research Program can make significant
contributions to the solution of highway transportation problems of are available from:
mutual concern to many responsible groups. The program,
however, is intended to complement rather than to substitute for or Transportation Research Board
duplicate other highway research programs. Business Office
500 Fifth Street, NW
Washington, DC 20001
and can be ordered through the Internet at:
Note: The Transportation Research Board of the National Academies, the
National Research Council, the Federal Highway Administration, the American
Association of State Highway and Transportation Officials, and the individual
http://www.national-academies.org/trb/bookstore
states participating in the National Cooperative Highway Research Program do
not endorse products or manufacturers. Trade or manufacturers’ names appear
herein solely because they are considered essential to the object of this report. Printed in the United States of America
The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished schol-
ars engaged in scientific and engineering research, dedicated to the furtherance of science and technology
and to their use for the general welfare. On the authority of the charter granted to it by the Congress in
1863, the Academy has a mandate that requires it to advise the federal government on scientific and techni-
cal matters. Dr. Bruce M. Alberts is president of the National Academy of Sciences.
The National Academy of Engineering was established in 1964, under the charter of the National Acad-
emy of Sciences, as a parallel organization of outstanding engineers. It is autonomous in its administration
and in the selection of its members, sharing with the National Academy of Sciences the responsibility for
advising the federal government. The National Academy of Engineering also sponsors engineering programs
aimed at meeting national needs, encourages education and research, and recognizes the superior achieve-
ments of engineers. Dr. William A. Wulf is president of the National Academy of Engineering.
The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the
services of eminent members of appropriate professions in the examination of policy matters pertaining
to the health of the public. The Institute acts under the responsibility given to the National Academy of
Sciences by its congressional charter to be an adviser to the federal government and, on its own initiative,
to identify issues of medical care, research, and education. Dr. Harvey V. Fineberg is president of the
Institute of Medicine.
The National Research Council was organized by the National Academy of Sciences in 1916 to associate
the broad community of science and technology with the Academy’s purposes of furthering knowledge and
advising the federal government. Functioning in accordance with general policies determined by the Acad-
emy, the Council has become the principal operating agency of both the National Academy of Sciences
and the National Academy of Engineering in providing services to the government, the public, and the
scientific and engineering communities. The Council is administered jointly by both the Academies and
the Institute of Medicine. Dr. Bruce M. Alberts and Dr. William A. Wulf are chair and vice chair,
respectively, of the National Research Council.
The Transportation Research Board is a division of the National Research Council, which serves the
National Academy of Sciences and the National Academy of Engineering. The Board’s mission is to promote
innovation and progress in transportation through research. In an objective and interdisciplinary setting,
the Board facilitates the sharing of information on transportation practice and policy by researchers and
practitioners; stimulates research and offers research management services that promote technical
excellence; provides expert advice on transportation policy and programs; and disseminates research
results broadly and encourages their implementation. The Board’s varied activities annually engage more
than 5,000 engineers, scientists, and other transportation researchers and practitioners from the public and
private sectors and academia, all of whom contribute their expertise in the public interest. The program is
supported by state transportation departments, federal agencies including the component administrations of
the U.S. Department of Transportation, and other organizations and individuals interested in the
development of transportation. www.TRB.org
www.national-academies.org
COOPERATIVE RESEARCH PROGRAMS STAFF FOR NCHRP REPORT 531
ROBERT J. REILLY, Director, Cooperative Research Programs
CRAWFORD F. JENCKS, Manager, NCHRP
EDWARD T. HARRIGAN, Senior Program Officer
EILEEN P. DELANEY, Director of Publications
NCHRP PROJECT 9-27 PANEL

Field of Materials and Construction—Area of Bituminous Materials
JAMES A. MUSSELMAN, Florida DOT (Chair)

MICHAEL E. BIENVENU, Troxler Electronic Laboratories, Inc., Research Triangle Park, NC
CHARLES DEAHL, Compaction America, Kewanee, IL
DONNA HARMELINK, FHWA—Colorado Region
DALE RAND, Texas DOT
MICHAEL D. REMILY, Oregon DOT
MILTON SIMMS, Maryland DOT
JOHN BUKOWSKI, FHWA Liaison Representative
LESLIE ANN MYERS, FHWA Liaison Representative
RONALD A. SINES, P.J. Keating Co., Lunenburg, MA, Other Liaison
FREDERICK HEJL, TRB Liaison Representative
This report presents recommended guidelines for hot mix asphalt pavement con-
FOREWORD struction to achieve satisfactory levels of in-place air voids and permeability. These
By Edward T. Harrigan guidelines were developed from the findings of a research project that examined the
Senior Program Officer relationship of air voids content to permeability and hot mix asphalt lift thickness. The
Transportation Research report will be of particular interest to materials and construction engineers in state high-
Board way agencies, as well as to materials supplier and paving contractor personnel respon-
sible for the production and placement of hot mix asphalt.
For satisfactory performance, hot mix asphalt (HMA) pavements must be con-
structed with adequate field density and impermeability to moisture. During the transi-
tion to the use of the Superpave mix design method since 1994, several states reported
problems with greater than expected permeability associated with the use of coarse-
graded mixes. In addition, there has been ongoing debate over the in-place air voids
content and layer thickness needed to ensure an impermeable pavement. Some state
highway agencies have addressed these issues by increasing their field density require-
ments, lift thickness requirements, or both, when coarse-graded mixes are used. Such
changes, however, entail increased expense. So other states have elected (1) to reduce
the nominal maximum aggregate size of given lifts (e.g., use of a 19.0-mm in place of
a 25.0-mm mix) or (2) to eliminate pavement layers (such as a binder layer) and
increase the thickness of the remaining layers to keep the total pavement thickness at
typically used levels. However, many agencies are reluctant to adopt any such change
without the support of specific research results that justify the increased cost or provide
evidence of satisfactory long-term performance.
Under NCHRP Project 9-27, “Relationships of HMA In-Place Air Voids, Lift
Thickness, and Permeability,” the National Center for Asphalt Technology (NCAT) at
Auburn University was assigned the tasks of (1) determining the minimum ratio of
layer thickness, t, to nominal maximum aggregate size, NMAS, needed to achieve
desirable pavement density levels, and thus impermeable pavements; (2) evaluating the
permeability characteristics of different thicknesses of compacted HMA; and (3)
assessing factors affecting the relationship between in-place air voids, permeability,
and lift thickness. To accomplish these tasks, the research team (1) conducted a criti-
cal review of the literature on the relationship of HMA lift thicknesses to in-place air
voids, the relationship of in-place air voids to permeability, and their effects on pave-
ment performance; (2) evaluated current state DOT guidelines and requirements for
minimum lift thickness and minimum in-place density; and (3) designed and carried
out coordinated laboratory and field experiments to establish relationships among air
voids, lift thickness, and permeability from which to develop practical field compaction
guidelines.
The NCAT project team found that the HMA pavement density that can be
obtained under normal rolling conditions is clearly related to the ratio t/NMAS of the
HMA. For improved compactibility, the agency recommended that t/NMAS be at least
3 for fine-graded mixes and at least 4 for coarse-graded mixes. The data for SMA mixes
indicate that the ratio should also be at least 4. Ratios less than these suggested values
can be used but a greater than normal compactive effort will generally be required in
these situations to obtain the desired in-place density.
The results of an experiment to evaluate the effect of mix temperature on the rela-
tionship between pavement density and t/NMAS found that the more rapid cooling of
the HMA is a key reason for low density in thinner sections (lower t/NMAS). Hence,
for thin HMA layers NCAT emphasized the importance of paving rollers staying very
close to the paving machine so that rolling can be accomplished prior to excessive cool-
ing.
The project team further identified the in-place air voids content as the most sig-
nificant factor impacting permeability of HMA mixtures, followed by coarse aggregate
ratio and VMA. As the coarse aggregate ratio increases, permeability increases, but it
decreases as VMA increases at constant air voids content. The variability of perme-
ability between various mixtures is very high; some mixtures are permeable in the
range of 8 to 10 percent air voids while others are not. However, to ensure that perme-
ability is not a problem NCAT recommends an in-place air voids content between 6
and 7 percent or lower. This appears to be true for a wide range of mixtures regardless
of NMAS and aggregate gradation.
The project final report presents detailed descriptions of the coordinated laboratory
(Task 3) and field (Task 5) experiments; a discussion of the research results from both
experiments; and the project findings, conclusions, and recommendations in five
volumes:
• Volume I: Task 3—Parts 1 and 2;

