Comparative Evaluation For The Performan

See
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/281826744
COMPARATIVE EVALUATION FOR THE

PERFORMANCE OF PAVING MATERIALS BY
USING MARSHALL AND...
Thesis · July 2006

DOI: 10.13140/RG.2.1.1649.9684
CITATIONS READS
0 6
1 author:
Noor Moutaz
Al-Mustansiriya University
10 PUBLICATIONS 2 CITATIONS
SEE PROFILE
All content following this page was uploaded by Noor Moutaz on 16 September 2015.
The user has requested enhancement of the downloaded file.

Ministry of Higher Education and Scientific Research
Al-Mustansiriya University
College of Engineering
Highway & Transportation Engineering Department
COMPARATIVE EVALUATION FOR THE

PERFORMANCE OF PAVING MATERIALS
BY USING MARSHALL AND SUPERPAVE
COMPACTION METHODS
A THESIS
SUBMITTED TO HIGHWAY AND TRANSPORTATION
ENGINEERING DEPARTMENT
COLLEGE OF ENGINEERING
AL-MUSTANSIRIYA UNIVERSITY
IN A PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF MASTER OF SCIENCE IN HIGHWAY AND
TRANSPORTATION ENGINEERING
BY
NOUR MOUTAZ ISMAIL ALAZAWY
(B.Sc), 2003
SUPERVISED BY
DR. NAMIR GHANI AHMED
MOHARRAM, 1426 FEBRUARY, 2006

‫اﻟﺮﱠﺣ ِﻴﻢ‬
‫ﱠﲪ ِﻦ ِ‬
‫ﺑِﺴِﻢ ِﷲ اﻟﺮ ْ‬
‫ْ‬
‫َﰊ‬
‫ﺎتِ ّر‬
‫َﻠِﻤ ِ‬
‫ْﺒَﺤﺮِﻣَﺪاداًِ ﻟّﻜ َ‬ ‫” ﻗُﻞ ْﻟﱠﻮﻛ َ‬
‫َﺎن اﻟ ُْ‬
‫َﰊوﻟَْﻮ‬
‫ﺎتِ ّر َ‬
‫َﻠِﻤ ُ‬
‫ْﺒَﺤﺮ ﻗـََﺒْﻞ أَن ﺗَﻨﻔََﺪﻛ َ‬
‫َﻨَﻔَﺪ اﻟ ُْ‬
‫ﻟِ‬
‫ِﺟْﺌـﻨَﺎ ﲟِِﺜْﻠِِﻪَﻣَﺪداً “‬
‫ﺻﺪق ﷲ اﻟﻌﻈﻴﻢ‬
‫) اﻵﻳﺔ ‪ ،109‬ﺳﻮرة اﻟﻜﻬﻒ(‬

‫‪ ‬أ‪ ฀‬اﻟﺬي ﻋﻠﻤﻨﻲ وﺳﺎﻧﺪﻧﻲ ‪...........‬واﻟﺪي‪...‬‬
‫‪ ‬أ‪ ฀‬ﻳﻨﺒﻮع اﻟﺤﻨﺎن واﻟﻌﻄﻒ ‪ ...............‬واﻟﺪﺗﻲ اﻟﻜﺮ ﻳﻤﻪ‪...‬‬
‫اﻟﺘﻲ راﻓﻘﺖ درﺑﻲ اﻟﻄﻮﻳﻞ ‪......‬إﺟﻼﻻ وﻋﺮﻓـﺎﻧﺎ ﺑﺎﻟﺠﻤﻴﻞ‬
‫‪ ‬أ‪ ฀‬ﻣﻦ ﻏﻤﺮوﻧﻲ ﺑﺎﺣﺘﺮاﻣﻬﻢ وﺣﺒﻬﻢ ‪ ...........‬أﺧﻮاﺗﻲ ‪...‬‬
‫إﻛﺮاﻣﺎ واﻋﺘﺰازا ﺑﻬﻢ‪.........‬‬

Supervisor Certificate
I certify that the preparation of this thesis entitled " Comparative

Evaluation for the Performance of Paving Materials by Using Marshall
and Superpave Compaction Methods " was prepared by the student
(Nour Moutaz Ismail Alazawy) under my supervision at the Highway and
Transportation Engineering Department, College of Engineering, AL-
Mustansiriya University, in a partial fulfillment of the requirements for
the Degree of Master of science in Highway and Transportation
Engineering .
Signature:
Name : Dr. Namir Ghani Ahmed
Date : / / 2006
In view of the available recommendations, I forward this thesis for debate

by Examining Committee.
Signature:
Name: Dr.Hamid Abdul Mahdi Faris
Chairman of Highway and Transportation Engineering Department
College of Engineering ;
AL-Mustansiriya University
Date: / / 2006
Certificate of the Examining Committee
We certify, as an examining committee , that we have read the thesis titled "
Comparative Evaluation for the Performance of Paving Materials by Using
Marshall and Superpave Compaction Methods " examined the student (Nour
Moutaz Ismail Alazawy) in its content and found that it meets the standard of a
thesis for the degree of Master of Science in Highway and Transportation
Engineering .
Signature: Signature:
Name: Dr.Gandhi .G.Sofia Name: Asst.Prof.Dr.Maher.B.Al-Fammani
(Member) (Member)
Date: / / 2006 Date: / / 2006
Signature: Signature:
Name: Dr.Namir G. Ahmed Name: Prof. Hamed M.H. Alani
(Supervisor) (Chairman)
Date: / / 2006 Date: / / 2006
Approved by the Dean the College of Engineering

Signature:
Name: Prof. Dr. Ali Al-Athari
Dean of the College of Engineering
AL-Mustansiriya University
Date: / / 2006
In the name of God , the most merciful , and before anything , thank to
God who enabled me to achieve this study .
I would like to express my deep gratitude and great appreciation to Dr.
Namir G. Ahmed, thesis advisor, whose advice, kindness, guideness and
help through the time of this study were indispensable.
Special thanks to Mr. Haeder for his grant assistance.
Many thanks to Mr.Zaid and Mr.Ahmed, from Civil Engineering
Department of Baghdad University, for their assistance.
I would also like to thank my colleagues who contributed in one way or
another in this work.
Final thanks and greeting to my lovely family for their continuous
guidance at all stages of this work.
I
The SHRP conducted a $ 50 million research effort to develop a Superpave mix design as
a new concept for the design of bituminous mixes. Although Superpave mixes have been
widely used by the developed countries for the last few years , the developing countries are
still working with the conventional mixes i.e. Marshall Mixes.
The main objective of the present study is the comparison between the conventional
Marshall design method and Superpave system design method, if applied in mix design of
the wearing course layers in flexible pavements.
A detailed experimental work is carried out to achieve the objectives of the study by
preparing (532) asphalt concrete Marshall and Superpave specimens through using
aggregate from AL-Taji Quarry and (40-50) grade of asphalt cement from Dourah refinery
and one type of cement as a mineral filler.
These specimens were tested by, Marshall, indirect tensile, creep and moisture damage
tests to evaluate the volumetric properties, tensile strength, creep strain and moisture
susceptibility for these mixes. The effect of mix type on the structural performance of the
wearing course was also examined by using MICHPAVE –Finite Element and PCPT
Programs. The effect of additives and film thickness of asphalt on the performance of
these mixes was also investigated.
From the analysis results, it is concluded that; the Superpave mixes are more economical
as compared with traditional Marshall mixes. In addition, adding carbon fiber and lime to
the mixes increases the tensile strength and reduces the creep strain. It is also noticed that,
Superpave Gyratory compactor is capable of achieving air void contents much lower than
that when using Marshall hammer. Furthermore, the Superpave mixes reflects better
resistance against permanent deformation and fatigue cracking , while , approximately the
same resistance to thermal cracking is shown by both mixes . Finally, the present study
recognizes the importance of using Superpave system instead of Marshall method in the
asphalt concrete mix design in Iraq.
II
Title Page No.
Acknowledgement I
Abstract II
Table of Contents III
List of Tables VIII
List of Figures XI
Symbols and Abbreviations VII
Chapter One : Introduction

1-1 Background 1
1-2 Problem Statement 2
1-3 Objectives of the Study 3
1-4 Scope of Thesis 3
1-5 Work Limitations 4
Chapter Two: Literature Review

2-1 Introduction 5
2-2 Marshall Mix Design 7
2-2-1 Material Selection 7
2-2-2 Aggregate Gradation 11
2-2-3 Specimen Fabrication 12
2-2-4 Volumetric Analysis 12
2-2-5 Stability and Flow Measurements 13
2-2-6 Optimum Asphalt Content 14
2-3 Superpave Mix Design 14
2-3-1 Gyratory Compactor 15
2-3-2 Material Selection 19
2-3-3 Design Aggregate Structure 24
2-3-4 Design Asphalt Content 30
III
2-3-5 Moisture Sensitivity of Design Mixture 31
2-4 Comparison of Marshall and Superpave 33
2-5 Effect of HMA Design Method on the Properties and Performance
of Pavement Structure 39
2-5-1 Elastic Layer Theory 42
2-5-1-1 Boussinesq's Half Space ( One Layer System ) 43
2-5-1-2 Burmister Theory ( Two Layer System ) 46
2-5-1-3 Multilayers System 47
2-5-2 Finite Element Method 48
2-5-2-1 Available Finite Element Computer Programs 49
Chapter Three: Materials and Methods of Testing

3-1 Materials 50
3-1-1 Asphalt Cement 50
3-1-2 Aggregate 51
3-1-3 Mineral Filler 51
3-1-4 Additives 54
3-2 Marshall Mix Design 55
3-2-1 Types of Mixes 55
3-2-2 Aggregate preparation and Gradation 55
3-3 Superpave Mix Design 56
3-3-1 Types of Mixes 56
3-3-2 The Design Aggregate Structure 56
3-3-4 Specimen Testing 58
3-4 Tests Methods 61
3-4-1 Resistance to Plastic Flow of Asphalt Mixture
(Marshall Test Method) 61
IV
3-4-2 Indirect Tensile Test 62
3-4-3 Creep Test 63
3-4-4 Standard Test for the Effect of Moisture on
Asphalt Concrete Paving Mixtures (Lottman Test) 64
3-5 Test Program of Asphalt Paving Mixture 65
Chapter Four: Results and Discussion

4-1 Introduction 68
4-2 Optimum Asphalt Content 68
4-2-1 Marshall Mixes 68
4-2-2 Superpave Mixes 69
4-3 Volumetric Properties 74
4-4 Indirect Tensile Test 77
4-5 Creep Test 83
4-6 Moisture Damage Test 96
4-7 Marshall Test 100
4-8 Asphalt Film Thickness 103
V
Chapter Five: Effect of Mix Design Method on the
Pavement Structural Performance
5-1 Introduction 108
5-2 Description of Mechanistic-Empirical Analysis Approach 109
5-3 FEM Approach to Mix Design 111
5-4 Factors Affecting Pavement Performance 115
5-5 Application of PCPT Program 116
5-6 Application of MICHPAVE Program 122
5-7 Comparison between the Mechanical Properties of
Each Mix 128
Chapter Six: Conclusions and Recommendations

5-1 Conclusions 133
5-2 Recommendations 134
5-3 Recommendations for Future Researches 135
References 136
Appendices
Appendix (A): Data Analysis for Marshall Mixes Design
Appendix (B): Data Analysis for Superpave Mixes Design
Appendix (C): Marshall Test Results and Superpave Mixes Analysis
Appendix (D): Creep Test Results for Mixes Design
Appendix (E): Indirect Tensile Strength Results for Mixes Design
Appendix (F): Moisture Damage Results for Mixes Design
Appendix (G): Data Results for MICHPAVE Program
VI
Figure No. Title Page No.
(2-1) Superpave Gyratory Compactor 15
(2-2) Mold Configuration 16
(2-3) Densification plot 17
(2-4) Superpave Mix Design 19
(2-5) 0.45 power gradation 24
(2-6) Superpave Aggregate Gradation 26
(2-7) Notations for Boussinesq's equations (Ullidtz, 1987) 44
(2-8) Two layers system (Burmister) 46
(2-9) Multilayer System 47
(3-1) Gradation of Wearing Course for two sections of
the Expressway No.1 in Iraq 53
(3-2) Local Superpave Gyratory Compactor 59
(3-3) Various steps of Superpave specimen fabrication 60
(3-4) Flow Chart of Testing and Evaluation Program 67
(4-1) Superpave Specimens 69
(4-2) Relationship between Asphalt content and Air void
of Superpave Specimen (for R1 Gradation) 70
(4-3) Relationship between Asphalt content and VMA
(4-4) Relationship between Asphalt content and VFA
(4-5) Relationship between Asphalt content and Air void
(4-6) Relationship between Asphalt content and VMA
(4-7) Relationship between Asphalt content and VFA
XI
(4-8) Effect of Mix design method on the Optimum asphalt content 73
(4-9) Effect of Mix design method on Volumetric properties 76
(4-10) Effect of Gradation on Volumetric properties 76
(4-11) Effect of Gradation on the Indirect tensile strength 78
(4-12) Effect of Additives on the Indirect tensile strength at 20C° test
Temperature for (R1)TRZ gradation 78
(4-15) Effect of Asphalt content on the Indirect tensile strength
(R1 gradation at 20 C° test temperature) 80
(4-16) Effect of Additives on the Indirect tensile strength
(Superpave with optimum of Superpave at 20C° test temperature) 81
(4-17) Effect of Testing Temperature on the Indirect tensile strength 83
(4-18) Strain –Time relationship for Marshall mixes (without additives) 84
(4-19) Strain – Time relationship for Marshall mixes
(with 0.5% carbon fiber) 85
(4-20) Strain – Time relationship for Marshall mixes
(with 1% carbon fiber) 85
(4-21) Strain – Time relationship for Marshall mixes (with 2% lime) 86
(4-22) Strain – Time relationship for Marshall mixes (with 4% lime) 86
(4-23) Strain-time relationship for R1 gradation of Marshall mixes
with different additives 87
(4-24) Strain-time relationship for R9 gradation of Marshall mixes
(4-25) Effect of gradation on permanent deformation for Marshall Mixes 88
(4-26) Effect of additives on permanent deformation for
Marshall Mixes (for R1 Gradation) 88
XII
Marshall Mixes (for R9 Gradation) 89
(4-28) Strain – Time relationship for Superpave mixes (without additives) 90
(4-29) Strain- Time relationship for Superpave mixes
(with 0.5% carbon fiber) 90
(4-30) Strain – Time relationship for Superpave mixes
(with 1% carbon fiber) 91
(4-31) Strain – Time relationship for Superpave mixes (with 2% lime) 91
(4-32) Strain – Time relationship for Superpave mixes (with 4% lime) 92
(4-33) Effect of gradation on permanent deformation for Superpave mixes 92
(4-34) Strain-time relationship for R1 gradation of Superpave mixes
(4-35) Strain-time relationship for R9 gradation of Superpave mixes
Superpave mixes (for R1 Gradation) 95
Superpave mixes (for R9 Gradation) 95
(4-38) Effect of Asphalt content on Moisture damage for Marshall and
Superpave mixes 96
(4-39) Effect of Mix design method on TSR Ratio 99
(4-40) Effect of Gradation on TSR Ratio 104
(4-41) Effect of Type of gradation on Stability values of Mixes 102
(4-42) Effect of Type of gradation on Flow values of Mixes 102
(4-43) Asphalt Film Thickness 105
(4-44) Effect of Mix Design Method on Asphalt Film Thickness 107
(5-1) Mechanistic and Empirical design and analysis 110
(5-2) Assumed pavement structure (Input Data) 114
(5-3) Low Temperature Crack in HMA 117
XIII
(5-4) Input Data PCPT Program for Marshall Mixes (R1 Gradation) 118
(5-5) Output Results PCPT Program for Marshall Mixes
(R1 Gradation) 118
(5-6) Input Data PCPT Program to for Superpave Mixes
(R1 Gradation) 119
(5-7) Output Results PCPT Program for Superpave Mixes
(R1 Gradation) 119
(5-8) Input Data PCPT Program for Marshall Mixes
(R9 Gradation) 120
(5-9) Output Results PCPT Program for Marshall Mixes
(R9 Gradation) 120
(5-10) Input Data PCPT Program for Superpave mixes
(R9 Gradation) 121
(5-11) Output Results PCPT Program for Superpave Mixes
(R9 Gradation) 121
(5-12) Resilient Modulus Model for Granular Soils 123
(5-13) Resilient Modulus Model for Cohesive Soils 124
(5-14) Typical Finite Element Mesh 125
(5-15) describe the current job for identification purposes 127
(5-16) Input the loading and design thresholds 127
(5-17) Specify pavement cross-section and material type 127
(5-18) Input to output sections, which displacements, stresses
and strains were computed 127
(5-19) Input of boundary conditions 128
(5-20) Illustration of propagation cracks 129
(5-21) Illustration of stress and strain within pavement layers 130
XIV
Table No. Title Page No.
(2.1) Superpave Design Gyratory Compactive Effort 18
(2.2) Superpave Coarse Aggregate Angularity Requirements 21