• Volume II: Task 3—Part 3;
• Volume III: Task 5;
• Volume IV: Appendices for Volumes I, II, and III; and
• Volume V: Executive Summary.
This report includes Volume V only; Volumes I through IV will be available online
at http://www4.trb.org/trb/onlinepubs.nsf/web/nchrp_web_documents as NCHRP Web
Document 68.
The recommended guidelines from Project 9-27 have been referred to the TRB
Mixtures and Aggregate Expert Task Group for its review and possible recommenda-
tion to the AASHTO Highway Subcommittees on Materials and Construction for revi-
sion of appropriate specifications and recommended practices.
CONTENTS 1 CHAPTER 1 Introduction and Problem Statement
2 CHAPTER 2 Objective
3 CHAPTER 3 Research Approach
5 CHAPTER 4 Test Results and Analysis
4.1 Part 1—Mix Designs for Specimens to Study the Effect of t/NMAS on
Density, 5
4.2 Evaluation of Effect of t/NMAS on Density Using Gyratory Compactor, 5
4.3 Evaluation of Effect of t/NMAS on Density Using Vibratory Compactor, 6
4.4 Evaluation of Effect of t/NMAS on Density from Field Study, 10
4.5 Evaluation of Effect of Temperature on the Relationship Between
Density and t/NMAS, 20
4.6 Evaluation of Effect of t/NMAS on Permeability Using Gyratory
Compactor, 22
4.7 Evaluation of Effect of t/NMAS on Permeability Using Vibratory
Compactor, 22
4.8 Evaluation of Effect of t/NMAS on Permeability From Field Study, 27
4.9 Part 2—Evaluation of Relationship of Laboratory Permeability, Density,
and Lift Thickness of Field Compacted Cores, 27
4.10 Controlled Laboratory Experiment to Evaluate Methods of Measuring
the Bulk Specific Gravity of Compacted HMA, 28
4.11 Field Validation of Relationships Between Permeability, Lift Thickness
and In-Place Density, 32
35 CHAPTER 5 Conclusions and Recommendations
37 CHAPTER 6 References
A-1 APPENDICES A Through E
1
CHAPTER 1
INTRODUCTION AND PROBLEM STATEMENT
Proper compaction of hot mix asphalt (HMA) mixtures is Numerous factors can potentially affect the permeability of
vital to ensure that a stable and durable pavement is built. For HMA pavements. In a study by Ford and McWilliams (5), it
dense-graded mixes, numerous studies have shown that initial was suggested that particle size distribution, particle shape,
in-place air voids should not be below approximately 3 percent and density (air voids or percent compaction) affect perme-
nor above approximately 8 percent (1). Lower percentages ability. Hudson and Davis (6) concluded that permeability is
of in-place air voids can result in rutting and shoving, while dependent on the size of air voids within a pavement, not just
higher percentages allow water and air to penetrate into the the percentage of voids. Research by Mallick et al. (7 ) has also
pavement, leading to an increased potential for water dam- shown that the nominal maximum aggregate size (NMAS)
age, oxidation, raveling, and cracking. Low in-place air voids and lift thickness for a given NMAS affect permeability.
are generally the result of a mix problem while high in-place Work by FDOT indicated that lift thickness can have an
voids are generally caused by inadequate compaction. influence on density and hence permeability (8). FDOT con-
Many researchers have shown that increases in in-place air structed numerous pavement test sections on Interstate 75
void contents have meant increases in pavement permeability. that included mixes of different NMAS and lift thicknesses.
Zube (2) showed in the 1960s that dense-graded pavements Results of this experiment suggested that increased lift thick-
become excessively permeable when in-place air voids exceed nesses could lead to better pavement density and hence lower
8 percent. Brown et al. (3) later confirmed this value during the permeability.
1980s. However, due to problems associated with coarse- Thus permeability, lift thickness, and air voids are all inter-
graded mixes (those with a gradation passing below the related. Permeability has been shown to be related to pave-
maximum density line), the size and interconnectivity of ment density (in-place air voids). Increased lift thickness has
air voids have been shown to greatly influence permeability. A been shown to allow desirable density levels to be more eas-
study conducted by the Florida Department of Transportation ily achieved. Westerman (9), Choubane et al. (4), and Mus-
(FDOT) (4) indicated that coarse-graded Superpave mixes selman et al. (8) have suggested that a thickness to NMAS
can sometimes be excessively permeable to water even when ratio (t/NMAS) of 4.0 is preferred. Most guidance recom-
in-place air voids are less than 8 percent. mends that a minimum t/NMAS of 3.0 be used (10). How-
Permeability is also a major concern in stone matrix asphalt ever, due to the potential problems of achieving the desired
(SMA) mixes that utilize a gap-graded coarse gradation. Data density, it is believed that this ratio should be further evalu-
have shown that SMA mixes tend to become permeable when ated based on NMAS, gradation, and mix type (Superpave
air voids are above approximately 6 percent. and SMA).
2
CHAPTER 2
OBJECTIVE
The objectives of NCHRP 9-27 were to (1) determine the meability characteristics of compacted samples at different
minimum t/NMAS needed for desirable impermeable pave- thicknesses, and (3) evaluate factors affecting the relationship
ment density levels to be achievable, (2) evaluate the per- among in-place air voids, permeability, and lift thickness.
3
CHAPTER 3
RESEARCH APPROACH
The laboratory evaluation of the relationship between thick- The third attempt at the effect of t/NMAS on compaction
ness, density, and permeability was divided into two parts. Part was to look at a field study during the rebuilding of the
1 evaluated the relationship of lift thickness, air voids, and per- National Center for Asphalt Technology (NCAT) test track.
meability in a controlled, statistically designed experiment. During this work, the layer thicknesses were varied and
This part looked at varying the lift thickness in the gyratory compacted under similar conditions. Seven mixes from the
compactor and determining density; the experimental vari- track were constructed on a paved surface adjacent to the
ables included three aggregates, four gradations, three nomi- track to look at the effect of layer thickness on density. A
nal aggregate sizes for Superpave mixes, and three nominal general description of these seven mixtures is provided in
aggregate sizes for SMA mixes. The aggregate properties are Table 2. For this part of the study, seven mixes were com-
shown in Table 1. Only one asphalt binder was used for this pacted at layer thicknesses varying from two to five times
study, a PG 64-22. After the mix designs were performed for the t/NMAS. For some of these seven mixes, one side was
these mixes, they were compacted in the Superpave gyratory compacted with a vibratory roller and the other sided was
compactor (100 gyrations) to heights of 2.0, 3.0, and 4.0 times compacted with vibratory and rubber tire rollers. The test
the t/NMAS. The effect of t/NMAS on density was then deter-
data were evaluated, as shown later, and provided reason-
mined. The plan was to select the t/NMAS that gave optimum
able results.
density; but, as will be shown later, the results from the Super-
Another part of the study for Part 1 looked at the effect of
pave gyratory compactor data did not provide a conclusive
lift thickness on permeability. The air voids were controlled
answer; hence, additional work was needed to better establish
at 7 percent and the thickness varied. The permeability results
the appropriate ratio.
It was then decided to look at many of the same mixes with were then determined. These variables were evaluated: two
a vibratory compactor, to establish whether the vibratory com- aggregate types, three gradations, two Superpave NMAS, three
pactor would better simulate field compaction and would SMA NMAS, and three t/NMAS.
provide more conclusive results The experimental variables Part 2 of Task 3 looked at the permeability of cores obtained
included two aggregates, three gradations, two nominal aggre- from the NCHRP 9-9 project. This project contained 40 sec-
gate sizes for Superpave, and three nominal aggregate sizes for tions with varying aggregate types, NMASs, thicknesses, and
SMA. These mixtures, which had already been designed in the design gyrations. The results were evaluated to determine the
first part, were compacted at three thicknesses using three effect of gradation, NMAS, thickness, and design gyration
compactive efforts with the vibratory compactor. The density on permeability. It was assumed that this information would
results were determined, and again the results did not identify help to determine the in-place air voids at which permeability
a definitive minimum ratio. It was then decided that additional would become a problem. Both field and lab permeability
work was needed if an acceptable answer was to be obtained. were measured.
4
TABLE 1 Physical properties of aggregate

Aggregate Type
Property Test Method Crushed
Granite Limestone
Gravel
Coarse Aggregate
Bulk Specific Gravity AASHTO T-85 2.654 2.725 2.585
Apparent Specific Gravity AASHTO T-85 2.704 2.758 2.642
Absorption (%) AASHTO T-85 0.7 0.4 0.9
19.0 mm 14, 0 10, 0 4, 0
Flat and
Elongated (%), 12.5 mm ASTM D4791 16, 0 6, 0 16, 2
3:1, 5:1
9.0 mm 9, 1 16, 3 19, 2
Los Angeles Abrasion (%) AASHTO T-96 37 35 31
Coarse Aggregate AASHTO
42.9 43.0 44.0
Angularity (%) TP56-99
Percent Crushed (%) ASTM D5821 100 100 80
Fine Aggregate
Bulk Specific Gravity AASHTO T-84 2.678 2.689 2.610
Apparent Specific Gravity AASHTO T-84 2.700 2.752 2.645
Absorption (%) AASHTO T-84 0.3 0.9 0.5
Fine Aggregate AASHTO T-33
49.4 45.7 48.8
Angularity (%) (Method A)
AASHTO T-
Sand Equivalency (%) 92 93 94
176
TABLE 2 Mix information for field density study

Section NMAS Gradation Asphalt Type Aggregate Type
1 9.5 mm Fine-Graded Unmodified Granite and
Superpave Limestone
2 9.5 mm Coarse-Graded Unmodified Limestone
Superpave
3 9.5 mm SMA Modified Granite
4 12.5 mm SMA Modified Limestone
5 19.0 mm Fine-Graded Unmodified Granite and
Superpave Limestone
6 19.0 mm Coarse-Graded Unmodified Granite
Superpave
7 19.0 mm Coarse-Graded Modified Limestone
Superpave
5
CHAPTER 4
TEST RESULTS AND ANALYSIS
4.1 PART 1—MIX DESIGNS FOR SPECIMENS est air void level. This figure indicates that for ARZ mixes,
TO STUDY THE EFFECT OF t/NMAS ON the two methods provide similar results. For the TRZ, BRZ,
DENSITY and SMA mixes, Figures 2 through 4 suggest that the bulk
specific gravity measurements derived from the two methods
Of the 36 mix designs, 27 were Superpave-designed mixes
moved farther apart as density decreased. The results also
and 9 were SMA mixes. The Superpave mixes were classified
indicate that, as the gradation became coarser, the difference
according to three gradations: above the restricted zone (ARZ),
in the test results for the two test methods increased. This
through the restricted zone (TRZ), and below the restricted
finding agrees with the research by Cooley et al. (11).
zone (BRZ). The optimum asphalt content, the effective
The apparent reasons for the different results according to
asphalt content (Pbe), voids in mineral aggregate (VMA),
the two test methods is loss of water during density measure-
voids filled with asphalt (VFA), percent theoretical maxi-
ment when using the T-166 method and the effect of surface
mum density at Ninitial (% Gmm at Nini), and ratio of dust to
texture. The loss of water when blotting in the T-166 method
effective asphalt content (P0.075/Pbe) for the Superpave mixes
causes a test error resulting in higher measured density. The
are summarized in Table 3. Data for SMA mixes are shown
in Table 4. The mix design information for both mix types surface texture can result in the vacuum seal device measur-
is presented in Appendix A. Optimum asphalt binder con- ing a lower density than the actual density. Because the vac-
tent was chosen to provide 4 percent air voids at the design uum seal device is more accurate in measuring the density
number of gyrations. However, for the 19-mm NMAS lime- of porous samples, it was used to determine density for this
stone SMA mix, 4 percent air voids could be achieved with research project.
5.7 percent asphalt content, which did not meet the mini- The main objective of this part of the study was to deter-
mum asphalt content requirement in accordance with the mine the minimum t/NMAS. To achieve this objective, rela-
“Standard Practice for Designing SMA,” AASHTO PP44-01. tionships of average air voids for the three aggregate types
Therefore, the minimum asphalt content of 6.0 percent was versus t/NMAS with respect to NMAS and gradation were
chosen, which resulted in 3.7 percent air voids at the design evaluated; the results are illustrated in Figures 5 through 10.
number of gyrations. Some designs did not meet the re- Originally it was intended to determine the t/NMAS at which
quirements of VMA, VFA, % Gmm at Nini, and/or dust/Pbe. the air voids began to level out and to pick that t/NMAS level
Efforts were made to redesign the respective mixes by as the minimum level recommended to achieve optimum
changing the gradation until the requirements were met or compaction. However much of the data in Figures 5 through
closely approximated. This is important in that the mixes 10 indicate that the air voids continue to drop with increasing
used in this project were intended to duplicate mixes uti- t/NMAS past typical t/NMAS values. These data therefore
lized in the field. No modification was made for the TRZ did not provide reasonable guidance for selecting a mini-
mixes that did not meet the requirements, as little could be mum t/NMAS. Hence an air void content of 7.0 percent was
done to modify these gradations and still pass through the selected as the criteria to determine the minimum t/NMAS.
restricted zone. This level of air voids was selected because compaction of
most pavements in the field is targeted at 92.0 to 94.0 per-
cent of theoretical maximum density. Because of the uncer-
4.2 EVALUATION OF EFFECT OF t/NMAS ON tainty in the relationship of average air voids to t/NMAS, as
DENSITY USING GYRATORY COMPACTOR indicated by the data, it was determined to compact some
laboratory samples with a vibratory compactor and also to
Before the evaluation was done, two methods of measuring compact some mixes in the field during reconstruction of
density, or bulk specific gravity, were compared: the AASHTO the NCAT test track. These two efforts, which are discussed
T166 (SSD) and the vacuum sealing (ASTM D6752-02a) later in the report, should provide sufficient information to
methods. All samples were measured using both methods. Fig- make reasonable conclusions concerning desired t/NMAS
ures 1 through 4 present these measurements for the three levels.
gradations of Superpave mixes and the SMA mixes. One potential problem with the Superpave gyratory com-
As shown in Figure 1, the air voids for ARZ mixes as mea- pactor is that it applies a constant strain to the mix during com-
sured by the two methods are approximately equal at low air paction and the force required varies as necessary to provide
voids and deviate by approximately 0.5 percent at the high- the desired strain. This is not the approach that is observed in
6
TABLE 3 Summary of mix design results for Superpave mixes