(2.3) Superpave Fine Aggregate Angularity Requirements 22
(2.4) Superpave Clay Content Requirements 22
(2.5) Superpave Flat, Elongated Particle Requirements 23
(2-6) The control points and restricted zone for a 12.5 mm
Superpave mixture 25
(2.7) Superpave Mixtures 28
(2-8) Boussinesq's for a point load (after Ullidiz, 1998) 44
(3-1) Physical Properties of Asphalt Cement 50
(3-2) Physical Properties of Al-Taji Quarry Aggregate 51
(3-3) Physical Properties of Filler (Cement) 52
(3-4) Job Mix Formula's for Wearing Course of the
Selected Sections 52
(3-5) Properties of Carbon Fiber 54
(3-6) Chemical Composition and Physical Properties of
Hydrated Lime 54
(4-1) Volumetric Properties for Marshall Specimens 74
(4-2) Volumetric Properties for Superpave Specimens 75
(4-3) Stability and Flow values for each Mix design 101
(4-4) Asphalt Film Thickness for both Mixes 107
(5-1) Input Data to MICHPAVE Program for Wearing Course 113
(5-2) Input Data of layers of pavement for MICHPAVE Program 113
(5-3) Output Results of the PCPT Program 117
(C-1) Data Analysis for Marshall Specimens and Superpave Specimens C1
(C-2) Marshall Test for Mix Design C2
VIII
(C-3) Asphalt Film Thickness for Marshall Mixes C3
(C-4) Asphalt Film Thickness for Superpave Mixes C4
(C-5) Data analysis for Superpave specimens at initial asphalt content for
R1gradation C5
(C-6) Data analysis for Superpave specimens at initial asphalt content for
R9gradation C6
(C-7) Data Analysis for Superpave Specimens at ±0.5% and +1% for
{(4.532) estimated asphalt content}(R1 (TRZ) gradation) C7
(C-8) Data Analysis for Superpave Specimens at ±0.5% and +1% for
{(4.59) estimated asphalt content}(R9 (ARZ) gradation) C7
(D-1) Creep test for Marshall Specimens with Optimum of Marshall for
(R9) ARZ Gradation D1
(D-2) Creep Test For Marshall Specimens With Optimum Of Marshall
for (R1) TRZ Gradation D2
(D-3) Creep Test For Marshall Specimens With Optimum Of Superpave
for R9 Gradation (ARZ) D3
(D-4) Creep Test of Marshall Specimens with Optimum of Superpave for
(R1) TRZ Gradation D4
(D-5) Creep test for Superpave Specimens with optimum of Superpave
for (R1) TRZ Gradation D5
(D-6) Creep test for Superpave Specimens with optimum of Superpave
for (R9) ARZ Gradation D6
(D-7) Creep test for Superpave Specimens with Optimum of Marshall for
(R1) TRZ Gradation D7
(D-8) Creep Test For Superpave Specimens with Optimum Of Marshall
For R9 (ARZ) Gradation D8
IX
(E-1) Indirect Tensile Strength (Kpa) for Marshall Specimens with
Optimum of Marshall E1
(E-2) Indirect Tensile Strength (Kpa) for Marshall Specimens with
Optimum of Superpave E2
(E-3) Indirect Tensile Strength (Kpa) for Superpave Specimens with
Optimum of Superpave E3
(E-4) Indirect Tensile Strength (Kpa) for Superpave Specimens with
Optimum of Marshall E4
(E-5) Temperature Susceptibility for Marshall Specimens E5
(E-6) Temperature Susceptibility for Superpave Specimens E6
(F-1) Lottman test for Marshall Specimens with
Optimum of Marshall F1
(F-2) Lottman test for Marshall Specimens with
Optimum of Superpave F2
(F-3) Lottman test for Superpave Specimens with
Optimum of Superpave F3
(F-4) Lottman Test for Superpave Specimens with
Optimum of Marshall F4
X
ARZ = Above Restricted Zone.
TRZ = Through Restricted Zone.
AASHTO =American Association of State Highway
and Transportation Officials .
AC = Asphalt Content .
ASTM = American Society for Testing and Materials .
AV = Air Voids .
D(40-50) = Dourah grade (40-50) penetration.
D/B = Dust Proportion ( Dust / Binder content ratio ).
FHWA = Federal Highway Administration.
HMA = Hot Mix Asphalt .
ITS = Indirect Tensile Strength .
JMF = Job Mix Formula.
Max. Theo. Sp. Gr. = Maximum Theoretical Specific Gravity.
MICHPAVE = Michigan–Pavement Software
PG = Performance Grade.
PCPT = Prediction of Critical Pavement Temperature .
SCRB = State Commission of Roads and Bridges.
SHRP = Strategic Highway Research Program .
SGC = Superpave Gyratory Compactor.
TRB = Transportation Research Board.
TSR = Tensile Strength Ratio .
TS = Tensile Susceptibility.
VTM = Voids in Total Mix.
VFA = Voids Filled with Asphalt .
VMA = Voids in Mineral Aggregate.
VII
Chapter One I ntroduction
Most of the hot mix asphalt (HMA) produced during the 50 years between the 1940
and mid 1990 were designed using the Marshall methods, and the increase in traffic
volumes and heavier loads became initiative for the Strategic Highway Research
Program (SHRP) in 1988. After five years of efforts, a new mix design, Superior
Performing Asphalt Pavements (Superpave), was developed. Superpave takes into
consideration the factors responsible for the typical distress on asphalt pavements,
rutting, fatigue, and thermal cracking. With the introduction of Superpave Mix
Design, the Marshall method of Mix design has become obsolete in highway
pavement, (Vasavi, 2002).
The Superpave technology was developed in the United States with proven success.
Superpave mixes have been widely used by developed countries over the last few
years. Superpave technology is replacing the Marshall method, which was used for
asphalt concrete mixture design for almost half a century. The Marshall method
was based mostly on experience and statistical analysis. The flexible pavement
sections designed using the Marshall method have had mixed success due to poor
understanding of mechanism of failure. The partial success has been mainly due to
very thick and uneconomical sections. The roads in Iraq are in a highly distressed
condition with pavement life much shorter than the expected. A new design
methodology, that is more thorough and comprehensive, is required. Superpave
technology can be rigorously tested under varying traffic and environmental
conditions.
1
This technology has a tremendous potential to be implemented in Iraq, which will

pay for itself with higher performance and longer lasting roads. Hence, there is
need to have a comprehensive study comparing the design of bituminous mixes
using both Superpave and the Marshall method of Mix Design.
Roads in Iraq are performing poorly with pavement life much shorter than the
expected. The high traffic intensity in terms of commercial vehicles, the serious
overloading of trucks and significant variation in daily and seasonal temperature of
the pavement have been responsible for early development of distress like rutting,
fatigue and thermal cracking on bituminous surfacing. One of the advantages of the
Marshall Mix Design method is that the performance of the mixes can be expected
for local materials and environmental impact.
Superpave technology as a new design methodology can be rigorously used under

varying traffic and environmental conditions. Although Superpave is recognized as
a significant system in the evaluation of asphalt concrete mixes, Iraqi agencies
continue to use Marshall Method as a unique mix design method in road projects.
Accordingly, an investigation is needed to compare analyze and investigate the
performance and the properties of Superpave and Marshall Mix Design methods.
There is international concern and interest in implementing Superpave in roads and
airport projects to investigate its impact on economic and performance of these
projects.
2
The main objective of the study is the comparison between traditional Marshall
design method and the Superpave system design method in the wearing course
mixes in flexible pavements. This process will be carried out by evaluating the
volumetric, mechanical properties and moisture susceptibility. Furthermore, the
resistance to the main distresses in flexible pavements (permanent deformation,
fatigue and thermal cracking) will be examined for these two types of mixes by
using MICHPAVE and PCPT software respectively.
This thesis is organized in six chapters and appendices. After the first chapter
(introduction), in chapter two, a summary of literature review and background into
Marshall and Superpave methods are outlined with information from related
studies. In chapter three, the research methodology and procedures for preparing
and testing specimens are presented. Chapter four deals with the results of the
experimental work and the analysis of the results. Chapter five studies the effect of
mix design method on pavement performance by using non-linear finite element
(MICHPAVE Software) and PCPT software to evaluate the resistance to thermal
cracking. Chapter six presents the conclusions and recommendations. Finally,
seven appendices are provided with this study, which are:
Appendix A presents a detailed information about data analysis for Marshall

Mixes design.
Appendix B presents a detailed information about data analysis for Superpave

mixes design.
Appendix C gives the Marshall test results and Superpave mixes analysis.
Appendix D contains the Creep test results for mixes design.
3
Appendix E presents the indirect tensile strength results for mixes design.
Appendix F presents the moisture damage results for mixes design.
Appendix G presents the results of MICHPAVE program application.
In this study, Superpave and Marshall Mixes will be designed for heavy traffic
level. Marshall and Superpave mixes were designed by considering locally
available materials and environmental impact. The experimental design used in this
study provides a comparison between these two types of mixes. The work was
limited to two types of gradation and one source of aggregate and asphalt cement.
One nominal maximum size aggregate (12.5 mm) was used in these mixes. The
work is limited to one traffic level and just laboratory testing, and field evaluation
could not be performed.
4
Chapter Two L iterature Review
In 600 B.C., the first asphalt road was paved in Babylon. During the last two
decades ,the amount of vehicle miles traveled per year and the amount of equivalent
single axle loads, ESAL, have increased by 75 and 60 percent respectively .As a
result, hot mix asphalt HMA , pavements have struggled to perform the intended
design life ,presenting rutting, fatigue and thermal cracking problems. This has
created a need to develop an enhanced hot mix asphalt concrete design procedure
(Amirkhanian, 2001).
For approximately the past 50 years , engineers have designed asphalt mixtures using
the Marshall or Hveem mix design methods , and over this period , different highway
agencies have modified these design procedures to better fit their particular needs.
Both methods have proven to be satisfactorily effective in aiding the design of
highways and interstates, but some problems exist.
The primary problem is that both the Marshall and Hveem design methods are
empirical – they do not produce samples that share the properties or performance of
the finished product. This makes it difficult to accurately predict how a particular
mix will perform in the field. (Khaled and Jason, 1998).
In U.S, SHRP was initiated in 1987 as a five – year, $50 million program designed
primarily to improve the performance and safety of roads in the United States. The
Superpave (Superior performing Asphalt Pavement) mix design method, is a product
of SHRP, and is being used for implementation into all American states.
Since its development, this mix design procedure has replaced the traditional
Marshall and Hveem methods in many applications.
٥
Superpave is more comprehensive and accurate system that it takes into account all
phases of mix design and performance including specification of asphalt binders and
mineral aggregates. Concurrent with the development of the Superpave mix design
method, Performance Grade asphalt binder specifications were introduced. The
Performance Grade specifications are more comprehensive than the previous asphalt
cement specifications. In addition, the asphalt is tested for performance with respect
to these distress types rutting,฀ fatigue and thermal Cracking. Additionally, the
Superpave specifications for aggregate are more stringent than aggregate requirement
in the Marshall method, particularly for mix designs where the twenty-year
cumulative equivalent single axle loads, ESAL, exceed 300,000 applications. These
requirements ensure sufficient interparticle friction to provide a stable asphalt
concrete mix .In most Superpave applications a “coarse” aggregate blended gradation
is selected to increase the rutting resistance in the mix. However, fine aggregate
blend gradations are generally, preferred for low volume roads to provide better
durability.
The States currently use Superpave for National Highways System projects. Based
on the success of these projects, the States are considering implementing Superpave
in all projects. Superpave covers five mix types based on the nominal maximum
aggregate size: 9.5, 12.5,19,25 and 37. The Superpave mix design method differs
from Marshall and Hveem mix design methods in using performance – based and
performance – related criteria to design the proper asphalt mix . This allows a direct
relationship to be drawn between the lab and field performance of the asphalt mix
(John P.Z., David D., 2004).
This review focuses on the Marshall and Superpave methods since they are
Currently used.
٦
Bruce Marshall, formerly the Bituminous Engineer with the Mississippi State
Highway Department, developed the original concept of the Marshall Method of
designing asphalt pavements. The present form of Marshall Mix design method
originated from an investigation started by the U.S Army Corps of Engineers in
1943. The purpose of Marshall method is to determine the optimum asphalt content
for a particular blend of aggregates and traffic level .The optimum asphalt content is
determined by the ability of a mix to satisfy stability ,flow ,and volumetric
properties,( Vasavi K. , 2002).
A) Asphalt cement
Before a good asphalt mix can be designed by using Marshall method, designers
must select the proper asphalt cement grade and determine its properties. They decide
on a proper asphalt cement grade by examining the type of asphalt mix being
designed and the geographical location of its use. After the asphalt cement is
selected, designers may determine its viscosity and whether the asphalt meets
specifications of flash point, penetration, ductility, and solubility. Once they
conclude that asphalt cement is acceptable, they find its specific gravity and create a
temperature – viscosity plot to determine its appropriate mixing and compaction
temperatures.
B) Aggregate
For the requirement of successful mix, the appropriate aggregate also must be
selected. When designers accept a particular aggregate, they test its gradation,
specific gravity, and absorption. Then, they determine the aggregate gradation to be
used in a mix.
٧
C) Filler
Asphalt mixtures have an optimum cohesion where maximum compaction will
occur. This cohesion can be affected by the amount of filler used in a mix. Santucci
and Schmidt, 1962, showed that if the binder volume (asphalt +filler) is held
constant, there is an optimum filler percentage where maximum compaction can
occur. A study by Heukelom, 1965 also showed that the amount of filler used in a
mix could influence how well a mix is compacted. For a given filler type, the ease of
compaction increases with the percentage of filler in the overall binder content. One
type of mineral filler (cement) has been used in this study.
D) Additives
Two types of additives (carbon fiber and lime) have been used in this study. Little
published information concerning carbon fiber modified asphalt is available. Most
studies that included fibers in asphalt mixtures have a limited number of trials. In this
study, the effects of carbon fiber modification are investigated. It is important to
understand other researches, which have utilized carbon fiber.
It is thought that the addition of carbon fiber to asphalt enhances material strength
and fatigue characteristics while adding ductility. Because of their inherent
compatibility with asphalt cement and excellent mechanical properties, carbon fibers
might offer an excellent potential for asphalt modification. With the new
developments in production, a carbon modified asphalt binder has become cost
competitive with polymer-modified binder. Based on results from other fiber-
modified composites, it was thought that the incorporation of carbon fibers into an
AC mixture would enhance its tensile strength properties, resulting in a decrease in
cracking due to cold temperatures and repeated loading at intermediate temperatures,
while stiffening the mixture at high temperatures, increasing its resistance to
permanent deformation. Modification of the asphalt binder is one approach taken to
improve pavement performance (M.Aren C. , 2000).
٨
The use of carbon fibers as a reinforcement in hot mix asphalt (HMA) has not been
the subject of much research. The majority of the present research has been
conducted on the addition of carbon fibers to concrete mixes.
In 1996, Serfass and Samanos investigated fiber-modified asphalts using Chrysotile,
rock wool, glass wool, and cellulose fibers. These modified asphalts were subjected
to a wide variety of tests on mastics (bitumen and fibers), mortars (bitumen, fibers,
and sand), and asphalt concrete. Common characteristic of all tested asphalts
includes resistance to thermal cracking, ageing, shearing, and aggregate dislodgment.
They concluded in their studies that the addition of fibers to asphalt concrete
improved the fixation of the asphalt binder in the mix. This relates to less bleeding
and improved skid resistance over unmodified mixtures of the same design. Fiber
modification also allowed for an increase in film thickness, resulting in less aging
and improved binder characteristics. The addition of fibers also resulted in the
reduction of temperature susceptibility of asphalt mixtures. Adding fibers enables
developing mixtures rich in bitumen [asphalt binder], and therefore displaying high
resistance to moisture, aging, fatigue, and cracking, (Serfass and Samanos 1996).
A two-phase study by (Aren C., 2000), investigated the behavior of carbon fiber
modified asphalt mixtures. The first phase focus on determining the feasibility of
achieving improvements in mechanical behavior with the addition of carbon fibers.
The second phase focus on investigating the factors that contribute to the new
behavior. Carbon fibers were found to create improvements in high temperature and
low temperature behavior. HMA samples containing 0.5%to 0.8% weight carbon
fiber in the asphalt cement binder showed a respective improvement in resistance to
repeated load deformation ranging of 38%to 82%.Fiber length taken after a pug mill
field trial by the research sponsor revealed a reduction in average final carbon fiber
length from 2.54 cm to between 0.2mm and 0.65mm. Potential problems identified
by this study were final fiber length, even distribution of fibers, and initial asphalt
quality (Cleven, 2000).
٩
Aren’s study formed the foundation for this current study due to the need to improve
carbon fiber length in the final asphalt cement by protecting it during mixing. An
estimate was made of a final length of 6mm for improved mechanical properties,
enough to arrest microcracks and reduce creep.
The second additive (Lime), hydrated lime has long been recognized as a highly
beneficial component of hot mix asphalt, based initially on its ability to reduce
stripping. More research and experience have demonstrated that lime’s benefits are
much broader, and include:
 Increased mix stiffness and reduced rutting.
 Reduced oxidation and age-hardening effects.
 Improved low-temperature cracking resistance (Dr Dallas Little, Jon Epps.,
2001).
Extensive research at the Western Research Institute (WRI) showed that age
hardening of asphalt can be reduced by the addition of hydrated lime (Petersen et
al.,(1987).
As little as one-half of one percent hydrated lime by dry weight is needed to achieve
a reduction in age hardening. This reduction in hardening has been confirmed in a
field study conducted by the Utah DOT (Jones, 1997).
Johannson, 1998 performed an extensive review of the literature of lime in bitumen
and conducted additional research on the reaction of hydrated lime with bitumen.
Some of Johannson’s most significant findings are:
1. Adding 20 percent hydrated lime by mass produces a significant increase in creep
stiffness but does not increase physical hardening. Furthermore, the lime-modified
bitumen demonstrates a greater potential for dissipating energy through deformation
(at low temperature) than the unmodified bitumen. This is a positive effect at low
temperatures because it reduces fracture potential.
2. Although the filler effect increases low temperature stiffness, fracture toughness is
substantially increased. Fracture toughness is the energy expended in fracturing a
١٠
material. Lesueur et al., (1998), also demonstrated that at low temperatures, lime
does not negatively affect relaxation but substantially increase fracture toughness.
3. Hydrated lime reduces the effects of age hardening more so at high temperatures
than at low temperatures.
Tarrer , 1996 investigated the bitumen-aggregate bond and concluded that, in the
field, the water at the surface of the aggregate has a high PH and therefore most
liquid antistrip agents remain at the surface because they are water soluble at high PH
levels. To overcome being washed away, the liquid antistripping agents must be
given time to cure (in excess of three hours).In contrast, hydrated lime cures rapidly
(within 15 to 30 minutes) and forms water insoluble compounds. Hydrated lime
creates a very strong bond between the bitumen and the aggregate, preventing
stripping at all PH levels. Tarrer also found that hydrated lime reacted with Silica
and alumina aggregates in a pozzolanic manner that added considerable strength to
the mixture.
It has been proved through laboratory and field-testing that hydrated lime in HMA
substantially reduces moisture sensitivity. Lime enhances the bitumen-aggregate
bond and improves the resistance of the bitumen itself to water-induced damage.
Recent surveys document the success and acceptance of lime in HMA (Dallas N.L.
and Jon A.E., 2001).
Aggregate gradation HMA is graded by the percentages of different – size aggregate

particles it contains. Aggregate gradation is a way of describing the proportions of
the various sizes of crushed stone, sand, and filler, by passing the aggregate through
a set of sieves and measuring the weight retained on each sieve. Aggregate gradation
is often plotted as a grading curve.
١١
For determining the design asphalt content for a particular blend of aggregates by the
Marshall method, a series of test specimens are required to include a range of asphalt
contents of at least 2.0%, at intervals not to exceed 0.5%.Three specimens are
required for each asphalt content used in the design. The standard method for
compacting the test specimens is to immediately compact them after mixing process
is completed.
A compaction effort of 75 blows per side is applied for heavy traffic levels and 50
blows per side for light traffic level.
At least one specimen is required at the estimated asphalt content to determine the
maximum specific gravity (AASHTO T209). These specimens are prepared at the
estimated asphalt content ratio for the mix.
The volumetric analysis focuses on five characteristics of the HMA and the influence
those characteristics are likely to have on HMA behavior. The Asphalt Institute
added these volumetric criteria to the method in 1973, the five characteristics are:
1) Mix Density;
2) Air Voids;
3) Voids in the Mineral Aggregate (VMA);
4) Voids Filled with Asphalt (VFA); and
5) Asphalt Content.
Before mix properties are discussed in detail, the engineer should understand the
paving mix properties and how the HMA will perform as a finished pavement.
 Mix density
The density of the compacted mix is its unit weight (the weight of a specific
volume of HMA). Density is important because proper density in the finished
product is essential for lasting pavement performance.
١٢
 Air voids
Air voids are small air spaces or pockets of air that occur between the coated
aggregate particles in the final compacted HMA. A certain percentage of air
voids is necessary in all dense-graded mixes to prevent the pavement from
flushing, shoving, and rutting.
 Voids in the mineral aggregate (VMA)
VMA are the void spaces that exist between the aggregate particles in the
compacted paving HMA, including the space filled with the binder.
 Voids filled with asphalt (VFA)
VFA are the void spaces that exist between the aggregate particles in the
compacted paving HMA that are filled with binder. VFA are expressed as a
percentage of the VMA that contains binder. Including the VFA requirement in a
mix, design helps prevent the design of HMA with marginally acceptable VMA.
The main effect of the VFA is to limit maximum levels of VMA and
subsequently maximum levels of binder content.
 Asphalt content
The proportion of binder in the HMA is critical. It must be accurately determined
in the laboratory, and then precisely controlled at the plant. The binder content
for a particular HMA is established by the mix design. The optimum binder
content of the HMA is highly dependent on aggregate characteristics such as
gradation and absorptiveness. Aggregate gradation is directly related to optimum
binder content.
Marshall Stability is defined as the maximum load carried by a compacted specimen

tested at 140° F (60° C) at a loading rate of 2 inches/minute.
The flow is measured at the same time as the Marshall stability. The flow is equal to
the vertical deformation of the sample (measured from start of loading to the point at
which stability begins to decrease) in hundredths of an inch.
١٣
The stability and flow measurements procedure (AASHTO T245) indicates that the
stability reading for a test specimen is only accurate if the test specimen measures
63.5mm in height. For test specimens that vary slightly from 63.5 mm, the stability
reading should be multiplied with a correlation ratio.
In this method the stability, flow, unit weight, air voids, VMA and VFA are plotted
versus the asphalt content. The optimum asphalt content of the mix is determined
from the data obtained from the plots. Consider that optimum asphalt content should
achieve the specification requirement from the volumetric properties point of view.
Superpave is an acronym for Superior Performing Asphalt Pavements. It is the

product of the Strategic Highway Research Program. Superpave includes a new
mixture design and analysis system based on performance characteristics of the
pavement. It is a multi-faceted system with a tiered approach to designing asphalt
mixtures based on desired performance. Superpave includes some old rules of thumb
and some new and mechanistic based features.
The Superpave mix design system is quickly becoming the standard system used in
the United States (US). The US was looking for a new system to overcome
pavement problems such as rutting and low temperature cracking that had become
common with the use of design systems such as Marshall and Hveem .The Superpave
system offers solutions to these problems through a rational approach. The
Superpave system builds from the simple to the complex. The extent to which the
designer utilizes the system is based on the traffic and climate for the pavement to be
built. The system includes an asphalt binder specification that uses new binder
physical property tests; a series of aggregate tests and specifications; a hot mix
asphalt (HMA)design and analysis system; and computer software to integrate the
١٤
system components. For low volume roads in moderate climates, a simple system
using materials selection and volumetric mix design is used. As the traffic level for
the road to be designed increases, the design requirements increase to improve
reliability. At the higher traffic levels, extensive performance testing is recommended
to a assure the highest reliability. A unique feature of the Superpave system is that its
tests are performed at temperatures and aging conditions that more realistically
represent those encountered by in-service pavements, (John A. D’Angelo, U.S. FHA).
The Superpave gyratory compactor (SGC), as shown in Figure (2-1) is used in

Superpave system to produce compacted specimens for volumetric analysis and
determination of mechanical properties. The equipment is capable of providing data
to indicate the trend of density variation throughout the compaction procedure. Large
aggregate can be accommodated, and compactability can be evaluated so that
potential tender mix behavior and similar compaction problems can be identified.
Finally, the equipment is portable and can be used in plant mix facilities as part of
quality control operations.
Figure (2-1) Superpave Gyratory Compactor
١٥
The SGC consists of the following components:

 Frame , rotating base , and motor
 Loading ram and pressure control
 Height measuring and recording system
 Mold and base plate
A loading system applies a load to the loading ram, which imparts A 600 KPa
compaction pressure to the specimen. A pressure gauge measures the ram loading to
maintain constant pressure during compaction. The SGC mold is cylindrical wall
(inside diameter of 150 mm) with a base plate at the bottom to provide confinement
during compaction. While the mold is positioned at a compaction angle of 1.25 ْ ◌,
Figure (2-2) shows the mold configuration during the compaction process.
Figure (2-2) Mold Configuration

Specimen height measurement is an important function of the SGC. By knowing the
mass of the material placed in the mold and the specimen height, an estimate of
specimen density can be made at any time throughout the compaction process.
Specimen density is computed by dividing the mass by the volume of the specimen.
Height is measured by recording the position of the ram throughout the test. By this
method, a compaction characteristic is developed as the specimen is compacted.
١٦
Figure (2-3) shows a densification plot of an asphalt mixture with increasing number
of gyrations .Three gyration levels, specified by the Superpave volumetric mixture
design procedure are of interest:
 Design number of gyration ( N DESIGN )
 Initial number of gyration ( N initial )
 Maximum number of gyrations ( N maximum)
Figure (2-3) Densification plot