Aggregate NMAS, Gradation Optimum Pbe, VMA VFA % Gmm P0.075/Pbe
mm Asphalt, % % % % at Nini
9.5 ARZ 6.7 6.2 18.4 76 89.0 0.8
9.5 BRZ 5.3 4.9 15.7 73 86.7 1.0
9.5 TRZ 5.4 5.0 15.6 75 88.9 1.0
19.0 ARZ 4.7 4.3 14.1 72 89.5* 1.2
Granite 19.0 BRZ 4.4 3.9 13.3 68 86.0 1.0
19.0 TRZ 4.0 3.6 12.5* 68 88.8 1.4*
37.5 ARZ 4.2 4.0 13.7 69 89.8* 0.8
37.5 BRZ 3.3 3.0 11.3 64 86.8 1.0
37.5 TRZ 3.6 3.3 12.0 65 88.1 0.9
9.5 ARZ 6.7 6.5 18.3 78* 88.4 0.8
9.5 BRZ 6.2 5.6 16.7 75 86.5 0.8
9.5 TRZ 6.0 5.4 16.3 75 87.7 0.9
19.0 ARZ 4.9 4.4 14.0 72 88.5 1.1
Gravel 19.0 BRZ 4.5 3.9 12.9* 69 86.3 1.3*
19.0 TRZ 4.4 3.8 12.8* 69 88.0 1.3*
37.5 ARZ 4.4 3.9 13.0 70 89.7* 0.8
37.5 BRZ 3.6 3.2 11.7 63 85.5 1.0
37.5 TRZ 3.9 3.5 12.0 66 85.6 0.9
9.5 ARZ 6.0 5.7 17.4 76 87.8 0.7
9.5 BRZ 5.0 4.6 15.3 72* 85.5 0.9
9.5 TRZ 4.4 4.2 14.4 70* 86 1.2
19.0 ARZ 4.1 3.5 12.6* 66 88.3 1.4*
Limestone 19.0 BRZ 4.7 4.4 14.3 71 85.5 0.7
19.0 TRZ 3.3 2.8 11.0* 62* 85.7 1.8*
37.5 ARZ 3.2 3.1 11.8 64 88.8 1.0
37.5 BRZ 2.7 2.6 10.6* 60* 86.0 1.2
37.5 TRZ 2.8 2.6 10.6* 61* 87.7 1.1
* Did not meet Superpave Design Requirements
the field where the stress is constant and the strain varies. 4.3 EVALUATION OF EFFECT OF t/NMAS ON
Hence, the Superpave gyratory compactor likely does not pro- DENSITY USING VIBRATORY COMPACTOR
vide a reasonable answer because the compaction provided by
this device is different from the field. The big problem with After obtaining the results for the Superpave gyratory com-
using this concept to establish a minimum t/NMAS is that pactor, it was concluded that more tests needed to be con-
the voids continue to increase significantly as the t/NMAS ducted to better simulate compaction in the field. The air voids
increases, making it impossible to select an optimum value. determined from the vacuum seal device were utilized in the
The optimum t/NMASs established using the Superpave analysis. To further evaluate the relationship between density
gyratory compactor vary from less than 2.5 up to approxi- and lift thickness, a similar study was conducted, but on a
mately 8. This wide range of numbers did not allow specific smaller scale, using the vibratory compactor as the compaction
criteria to be established. Hence, additional testing was per- mode. This was not part of the original proposed work, but
formed using the laboratory vibratory compactor and field it was believed that the vibratory compactor might provide
test section. compaction that has more typical of in-place compaction.
TABLE 4 Summary of mix design results for SMA mixes

Aggregate NMAS, Optimum Pbe, VMA, VFA, VCAmix,a VCAdrc,b
mm Asphalt, % % % % % %
9.5 7.2 6.6 18.7 78 30.9 41.9
Granite 12.5 6.6 6.4 18.8 77 30.3 42.7
19.0 6.4 5.9 17.6 77 29.6 42.0
9.5 7.3 6.5 18.6 77 30.4 41.8
Gravel 12.5 6.8 6.1 17.7 77 31.1 42.1
19.0 6.7 6.2 17.8 76 29.3 42.0
9.5 6.2 5.8 17.4 76 30.7 38.4
Limestone 12.5 7.4 7.0 19.6 80 31.1 38.9
19.0 6.0 5.6 16.8c 77 29.8 40.3
a
VCA = Voids in Compacted Aggregate
b
drc = dry-rodded compacted
c
Did not meet SMA Design Requirements
7
16.0
14.0
Air Voids, % - Vacuum Seal Method
y = 1.0576x + 0.0992
12.0 2
R = 0.9887
10.0
8.0
6.0
4.0
2.0
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
Air Voids, % - AASHTO T166
Figure 1. Relationship between air voids for ARZ mixes.
The vibratory compactor used compacted beam samples for of the vibratory specimen that could be obtained was 75.0 mm,
the wheel-tracking device. which would only be 2.0 t/NMAS. For the SMA mixes, three
Of the 36 mix designs analyzed for Part 1, 14 mixes were NMASs were selected (9.5 mm, 12.5 mm, and 19 mm). The
selected for further study. Two types of aggregates, granite t/NMAS ratios utilized were 2.0, 3.0, and 4.0. The compactive
and limestone were used. For Superpave designed mixes, effort for each t/NMAS was varied over a range including
two gradations were utilized (ARZ and BRZ) along with two 30 sec, 60 sec, and 90 sec of compaction. The range of com-
NMASs (9.5 mm and 19.0 mm). The 37.5-mm NMAS mix pactive efforts was selected for two reasons: (1) there is no
was excluded from the study because the maximum thickness standard compactive effort for the vibratory compactor and
20.0
18.0
16.0 y = 1.1074x + 0.3893

2
R = 0.978
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0
Figure 2. Relationship between air voids for TRZ mixes.

8
20.0
18.0

16.0
y = 1.201x + 0.4379
2
14.0 R = 0.9264
12.0
10.0
8.0
6.0
4.0
2.0
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0
Figure 3. Relationship between air voids for BRZ Mixes.
(2) the effects of compactive effort on density at different ferent t/NMAS values. However, in a few cases there was a
thicknesses could be evaluated. After compaction, the bulk difference. Also, in many cases the best t/NMAS was 2.0,
specific gravity was measured and the data were analyzed to which is significantly lower than that observed on field proj-
provide recommendations concerning the minimum t/NMAS. ects. Typically, it was assumed that coarse graded mixes
To determine the minimum t/NMAS, relationships between would have a desired t/NMAS greater than fine-graded
average air voids for the two types of aggregates and t/NMAS mixes. The results did not always follow that trend. It was
were plotted for each NMAS, compaction time, and grada- judged that some fieldwork was necessary to validate the
tion, as shown in Figures 11 through 17. In many cases there results with the Superpave gyratory compactor and with the
was very little difference between the densities for the dif- vibratory compactor.
22.0
20.0
18.0
y = 1.6583x - 0.9272
16.0 2
R = 0.7185
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0
Figure 4. Relationship between air voids for SMA mixes.

9
18.0
TRZ ARZ
16.0 BRZ y = 32.71x -0.9062
y = 31.16x -0.8913 R 2 = 0.9998 BRZ
14.0 R2 = 0.9807
Average Air Voids, %

TRZ
12.0
10.0
8.0
6.0
ARZ
4.0 y = 22.729x -0.8484
R 2 = 0.9771
2.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
t/NMAS
Figure 5. Relationships between air voids and t/NMAS for 9.5-mm

Superpave mixes.
12.0
ARZ TRZ
y = 24.593x -1.2151 BRZ
y = 17.879x -1.0631 -1.187
10.0 2
R = 0.9772 R 2 = 0.9963 y = 26.047x
2
R = 0.9954
8.0
6.0 ARZ
BRZ
4.0 TRZ
2.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
t/NMAS
Figure 6. Relationships between air voids and t/NMAS for

19.0-mm Superpave mixes.
9.0
TRZ BRZ
-0.7561 -0.841
8.0 y = 10.185x y = 14.404x
2 2
R = 0.9053 R = 0.9874
7.0
6.0
ARZ
5.0
BRZ
4.0
TRZ
3.0 ARZ
-0.2991
y = 6.2838x
2
2.0 R = 0.8546
1.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
t/NMAS

Superpave mixes.
10
20.0
18.0
-0.8298
y = 36.496x
16.0 2
R = 0.9988
Average Air Voids, % 14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
t/NMAS
Figure 8. Relationships between air voids and t/NMAS for 9.5-mm SMA mixes.
4.4 EVALUATION OF EFFECT OF t/NMAS ON 4.4.1 Section 1

DENSITY FROM FIELD STUDY
Section 1 was constructed on July 18, 2003, and con-
The field test sections consisted of 7 mixes that were to be sisted of a 2.0 to 5.0 t/NMAS overlay of an existing HMA
placed on the test track. These mixes had to be verified before layer. This construction was performed adjacent to the NCAT
placing on the track; hence, these mixes could be placed Test Track. The mix was a 9.5-mm NMAS fine-graded mix-
and tested without significant costs. Some of the mixes did ture. The length of the section was about 40 m, and the width
not meet volumetrics and other requirements, but they were was about 3.5 m. On some of the sections the placement
judged sufficient for this part of the study because determin- began on the thick side and in some cases the placement began
ing the desired thickness range was a relative value based on on the thin side. This technique was used so that there would
t/NMAS. be no bias due to the placement of the HMA. On this sec-
20.0
18.0
16.0
y = 29.982x -0.7232
14.0
R2 = 0.9945
12.0
10.0
8.0
6.0
4.0
2.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
t/NMAS
11
16.0
14.0
12.0
y = 27.14x-0.9206
Average Air Voids, % R2 = 0.9974
10.0
8.0
6.0
4.0
2.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
t/NMAS
tion the paving began with the thicker portion of the section meet desired requirements for weight and tire pressure, and
and the thickness was slowly decreased as the paver moved thus the data generated for the rubber tire roller compacted
down the test lane. The desired mat thickness was achieved mixture were omitted from the analysis for this section. The
by gradually adjusting the screed depth crank of the paver breakdown rolling was performed with one pass in the static
during the paving operation. The weather conditions during mode on the mat at a temperature of about 300°F. This was fol-
the paving were 84°F, overcast, with calm wind. The existing lowed by three passes in the vibratory mode at low amplitude
surface temperature prior to overlay was also 84°F. and high frequency (3800 vibrations per minute [vpm]) and
The roller utilized in this section was an 11-ton steel roller one pass in the static mode. It was determined that this com-
HYPAC C778B with a 78-in. wide drum that could operate in paction effort reached the peak density; hence, additional
vibratory or static mode. The rubber tire roller available did not rolling was not performed.
7
30 sec
6
60 sec
5 90 sec
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
Figure 11. Relationships between air voids and t/NMAS for 9.5-mm ARZ mixes.
12
12
30 sec
10
60 sec
90 sec
8
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
Figure 12. Relationships between air voids and t/NMAS for 9.5-mm BRZ mixes.
A total of 16 cores were obtained from this section and the the air voids increased by 0.5 percent (less than 0.5 percent
test results of the cores are presented in Figure 18. The results were considered insignificant). This number is somewhat
include the thickness of cores, t/NMAS, and the air voids arbitrary, but it is realistic. Therefore, as shown in Figure 18,
determined from the vacuum seal device. the recommended t/NMAS range for 9.5-mm fine-graded
A review of the data indicated that a polynomial function mix was 3.4 to 5.8. This does not mean that satisfactory com-
provided the best fit line. The best-fit line indicates that the paction cannot be obtained outside of these limits, but it does
air voids decreased as the t/NMAS increased to a point where indicate that more compactive effort would be needed. So
additional thickness resulted in increased air voids. The rec- this recommended range should only be used as a guide and
ommended thickness range was selected as the point(s) where should not be a rigid requirement. The effect of t/NMAS on
30 sec
7
60 sec
6
90 sec
5
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
Figure 13. Relationships between air voids and t/NMAS for 19.0-mm ARZ mixes.
13
12
10 30 sec
Average Air Voids, % 60 sec
8 90 sec
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
Figure 14. Relationships between air voids and t/NMAS for 19.0-mm BRZ mixes.
the measured density was determined from Figure 18. Data The paving started from the thick portion of the mat and pro-
in the figure indicate that the lowest air voids (7.0 percent air gressed toward the thinner portion. The weather conditions
voids) occurred at t/NMA 4.4. Table 5 shows the air voids at during the paving were 82°F, overcast, with calm wind. The
various t/NMAs as related to this minimum. existing surface temperature was 96°F.
The roller utilized in this section was an 11-ton steel drum
4.4.2 Section 2 roller HYPAC C778B with a 78-in. wide drum that could
operate in vibratory or static mode. The rubber tire roller was
Section 2 was constructed on August 7, 2003, and the a 15-ton HYPAC C560B with a tire pressure of 90 psi. For the
t/NMAS for this overlay ranged from 2.0 to 5.0. The mixture side of the mat utilizing only the steel drum roller, the initial
was a 9.5-mm NMAS coarse-graded mixture. The length of rolling was performed with four passes in the vibratory mode
the section was about 40 m, and the width was about 3.5 m. at low amplitude and high frequency (3800 vpm) at a mix tem-
12
30s
10
60s
90s
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
14
12.0
30 s
10.0 60 s
90 s