In Superpave, asphalt mixtures are designed at a specified level of compactive effort,
identified by N design. As a function of the traffic level, N design is used to vary the
compactive effort of the design mixture. Traffic is represented by the design
equivalent single axle loads (ESALs). Currently, N design ranges from 68 to 172.
However, recent research suggests that the range and the number of gyration may
need to be modified depending on traffic. The test specimens are compacted to the
maximum level using N maximum gyrations. At N maximum, the density is not
allowed to exceed 98 percent of maximum theoretical density. Specifying this
maximum density requirement at N maximum prevents the design of mixture that is
susceptible to excessive compaction under the design traffic. Such a mix is prone to
١٧
excessive rutting. N maximum is calculated using N design in the following

relationship:
Log N max = 1.10 Log (N des)
The compatibility of the mixture is estimated at N initial. The density is not to exceed
89 percent of Gmm at N initial. Specifying this maximum density requirement at N
initial prevents the design of a mixture that has a weak aggregate structure and low
internal friction, which are sometimes indicators of a tender mix. N initial is
calculated using N design through the following relationship:
Log N initial = 0.45 Log (N design)
Currently, the values of N maximum range from 104 to 288 and the values of N
initial range from 7 to 10, (Mansour S. et al, 1999).
Table 2-1 Superpave Design Gyratory Compactive Effort
Design Average Design High Air Temperature

ESALs ( 43-44ºC)
( millions) Nini Ndes Nmax
<0.3 7 82 127
0.3-1 8 93 146
1-3 8 105 167
3-10 9 119 192
10-30 9 135 220
30-100 10 153 253
>100 10 172 288
١٨
The Superpave mix design procedure involves careful material selection and
volumetric proportioning as a first approach in producing a mix that will perform
successfully. The four basic steps of Superpave asphalt mix design are materials
selection, selection of the design aggregate structure, selection of the design asphalt
binder content and evaluation of the mixture for moisture sensitivity.
Figure (2-4) Superpave Mix Design
1. Selection of Binder
The design process starts with material selection. The key aspect to the performance
of any asphalt mixture is the selection of the optimum materials that will be used in
the mixture. One of the key components of Superpave is materials selection. The
binders are selected using the performance based binder specification and the
١٩
aggregates are selected using performance related aggregate criteria. The asphalt
binder will affect various performance aspects of the asphalt mixture such as
permanent deformation, fatigue cracking and low temperature cracking. The
Superpave binder specification is intended to select the binder to optimize its effect
on the performance of the pavement. The binder is selected based on the climate of
the pavement where it will be used, the expected traffic and the location in the
pavement structure. The binders are evaluated at the expected highest pavement
temperature and the lowest pavement temperatures. The average 7-day high
temperature is used to determine the critical maximum pavement temperature at a
depth of 20 mm in the pavement. Using pavement temperatures to select the binder
allows for the selection of a binder that will meet both high and low temperature
needs for the pavement being placed,( John A .D'Angelo).
2. Selection of Aggregates
The next step in the Superpave design process is the selection of the aggregate to be
used in the mix. Aggregates are the major components of hot mix asphalt. The
quality of the aggregates is critical to the performance of the asphalt mixes.
Aggregates make up 80 to 85% of the mixture by volume. Aggregate characteristics
are a major factor in the performance of an asphalt mixture. In the Superpave mixture
design system, many aggregate criteria are included to assure the performance of the
asphalt mix. These criteria included consensus aggregate properties and source
aggregate properties. Consensus properties include coarse aggregate angularity,
uncompacted voids in fine aggregate or fine aggregate angularity (FAA), flat and
elongated particles and clay content. Source properties include toughness, soundness,
deleterious materials, and gradation parameters. The recommended limits set by
SHRP on these aggregate criteria were established by consensus of a group of
experts based on years of previous research and experience by the Modified Delphi
group (Cominsky, 1994).
٢٠
Coarse aggregate angularity is defined as the percent by weight of aggregates larger

than 4.75 millimeters with one or more fractured faces. Table (2-2) presents the
Superpave coarse aggregate angularity requirements.
Table (2-2) Superpave Coarse Aggregate Angularity Requirements
Superpave Coarse Aggregate Angularity Requirements
Percent minimum
Traffic , million ESALs Depth from surface
< 100 mm > 100 mm
<0.3 55/- -/-
0.3 to < 3 75/- 50/-
3 to < 10 85/80 60/-
10 to < 30 95/90 80/75
≥ 30 100/100 100/100
Note: "85/80" means that 85% of the coarse aggregate has one
fractured face and 80% has two fractured faces.
Fine aggregate angularity is defined as the percent of air voids present in loosely
compacted aggregates smaller than 2.36 millimeter. Table (2-3) presents the
Superpave Fine aggregate angularity requirements.
٢١
Table (2-3) Superpave Fine Aggregate Angularity Requirements
Superpave Fine Aggregate Angularity Requirements
Percent minimum
Traffic , million ESALs Depth from surface
< 100 mm > 100 mm
<0.3 - -
0.3 to < 3 40 40
3 to < 10 45 40
10 to < 30 45 40
≥ 30 45 45
Note: Criteria are presented as percent air voids in loosely

compacted fine aggregate.
Clay content is the percentage of clay material contained in the aggregate fraction
that is finer than a 4.75 mm sieve. Table (2-4) shows the Superpave clay content
requirements.
Table (2-4) Superpave Clay Content Requirements
Superpave Clay Content Requirements
Traffic , million ESALs Sand equivalent ,

minimum
<0.3 40
0.3 to < 3 40
3 to < 10 45
10 to < 30 45
≥ 30 50
٢٢
Flat and elongated aggregate can also affect performance. This characteristic is the
percentage by weight of coarse aggregates that have a specified maximum to
minimum dimension ratio. Table (2-5) shows the Superpave flat elongated particle
requirements.
Table (2-5) Superpave Flat, Elongated Particle Requirements
Superpave Flat Elongated Particle Requirements
Traffic , million ESALs Percent , maximum
<0.3 -
0.3 to < 3 10
3 to < 10 10
10 to < 30 10
≥ 30 10
Note: Criteria are presented as maximum percent by weight of flat

and elongated particles.
٢٣
Chapter Two Literature Review
To specify gradation, Superpave uses a modification of an approach already used by

some U.S agencies. It uses the 0.45 power gradation chart to define a permissible
gradation as presented in Figure (2-5).
Figure (2-5) 0.45 power gradation

Nijboer L. W. 1948, presents the basis concept of the 0.45 power gradation chart.
Nijboer employs a double logarithmic gradation chart in order to study the influence
of aggregate gradation on mineral voids (VMA). He uses both round gravel and an
angular coarse aggregate to show that for a gradation having a slope of 0.45, the
aggregates exhibit minimum voids in VMA. An important feature of the 0.45 power
chart is the maximum density gradation. This gradation is plotted as a straight line
from the maximum aggregate size through the origin. Superpave uses a standard set
of ASTM sieves and the following definitions with respect to aggregate size:
• Maximum Size: One sieve size larger than the nominal maximum size.
• Nominal Maximum Size: One sieve size larger than the first sieve to retain more
than 10 percent.
٢٤
The maximum density gradation represents a gradation in which the aggregate

particles fit together in their densest possible arrangement. Clearly, this gradation is
to be avoided because there would be very little aggregate space within which to
develop sufficiently thick asphalt films for a durable mixture. Figure (2-5) shows a
0.45 power gradation chart with a maximum density gradation for a 19 mm
maximum aggregate size and 12.5 mm nominal maximum size. To specify aggregate
gradation, two additional features are added to the 0.45 power chart: control points
and a restricted zone. Control points function as master ranges through which
gradations must pass. They are placed on the nominal maximum size, an intermediate
size (2.36 mm), and the dust size (0.075 mm). Table (2-6) illustrates the control
points and restricted zone for a 12.5 mm Superpave mixture.
Table (2-6): The control points and restricted zone for a 12.5 mm Superpave mixture.
12.5 Nominal Size
Control Points Restricted Zone Boundary

Sieve , mm Minimum Maximum Minimum Maximum
19 100
12.5 90 100
9.5 90
4.75
2.36 28 58 39.1 39.1
1.18 25.6 31.6
0.6 19.1 23.1
0.3 15.5 15.5
0.15
0.075 2 10
٢٥
The restricted zone resides along the maximum density gradation between the
intermediate size (either 4.75 or 2.36 mm) and the 0.3mm size. It forms a band
through which gradations should not pass. The Superpave aggregate gradation is
illustrated in Figure (2-6).
Figure (2-6) Superpave Aggregate Gradation

The concept of a restricted zone around the maximum density line near the 0.6-mm
sieve size can probably be indirectly traced back to (Goode and Lufsey, 1962). Based
on the work by Nijboer, Goode and Lufsey present a 0.45 power grading chart for
plotting aggregate gradations. This grading chart uses the sieve size (in microns)
raised to the 0.45 power as the horizontal axis and the percent passing (by mass) in
arithmetical scale as the vertical axis.
They also presented their method for identifying the maximum density line. Their
interpretation consists of drawing a straight line from the origin of the chart to the
percentage point plotted for the largest sieve with material retained. They use the
term “effective aggregate size.”
To use the newly developed gradation chart, they evaluated 24 gradations to observe
the effect of sand content on the stability of mixtures. The effect of sand content was
٢٦
evaluated by producing gradations with different “hump” characteristics. These

included gradations from three parts of the country that had exhibited the tenderness
problem.
Two of the gradations showed humps beginning below their maximum density line,
extending above the maximum density line at the 1.18-mm sieve, and falling back
below the maximum density line at the 0.15-mm sieve. The other gradation started
below the maximum density line, crossed over at the 0.25-mm sieve, and showed the
hump between the 1.18-mm and 0.15-mm sieves. They use three different types of
mixture in their study: low in total sand, medium in total sand, and high in total sand.
A gravel coarse aggregate was used that has its first material retained on the 12.5-
mm sieve. The Superpave defines maximum density line with Superpave restricted
zone superimposed on the figure. Goode and Lufsey found that, in general,
gradations that show appreciable humps above the maximum density line at about the
0.6-mm sieve produce higher VMA and lower Marshall stabilities than do gradations
plotted with lesser humps. They used rounded gravel, sand, and commercial
limestone filler as the aggregates for this study; hence, they recommend that
gradations avoid humps near the 0.6-mm sieve and do not pass above the maximum
density line. This recommendation can be viewed as the beginning of the concept of
restricted zone. They used uncrushed coarse and fine aggregate. Their conclusions
may not be valid for aggregate gradations using crushed materials. Gradations that
pass through the restricted zone have often been called “humped gradations” because
of the characteristic hump in the grading curve that passes through the restricted
zone. In most cases, a humped gradation indicates a mixture that possesses too much
fine sand in relation to total sand. This gradation practically always results in tender
mix behavior, which is manifested by a mixture that is difficult to compact during
construction and offers reduced resistance to permanent deformation during its
performance life.
٢٧
Gradations that violate the restricted zone may possess weak aggregate skeletons that
depend too much on asphalt binder stiffness to achieve mixture shear strength. These
mixtures are also very sensitive to asphalt content and can easily become plastic. The
term used to describe the cumulative frequency distribution of aggregate particle
sizes is the design aggregate structure. A design aggregate structure that lies between
the control points and avoids the restricted zone meets the requirements of Superpave
with respect to gradation.
Superpave defines five mixture types by their nominal maximum aggregate size as
presented in Table (2-7).
Table (2-7) Superpave Mixtures
Superpave Mixtures
Superpave mixture Nominal maximum Maximum

designation size , mm size , mm
37.5 mm 37.5 50
25 mm 25 37.5
19 mm 19 25
12.5 mm 12.5 19
9.5 mm 9.5 12.5
According to the Superpave recommendations, the gradation line should pass