8.0
6.0
4.0
2.0
0.0
0 0 .5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
SMA mixes.
12.0
30s
10.0 60s
90s
8.0
6.0
4.0
2.0
0.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
t/NMAS
SMA mixes.
10. 0
9.0
R2 = 0 . 6 3 9 2
8.0
7.0
Air Voids, %
6.0
5.0
4.0
3.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
t/NMAS
Figure 18. Relationships of air voids and t/NMAS for 9.5-mm

fine-graded mix.
15
TABLE 5 Relationship of air voids and tire roller. The effect of t/NMAS on the measured density
t/NMAS for 9.5-mm fine-graded HMA was determined from Figure 19. Data in the figure indicate
compacted with steel roller
that the lowest in-place air voids (10 percent air voids for
t/NMA Percentage points the steel wheel roller only and 10.5 percent air voids for the
above lowest
steel and rubber tire rollers) occurred at t/NMAS of 4.7 for
4.4 (lowest air voids, 7.0 %) 0.0
2 2.5 the steel wheel roller and 3.8 for the rubber and steel wheel
3 1.0 roller. Table 6 shows the air voids at various t/NMAs as related
4 0.1 to this minimum.
5 0.1
4.4.3 Section 3
perature of about 300°F. This was followed with four passes
in the static mode. For the side of the mat that used a rubber Section 3 was constructed on July 25, 2003, and consisted
tire roller as an intermediate roller, the breakdown rolling was of a 2.0 to 5.0 t/NMAS overlay of an existing HMA layer.
performed with four passes in the vibratory mode operated at The mix was a 9.5-mm NMAS SMA. The length of the sec-
low amplitude and high frequency (3800 vpm). This was fol- tion was about 40 m, and the width was about 3.5 m. The
lowed with five passes of the rubber tire roller and one pass paving started from the thick portion of the mat and pro-
of the steel roller in the static mode. gressed to the thinner portion. The desired mat thickness
A total of 15 cores were obtained from the side that uti- was achieved by gradually adjusting the screed depth crank
lized only a steel drum roller and 16 cores from the side that of the paver during the operation. The weather conditions
used the rubber tire roller. The relationship of air voids during the paving were 95°F, partly cloudy, with calm wind.
measured from the vacuum seal device and t/NMAS was The existing surface temperature was 115°F.
evaluated for each rolling pattern. The results are illustrated The roller utilized in this section was an 11-ton steel drum
in Figure 19. roller HYPAC C778B with a 78-in. wide drum that could
A review of the data indicated that a polynomial function operate in vibratory or static mode. The rubber tire roller was
provided the best fit. As the thickness increased, the air voids a 15-ton HYPAC C560B with a tire pressure of 90 psi. For
decreased until a point where additional thickness resulted in the side of the mat utilizing only the steel drum roller, the ini-
increased air voids. The plots also suggest that the side uti- tial rolling was performed with one pass in the static mode
lizing only a steel drum compactor had better compaction. followed by five passes in the vibratory mode operated in low
To determine the desired thickness, it was decided to use air amplitude and high frequency (3800 vpm) on the mat having
voids 0.5 percent larger (a void level less than 0.5 percent dif- a mix temperature of about 320°F. This was followed with
ferent was not considered significantly different) than the two passes in the static mode for the finish rolling. For the
minimum air voids from the best-fit line. Therefore, as shown side of the mat that used a rubber tire roller as an intermediate
in Figure 19, the desired t/NMAS range for 9.5-mm coarse- roller, the breakdown rolling was performed with one pass in
graded mix was 3.5 to 5.9 for compaction with a steel wheel the static mode and four passes in the vibratory mode oper-
roller and 2.9 to 4.6 for compaction with the steel and rubber ated in low amplitude and high frequency (3800 vpm). This
14.0
13.0
12.0
Steel/Rubber Tire Roller
2
11.0 R = 0.5115
Air Voids, %
10.0
9.0
Steel Roller
2
R = 0.68
8.0
7.0
6.0
5.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
t/NMAS
Figure 19. Relationships of air voids and t/NMAS for 9.5-mm coarse-graded mix.
16
TABLE 6 Relationship of air voids and t/NMAS for 9.5-mm coarse-graded

HMA compacted with steel roller and with steel and rubber tire rollers
Steel roller Steel and rubber tire rollers
t/NMA Percentage t/NMA Percentage
points above points above
lowest lowest
4.7 (lowest air voids, 10.0 %) 0.0 3.8 (lowest air voids, 10.5 %) 0.0
2 2.5 2 2.0
3 1.0 3 0.5
4 0.5 4 0.0
5 0.0 5 1.0
was followed with eight passes of the rubber tire roller and 4.4.4 Section 4
two passes of the steel wheel roller in the static mode.
A total of 12 cores were obtained from the side that utilized Section 4 was constructed on August 12, 2003, and con-
only the steel drum roller and another 12 cores from the side sisted of a 2.0 to 5.0 t/NMAS overlay of an existing HMA
that used the rubber tire roller. To determine the range of rec- layer. The mix was a 12.5-mm NMAS SMA. The length of
ommended t/NMAS for this mix, the relationship of air voids the section was about 40 m, and the width was about 3.5 m. The
from the vacuum seal device and t/NMAS was evaluated for paving started from the thinner portion and proceeded toward
each rolling pattern. The results are illustrated in Figure 20. the thicker portion of the mat. The weather conditions during
The best-fit lines indicate that the air voids decreased as the paving were 80°F, overcast, with calm wind. The existing
the thickness increased to a point where additional thickness surface temperature was 85°F.
resulted in increased air voids. The plots also suggest that the The roller utilized in this section was an 11-ton steel drum
side utilizing only the steel drum compactor had higher den- roller HYPAC C778B with a 78-in. wide drum that could
sity. Rubber tire rollers are not used on SMA mixtures and operate in vibratory and static modes. The rubber tire roller
these data confirm that there is no need to use the rubber tire was a 15-ton HYPAC C560B with a tire pressure of 90 psi.
roller. As shown in Figure 20, the recommended range for For the side of the mat utilizing only the steel drum roller,
t/NMAS for the 9.5-mm SMA mix is 3.8 to 5.3 for the com- the initial rolling was performed with four passes in the
paction with a steel wheel roller and 2.6 to 5.1 for compaction vibratory mode operated at low amplitude and high frequency
with a steel and rubber tire roller. The effect of t/NMAS on (3800 vpm). The mat temperature was approximately 320°F.
the measured density was determined from Figure 20. Data in This was followed with three passes in the static mode includ-
the figure indicate that the lowest in-place air voids (8.5 per- ing finish rolling. For the side of the mat that used a rubber
cent air voids for the steel wheel roller only and 10.3 percent tire roller as an intermediate roller, the initial rolling was per-
air voids for the steel and rubber tire rollers) occurred at formed with four passes in the vibratory mode operated at
t/NMAS of 4.5 for the steel wheel roller and 3.8 for the rubber low amplitude and high frequency (3800 vpm). This was fol-
and steel wheel roller. Table 7 shows the air voids at various lowed with four passes of the rubber tire roller and one pass
t/NMAs as related to this minimum. of the steel roller in the static mode.
14.0
13.0
12.0 2
R = 0.1864
11.0
10.0
Air Voids, %
9.0
8.0
Steel Roller
7.0 2
R = 0.8335
6.0
5.0
4.0
3.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
t/NMAS
Figure 20. Relationships of air voids and t/NMAS for 9.5-mm SMA mix.
17
TABLE 7 Relationship of air voids and t/NMAS for 9.5-mm SMA mix
compacted with steel roller and with steel and rubber tire rollers
points points
above above
lowest lowest
2 5.5 2 1.2
3 2.0 3 0.2
4 0.2 4 0.0
5 0.2 5 0.5
A total of 21 cores were obtained from the side that uti- rollers) occurred at t/NMAS of 4.5 for the steel wheel roller
lized only a steel drum roller and 21 cores from the side that and 4.8 for the rubber and steel wheel rollers. Table 8 shows
used the rubber tire roller. To determine the recommended the air voids at various t/NMAs as related to this minimum.
t/NMASs for this mix, the relationship of air voids from the
vacuum seal device and t/NMAS was evaluated for each
rolling pattern. The results are illustrated in Figure 21. 4.4.5 Section 5
The best-fit lines indicate that the air voids decreased as the
thickness increased to a point where additional thickness Section 5 was constructed on July 16, 2003, and consisted
resulted in increased air voids. The plots also suggest that the of a 2.0 to 5.0 t/NMAS overlay of an existing HMA. The mix
side utilizing only the steel drum compactor had higher den- consisted of a 19.0-mm NMAS fine-graded HMA. The length
sity. As shown in Figure 21, the suggested minimum t/NMAS of the section was about 40 m, and the width was about 3.5 m.
for 12.5-mm SMA mix is 3.8 for compaction with steel wheel The paving started on the thin end of the section and pro-
roller and 4.6 for compaction with steel and rubber tire roll- ceeded to the thicker portion. The desired mat thickness was
ers. For these mixes, the density increased as the t/NMAS achieved by gradually adjusting the screed depth crank of
increased even at the thicker portions. Also the curve did not fit the paver during the operation. The weather conditions dur-
the data as well as desired, so the data points were actually used ing the paving were 90°F, clear, with calm wind. The existing
to select the suggested t/NMAS number. Note in the plots that surface temperature was 96°F.
the data points continue downward with increasing t/NMAS to The roller utilized in this section was an 11-ton steel roller
a point and then the air voids remain relatively constant as HYPAC C778B with a 78-in. wide drum that operated in
the t/NMAS increased. vibratory and static modes. The rubber tire roller used did not
The effect of t/NMAS on the measured density was deter- meet the tire pressure requirements and the results were omit-
mined from Figure 21. Data in the figure indicate that the low- ted from the analysis for this section. The breakdown rolling
est in-place air voids (4.7 percent air voids for the steel wheel was performed with four passes in the vibratory mode oper-
roller only and 7.5 percent air voids for the steel and rubber tire ated in low amplitude and high frequency (3800 vpm). The
19.0
18.0
17.0
16.0
15.0
14.0
13.0
12.0 Steel/Rubber Tire Roller
Air Voids, %
11.0 2
R = 0.77
10.0 Steel Roller
9.0 2
R = 0.87
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
t/NMAS
Figure 21. Relationships of air voids and t/NMAS for 12.5-mm SMA Mix.
18
TABLE 8 Relationship of air voids and t/NMAS for 12.5-mm SMA mix
compacted with steel roller and with steel and rubber tire rollers
points points
above above
lowest lowest
2 11.3 2 6.5
3 3.3 3 3.5
4 0.3 4 0.5
5 0.5 5 0.0
mat temperature was approximately 300°F. Three passes in 3.5 m. The paving started from the thinner portion of the mat
the static mode and one pass for finish rolling followed this and proceeded to the thicker portion. The weather conditions
initial rolling. during the paving were 79°F, cloudy, with calm wind. The
A total of 20 cores were obtained from this section. To existing surface temperature was 84°F.
determine the minimum t/NMAS for this mix, the relationship The roller utilized in this section was an 11-ton steel drum
between air voids (from the vacuum seal device) and thickness roller HYPAC C778B with a 78-in. wide drum that could
was evaluated. The results are illustrated in Figure 22. operate in vibratory and static mode. The rubber tire roller was
The best-fit line indicated that the air voids decreased as a 15-ton HYPAC C560B with a tire pressure of 90 psi. For the
the thickness increased to a point where additional thickness side of the mat utilizing only the steel drum roller, the initial
resulted in increased air voids. As shown in Figure 22, the rec- rolling was performed with four passes in the vibratory mode
ommended t/NMAS range for the 19.0-mm fine-graded mix operated at low amplitude and high frequency (3800 vpm).
was 3.1 to 4.6. The effect of t/NMAS on the measured density The mat temperature was approximately 300°F. This initial
was determined from the figure. Data in the figure indicate that rolling was followed with six passes in the static mode. For the
the lowest in-place air voids (6.2 percent air voids) occurred at side of the mat that used a rubber tire roller as the intermedi-
t/NMAS of 3.8. Table 9 shows the air voids at various t/NMAs ate roller, the initial rolling was performed with four passes in
as related to this minimum. the vibratory mode operated in low amplitude and high fre-
quency (3800 vpm). This initial rolling was followed with four
4.4.6 Section 6 passes of the rubber tire roller and two passes with a steel
wheel roller in the static mode.
Section 6 was constructed on August 6, 2003, and consisted A total of 22 cores were obtained from the side that utilized
of a range of 2.0 to 5.0 t/NMAS overlay of an existing HMA. only a steel drum roller and 16 cores from the side that used
The mix was a 19.0-mm NMAS coarse-graded HMA. The the rubber tire roller. To determine the minimum t/NMAS for
length of the section was about 40 m, and the width was about this mix, the relationship between air voids from vacuum seal
10.0
9.0
2
R = 0.77
8.0
Air Voids, %
7.0
6.0
5.0
4.0
3.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
t/NMAS
Figure 22. Relationships of air voids and t/NMAS for 19.0-mm fine-graded mix.
19
TABLE 9 Relationship of air voids and t/NMAS Table 10 Relationship of air voids and t/NMAS
for 19.0-mm fine-graded mix compacted with for 19.0-mm coarse-graded mix compacted with
steel roller steel and rubber tire roller*
t/NMA Percentage points t/NMA Percentage points
above lowest above lowest
2 3.1 2 1.8
3 0.6 3 0.6
4 0.0 4 0.1
5 1.3 5 0.1
*The steel wheel roller alone was not used because it
produced too much scatter in the data
device and thickness was evaluated for each rolling pattern.
The results are illustrated in Figure 23. The best-fit lines indi-
cate that the air voids decreased as the thickness increased to HMA and utilized a modified asphalt. The length of the sec-
a point where additional thickness resulted in increased air tion was about 40 m, and the width was about 3.5 m. The
voids. The plots also suggest that the side utilizing the rubber paving started from the thicker portion of the mat and pro-
tire roller had higher density. As shown in Figure 23, the rec- ceeded to the thinner portion. The weather conditions dur-
ommended minimum thickness for 19.0-mm coarse-graded ing the paving were 90°F, clear, with calm wind. The existing
mix was 3.0 for compaction with the steel and rubber tire surface temperature was 120°F.
rollers. There is too much scatter in the data to make a good The roller utilized in this section was an 11-ton steel
selection of a recommended value for compaction with a steel drum roller HYPAC C778B with a 78-in. wide drum that
wheel roller. could operate in the vibratory and static modes. The rubber
The effect of t/NMAS on the measured density was deter- tire roller was a 15-ton HYPAC C560B with a tire pressure
mined from Figure 23. Data in the figure indicate that the low- of 90 psi. For the side of the mat utilizing only the steel
est in-place air voids (5.7 percent for the steel and rubber tire drum roller, the initial rolling was performed with four passes
roller, the steel wheel roller alone was not used because it pro- in the vibratory mode operated in low amplitude and high fre-
duced too much scatter in the data) occurred at t/NMAS of 4.5. quency (3800 vpm). The mat temperature was about 330°F.
Table 10 shows the air voids at various t/NMAs as related to This was followed with another five passes in the vibra-
this minimum. tory mode operated at low amplitude and high frequency
(3800 vpm). There was one additional pass with the steel
4.4.7 Section 7 wheel roller in the static mode to finish the mat. For the side
of the mat that used a rubber tire roller as an intermediate
Section 7 was constructed on August 14, 2003, and con- roller, the initial rolling was performed with two passes in
sisted of a range of 2.0 to 5.0 t/NMAS overlay of an existing the vibratory mode operated at low amplitude and high fre-
HMA. The mix consisted of a 19.0-mm NMAS coarse-graded quency (3800 vpm). This was followed with ten passes with
11. 0
10. 0
Steel Roller
R2 = 0.1601
9 .0
8 .0
7 .0
Air Voids, %
6 .0
5 .0
4 .0
3 .0
R2 = 0.4489
2 .0
1 .0
0 .0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
t/NMAS
Figure 23. Relationships of air voids and t/NMAS for 19.0-mm coarse-graded mix.
20
the rubber tire roller and two passes of the steel wheel roller The t/NMAS should be set at 4.0 for coarse-graded mixes due
in the static mode. to the significant increase in voids when reducing the t/NMAS
A total of 23 cores were obtained from the side that utilized from optimum down to 3.0.
only the steel drum roller and 26 cores from the side that used
the rubber tire roller. To determine the minimum t/NMAS
for this mix, the relationship of air voids from the vacuum seal 4.5 EVALUATION OF THE EFFECT OF
device and t/NMAS was evaluated for each rolling pattern. TEMPERATURE ON THE RELATIONSHIP
The results are illustrated in Figure 24. BETWEEN DENSITY AND t/NMAS
The best-fit lines indicate that the air voids decreased as
the thickness increased to a point where additional thickness Three locations were selected for temperature measure-
resulted in increased air voids. The plots also suggested that ments for each section in the field experiment; one near
the side utilizing only the steel drum compactor had higher the beginning of the section, one near the middle, and one
density. As shown in Figure 24, the minimum t/NMAS range near the end of the section. To determine the effect of mix
for 19.0-mm coarse-graded with modified asphalt mix was temperature on the density, the temperature at 20 minutes
3.4 to 4.8. The effect of t/NMAS on the measured density after placement of the mix at each location was selected
was determined from Figure 24. Data in the figure indicate because this provides a reasonable compaction time. Because
that the lowest in-place air voids (5.6 percent air voids for the mixes in this study used two different types of asphalt
the steel wheel roller only and 7.4 percent air voids for the binder, PG 67-22 and PG 76-22, the temperatures at 20 min-
steel and rubber tire rollers) occurred at t/NMAS of 4.2 for utes were normalized by subtracting the high temperature
the steel wheel roller and 5.3 for the rubber and steel wheel grade of the asphalt type from the temperatures at 20 min-
roller. Table 11 shows the air voids at various t/NMAs as utes. Table 13 presents the t/NMAS, the average tempera-
related to this minimum. ture readings at 20 minutes, the asphalt high temperature
grade, and the difference between mix temperature and high
temperature grade. The differences in temperature were plot-
4.4.8 Summary ted against the t/NMAS together with the core densities for
each section, as shown in Figures 25 through 31.
In summary, the data for the seven sections appear to be The relationship between density and t/NMAS for all
reasonable and to match past experience. A summary of the sections is shown in Figure 32. The best-fit line has an R2
results compared to the t/NMAS for lowest voids is provided of 0.26 and indicates that the density increased as the thick-
in Table 12. These results indicate that the t/NMAS should be ness increased to a point where additional thickness resulted
somewhere between 3 and 5 for best results. Based on the lim- in a decrease in density. The effect of the layer thickness and
ited data, a t/NMAS of 3 is probably reasonable for fine-graded cooling time on mix temperature is provided in Figure 33.
mixes, because there is less than 1 percentage point change in The data were obtained from the thermocouples installed in
density when the t/NMAS is reduced from optimum to 3.0. the pavement. This plot indicates that, during hot weather,
15.0
14.0
13.0
2
12.0 R = 0.81
11.0
Air Voids, %
10.0
9.0 Steel Roller

2
R = 0.65
8.0
7.0
6.0
5.0
4.0
3.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
t/NMAS
Figure 24. Relationships of air voids and t/NMAS for 19.0-mm coarse-graded mix with
modified asphalt.
21
TABLE 11 Relationship of air voids and t/NMAS for 19.0-mm coarse-graded

mix with modified asphalt compacted with steel roller and with steel and
rubber tire rollers
points points
above above
lowest lowest
2 4.9 2 6.1
3 1.3 3 3.4
4 0.0 4 0.8
5 0.8 5 0.0
compaction time for a layer thickness of 1.5 in. is approxi- same amount of compactive effort on an HMA mixture
mately twice that for a 1-in. layer. This clearly shows that prior to cooling to some defined temperature will take twice
one of the problems in obtaining density is layer thickness as many rollers at a 1-in. thickness as that required for a 1.5-in.
regardless of the t/NMAS. If the amount of compaction surface. It is likely to be significantly more difficult to compact
time is reduced by 50 percent, it may be very difficult to a 1-in. layer than to compact a 1.5-in. layer simply because of
compact the mixture to an adequate density. To place the the cooling rate.
TABLE 12 Effect of t/NMAS on compactibility of HMA

Increase in Air Increase in Air Increase in Air Increase in Air
Description of Voids for Voids for Voids for Voids for
Mix t/NMAS=2 t/NMAS=3 t/NMAS=4 t/NMAS=5
Section 1-9.5mm
Fine Graded— 2.5% 1.0% 0.1% 0.1%
Steel Roller
Section 2-9.5mm
Coarse Graded- 2.5% 1.0% 0.5% 0.0%
Steel Roller
Section 2-9.5mm
Coarse Graded-
2.0% 0.5% 0.0% 1.0%
Steel and Rubber
Roller
Section 3-9.5mm
SMA(mod AC) 5.5% 2.0% 0.2% 0.2%
Steel Roller
Section 3-9.5mm
SMA(Mod AC)
1.2% 0.2% 0.0% 0.5%
Steel & Rubber
Roller
Section 4-
12.5mm SMA
11.3% 3.3% 0.3% 0.5%
(mod AC) Steel
Roller
Section 4-
12.5mm SMA
6.5% 3.5% 0.5% 0.0%
(mod AC) Steel
& Rubber Roller
Section 5-19mm
Fine Graded 3.1% 0.6% 0.0% 1.3%
Steel Roller
Section 6-19mm
Coarse Graded
1.8% 0.6% 0.1% 0.1%
Steel and Rubber
Roller
Section 7-19mm
Coarse Graded
4.9% 1.3% 0.0% 0.8%
(mod AC) Steel
Roller
Section 7-19mm
Coarse Graded
(mod AC) Steel 6.1% 3.4% 0.8% 0.0%
& Rubber Roller
22
TABLE 13 t/NMAS, temperature in C at 20 min., gyrations were increased up to 300 gyrations. This shows the
asphalt high temperature grade, and difference in difficulty of compacting mixes at thinner lifts in the gyratory
temperature mold. Permeability testing was only performed on specimens
Section/Mix Temp. at Asphalt Difference that met the desired air voids. The results were very limited,
20 min., °C Grade, PG but, did show that generally the coarser mixes (larger maxi-
1 2.5 60 67 -7
mum aggregate size or higher percentage of coarse aggregate)
9.5mmFG 3.6 82 67 15
5.1 95 67 28 had higher permeabilities.
2 2.1 64 67 -3
9.5mmCG 2.4 72 67 5
5.1 105 67 38
3 2.2 65 76 -11 4.7 EVALUATION OF EFFECT OF t/NMAS
9.5mmSMA 3.7 100 76 24 ON PERMEABILITY USING
5.2 112 76 36 VIBRATORY COMPACTOR
4 2.2 72 76 -4
12.5mmSMA 3.1 118 76 42 All specimens compacted at t/NMAS of 2.0, 3.0, and 4.0 did
3.8 120 76 44 achieve the target air void content, which was 7 ± 1.0 percent.
5 2.6 124 67 57
19mmFG 3.0 122 67 55
Figure 34 shows the relationship between average permeabil-
5.2 130 67 63 ity for the two aggregate types and t/NMAS. In general, the
6 2.1 82 67 15 permeability decreased as t/NMAS increased. Most of the
19mmCG 3.2 120 67 53 mixes had permeability values fewer than 50 × 10−5 cm/sec.
5.1 118 67 51
7 2.7 86 76 10
However, at t/NMAS equal to 2.0, the 9.5-mm and 12.5-mm
19mmCG 3.8 120 76 44 NMAS SMA mixes had average permeability values of 173
5.2 142 76 66 × 10−5 cm/sec and 196 × 10−5 cm/sec, respectively. These
values for the SMA exceed the recommended maximum
permeability value of 125 × 10−5 cm/sec. It appears from these
4.6 EVALUATION OF EFFECT OF t/NMAS
data that a specification requirement of 7 percent air voids
ON PERMEABILITY USING
GYRATORY COMPACTOR would be acceptable for all of the mixes if the t/NMAS is
3 or greater. The likely reason that the thinner samples have
Specimens were compacted to 7.0 ± 1.0 percent air void high permeability is that the voids are more likely to be inter-
content at t/NMAS of 2.0, 3.0, and 4.0. For most mixes, spec- connected all the way through the samples when the samples
imens could not achieve the target air voids even when the are thinner. Hence when mixes are placed thin, in this case
97.5 40
97.0
Steel Wheel Roller 30