between the control points and avoid crossing the restricted zone. Studies have been
conducted to evaluate different aggregate gradations and to compare between
conventional method of mix design and Superpave method.
Rommel N. Y.2004, studied the influence of avoiding the Superpave restricted zone
on the asphalt concrete performance. He studied the gradations for asphalt concrete
wearing, leveling and base courses. Selected gradation, through restricted zone,
boundaries of restricted zone, below and above boundaries are going to be
٢٨
investigated. Based on the results of this study: it has been concluded that the
gradation passing below restricted zone produces lower temperature susceptibility for
paving mixture. The selected gradation above restricted zone shows less
susceptibility of mixture to moisture damage .Gradation passing through the
restricted zone shows more resistance to plastic flow for asphalt paving materials.
Kandhal et al, 1998 studied two asphalt concrete mixtures with nominal maximum
sizes of 12.5 mm and 19 mm. The aggregate gradations were used with these two
mixes above, through, and below the restricted zone. The VMA for mixtures with
various aggregate gradations are calculated based on the asphalt film thickness. The
results show that the aggregate gradation changes the minimum desirable VMA.
Gradations below the restricted zone had the lowest VMA in both mixes with 12.5
mm and 19 mm aggregate nominal maximum sizes. The study recommends that a
minimum average asphalt film thickness should be used to ensure mix durability
instead of a minimum VMA.
Collins et al .1997 studied the effect of aggregate degradation on the design
gradation and final volumetric properties of the asphalt mix compacted by the
Superpave gyratory compactor (SGC) and Astec vibratory compactor. Aggregate
with high and low Angeles of abrasion resistance on gradation change and
volumetric properties.
The change in the 0.075 mm materials caused by aggregate degradation during
compaction was significant to prevent the specimens from meeting the dust
proportion requirements. It was concluded that when high abrasion loss aggregate is
used, the gradation should be designed to stay sufficiently below the restricted zone
to compensate for the effect of degradation during compaction.
Stephen A. et al, 1999, studied the effects of gradation on the performance of asphalt
mixtures. Based on the results of this study and considering the materials tested, the
following conclusions are warranted.
1. The fine gradation is stronger than the coarse gradation as measured by
indirect tensile strength and permanent deformation.
٢٩
2. The fine gradation is more durable than the coarse gradation as measured by
fatigue life.
3. The fine gradation is more resistant to moisture-induced damage as measured
by the wet rutting test and air permeability.
4. The fine mixture is less sensitive to a coarsening in gradation as might be
experienced by segregation, at the same level of segregation, than the coarse
mixture.
Prithvis S. and L Allen., 2002 studied the coarser versus fine graded Superpave
mixtures and their resistance to rutting. Based upon the results of this study, mix
design should not be limited to Superpave mixes on the coarse or fine side of the
restricted zone. Mixes having either gradation type can perform well. Therefore, it is
recommended that gradation specification utilizes both coarse and fine graded
mixes.
In summary, the literature shows that some researches have been conducted to
evaluate the effect of aggregate gradation on the properties of Superpave mixtures. It
can be also seen that research is still needed in order to obtain better understanding of
the Superpave mixture to be conducted in the next few decades in that area until the
Superpave practice becomes more established and accepted by various highway
agencies.
The design asphalt content is defined by Superpave as the asphalt content that
produces 4% air voids at Nd and meets all other criteria. A good estimate of design
asphalt content is established from the design aggregate blend. Two specimens are
prepared at an estimate asphalt content, the specimens are compacted to Nd gyrations
and volumetric analysis is performed. Then, asphalt content corresponding to 4% air
void is determined with other volumetric properties. The design mixtures must meet
the requirement of VTM, VMA, and VFA. Two specimens are prepared with four
levels of asphalt content PEST %, Pest ±0.5%, and Pest +1% the samples are compacted
to Nd gyrations, a volumetric analysis is performed and the results are plotted.
٣٠
Moisture damage has been a significant problem that results in premature pavement
failure. Environmental factors such as temperature and moisture can have a profound
effect on the durability of hot mix asphalt pavements. When critical environmental
conditions are coupled with poor materials and traffic, premature failure may result
because of stripping of the asphalt binder from the aggregate particles.
A significant amount of research effort has been directed at this problem in the past
and more will be anticipated in the future. Numerous test methods, both qualitative
and quantitative, have been developed and used in the past to assess the moisture
susceptibility of HMA mixes, (Randy C.W. et al., 2004)
The adhesion between the asphalt and aggregate is an important, yet complex and not
well understood, property that helps to ensure good pavement performance. The loss
of bond, or stripping, caused by the presence of moisture between the asphalt,
aggregate is a problem in some areas, and can be severe in some cases. Many factors
such as aggregate characteristics, asphalt characteristics, environment, traffic,
construction practices and drainage can contribute to stripping.
The moisture susceptibility test (AASHTO – T283) is used to evaluate HMA for
stripping.” This test is not a performance based test but serves two purposes. First, it
identifies whether a combination of asphalt binder and aggregate is moisture
susceptible. Second, it measures the effectiveness of anti-stripping additives.
Chad W. H., 2004, conducted a study to evaluate the use of gyratory compacted
asphalt specimen for tensile strength ratio (TSR), and to compare the test result
obtained from the smaller diameter Marshall specimens with the larger diameter
gyratory specimens for the TSR. Two performance grades of liquid asphalt binder
(PG 64-22 and PG 76-22) were used in the mixes preparation .Based on the study
results; the gyratory specimens have higher average wet and dry strengths. The
higher average wet and dry strengths cause the TSR values to increase, but they
remain below the minimum 85%except for one mixture with limestone as the
aggregate. The percent visual stripping was fairly consistent from the Marshall to the
٣١
gyratory specimens. The mixtures that had an antis tripping agent had very low
visual stripping percentages. The mixtures that did not have an anti-stripping agent
had high visual stripping percentages except for one mixture. This mixture was
composed of limestone and did not show much visual stripping because of its
aggregate classification. The specified minimum wet strength is 65 psi for all indirect
tensile strength tests. The gyratory mixes without lime had higher wet indirect tensile
strengths than 65 psi. The average wet tensile strength for the gyratory mixtures
without lime was 93 psi. The average wet tensile strength for the same gyratory
mixtures with lime was 129 psi. Some wet tensile strengths for the gyratory mixtures
without lime were as high as 109.2 psi while others were as low as 78.2 psi. Some
wet tensile strengths for the gyratory mixtures with lime were as high as 229.8 psi
while others were as low as 104.7 psi.
٣٢
Since then, several mixture design methods have been developed to improve the
quality of asphalt concrete mixtures. These include methods such as Hveem,
Marshall, and Superpave (Roberts, 1996). The Marshall method was developed
during 1930's and has undergone several refinements over the years. The Superpave
method was developed by the SHRP in the late 1980 .In Iraq and most of the Middle
East countries , the Marshall mix design procedure is used in the majority of the
pavement design requirement. However, most of the modern countries use the
Superpave method for high traffic volume roads.
Asphalt concrete specifications include the control on the materials (asphalt, and
aggregate), and the control on the mixture. The Marshall method uses the penetration
grade for asphalt while the Superpave uses the PG specification for mix design
method. The specification and selection rules are independent of the mix design
method. One of the primary differences between the Marshall and Superpave
method is the aggregate specification .The aggregate specification and mix design
specification do vary between the two design methods . Aggregate specifications
vary in both gradation requirements and aggregate evaluation tests (John P.Z and
Jason N.2003).
The Marshall specifications don't differentiate between source properties, however
Superpave requires consensus properties that are not considered in Marshall Method.
Superpave requires avoiding the restricted zone and it should does not exceed the
control points that are not considered in Marshall Mix design. The mix design criteria
for both the Marshall and Superpave methods include VMA, VFA, VTM, and D/B.
The Marshall requirements for VTM show a range of 3-5% and the Superpave
requirements are exactly 4%. The Marshall method has requirement for stability and
flow that are not considered in Superpave. The Superpave method has requirement
for the initial and maximum compaction densities that are not considered in the
Marshall method. The Marshall and Superpave mix design methods have different
٣٣
criteria depending on the level of traffic anticipated on project .The Superpave mix
design method accommodates more traffic than the Marshall method . In addition,
Superpave requires evaluation of the mix for the moisture susceptibility, which is not
a requirement for the Marshall method.
The literature review identifies studies that compare between the Marshall and
Superpave methods. It is important to note that Superpave compaction requirements
have been recently modified and reduced to the current four levels (Brown and
Buchanan, 2001).
A study was conducted at the University of Wyoming by Dr. Khaled ksaibati. and
Jason Stephen ,1998, in which the researchers evaluated the performance of asphalt
mixes prepared using the Marshall mix design method and the Superpave level one
mix design method . The Georgia loaded wheel tester and the thermal stress
restrained specimen tester were used to test the rut-resistance and low-temperature
cracking of asphalt mixes. This study use of aggregate gradation for Superpave is
close to the gradation used for the Marshall Mix design, but the 0.45 power gradation
plot of aggregate used for Marshall Mix design cross into the restricted zone
established by Superpave. It was found that the optimum asphalt contents determined
by the Marshall and Superpave mix design are similar. This shows that in some cases
Marshall and Superpave produce nearly identical mix designs when the same
materials are used and the aggregate gradation are similar in both designs. The
Superpave was tested in the GLWT ruts slightly more than Marshall Samples, though
both mix designs produce samples that do not come close to failure at rut depth of
more than 7.62 mm after 8000 cycles. The Superpave samples tested in the TSRST
fractured at slightly higher pressure and lower temperature than the Marshall
samples.
N.Paul k. and Sachiyo k., 2000, conducted a study to make comparative evaluation of
design and performance of Marshall and Superpave mixes. In this study a detailed
laboratory investigation into the use of two mix design procedures was performed for
comparison .It also considered the effectiveness of the restricted zone for aggregate
٣٤
gradation introduced by the Superpave mix design , three aggregate gradations which
run through , above , and below the restricted zone which satisfy the gradation
criteria and the Superpave criteria of control point , were selected to study the effect
of restricted zone .
Based on the finding of this study, the following specific conclusion can be drawn:
1) Although the Superpave system recommends having gradations below the
restricted zone for heavy traffic loads, the gradations passing through and above the
restricted zone in this study meet all of the Superpave requirements. This indicates
that these gradations with crushed and angular aggregates can be expected to perform
adequately in the field.
2) The estimated asphalt binder content for the Superpave mix design is lower than
that of the Marshall Mix method. This indicates that the Superpave mix design yields
a more economical mixture.
3) The average asphalt film thickness of the Marshall Mix at its optimum asphalt
content is determined to be higher than that of the Superpave mix design. This
illustrates that the Marshall mixtures might be more durable. However, the
Superpave mixtures meet the recommended values for film thickness and were
determined to be adequate to ensure the durability of the mixture.
4) The repeated shear at constant height test suggests that the Superpave mix design
method provides stronger mixtures than the Marshall method. The Superpave
mixtures were found to be more resistant to permanent deformation.
5) The results from the frequency sweep at constant height test indicate that the
mixtures prepared by the Superpave system provide stiffer and stronger mixtures
than those prepared by the Marshall method.
6) The indirect tension test also illustrates that the Superpave mixtures are much
stronger than the Marshall mixtures in terms of their tensile strength values.
7) The fatigue analyses of asphalt mixtures indicate that the Superpave mixtures have
higher mix resistance to fatigue distress.
٣٥
8) The rutting analyses of asphalt mixtures suggest that both mix design methods
provide good performance mixtures. They also indicate that the Superpave mixtures
are more resistant to rutting than the Marshall mixtures.
D’Angelo, et al. 1995 studied five asphalt mixes designed in the Superpave and the
Marshall Compaction procedures. Two of the mixes were designed first using the
SGC at Ndes levels of 86 and 100 gyrations and later evaluated with the Marshall
hammer using 112 blows and 50 blows. The 112 blow Marshall Compaction was
used with 6 in. Marshall molds. Three other mixes were designed first using the
Marshall hammer with 112, 75 and 50 blows and then evaluated with the SGC at
Ndes levels of 126,109, and 100 gyrations, respectively. Conclusions from this study
are demonstrated below:
 Samples compacted with the SGC had slightly less variability in air voids than
did the Marshall samples.
 Based on air voids alone, the SGC and the Marshal hammer could both be
expected to perform well in quality control applications. VMA distinguishes
the two-compaction devices. The results show that for every mixture tested,
the SGC samples had lower VMA than Marshall Samples. The general trend
of lower VMA with SGC indicates that the compaction effort obtained with
the SGC is greater than with the Marshall hammer.
 The overall conclusion of the study is that the SGC is better able to track plant
production variability than the Marshall hammer.
Another research project was conducted in 1998 by the Kansas Department of
Transportation, KDOT, to compare the Superpave and Marshall Mix designs for low
volume roads and paved shoulders (Habib, et al; 1998). In this research, five blends
were compacted by using the Superpave gyratory and the Marshall hammer. Mixes
studied were 19 mm nominal maximum size with an AC-10 binder. This binder also
meets the PG 58-22 requirements. Bulk densities and maximum theoretical specific
gravity were measured for each blend and design volumetric parameters were
calculated and analyzed to 4 percent design air content. Superpave samples were
٣٦
designed for less than 0.3 million ESALs with Nini = 7, Ndes=68 and Nmax=104
gyrations. Note that the number of gyrations used in the above-mentioned research is
not the same as number of gyrations in the current Superpave specification. Marshall
Samples were compacted to 50 blows per face. Results obtained from this research
can be listed as follows:
 Superpave mix design for low volume roads and shoulders results in lower
optimum asphalt content compared with the Marshall method. Hence,
Superpave mixtures would be economical in these applications.
 The VMA and VFA values are also lower than those for the Marshall mixes
 River sands appear to have the potential to be used as fine aggregates in the
Superpave mixes for low volume pavements and shoulders. However, the use
of coarse river sand should be minimized because it increases the optimum
asphalt content and could result in a weaker aggregate structure.
 Superpave requirements for VFA for low volume traffic, less than 0.3
millions ESALs appear to be too high.
 Lowering Ndes would result in increased asphalt requirement for a Superpave
mixture with a given gradation.
The Virginia Transportation Research Council compared several asphalt design
methods and found differences in the optimum asphalt contents obtained by Marshall
and Superpave methods (Maupin, 1998). In this research, six 19 mm NMAS mixes
were tested using the 50-blow Marshall design, the 75-blow Marshall design, two
brands of 55 SHRP gyratory compactors, Pine and Troxler, and the U.S. Army Corps
of Engineers' gyratory testing machine, GTM. For purposes of this research, only
results from Superpave and Marshall Designs are described.
The Superpave criteria for this study are based on a traffic level of 3 to 10 million
ESALs and an average high temperature less than 39C°. The corresponding
compaction levels are Nini = 8, Ndes = 96 and N max =152. This compaction level is
slightly less than the current Ndes requirement of 100 revolutions for 3 to 30 million
ESALs. Air void contents for Marshall were 4.0 and 4.5 percent and for Superpave
٣٧
only 4 percent. The optimum asphalt content of 96 gyrations Superpave mixes was
less consistent than with 75 blows of Marshall mixes. This implies that at 96
gyrations the SGC is compacted mixes of more than 75 blows of Marshall method.
Research was conducted at West Virginia University compared mix designs prepared
using the Marshall and Superpave methods for 19 mm base mixtures using a PG 70-
22 binder (Kanneganti, 2002). Mix designs were prepared with limestone aggregates
for three traffic levels. Mix performance was evaluated with the APA. Statistical
evaluation of the data indicates that there is no statistically significant difference
between the optimum asphalt and performance of the mixes at a 95 percent
confidence level. Mix designs prepared under the Superpave criteria were evaluated
under the Marshall method and found to pass all criteria. Similarly, mixes prepared
under the Marshall criteria passed all Superpave criteria when compacted with the
SGC.
JohnP. Z., Jason N. 2003, evaluated the differences between the mix designs from
these two design methods for asphalt concrete wearing courses. These are the
WVDOH wearing I mix for the Marshall method and the 9.5 mm design for the
Superpave method, mixes were developed for light, medium, and heavy traffic. From
the study results, the differences in asphalt contents between the two-mix design
methods range from 0.2 to 0.8 percent. While, for the 100 percent lime stone mixes,
the asphalt content was higher for the Superpave mixes. For the mixes with 13
percent natural sand, the Marshall mixes require more asphalt. On the other hand, the
Marshall Mix with 13 percent sand showed greater rutting potential than the 100
percent lime stone mix. The Superpave mixes for low traffic roads designed with
high sand contents displayed very high rutting potential. The results of this research
indicated that the performance of Marshall and Superpave mixes is comparable with
respect to rutting performance. This demonstrates that correctly applying the
metholodgy and criteria to a mix design method may be more important than when
mix design method is used.
٣٨
The material properties and changes caused by loading and the environment are
required to predict the characteristics and performance of the pavement. The primary
characteristics (mechanistic properties) used to evaluate the performance of
pavement materials under various loading and environmental conditions are the
resilient modulus (E) and Poisson’s ratio (µ) of materials. The mechanistic properties
of the pavement materials and subgrade are used to calculate the stresses, strains and
displacements within the pavement under vehicular loading (Chris O. and David H.,
2004)
One of most important properties of HMA is the elastic modulus. The elastic
modulus has many benefits over other index properties such as AASHTO layer
coefficients, R-value, and CBR since it has a direct effect on the analytical models
used to predict the state of stress. Despite this key advantage, there are some
significant problems associated with its use. First, bituminous pavement materials are
not elastic. Accordingly, a surrogate for elastic modulus (resilient modulus) is used to
characterize a given layer material is bending resistance under the state of stress that
it will experience in-situ. Another problem concerns the difficulty in accurately
measuring resilient modulus in the laboratory. Although improvements on the
laboratory-based resilient modulus test method are anticipated, a second method
involving the use of nondestructive testing and backcalculation analysis also holds
promise. In this latter approach, measurements of surface deflection are obtained
nondestructively in the field and then evaluated mechanistically (using a
computerized process known as backcalculation) to determine each layer’s in situ
resilient modulus. This process is especially useful for rehabilitation design, but it
also has some applications to new pavement design if the nondestructive test
measurements are obtained along the planned road alignment. Another important
consideration that the engineer should recognize is the effect of thickness of HMA on
٣٩
pavement performance, the fact that the majority of all rutting in the HMA layer will
generally occur within the top 3-to 5-in. Thus, if a poor quality HMA mixture is
being used, increasing the thickness of this poor quality layer will not decrease the
rutting in the HMA layer. In fact, in all likelihood, the rutting will be increased.
Thus, increasing the thickness of a HMA layer, of poor quality, will provide
absolutely no benefit of having the total pavement rut depth decreased. If the
engineer is convinced that the HMA mix design is adequate, increases in the HMA
layer thickness may be evaluated to ascertain to what degree, the potential HMA
layer rut can be decreased. In general, this decrease may not be significant.
However, increasing the HMA thickness will definitely provide benefits by
decreasing the layer rut depth in the unbound base, sub base and particularly, the
subgrade layer.
Generally, if any of the layers within the base /subbase are of poor quality; increasing
the thickness of the poor layer will only tend to increase the rutting and not decrease
it. Finally, the presence material, as the modulus is decreased, the resilient strain
significantly increases and consequently, the rutting greatly increased. This
magnifies the need to have highly drainable base /subbase systems present in any
design.
In general, the use of added HMA layers accomplishes two significant factors. First,
the increased modulus of the thicker layer will result in a significant increase in the
layer “relative stiffness”. This will cause a reduction in the stress and strain states in
the subgrade, which will reduce the rut depth magnitude.
Finally, the effective way in which the rutting in the subgrade can be reduced is by
increasing the thickness of the unbound subbase layer. This effectively, reduces the
stress (strain) states in the subgrade layers that leads to reduced rut depth magnitude
in the subgrade.
Propagation of the cracks throughout the entire layer thickness will allow water to
seep into the lower unbound layers, weakening the pavement structure and reducing
the overall performance. This will result in a significant loss in smoothness causing a
٤٠
decrease in pavement rideability. This phenomenon of crack initiation and then