96.5 Difference in Temperature
Difference in Temperature, oC
96.0 20
%Lab Density
95.5
10
95.0
%Lab Density Difference in Temperature

94.5 Steel Roller 2
R = 0.9821 0
2
R = 0.6392
94.0
-10
93.5
93.0 -20
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
t/NMAS
Figure 25. Relationships between density, t/NMAS, and temperature for Section 1.
23
96.0 50.0
Difference in Temperature.
2
95.0 R = 0.998
40.0
94.0
30.0
Difference inTemperature, C
o
93.0
%Lab Density
20.0
92.0
10.0
91.0
%Lab Density
0.0
%Lab Density Steel Roller
90.0
Steel/Rubber Tire Roller 2
R = 0.6796
2
R = 0.5115
-10.0
89.0
88.0 -20.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
t/NMAS
100.0 50
Difference in Temperature
Steel Wheel Roller 2
99.0 R = 0.976
Rubber Tire Roller 40
%Lab Density
Difference in Temperature Steel Roller
98.0
2
R = 0.8335
30
97.0
Difference in Temperature, C
o
20
96.0
%Lab Density
95.0 10
94.0
%Lab Density 0
93.0 2
R = 0.1864
-10
92.0
-20
91.0
90.0 -30
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
t/NMAS
24
104.0 100
%Lab Density
103.0 Steel Roller
102.0 Steel Wheel Roller 2
R = 0.8711
Rubber Tire Roller 80
101.0
Difference in Temperatur e
100.0
99.0
98.0 60
o
97.0
96.0
%Lab Density
40
95.0
94.0 %Lab Density
93.0 Steel/Rubber Tire Roller
2 20
92.0 R = 0.7651
91.0
90.0 0
2
R = 0.8884
88.0
87.0 -20
86.0
85.0
84.0 -40
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
t/NMAS
105.0 70
Different in Temperature @ 20 Min.
2
R = 0.8191
104.0
60
103.0
50
102.0
o
101.0
%Lab Density
40
100.0
30
99.0
%Lab Density
98.0 Steel Roller 20
2
R = 0.7736
97.0
10
96.0
95.0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
t/NMAS
25
100 70
Difference in Temperature
2
R = 0.6816
99 60

98 Difference in Temperature 50
Difference in Temperature, °C
%Lab Density
97 40
96 30
95 20

2
94 R = 0.4489 10
93 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
t/NMAS
100.0 70
Steel Roller
99.0 60
50
97.0
40
96.0 %Lab Density
Difference in Temperature,°C
Steel Roller
2 30
95.0 R = 0.6529
%Lab Density
94.0 20
93.0 10
92.0 Difference in Temperature %Lab Density 0

2
R = 0.9904 Steel/Rubber Tire Roller
91.0 2
R = 0.8092
-10
90.0
-20
89.0
88.0 -30
87.0 -40
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
t/NMAS
26
104.0
102.0
100.0
98.0 2
y = -0.8216x + 7.2531x + 80.848
2
R = 0.2562
%Lab Density
96.0
94.0
92.0
90.0
88.0
86.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
t/NMAS
Figure 32. Relationships between density and t/NMAS for all sections.
180
25 mm (1.0")
32 mm (1.25")
160 38 mm (1.5")
44 mm (1.75")
140 51 mm (2.0")
64 mm (2.5")
89 mm (3.5")
Temperature of Mix, °C
120
100
80
60
40
20
0
0 10 20 30 40 50 60 70
Time, min.
Figure 33. The effect of layer t/NMAS and cooling time on mix temperature.
27
250
Average Permeability x E-05 cm/sec 200
Recommended maximum
150
permeability, 125 x 10-5
100
50
0
2:1 3:1 4:1 2:1 3:1 4:1 2:1 3:1 4:1 2:1 3:1 4:1 2:1 3:1 4:1 2:1 3:1 4:1 2:1 3:1 4:1
9.5 ARZ 9.5 BRZ 9.5 SMA 12.5 SMA 19.0 ARZ 19.0 BRZ 19.0 SMA
t/NMAS @ each mix
Figure 34. Relationships between permeability and t/NMAS.
less than a 3:1 t/NMAS, the air voids have to be lower to 4.9 PART 2—EVALUATION OF RELATIONSHIP
ensure that the mixes are impervious. OF LABORATORY PERMEABILITY,
DENSITY AND LIFT THICKNESS OF FIELD
COMPACTED CORES
4.8 EVALUATION OF EFFECT OF t/NMAS ON
PERMEABILITY FROM FIELD STUDY The average thickness, the average air void content by
the vacuum seal device method, and the average laboratory
Permeability tests were conducted on the seven HMA sec- permeability values were determined for each of the cores
tions that were evaluated in the field. These tests were con- obtained from the work under NCHRP Project 9-9 (1). Figures
ducted in-place with the field permeameter and in the labora- 35 through 37 present the plots of in-place air voids versus
tory with the lab permeability test. Cores were taken from the permeability for each NMAS mix. The relationship between
in-place pavement for measurement of density and for mea- in-place air voids and permeability for 9.5-mm NMAS is illus-
surement of lab permeability. The field permeability values trated in Figure 35. The R2 values for both coarse-graded
were determined adjacent to the location where the cores were and fine-graded mixes were relatively high (0.70 and 0.86,
taken for density and for lab permeability. The results of these respectively) and both relationships are significant (p-value
tests for the 7 sections are provided in Table 14. = 0.000). At 8 percent air voids, the pavement is expected to
In summary, the coarse-graded mixes had permeability have a permeability of 60 × 10−5 cm/sec for coarse-graded
values that exceeded the recommended value when the air mix and 10 × 10−5 cm/sec for fine-graded mix. Because there
voids exceeded about 8 percent. The fine graded mixes never are only a couple of data points for fine-graded mix above
exceeded the recommended value even up 9 to 10 percent approximately 10 percent air voids, this model should not
air voids. be used to predict permeability at these higher void levels. At
TABLE 14 Comparison of laboratory and field permeabilities

Section Mix Type In-Place Air Field Lab
Number Voids Permeability Permeability
(percent) (cm/s x 10-5) (cm/s x 10-5)
1 9.5mm FG 6.6 to 8.8 1 to 28 1 to 35
2 9.5mm CG 9.0 to 12.6 14 to 632 107 to 1070
3 9.5mm SMA 7.7 to 12.6 110 to 651 29 to 168
4 12.5mm SMA 4.1 to 17.9 3 to 1778 0.1 to 5850
5 19.0mm FG 5.7 to 9.5 38 to 161 1 to 77
6 19.0mm CG 5.3 to 9.8 10 to 1760 1 to 141
7 19.0mm CG 4.8 to 15.2 72 to 3030 0 to 1203
28
4000
3500 Coarse-Graded
Fine-Graded
Lab Permeability, E-5 cm/sec

Coarse-graded
3000 y = 2.279e0.4225x
R2 = 0.6942
2500
2000
1500
1000 Fine-graded
y = 0.0309e0.7457x
500 R2 = 0.8602
0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0
In-place Air Voids (Corelok), %
Figure 35. Plot of permeability versus in-place air voids for 9.5-mm NMAS mixes.
lower void levels the coarse-graded mixes are more permeable is significant (p-value = 0.000). Based on the trend line, per-
than fine-graded mixes. meability is very low at air void contents less than 8 percent.
The relationships for the coarse-graded and fine-graded At air void contents above 8 percent, the permeability begins
12.5-mm NMAS mixes are shown in Figure 36. For these to increase rapidly with a small increase in in-place air void
projects there was no significant difference between fine and content. At 8 percent air voids, the fine-graded 19.0-mm
coarse graded mixes. The relationships between in-place air NMAS mix has a permeability value of 16 × 10−5 cm/sec.
voids and permeability for both gradation types were reason-
able and significant with an R2 of 0.61 for coarse-graded mixes
(p-value = 0.000) and 0.58 for fine-graded mixes (p-value 4.10 CONTROLLED LABORATORY
= 0.000). As shown by the best-fitted lines, the permeability EXPERIMENT TO EVALUATE METHODS
values for both gradation types were basically the same at a OF MEASURING THE BULK SPECIFIC
GRAVITY OF COMPACTED HMA
given air void content. The permeability at 8.0 percent air voids
for coarse-graded and fine-graded mixes was approximately 4.10.1 Introduction and Problem Statement
30 × 10−5 cm/sec.
Figure 37 illustrates the relationship between in-place A major concern of the HMA industry is the proper mea-
air voids and permeability for fine-graded 19.0-mm NMAS surement of bulk specific gravity (Gmb) for compacted samples.
mixes. The R2 value for this figure is 0.59 and the relationship This issue has become a bigger problem with the increased
1000
12.5 mm Coarse-Graded
900
12.5 mm Fine-Graded
800
700
600
500
400
300
200
100
0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0
Figure 36. Plot of permeability versus in-place air voids for 12.5-mm NMAS Mixes.
29
140
120

100
80
0.7362x
y = 0.0437e
2
R = 0.5923
60
40
20
0
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Figure 37. Plot of permeability versus in-place air voids for 9.5-mm NMAS mixes.
use of coarse gradations. Bulk specific gravity measurements gradations passing close to or above the Superpave defined
are the basis for volumetric calculations used during HMA maximum density line (i.e., fine-graded). However, since the
mix design, field control, and construction acceptance. Dur- adoption of the Superpave mix design system and the increased
ing mix design, volumetric properties such as air voids, voids use of SMA, mixes are being designed with coarser gradations
in mineral aggregates, voids filled with asphalt, and percent than in the past.
theoretical maximum density at a certain number of gyrations The potential problem in measuring the Gmb of mixes like
are used to evaluate the acceptability of mixes. All of these coarse-graded Superpave and SMA using the SSD method
properties are based upon Gmb. comes from the internal air void structure within these mix
In most states, acceptance of HMA construction by the types. These types of mixes tend to have larger internal air
owner is typically based upon percent compaction (density voids than the finer conventional mixes, at similar overall
based upon Gmb and theoretical maximum density). Whether air void contents. Mixes with coarser gradations have a much
nondestructive (e.g., nuclear gauges) or destructive (e.g., cores) higher percentage of large aggregate particles. At a certain
tests are used as the basis of acceptance, Gmb measurements overall air void volume, which is mix specific, the large
are equally important. When nondestructive devices are uti- internal air voids of the coarse mixes can become inter-
lized, each device first has to be calibrated to the Gmb of cores. connected. During Gmb testing with the SSD method, water
If the Gmb measurements of the cores are inaccurate in this can quickly infiltrate into the sample through these intercon-
calibration step, then the nondestructive device will provide nected voids. However, after removing the sample from the
inaccurate data. Additionally, pay factors for construction, water bath to obtain the saturated-surface dry condition the
whether reductions or bonuses, are generally based upon water can also drain from the sample quickly. This drain-
percent compaction. Thus, errors in Gmb measurements can ing of the water from the sample is what causes errors when
potentially affect both the agency and producer. using the SSD method.
For many years, the measurement of Gmb for compacted Because of the potential errors noted with the saturated
HMA has been accomplished by the water displacement surface-dry test method of determining the bulk specific grav-
method using saturated-surface dry (SSD) samples. This ity of compacted HMA, the primary objective of this task was
method consists of first weighing a dry sample in air, then to compare AASHTO T166 with other methods of measur-
obtaining a submerged mass after the sample has been placed ing bulk specific gravity to determine under what conditions
in a water bath for a specified time interval. Upon removal AASHTO T166 is accurate.
from the water bath, the SSD mass is determined after The plan for this part of the study was to evaluate two sep-
patting the sample dry using a damp towel. Procedures for arate sample types: laboratory compacted and field compacted.
this test method are outlined in AASHTO T166 (ASTM Laboratory compacted mixtures having various aggregate
D2726). types, nominal maximum aggregate sizes, gradation shapes,
The SSD method has proven to be adequate for conven- and air void levels were prepared. Each of the prepared
tionally designed mixes, such as those designed according to samples was tested to determine bulk specific gravity by four
the Marshall and Hveem methods, that generally utilized fine- different test methods: water displacement, vacuum-sealing,
graded aggregates. Historically, mixes were designed to have gamma ray, and dimensional.
30
For the field compacted samples, cores obtained during Statistically, results of the DMRT comparisons show that
the field validation portion of this study were subjected to the all methods produced statistically different air void contents.
same four bulk specific gravity test methods. Because cores However, vacuum-sealing and gamma ray bulk specific grav-
have a different surface texture than laboratory compacted ity methods provided similar results given a difference of
samples, it was necessary to evaluate them also. Testing also 0.24 percent air voids. On average, the dimensional method
conducted on core samples included laboratory permeabil- resulted in the highest air void contents, followed by the
ity tests and effective air void content using the vacuum- vacuum-sealing and gamma ray methods, respectively. Air
sealing device. void contents determined from AASHTO T166 resulted in
the lowest air void contents. None of the alternative meth-
ods provided similar results to AASHTO T166.
The results for SMA mixtures are provided in Figure 39. As
4.10.2 Field Compacted Samples with the Superpave mixes, the vacuum-sealing and gamma
ray methods resulted in similar air void contents. The dimen-
Each of the cores obtained during the Task 5 field valida- sional method again resulted in the highest air voids and
tion were tested to determine bulk specific gravity using the the AASHTO T166 method resulted in the lowest air voids.
same four tests as the laboratory experiment: water displace- Analysis of both the Superpave and SMA data indicated that
ment, vacuum sealing, gamma ray, and dimensional analysis. the four methods of measuring bulk specific gravity signif-
Because of the differences in surface texture between labora- icantly affected resulting air voids. For both mix types, the
tory compacted samples (surface texture around entire sam- vacuum-sealing and gamma ray methods provided similar air
ple) and field compacted samples (surface texture only on top voids; however, the dimensional method provided significantly
of sample because of core bit and sawing), the experiment higher air voids and AASHTO T166 provided significantly
was also extended to core samples. lower air void contents.
Because of the differences in resulting air voids for the Theoretically, the dimensional method should provide the
four methods of measuring bulk specific gravity, a Duncan’s highest measured air void content, as this method includes
multiple range test (DMRT) was conducted to determine both the internal air voids and the surface texture of the sam-
which methods, if any, provided similar results. This analy- ple. Therefore, the results in Figures 38 and 39 pass the test
sis method provides a ranking comparison between the dif- of reasonableness for the vacuum-sealing, gamma ray, and
ferent methods. The range of sample means for a given set of AASHTO T166 methods as all three provided air void content
data (method) can be compared to a critical valued based on lower than the dimensional method.
the percentiles of the sampling distribution. The critical value Because it was assumed that the T-166 method would be
is based on the number of means being compared (four, rep- accurate at low water absorption levels, it was decided to test
resenting the different methods) and number of degrees of the mixes with low absorption, less than 0.5 percent, to see
freedom at a given level of significance (0.05 for this analy- which mixes provided results similar to the T-166 method.
sis). Results of the DMRT analysis for the Superpave mixes The results are provided in Figure 40. This figure shows that
are illustrated in Figure 38. the vacuum-sealing and AASHTO T166 methods provided
11 Letters represent results of Duncan's

A Multiple Range Test for air voids
10 9.39 resulting from the bulk specific gravity
methods. Methods with the same letter
9 ranking are not significantly different.
Average Air Void Content, %
B C
8 7.50
7.26
D
7
6.22
6
0
Dimensional Vacuum-Sealing Gamma Ray AASHTO T166
Test Method
Figure 38. Average air voids and DMRT results for Superpave mixes.
31
12
Letters represent results of Duncan's
A Multiple Range Test for air voids
10.11 resulting from the bulk specific gravity
10 ranking are not significantly different.
Average Air Void Content, % B

B
8 7.24 7.09
C
6
4.97
0
Dimensional Vacuum-Sealing Gamma Ray AASHTO T166
Test Method
Figure 39. Average air voids and DMRT results for SMA mixes.
similar results and that both were significantly different than to the gamma ray method may make this method a viable
the dimensional and gamma ray methods. The dimensional option in the future.
method provided the highest air void content, as expected.
The AASHTO T166 method is accurate for low water absorp-
tion mixes and at these low void levels provide similar den- 4.10.3 Analysis of Field Compacted Samples
sity values to that of the vacuum seal method. These results
suggest that the vacuum-sealing method provides an accurate Included within this portion of the study were the cores
density for low voids, which indicates that it also provides an obtained during the Task 5 field validation experiment. Only
accurate density at higher void levels because the plastic seal the vacuum-sealing and AASHTO T166 test methods were
will clearly prevent water from being absorbed into the mix- analyzed, as they were shown most accurate during the labo-
ture. Figures 38 and 39 suggest that the gamma ray method ratory phase of this experiment. Figure 41 illustrates the rela-
does an overall adequate job of estimating bulk specific grav- tionship between air voids determined from the two methods
ity; however, Figure 40 suggests that it is not as accurate as for all field cores obtained from the 20 field projects during
AASHTO T166 or the vacuum-sealing methods. Refinements Task 5. This figure illustrates that when air void content is less
7.0
Letters represent results of Duncan's
Multiple Range Test for air voids
6.0 resulting from the bulk specific gravity
A
B ranking are not significantly different.
4.9
5.0
Air Void Content, %
4.3 C C
4.0
4.0 3.8
3.0
2.0
1.0
0.0
Dimensional Gamma Ray Vacuum-Sealing AASHTO T166
Test Method
Figure 40. Comparison of test methods, mixes with low water absorption level.
32
24.0
22.0
20.0
18.0
Air Voids (AASHTO T166), %
16.0
14.0 0.842
y = 1.2486 x
12.0 2
R = 0.867 6
10.0
8.0
6.0
4.0
2.0
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 20.0 22.0 24.0 26.0
Air Voids (Vacuum-Sealing), %
Figure 41. Comparisons between AASHTO T166 and vacuum-sealing methods, field projects.
than about 5 percent, the two methods provided approximately tigated for 12.5 and 19.0 mm NMASs. The effect of lift thick-
similar results. Above 5 percent air voids, the vacuum-sealing ness was evaluated within the 9.5, 12.5, and 19.0 mm NMASs.
method resulted in higher air void contents. As air voids To determine if a general trend occurred between in-place air
increased, the two methods diverged and it is believed that the voids and t/NMAS, a regression was performed on the com-
reason for this divergence is the loss of water during the SSD bined data. Figure 42 illustrates this general relationship. From
method. Hence, at low air voids, both methods should be close this regression, a low R2 of 0.09 was found. The trendline sug-
to correct; however, at higher air voids the vacuum-sealing gested that as the ratio of lift thickness to NMAS increased,
method should be more correct. in-place air voids decreased.
To determine if the relationship between in-place air
voids and the t/NMAS ratio was significant, an ANOVA
4.11 FIELD VALIDATION OF RELATIONSHIPS
BETWEEN PERMEABILITY, LIFT was conducted on the regression. For the combined data,
THICKNESS, AND IN-PLACE DENSITY the p-value was 0.014, which indicated that the overall rela-
tionship was significant. Then the data were separated into
The main objective of the field portion of NCHRP 9-27 the three mix types. When an ANOVA was conducted on the
(Task 5) was to provide a field validation of the relationships regressions for the mix types, it was found that the relation-
between permeability, lift thickness, and in-place density so ship was not significant for any of the mix types (p-values
the overall objectives of the study could be accomplished. In of 0.956, 0.994, and 0.107 for fine-graded, coarse-graded,
order to field verify the relationships between air voids, lift and SMA, respectively). There is a lot of scatter in the data, but,
thickness, and permeability, 20 HMA construction projects as can be seen in Figure 42, every increase of 1 in the t/NMAS
were visited. Testing at these projects included tests on plant- results in a decrease in voids of approximately 0.6 per-
produced mix and on the compacted pavement. Testing of cent. This finding involves average numbers, and it must
the plant produced mix included compacting samples to both be realized that many other factors affect the density of these
the design compactive effort and to a specified height. Test- field projects.
ing on the compacted pavement included performing field Another factor to consider for these projects is the specifi-
permeability tests with the NCAT Field Permeameter. Selec- cation requirements were approximately the same for all of
tion of the 20 projects was based upon the following factors: these mixes. Hence, the contractor was trying to compact all
NMAS, gradation type (fine-graded, coarse-graded, and mixes to a low void content. Even with the same target density
SMA), and the lift thickness to NMAS ratio (t/NMAS). Table the t/NMAS affected the results.
15 presents the 20 projects evaluated. For Figure 43, a best-fit line was produced on the com-
Table 15 shows that both fine- and coarse-graded Superpave bined data for the 12.5-mm NMAS mixes. A low correlation
designed mixes were investigated for each of four NMAS, was also found for this regression (0.19), but the general
ranging from 9.5 to 25.0 mm NMAS. SMA mixes were inves- trend suggested that in-place air voids decreased as the lift
33
TABLE 15 Field project summary information

Fine or Average Lift Actual Lift AC
Project Coarse Thickness Thickness/ Performance
ID NMAS Gradation (mm) NMAS Ratio Grade Ndesign
1 9.5 Fine 48.7 5.1:1 70-22 65
2 19.0 Coarse 65.7 3.5:1 64-22 65
3 9.5 Coarse 32.3 3.4:1 64-22 65
4 12.5 Fine 68.6 5.5:1 * 75
5 9.5 Fine 41.0 4.3:1 70-22 100
6 12.5 Coarse 50.3 4.0:1 58-28 75
7 9.5 Fine 40.6 4.3:1 64-28 75
8 19.0 Coarse 58.9 3.1:1 64-22 100
9 19.0 Coarse 96.4 5.1:1 64-22 100
10 19.0 Coarse 70.9 3.7:1 64-34 100
11 19.0 Coarse 38.0 2.0:1 64-34 125
12 25.0 SMA 42.6 1.7:1 76-22 50
13 25.0 Fine 70.0 2.8:1 67-22 100
14 9.5 SMA 26.8 2.8:1 76-22 75
15 19.0 Coarse 50.4 2.7:1 76-22 100
16 12.5 Coarse 43.8 3.5:1 67-22 86
17 12.5 Fine 43.3 3.5:1 64-22 75
18 12.5 Coarse 44.5 3.6:1 67-22 75
19 9.5 Fine 41.5 4.4:1 67-22 75
20 12.5 Fine 34.5 2.8:1 67-22 80
* Designated RA295 by the agency
thickness increased. An ANOVA conducted for the com- NMAS mixes, as well as for the individual mix types. For the
bined regression indicated that the relationship was signifi- combined data, the regression produced a low R2 value (0.09).
cant (p-value = 0.001). The data were then separated into the An ANOVA performed on the regression determined that the
different mix types to see if the relationship was significant relationship between t/NMAS and in-place air voids for the
for each mix type. For the fine-graded mixes, the relationship 19.0-mm NMAS mixes was significant (p-value of 0.000).
was significant (p-value = 0.000). The coarse-graded mixes The data indicate that an increase of 1 for the t/NMAS results
did not have a significant relationship between in-place air in an average decrease of 1.0 in the air voids.
voids and t/NMAS (p-value = 0.932). These data indicate In summary, even though there is a large amount of scatter
that an increase of 1 for the t/NMAS resulted in an average in the data for the three NMAS mixes, the results suggest that
decrease in air voids of 0.5 percent. the air voids dropped 0.5 to 1.0 percent for each increase of 1
Figure 44 shows the relationship between lift thickness and in the t/NMAS. This shows the importance of making sure
in-place air voids for the combined data set for the 19.0-mm that the t/NMAS is sufficiently high.
16.0
All Data y = -0.5809x + 10.792