propagation through the entire layer occurs not only in the surface layer but also in
all the stabilized layers underneath. Cracking in the underlying layer, such as the
cement stabilized, reduces the overall structural capacity and may induce reflective
cracking in the upper layers. Perhaps the most important fundamental conclusion that
can be drawn, is that for good performance, the proper thickness of HMA layers must
be either as thin as practical or as thick as possible. The fact clearly indicates that the
greatest potential for fatigue fracture is associated with HMA layers that are typically
in the 3-to 5-in thickness range.
It should be intuitive to the reader, that as the HMA thickness increases beyond 4
inches, the tensile strains generated at the bottom of the HMA layer are reduced.
Thus, it is logical that as the HMA thickness is increased beyond a 4-inch layer, the
fatigue life is directly increased due to a smaller tensile strain value occurring in the
pavement system. Nonetheless, the real important fact that must be recognized is that
the magnitude of the tensile strain does not necessarily increase proportionately to
cause a decrease in HMA thickness. In fact, as the HMA thickness is reduced below
the "maximum cracking level of 3-to 5-in", the tensile strains actually start to
decrease and, in fact, may actually become compressive in nature. Thus, at very thin
HMA layers, there is little or no tensile stresses or strains at the bottom of the HMA
layer. This clearly explains why, fatigue behavior may improve with the decreasing
levels of HMA thickness.
It is important to recognize that the variable of HMA layer thickness and HMA mix
stiffness are directly integrated together to achieve optimal mix fatigue resistance. If
thin HMA layers are used, it is highly desirable to have a low stiffness (low E*)
material. The presence of thin, very stiff HMA layer is highly susceptible to alligator
cracking. On contrast, as thicker HMA layers are used, the pavement engineer should
try to utilize the highest stiffness (high E*) HMA possible. This will tend to decrease
the critical tensile strains at the bottom of the HMA layer and enhance the structure’s
resistance to alligator cracking.
٤١
While Boussinesq's equations represent an elastic solution to a one-layer system,

Burmister (1943, 1945) developed solutions first for a two-layer and later for a three-
layer system, which have advanced mechanistic pavement analysis considerably.
The next significant breakthrough came in 1962 when Schiffman presented his
solution to the analysis of stresses and displacements in a multilayered elastic
system.
In combination with the development of electronic computers, multilayered elastic
pavement analysis soon became popular. Based on Schiffman's work several
computer programs have been developed, of which the earliest programs for the
determination of stresses, strains, and displacements at any position in a layered
pavement structure are CHEVRON (Michelow, 1963), and BISTRO (Peutz, et al.,
1967), of which the latter succeeded by BISAR (De Jong, et al., 1973).Both
programs have been used worldwide by pavement researchers and engineers for
linear-elastic analysis of multi-layered pavement structures. The use of layered
elastic theory instead of empirical methods allows realistic modeling of the pavement
system and surface loading as well as seasonal variations.
In the formulation of elastic theory for an isotropic material in two-dimensional
loading (plane stress) three assumptions are made:
 Equilibrium
 Compatibility between normal and shear strains
 Stress and strain are related according to Hooke's law.
The first assumption is a static condition, which relates normal and shear stresses in
the two coordinate directions. The static condition implies that loads are static and
material weightless. The second assumption is a geometric condition, which assures
continuity between normal and shear strains, while the third assumption is a physical
condition. The formulation results in a fourth-order differential equation can be
solved with regard to boundary conditions like stiff or infinite bottom layer where the
٤٢
deflection is defined as zero. Furthermore, loading conditions are included. In

addition, the geometric condition is difficult to fulfill for granular materials, and that
the physical condition is not valid for many pavement materials. The latter problem
is often considered by the use of nonlinear constitutive material modeling.
The simplest way of describing a pavement system is that of a half space. Boussinesq
(1885) presents a procedure for the determination of stresses, strains, and deflections
in a homogeneous, isotropic, linear elastic half-space. The half-space covers an
infinitely large area; an infinite depth and loads are applied at its top plane.
Originally, Boussinesq's equations were developed for a static point load, but by
integration, equations for a static distributed load can be derived. Figure (2-7)
presents a notation for Boussinesq's equations for a point load P in polar coordinates,
where z is the depth and r is horizontal distance to the point where response is
desired.
As Boussinesq's equations for stresses (σ and τ),strains (ε and δ) and displacements
(d) for a point load are used extensively throughout the report, the equations are
provided in Table (2.8),where R 2 =z 2 +r 2 .
For the three-dimensional case, a set of coordinate directions is defined where no
shear forces exist. These are called the directions of the principal stresses σ 1, σ2, and
σ3.
Boussinesq's equations represent the first simple approach to mechanistic pavement
analysis with the limitation of not taking into account practical aspects like layered
pavement structures, nonlinearity, and non-homogeneity and anisotropy of materials.
٤٣
Figure (2-7) Notations for Boussinesq's equations (Ullidtz, 1987)
Table (2-8) Boussinesq's for a point load (after Ullidiz, 1998).
 1  2. 
r     
2. .R 2 1  cos  
P 2
3 . cos . sin
(1  2. ).P  
t     
2. .R 1  cos  
1

2
. cos
  
2 . . R
3 .P 3
z 2
. cos
(1   ).P  1  2. 
r        
2. .R 2 .E  1  cos  
3
. 3 . cos ( 3 2 . ). cos
( 1   ). P  1  2 . 
    
2 . . R 2 . E  1  cos  
t cos
٤٤
z 
(1   ). P
2 . . R . E
2

3 . cos 3
  2 . . cos  
 rz  . cos 2  . sin 
2 . .R
3.P
2
(1   ). P
 rz  . cos 2  . sin 
 .R 2
(1   ). P  (1  2 . ) sin  
dr      
2 . . R . E  1  cos  
cos . sin
dZ 
(1   ). P
2 . . R . E

. 2 .( 1   )  cos 2
 
Where:
σz = Vertical Stress
σr = Radial Stress
σt = Tangential Stress
Trz = Shear Stress
εz = Vertical Strain
εr = Radial Strain
εt = Tangential Strain
γrz = Shear Strain
٤٥
Typical flexible pavements are composed of layers so that the moduli of elasticity
decrease with depth. The effect is to reduce stress and deflections in the subgrade
from those obtained for the ideal homogeneous. In the solution of the two-layer
problem, certain essential assumptions are made regarding boundary and continuity
conditions. The materials in the layers are assumed homogenous, isotropic, and
elastic. The surface layer is assumed infinite in extent in both the horizontal and
vertical directions. On the other hand, the underlying layer is infinite in both the
horizontal and vertical directions. Stress and deflection values as obtained by
Burmister are dependent upon the strength ratio of the layers, E1/E2, where E1 and
E2 are the moduli of the reinforcing and subgrade layers respectively. Burmister
(Two layer system) is presented in Figure (2-8) (Yoder, and Witczak ,1975).
Figure (2-8) Two layers system (Burmister).
٤٦
Fox and Acum, produced the first extcutive tabular summary of normal and radial
stress in three layer systems at the intersection of the plate axis with the layer
interface. Figure (2-9) shows the pavement multilayer system.
Figure (2-9) Multilayer System.
٤٧
The finite element method is a highly sophisticated tool, which can be used for the
analysis of stress, strain and displacement in a pavement structure. It is broken down
into a mesh consisting of a number of finite elements connected by nodal points.
When one uses different constitutive material models, the Finite Element program
calculates displacements in the nodes until the user-specified convergence criterion
(corresponding to a desired level of accuracy) is reached. Based on the nodal
displacements, the method in turn determines strain and stress.
The advantage of the finite element method is that it allows modeling of pavement
response for both static and dynamic (i.e., time dependent) loading for different
geometrical structures, which may include consideration of cracking. Furthermore,
finite element modeling allows the use of several different constitutive material
models, which can describe nonlinear elastic, visco-elastic, or plastic behavior.
The finite element method is a numerical method, which does not provide an exact
solution, and the material models are based on continuum mechanics, thus validation
against real data is needed. Many facilities in finite element programs come at a
price: most programs are complicated in use and processing time is high. The
complicated nature of finite element programs makes them suited only for forward
analysis of pavement structures (i.e., determination of response), while back
calculation of Young's modulus, based on surface deflection is not possible with
most programs. Numerous commercial all-purpose finite element programs exist and
have been used by pavement researchers for the analysis of pavement response.
Programs especially dedicated to pavement analysis also exist through: 2002 design
guide, ILLI-PAVE, and MICHPAVE.
٤٨
The program was developed at the University of Illinois by Raad and Figueroa , is a
finite element computer program for flexible pavement analysis originally developed
by Wilson and later modified by Duncan and others and by Raad and Figueroa
served as the main research tool. The recent modification introduced by Raad and
Figueroa provides a more rational assessment of the state of stress of pavement
materials approaching failure and consequently their moduli values according to the
Mohr-Coulomb theory of failure (Abdul Haqh, 2000).
It is developed at the Michigan State of University by (Harich and Ran et al), and
used for nonlinear analysis. The Michpave program is very similar to ILLI-Pave and
uses the same methods to characterize granular materials and fine-grained soils and
the same Mohr-Coulomb Failure criterion to adjust the state of stress (Abdul Haqh,
2000).
CIRCLY software is for the mechanistic analysis and design of road pavements.
CIRCLY uses state-of-the-art material properties and performance models and is
continuously developed and extended. The first mainframe version of CIRCLY was
released in 1977. The system calculates the cumulative damage induced by a traffic
spectrum consisting of any combination of user-specified vehicle types and load
configurations. As well as using the usual 'equivalent' single wheel and axle load
approximations, optionally the contribution of each vehicle/load configuration can be
explicitly analyzed.
٤٩
Chapter Three M aterials and M ethods of Testing
The Materials used in this study are locally available and selected from the currently
used materials in road construction in Iraq.
One type of asphalt cement (40-50) penetration graded was used in this study, which
represents PG (64-22) as classified by the Superpave system. It is obtained from
Dourah refinery. The physical properties of this type of asphalt cement are shown in
Table (3-1).
Table (3-1): Physical Properties of Asphalt Cement.
Test Unit ASTM Results

D ( 40-50 )
Penetration 25°C,100 gm , 5 sec. 1/10 mm D5 42
Absolute Viscosity at 60°C (*) Poise D2171 ٢٠٧٠

Kinematics' Viscosity at ٦٠°C (*) C St. D2170 370
Ductility (25°C, 5 cm/min.) Cm. D 113 >100
Softening Point ( Ring & Ball ) C° D 36 51.0
Specific Gravity at 25°C (*) ……. D 70 1.04
Flash Point C° D 92 332
(*) The test was conducted in Dourah refinery
٥٠
One type of crushed aggregate was used in this study, which was brought from
Amanat Baghdad. The source of this type of aggregate is from Al-Taji quarry. The
physical properties of the aggregate are shown in Table (3-2), and the aggregate
gradation was taken from gradation of expressway No.1.
One nominal maximum size was selected (12.5) with two aggregate gradations (R1
and R9). The gradation R9 is passing through the Superpave limitation control points
and restricted zone, while, the gradation R1 is located out of the Superpave restricted
zone requirement. These two gradations were selected to compare the effect of
restricted zone on the mix performance. Mix design was prepared for heavy traffic
level using the Superpave methodology and the traditional Marshall methodology.
The Marshall mix design was evaluated under the Superpave criteria and vice versa.
These gradations are shown in Figure (3-1) and presented in Table (3-4).
Table (3-2): Physical Properties of Al-Taji Quarry Aggregate.
Property Coarse Aggregate Fine Aggregate