14.0
Fine y = -0.0017x + 7.9818
Coarse y = -0.0488x + 9.1967
12.0
SMA y = -3.0894x + 16.837
In-place Air Voids, %
SMA Coarse
10.0
Fine
8.0
6.0
All Data
All Data: R2 = 0.09, p-value = 0.014
4.0 Fine: R2 = 0.00, p-value = 0.956
Coarse: R2 = 0.00, p-value = 0.994
2.0 SMA: R2 = 0.17, p-value = 0.107
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Thickness-to-NMAS Ratio
Figure 42. Relationship between t/NMAS and in-place air voids—9.5 mm, all data.
34
16.0
All Data y = -0.5205x + 10.079
14.0 Fine y = -0.6277x + 11.1
Coarse y = 0.0412x + 7.2391
12.0
10.0
Fine
8.0
6.0 Coarse
All Data
2
4.0 All Data: R = 0.19, p-value = 0.000
Fine: R2 = 0.41, p-value = 0.000
2.0 Coarse: R2 = 0.00, p-value = 0.932
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Thickness-to-NMAS Ratio
Figure 43. Relationship between t/NMAS and in-place Air Voids—12.5 mm NMAS.
16.0
14.0
12.0
10.0
8.0
6.0 All Data (All Coarse)
y = -1.0455x + 11.09
4.0 R2 = 0.0913
p-value = 0.020
2.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Thickness-to-NMAS-Ratio
Figure 44. Relationship between t/NMAS and in-place air voids—19.0 mm NMAS.
35
CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS
The density that can be obtained under normal rolling con- of the sample, an error occurs. The allowable absorption
ditions is clearly related to the t/NMAS. For improved com- level to use the displacement test method is specified as
pactibility, it is recommended that the t/NMAS be at least 3 2 percent in AASHTO T166, but this level of absorption
for fine-graded mixes and at least 4 for coarse-graded mixes. can create accuracy problems, as shown in this report. It
The data for SMA indicate that the ratio should also be at least is recommended that the absorption limit for the displace-
4. Ratios less than these suggested numbers could be used, but ment test method be reduced to 1 percent. If the vacuum-
more compactive effort would generally be required to obtain seal method is adopted on a project, the measured voids may
the desired density. In most cases, a t/NMAS of 5 does not now be somewhat higher than with the water displacement
result in the need for more compactive effort to obtain maxi- method.
mum density. However, care must be exercised when the The water displacement method was accurate for all water
thickness gets too large to ensure that adequate density is absorption levels encountered for mixes that were fine-graded
obtained. (ARZ gradations). For mixes having gradations near the max-
The results of the evaluation of the effect of mix tempera- imum density line (TRZ) or coarser (BRZ and SMA), the
ture on the relationship between density and t/NMAS indi- level of water absorption at which AASHTO T166 began to
cate that one of the reasons for low density at thinner sections lose accuracy was between 0.2 and 0.4 percent.
(lower t/NMAS) is the more rapid cooling of the mixture. For mix design samples and other laboratory samples that
Hence, for thinner layers it is even more important that rollers are compacted to relatively low voids, the displacement method
stay very close to the paver so that rolling can be accomplished will provide reasonably accurate answers. However, for field
prior to excessive cooling. For the conditions of this study, the samples where the void levels will typically be 6 percent or
mixes placed at the NCAT test track at 25-mm thickness higher, it is important to evaluate absorption to determine if the
cooled twice as fast as mixes placed at 37.5-mm thickness. For vacuum-seal method needs to be used.
thicker sections (larger t/NMAS), the rate of cooling is typi- Care must be used when using the vacuum sealing method
cally not a problem. to measure density. Many times the plastic bag develops a leak
The in-place void content is the most significant factor during the test, leading to an error in the result. Weighing the
impacting permeability of HMA mixtures. This is followed sample in air after measuring the submerged weight will indi-
by coarse aggregate ratio and VMA. As the values of coarse cate if a leak has developed. If a leak is identified, the test must
aggregate ratio increases, permeability increases. Permeability be repeated until an acceptable test is achieved.
decreases as VMA increases for constant air voids. There appears to be a need for a correction factor for the
The variability of permeability between various mixtures vacuum-sealing and water displacement methods to provide
is very high. Some mixtures are permeable at the 8 to 10 per- equal measured air void contents even when the air void level
cent void range and others do not seem to be permeable at is low. The correction factor for the mixtures evaluated in
these higher voids. However, to ensure that permeability is this report was approximately 0.2 percent air voids. A better
not a problem, the in-place air voids should be between 6 and determination of the correction factor can be made for specific
7 percent or lower. This appears to be true for a wide range dense graded mixes by compacting samples in the Superpave
of mixtures regardless of NMAS and grading. gyratory compactor to approximately 4 percent air voids
When laboratory prepared samples having low levels of (design air void content) and testing using the two test methods.
water absorption were evaluated, the dimensional method The difference between these two tests will be the correction
resulted in the highest air void contents followed by the gamma factor for the mix.
ray method. The vacuum-sealing and water displacement The in-place air voids of the 20 field projects were high.
(AASHTO T166) methods resulted in similar air void con- Fourteen of the 20 mixes tested had average in-place air voids
tents when the water absorption level was low. The vacuum above 8 percent and seven of the mixes had average air voids
seal method is an acceptable method to use for low and high over 10 percent (based on test results with the vacuum-seal
void levels. method). This low density on a high percentage of random
At low levels of water absorption, the water displace- projects is disturbing because this lower density will most
ment method is an accurate measure for bulk specific grav- certainly lead to significant loss in pavement life.
ity. The error develops when removing the sample from More emphasis must be placed on obtaining adequate
water to determine the SSD weight. When water flows out density. Regardless of the method of density measurement
36
used, some cores have to be taken and tested for calibration. Even though there is a lot of scatter within and between
The most reliable way to measure density is to take cores projects, most field results support the finding that higher
for density testing. If the amount of absorption during den- t/NMAS ratios generally provide lower void levels. Coarse-
sity measurement exceeds 1 percent, the T166 method will graded mixtures generally have higher permeability values
likely provide a higher measured density than the true den- than the fine-graded mixtures for a given air void level. Air
sity. The vacuum seal method is one approach to measure a voids were clearly shown to be a key determinant of perme-
density more accurately when the water absorption exceeds ability. However, many times the air voids were reasonably
1 percent. low (5 to 7 percent) and the permeability was still high.
37
CHAPTER 6
REFERENCES
1. Roberts, F. L., P. S. Kandhal, E. R. Brown, D. Y. Lee, and T. W., 7. Mallick, R. B., L. A. Cooley, Jr., and M. Teto. “Evaluation of
Kennedy. Hot Mix Asphalt Materials, Mixture Design, and Con- Permeability of Superpave Mixes in Maine, Final Report.” Tech-
struction. NAPA Education Foundation, Lanham, MD. Second nical Report ME-001, November 1999.
Ed., 1996. 8. Musselman, J. A., B. Choubane, G. C. Page, and P. B. Upshaw.
2. Zube, E. “Compaction Studies of Asphalt Concrete Pavements “Superpave Field Implementation: Florida’s Early Experience.”
as Related to the Water Permeability Test.” Highway Research Transportation Research Record 1609, 1998.
Board, Bulletin 358, 1962. 9. Westerman, J. R. AHTD’s Experience with Superpave Pave-
3. Brown, E. R., R. Collins, and J. A. Brownfield. “Investigation ment Permeability. Presented at Arkansas Superpave Sympo-
of Segregation of Asphalt Mixtures in the State of Georgia.” sium, January 21, 1998.
Transportation Research Record 1217, 1989. 10. “Superpave Construction Guidelines.” National Asphalt
4. Choubane, B., G. C. Page, and J. A. Musselman. “Investigation Pavement Association. Special Report 180. Lanham, Mary-
of Water Permeability of Coarse-Graded Superpave Pave- land. October 1997.
ments.” Journal of the Association of Asphalt Paving Technolo- 11. Cooley, Jr., L. A., B. D. Prowell, M. R. Hainin, M. S. Buchanan,
gists, Volume 67, 1998. and J. Harrington. Bulk Specific Gravity Round Robin Using the
5. Ford, M. C., and C. E. McWilliams. “Asphalt Mix Permeabil- Corelok Vacuum Sealing Device. National Center for Asphalt
ity.” University of Arkansas, Fayetteville, AR, 1988. Technology (NCAT) Report No. 02-11. November 2002.
6. Hudson, S. B., and R. L. Davis. “Relationship of Aggregate
Voidage to Gradation.” Proceedings of the Association of
Asphalt Paving Technologists, Volume 34, 1965.
A-1
APPENDICES A THROUGH E
UNPUBLISHED MATERIAL
Appendices A, B, C, D, and E as submitted by the research Appendix C: Lift Thickness Versus Density Data Using
agency are not published herein. For a limited time, they are Vibratory Compactor
available for loan on request to NCHRP. Their titles are as Appendix D: Lift Thickness Versus Permeability Data
follows: Appendix E: Factors Affecting Permeability Data Using
Field Core Samples
Appendix A: Mix Design Summary Information for Part 1
Appendix B: Lift Thickness Versus Density Data Using
Gyratory Compactor
Abbreviations used without definitions in TRB publications:
AASHO American Association of State Highway Officials

AASHTO American Association of State Highway and Transportation Officials
APTA American Public Transportation Association
ASCE American Society of Civil Engineers
ASME American Society of Mechanical Engineers
ASTM American Society for Testing and Materials
ATA American Trucking Associations
CTAA Community Transportation Association of America
CTBSSP Commercial Truck and Bus Safety Synthesis Program
FAA Federal Aviation Administration
FHWA Federal Highway Administration
FMCSA Federal Motor Carrier Safety Administration
FRA Federal Railroad Administration
FTA Federal Transit Administration
IEEE Institute of Electrical and Electronics Engineers
ITE Institute of Transportation Engineers
NCHRP National Cooperative Highway Research Program
NCTRP National Cooperative Transit Research and Development Program
NHTSA National Highway Traffic Safety Administration
NTSB National Transportation Safety Board
SAE Society of Automotive Engineers
TCRP Transit Cooperative Research Program
TRB Transportation Research Board
U.S.DOT United States Department of Transportation

NCHRP RPT 531

Uploaded by

Copyright:

Available Formats

NCHRP RPT 531

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NCHRP RPT 531

Uploaded by

Copyright:

Available Formats

NATIONAL

NCHRP REPORT 531

Relationship of Air Voids,

MARION C. BLAKEY, Federal Aviation Administrator, U.S.DOT (ex officio)

NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM

NCHRP REPORT 531

TRANSPORTATION RESEARCH BOARD

NCHRP PROJECT 9-27 PANEL

JAMES A. MUSSELMAN, Florida DOT (Chair)

• Volume I: Task 3—Parts 1 and 2;

INTRODUCTION AND PROBLEM STATEMENT

TABLE 1 Physical properties of aggregate

TABLE 2 Mix information for field density study

TEST RESULTS AND ANALYSIS

TABLE 3 Summary of mix design results for Superpave mixes

TABLE 4 Summary of mix design results for SMA mixes

Figure 1. Relationship between air voids for ARZ mixes.

16.0 y = 1.1074x + 0.3893

Figure 2. Relationship between air voids for TRZ mixes.

Air Voids, % - Vacuum Seal Method

Figure 3. Relationship between air voids for BRZ Mixes.

Figure 4. Relationship between air voids for SMA mixes.

Average Air Voids, %

Figure 5. Relationships between air voids and t/NMAS for 9.5-mm

Figure 6. Relationships between air voids and t/NMAS for

Figure 7. Relationships between air voids and t/NMAS for 37.5-mm

4.4 EVALUATION OF EFFECT OF t/NMAS ON 4.4.1 Section 1

Average Air Voids, %

Figure 18. Relationships of air voids and t/NMAS for 9.5-mm

TABLE 6 Relationship of air voids and t/NMAS for 9.5-mm coarse-graded

9.0 Steel Roller

TABLE 11 Relationship of air voids and t/NMAS for 19.0-mm coarse-graded

TABLE 12 Effect of t/NMAS on compactibility of HMA

Steel Wheel Roller 30

%Lab Density Difference in Temperature

Steel/Rubber Tire Roller

Steel/Rubber Tire Roller

92.0 Difference in Temperature %Lab Density 0

Average Permeability x E-05 cm/sec 200

Figure 34. Relationships between permeability and t/NMAS.

TABLE 14 Comparison of laboratory and field permeabilities

Lab Permeability, E-5 cm/sec

Lab Permeability, E-5 cm/sec

In-place Air Voids (Corelok), %

11 Letters represent results of Duncan's

Average Air Void Content, % B

Air Voids (Vacuum-Sealing), %

TABLE 15 Field project summary information

All Data y = -0.5809x + 10.792

6.0 All Data (All Coarse)

CONCLUSIONS AND RECOMMENDATIONS

AASHO American Association of State Highway Officials

You might also like