R1 R9 R1 R9
Bulk specific gravity 2.518 2.5189 2.615 2.6225

ASTM C 128
Apparent specific gravity 2.553 2.554 2.662 2.689
ASTM C127 and C128
Percent water absorption 0.556 0.56 0.68 0.94
ASTM C 127 and C 128
One type of mineral filler (Ordinary Portland Cement) has been used in this study,
which is obtained from Badoush factory. The physical properties are shown in Table
(3-3).
٥١
Table (3-3): Physical Properties of Filler (Cement).
Property Results
Specific Gravity 3.12

% Passing sieve No.200 ASTM C117 95
Table (3-4): Job Mix Formula's for Wearing Course of the Selected Sections (*).
Sieve Percent Passing

opening Gradation Shape
( mm ) TRZ ( R1 ) ARZ ( R9 )
19 100 100
12.5 92 89.5
10 83.1 77.8
4.74 66.9 55.3
2 41.5 40.2
1 28.2 32
0.63 21.4 25
0.25 14.4 15.1
0.125 11.6 12.2
0.075 9.8 9.8
Specification Requirements:
Stability , Kg 1000 1000

Flow , mm 2-6mm 2-5mm
Air Voids % 3-5% 2-5%
Asphalt 4.7±0.3 4.63±0.3
Compaction >98 >98
(*) Data from the SCRB documents
٥٢
100.00
Control Point of Iraq Specification
Control Point of Superpave System

80.00
Above ARZ (R9)
Through TRZ (R1)

Percent Passing
60.00
40.00
20.00
0.00
0.01 0.10 1.00 10.00 100.00

Sieve Size,mm,Log Scale
Figure (3-1) Gradation of Wearing Course for two sections of the Expressway No.1 in Iraq.
٥٣
Two types of additives (carbon fiber and lime) have been used in this study. The
physical properties of additives are shown in Tables (3-5), and (3-6). Two
proportions of carbon fiber (1% and 0.5%) by weight of asphalt cement and two
proportions of lime (2% and 4%) by weight of aggregate were used in this study.
Table (3-5) Properties of Carbon Fiber (*).
Properties Results
Form 2.54 cm cut
Density 1.8 gm/cm 3
Tensile modulus 29 psi
(*) Results from Al-Furat Beirut
Table (3-6) Chemical Composition and Physical Properties of Hydrated Lime (*).
Chemical composition Hydrated lime
Sulfuric anhydride (SO3) 0.82

Ca(OH)2 93.88
Total 94.70
Physical properties
Apparent specific gravity 2.343
(*) This test from lime factory in Karbala.
٥٤
Two types of mixes were prepared with two-type gradation. Three specimens for
each mix were prepared, and the average of results was reported. In order to
compare the two-mix design directly, the following two types of mixes were
prepared with two optimum asphalt contents of mix design:
 Marshall mixes with optimum asphalt content determined by Marshall
Method.
 Marshall mixes with optimum asphalt content determined by Superpave
system.
The aggregate were obtained from Al-Taji Quarry. The aggregates were
processed by washing, oven drying and sieving. Dried aggregate were separated
with a set of sieves, consisting of the following sieve openings; 19, 12.5, 10,
4.74, 2, 1.18, 0.6, 0.25, 0.15, 0.075 mm, and the material retained on each sieve
and pan was placed in storage pans. Then three samples of each aggregate
gradation were prepared. The steps followed in determining the Marshall mix
design is explained as follows.
Aggregate were heated to a temperature of 155C° before mixed with asphalt

cement. Asphalt cement was heated to the temperature producing a kinematics
viscosity of (170 ± 20) centistokes (up to 163 C° as an upper limit). Then, the
desired amount was added to the heated aggregate and mixed thoroughly until
all aggregate particles were coated with asphalt, then the mix was compacted in
accordance with the method stated in ASTM 1559.
The prepared mix was placed in preheated mold of (4) in , (101.6mm) in
diameter by (3) in (76.2mm) in height , and compacted with 75 blows/end with a
٥٥
hammer of 10 Ib ( 4.536 kg) sliding weight , and a free fall of ( 18 ) in,

(457.2mm) on the top and bottom of each specimen . The specimens were then
left to cool at room temperature for 24 hours.
Two types of mixes were prepared by using two types of gradation. The first
type of mixes was prepared with ARZ aggregate gradation and the other mix
with TRZ aggregate gradation. Two specimens of each mix were prepared and
the average results were reported. For the comparison requirements, the
following two types of mixes were prepared with the optimum asphalt content:
 Superpave mixes with optimum asphalt content determined by Superpave
system.
 Superpave mixes with optimum asphalt content determined by Marshall
Method.
Two specimens for each trial blend at their corresponding initial asphalt content
were compacted to 135 gyrations, which represent the value of Nd for heavy
traffic level. The bulk specific gravity and volumetric properties of the
compacted specimen were determined and the average values were presented
and then the estimated binder content was determined. For the purpose of
comparison process, the design aggregate structure used for Superpave system
was the same as that used for the Marshall system. To determine the design
asphalt content, the first procedure is to measure Gmm. Once the design
aggregate structure is identified, the design asphalt content must be determined.
Starting with the design aggregate structure and the estimate asphalt content,
specimens are prepared at four levels of asphalt content:
٥٦
 P b est -0.5%
 P b est
 P b est +0.5%
 P best +1%
Two compaction specimens were prepared for each asphalt content .This
produces the data for the volumetric analysis, which is identical to the analysis
performed to evaluate the design aggregate structure. The asphalt content that
produces four percent air voids and meets all the other Superpave criteria
represents the design asphalt content. The specimens used for the purpose of
design asphalt content determination were compacted to the design level of
revolutions. The volumetric properties of the specimens determined by
Superpave Procedure are presented in the next chapter. The optimum asphalt
content is determined by selecting the asphalt content that produces four percent
air voids and meets all other mix design criteria. To ensure the mix will not over
densify under traffic, two specimens were prepared with the design aggregate
structure, optimum asphalt content and compacted to the maximum level of
revolutions. The void content of these specimens was determined and compared
with the Superpave criteria.
This procedure outlines the preparation of HMA test specimens using the
Superpave gyratory compactor (SGC). It includes guidelines for mixing and
compacting test specimens.
A batching sheet is prepared that contains the batch weights of each aggregate
component and the asphalt content. The proper weights of each aggregate
component are weighted into pans and the asphalt is heated to the desired
٥٧
mixing temperature. Meanwhile, all mixing tools used such as spatulas, mixing
bowl, and other tools are heated also.
The hot mixing bowl is placed on a scale and zero the scale. Then, the mixing
bowl is charged with the heated aggregates and dry mix thoroughly. A crater is
formed in the blended aggregate and the required asphalt is weighted into the
mixture to achieve the desired batch weight. The mixing bowl is removed from
the scale and the asphalt and aggregate are mixed using a mechanical mixer. The
mixing continued until the aggregate is thoroughly coated with asphalt. The mix
is placed in a flat shallow pan at an even thickness of 21-22 Kg/m 2 and the pan
is placed in the forced draft oven at 135C°. Short term age the specimen is 4
hours.
Prepare the compactor while mixing specimen in the short-term aging. This
includes verifying the compaction pressure, the compaction angle and speed of
gyration are set to their proper values, and the desired number of gyrations is set
to, Nmax. Approximately 45-60 minutes before compaction of the first
specimen, place the compaction molds and base / top plates in an oven set at the
compaction temperature. After the short term aged mix reaches compaction
temperature, place it in the mold, level the mix and place a paper disk on top of
the leveled mix. The top of the uncompacted specimen should be slightly
rounded. The mold is placed in the compactor and centered under the ram.
The ram is then lowered until it contacts the mixture and the resisting pressure is
600 kPa (± 18 kPa).
The angle of gyration (1.25 ° ± 0.02 °) is then applied and the compaction
process begins. When Nmax has been reached, the compactor automatically
stops. After the angle and pressure are released, the mold containing the
compacted specimen is then removed. After a suitable cooling period, the
specimen is extruded from the mold. Figure (3-3) illustrates the various steps in
specimen preparation and compaction, Figure (3-2) shows Iraqi Superpave
٥٨
Gyratory Compactor which is manufactured by Abbas F. Jasim , M.SC student/

Highway and Transportation Engineering.
The bulk specific gravity of test specimens measured using AASHTO T166.
Maximum theoretical specific gravity measured using AASHTO T 209.
Figure (3-2) Local Superpave Gyratory Compactor.
٥٩
Figure (3-3) Various steps of Superpave specimen fabrication.
٦٠
3-4-1 Resistance to Plastic Flow of Asphalt Mixture

(Marshall Test Method)
This method covers the measurement of the resistance to plastic flow of
cylindrical specimen of bituminous paving mixtures loaded on the lateral surface
by means of the Marshall apparatus according to ASTM (D 1559). After the
required specimen is prepared, the Marshall stability and flow tests are
performed on each specimen. The cylindrical specimen is placed in water bath at
60C° for 30 to 40 minutes, and then compressed on the lateral surface at
constant rate of 2in/min (50.8mm/min) until the maximum load (failure) is
reached. The maximum load resistance and corresponding flow are recorded.
Three specimens for each combination are prepared and the average results are
reported. The Marshall stiffness is then calculated from the formula shown
below:
Marshall Stiffness = Marshall stability/ Marshall flow ……… (3-1)
The bulk specific gravity and density ASTM (D2726), theoretical (maximum)
specific gravity of void less mixture are determined in accordance with ASTM
(D 2041).
The percent of air voids is then calculated from the formula shown below:
%Air Voids = {1-bulk SP.Gr. / Max.Theo.Sp.Gr.} *100…….( 3-2)
٦١
3-4-2 Indirect Tensile Strength .

The indirect tensile strength is determined according to the method described by
(ASTM D4123, 1989). The specimens are prepared in accordance with (ASTM
D 1559, 1989), left to cool at room temperature for 24hours and then placed in
water bath at different test temperatures (20, 40, 60 C°) for 30 minutes. Then
they are tested by Versa-Tester using a 1/2 in ( 12.5mm) wide curved, stainless
steel loading strip on both the top and bottom , running parallel to the axis of the
cylindrical specimen which are loaded diametrically at a constant rate of 2
in/min (50.8mm/min) until reaching the ultimate loading resistance. Three
specimens were prepared for each tested mixtures, and the average results are
reported. The indirect tensile strength (I.T.S) is calculated, as follows:
I.T.S = 2P/пDT……………..(3-3)
where:
I.T.S =tensile strength, psi (Kpa)
P = ultimate load to fail the specimen, Ibs(Newtons)
D = diameter of the specimen, mm
T = thickness of specimen, mm
The temperature susceptibility is calculated as follows:
TS= {(ITS)t1-(ITS)t2} / (t2-t1) …………..( 3-4)
(ITS)t1 = indirect tensile strength at t1 C°, t1 =20
(ITS)t2 = indirect tensile strength at t2 C° , t2 =60
٦٢
3-4-3 Creep Test

The diametric – indirect tensile creep test has been used to determine the
stiffness of asphalt mixtures by measuring strain – time values. The diametric –
indirect creep tests are performed on Marshall specimens at corresponding
optimum asphalt content (o.a.c) for various mix types under constant stress of
0.1 MPa . The specimens are immersed in a water bath for 30 min .at the desired
temperature of (25 C°). The specimen is loaded to static stress of 0.1 MPa for 1
hour , and the deformation is recorded at certain time increments ( 0.1 , 0.25 ,
0.5 , 1 ,2 ,4 ,8 ,15 ,30 ,45 , and 60 min ) . The load is then released, and the
recovered strain for 1 hour is recorded, at the same periods.
The vertical strain is calculated by using the following formula:
Єmix = ∆H/Do (mm/mm)………… (3-4)
where:
∆H = the total measured vertical deformation at a certain loading time (mm),
and
Do = the original diameter of specimen (mm)
The stiffness modulus of the mixture is calculated by:
Smix = δ/Єmix (N/mm2) ………….. (3-5)
where:
Smix = stiffness modulus (N/mm2),
δ = applied stress (N/mm2), and
Єmix = vertical strain in the mix specimen.
Three specimens are prepared for each mix combination.
٦٣
3-4-4 Standard Test for the Effect of Moisture on Asphalt Concrete Paving
Mixtures ( LOTTMAN TEST)
This test method covers procedures for preparing and testing asphalt concrete
specimens to measure the effect of water on the tensile strength of the paving
mixture. This test can be used to evaluate the effect of moisture with or without
antistripping additives including lime, Portland cement or carbon fiber. The
tested specimens are prepared by using the optimum asphalt content. Each set of
specimen is divided into two subsets. One subset is tested in dry condition to
determine the indirect tensile strength, and the other subset is subjected to
vacuum saturation followed by a freeze and warm water soaking cycle. Then,
the subset is tested for indirect tensile strength, making at least six specimens for
each test, compacted to 7±1% air voids when using Marshall apparatus. For
gyratory mixes, at least four specimens at least were prepared and compacted to
7±0.5% air voids where two specimens were tested dry and two tested after
partial saturation and conditioning.
The dry tensile strength is calculate as follows:
ITS(dry) = 2P/пDT
where:
ITS (dry) = tensile strength, psi (Kpa)
P = maximum load, Ibs (Newtons)
D = specimen diameter, inches (mm)
T = specimen height immediately before tensile test, inches (mm).
The wet tensile strength is calculate as follows:
ITS (wet) = 2P/пhd
where:
ITS (wet) = tensile strength, psi (Kpa)
P = maximum load, Ibs (newtons)
d = new specimen diameter after conditioning, inches (mm)
٦٤
h = specimen height, after conditioning and immediately before tensile test,

inches (mm).
The tensile strength ratio is calculate as follows:
TSR = (ITS (WET)/ITS (DRY))*100
where:
TSR = tensile strength ratio, %
ITS (wet) = wet strength or average tensile strength of the moisture –
conditioned subset, psi (Kpa), and
ITS (dry) = dry strength or average tensile strength of the dry subset, psi (Kpa).
The recommended minimum tensile strength ratio is 80 and 70 percent, for
Superpave and Marshall respectively.
The following variables have been selected in preparing the asphalt concrete
mixtures for different tests:
1) One nominal maximum size (1/2 in. (12.5 mm)) has been selected each
with two gradation curve (Above Restricted Zone (ARZ) and Through
Restricted Zone (TRZ)).
2) One type of crushed aggregate from one source (Al-Taji Quarry).
3) One grade of asphalt cement (40-50) penetration graded from Dourah
refinery.
4) Two types of additives (carbon fiber and lime) are used with (1%,0.5%)
for carbon fiber , and (4% , 2%) for lime , by weight of asphalt paving
mixtures.
٦٥
5) Two different asphalt cement contents (optimum of Marshall and

optimum of Superpave ) are used as a percentage by weight of total
mixture , including :
1/2 in. (12.5mm) nominal maximum size.
 ARZ gradation ( 4.63 opt. of Marshall , 4.54 opt. of Superpave )
 TRZ gradation ( 4.7 opt. of Marshall , 4.42 opt. of Superpave )
6) One type of filler (cement) is used.
7) One compaction effort ( 75 ) blows/end using Marshall test method and
one compactive effort ( Ninit = 9 , Ndes=135 , Nmax = 220 ) using
Superpave test method for the preparation of specimens for creep test ,
Indirect tensile test and Lottman test .
8) Three testing temperatures (20, 40, 60C°) were used in the indirect
tensile strength test.
9) MICHPAVE and PCPT Programs are used theoretically for comparison
of performance between the pavements those prepared by Marshall and
Superpave mixes.
Figure (3-4) shows the flow chart of this work.
٦٦
Asphalt Aggregate
Gradation
ARZ TRZ
Marshall Superpave
Volumetric Properties Volumetric Properties

Optimum asphalt
Stability , Stiffness Optimum Asphalt
Marshall with opt. Marshall Superpave with opt. Superpave

Superpave with opt. Marshall Marshall with opt. Superpave
Indirect tensile test Lottman test Creep test

(280 specimens) (100 specimens) (120 specimens)
Data analysis
Evaluating the performance of asphalt concrete mixtures
Analysis with Finite Element Software (MICH-PAVE)

Evaluation of (Permanent deformation and Fatigue cracking)
Analysis by using PCPT Program(Evaluation of Thermal cracking)

Figure (3-4) Flow Chart of Testing and Evaluation Program
٦٧
Chapter Four Results and D iscussion
As stated in chapter one , the main objective of this research is to

compare between the performance of Superpave Mix design and Marshall
Mix design . In the present chapter, mix design prepared for heavy traffic
levels using the Superpave and Marshall Methodologies. The results of
Marshall mix design are evaluated under the Superpave criteria and vice
versa. The effectiveness and role of restricted zone on the aggregate
gradation were considered in the Superpave mix design. Two types of
gradation through and above restricted zone were selected to study the
effect of restricted zone on the mixes performance.
A comparison in mix characteristics and design criteria has been made
between Superpave and Marshall mix design.
The optimum asphalt content of the HMA is highly dependent on the

aggregate characteristics such as gradation and absorption. The
relationship between the aggregate surface area and the optimum asphalt
content is most pronounced where as very fine aggregate fractions which
pass sieve the No.200 is involved. Fine material in HMA can act as an
asphalt extender resulting in lower air voids and possible flushing. If the
asphalt content is reduced to stop the flushing, HMA may become dry
and brittle due to the increase in the viscosity of asphalt and the change
in its rheological properties.
68
In Marshall mix design , the optimum asphalt content was to be 4.7% for
(R1) TRZ gradation and 4.63% for (R9) ARZ gradation as per the job
mix formula for expressway No.1 (SCRB documents).
In the Superpave mix design, Figure (4-1), shows Superpave specimen as

fabricated by the Superpave Gyratory Compactor (SGC).
Figures (4-2) to (4-7) show that the optimum asphalt content is found to
be 4.42% for (R1) TRZ gradation and 4.54% for (R9) ARZ gradation.
Figure (4-1) Superpave Specimens.
69
R1 Gradation (TRZ)
6.0
5.5
%Air void
5.0
4.5
4.0
3.5
3.0
4.0 4.5 5.0 5.5
%Asphalt content
Figure (4-2) Relationship between Asphalt content and Air void of

Superpave Specimen, (for R1 Gradation).
R1 Gradation (TRZ)
15.2
15.0
14.8
14.6
VMA
14.4
14.2
14.0
13.8
13.6
4.0 4.5 5.0 5.5 6.0
%Asphalt content
Figure (4-3) Relationship between Asphalt content and VMA of

Superpave Specimen, (for R1 Gradation).
70
R1 Gradation (TRZ)
85
80
75
VFA
70
65
60
4.0 4.5 5.0 5.5 6.0
%Asphalt content
Figure (4-4) Relationship between Asphalt content and VFA of

Superpave Specimen,(for R1 Gradation).
R9 Gradation (ARZ)
6.0
5.5
5.0
%Air void
4.5
4.0
3.5
3.0
2.5
2.0
4.0 4.5 5.0 5.5 6.0
% Asphalt content
Figure (4-5) Relationship between Asphalt content and Air void of

71
R9 Gradation (ARZ)
15.1
15.0
14.9
14.8
14.7
VMA
14.6
14.5
14.4
14.3
14.2
14.1
4.0 4.5 5.0 5.5 6.0
% Asphalt content
Figure (4-6) Relationship between Asphalt Content and VMA of

R9 Gradation (ARZ)
90.0
85.0
80.0
VFA
75.0
70.0
65.0
60.0
4.0 4.5 5.0 5.5 6.0
%Asphalt content
Figure (4-7) Relationship between Asphalt Content and VFA of

72
Figure (4-8) shows that the Superpave mixes have lower optimum asphalt
content than those of Marshall mixes. The lower optimum asphalt content
of the Superpave mixes indicates that SGC at 135 gyrations for Ndes
applies more compaction energy than the Marshall hammer of 75 blows.
From the presented results and as shown in Figure (4-8), it can be seen
that the, ARZ and TRZ gradations have similar optimum asphalt contents,
and TRZ gradation has the lowest optimum asphalt contents. This can be
expected since the TRZ gradation is close to the maximum density line
and less asphalt is needed to fill up the voids. The optimum asphalt
content from each of the mix design method was determined to be
slightly different from each other. Based on the design asphalt contents ,
the Superpave mix design yields a more economical mixture since it
prescribes a lower asphalt content .
Marshall mixes
4.75
Superpave mixes
4.7
4.65
% Optimum asphalt content
4.6
4.55
4.5
4.45
4.4
4.35
4.3
4.25
TRZ (R1) ARZ (R9)
Type of gradation
Figure (4-8) Effect of Mix Design Method on the Optimum Asphalt

Content.
73
The HMA specimens were prepared in the laboratory; they were analyzed
to determine the probable performance of the mixes. The analysis focuses
on volumetric properties of the HMA and the influence of these
characteristics on the HMA behavior.
It can be seen from Table ( 4-1 ) , that these volumetric properties are
checked against the requirements . All the requirements are satisfied .
Table ( 4-1 ) , Volumetric Properties for Marshall Specimens .
Volumetric Properties for Marshall Specimens
R1 Gradation ( TRZ ) R9 Gradation ( ARZ )
A.C 4.7 A.C 4.63
VMA 14.5 VMA 14.85
VTM 3.9 VTM 4.49
VFA 73.1 VFA 69.76
Density 2.353 Density 2.335
It can be seen from Table (4-2), that the volumetric properties of the
Superpave specimens and the gradation meet the criteria of Superpave
mix design. Consider that the Superpave system prohibits the gradation to
be passing through the restricted zone and recommends the gradation to
be below the restricted zone for heavy traffic loads. Figure (4-9 ) shows
that VTM,VMA and VFA values are lower if compared with Marshall
mix design values .
74
The shearing action during the operation of SGC is efficiently orienting

the aggregate into a dense configuration. This may explain the lower
value volumetric properties.
The effect of the selected gradation relative to restricted zone on the
volumetric properties at the corresponding selected optimum asphalt
content is studied. The above mentioned results indicate that the
gradation passing through the Superpave restricted zone produces higher
bulk density , lower VMA and VTM as compared with the gradation
passing above restricted zone as shown in Figure (4-10) .
Table (4-2), Volumetric Properties for Superpave Specimens.
Volumetric Properties For Superpave Specimens
R1 Gradation ( TRZ ) R9 Gradation ( ARZ )
A.C 4.42 A.C 4.54
VMA 14 VMA 14.32
VTM 4 VTM 4
VFA 71.42 VFA 72
Density 2.362 Density 2.349
75
Volumetric properties
for Marshall Mixes
80
Volumetric propertie
70 for Superpave Mixes
60
% Percent
50
40
30
20
10
0
VMA VTM VFA
Volumetric properties
Figure (4-9) Effect of Mix Design Method on Volumetric Properties.
VMA
80 VFA
70
60
% Percent
50
40
30
20
10
0
TRZ (R1) ARZ (R9)
Type of gradation
Figure (4-10) Effect of Gradation on Volumetric properties .
76
The indirect tensile test was conducted according to ASTM D 4123 to

determine the tensile strength of specimen. The indirect tensile test is the
most widely used test for determining the tensile properties of the
highway asphalt materials. Thus , the tensile strength is one of the critical
parameters to be always taken into consideration for the performance
evaluation . The evaluation of the fatigue life of a pavement is based on
the flexural stiffness measurements. Tensile strain at the bottom of the
asphalt concrete layer in a pavement is an important parameter in the
measurement of fatigue life of a mixture. The bottom of asphalt concrete
layer has the greatest tensile stress and strain. Cracks are initiated at the
bottom of this layer and later propagate due to the repeated stresses in
tension of asphalt concrete pavements caused by bending beneath the
wheel loads . Ultimately, the cracks which appear on the surface in the
wheel paths are characterized as fatigue cracking ( N.Paul K. ,2005 ) .
As mentioned in chapter three, three different testing temperatures have
been conducted to evaluate the resistance of mixture to variations in
temperature, (20C °, 40C °, 60C°) with the selected optimum asphalt
content for each mix design. From Figure (4-11), the results indicate that
Marshall mixes with optimum of Marshall have a tensile strength higher
than that of Marshall mixes with optimum of Superpave. This is true, in
tensile stress state; the mixture strength depends on the cohesion element
(asphalt) in resisting stresses, and the increase in the asphalt content up to
a certain limit causes an increase in the surface area of aggregate coated
with binder. As a result, this will increase the strength of mixture. Figures
(4-12) to (4-14), show the effect of additives on indirect tensile strength
for Marshall mixes.
77
Marshall Mixes
1400
Superpave
1200 Mixes
Tensile strength (Kpa)

1000
800
600
400
200
0
TRZ ARZ
Type of gradation
Figure (4-11) Effect of Gradation on the Indirect Tensile Strength.
Marshall with opt.of Marshall

Marshall with opt.of Superpave
1400
1200
1000
800
600
400
200
0
R1 w ithout R1 w ith 2% R1 w ith 4% R1 w ith R1 w ith 1%
additives lim e lim e 0.5% fibe r fiber
Figure (4-12) Effect of Additives on the Indirect Tensile Strength at 20C°

Test Temperature for TRZ (R1) Gradation.
78

Marshall with opt.of Superpave
600
500
400
300
200
100
0
R1 w ithout R1 w ith 2% R1 w ith 4% R1 w ith 0.5% R1 w ith 1%
additive s lim e lim e fibe r fibe r
Figure (4-13) Effect of Additives on the Indirect Tensile Strength at 40C°

Test Temperature for TRZ (R1) Gradation.

Marshall with opt.of superpave
160
140
120
100
80
60
40
20
0
R1 without R1 with 2% R1 with 4% R1 with 0.5% R1 with 1%
additives lime lime fiber fiber
Figure (4-14) Effect of Additives on the Indirect Tensile Strength at

60 C° Test Temperature for TRZ (R1) Gradation .
79
The tensile strength is primarily a function of the binder properties. The

amount of asphalt binder in a mixture and its stiffness influence the
tensile strength. Tensile strength also depends on the absorption capacity
of the aggregates used. At given asphalt content, the film thickness of
asphalt on the surface of aggregate and particle-to-particle influences the
adhesion or tensile strength of a mixture. Various studies have reported
and proved that the tensile strength increases with decreasing air voids.
The tensile strength of a mixture is strongly influenced by the consistency
of the asphalt cement, which influences rutting. Thus , tensile strength
plays an important role as a design and evaluation tool for Superpave
mixes .
From Figure (4-15), it can be seen, that the Superpave mixes with
optimum of Marshall have tensile strength higher than that of Superpave
mixes with optimum of Superpave.
Marshall mixes
Superpave mixes
1600
1400
1200
1000
800
600
400
200
0
with opt.of Marshall with opt.of Superpave
Figure (4-15) Effect of Asphalt Content on the Indirect Tensile Strength

( R1 Gradation at 20 C° Test Temperature ).
80
The test indicates that the Superpave mixtures have higher strength than
the Marshall mixtures due to differences in compaction technique. The
SGC rotates at a constant rate during the compaction , and this
characteristic provides a better orientation of aggregate particles and the
aggregate interlock and this process simulates the field compaction
closely . On the other hand , Marshall compaction hammer provides only
the vertical movement .
To evaluating the effect of restricted zone on strength of mixes , it can be
seen , that TRZ gradation has higher tensile strength as compared to ARZ
gradation. This behavior is true , since the TRZ mixture has low air voids
with more interactive surface area which can carry more tensile stress .
Figure (4-16) shows the effect of additives on the indirect tensile strength
(ITS) values corresponding to the selected gradation.
Gradation (TRZ) R1
Gradation (ARZ) R9
1600
1400
1200
1000
800
600
400
200
0
without with 2% lime with 4% lime with 0.5% fiber with 1% fiber
additives
Figure (4-16) Effect of Additives on the Indirect Tensile Strength

(Superpave with Optimum of Superpave at 20C° Test Temperature).
81
Figure (4-16) also indicates that, (ITS) value increases with an increase in
the amount of lime additives from (2% to 4%) in a mixture. So, hydrated
lime reduces asphalt cracking that can result from causes other than
aging, such as fatigue and low temperatures, although, in general, stiffer
asphalt mixes result in more cracks. Accordingly, the addition of lime
improves fatigue characteristics and reduces cracking. Cracking often
occurs due to the formation of micro cracks. These micro cracks are
intercepted and deflected by tiny particles of hydrated lime. Lime reduces
cracking more than inactive fillers because of the reaction between the
lime and the polar molecules in the asphalt cement, which increases the
effective volume of the lime particles by surrounding them with large
organic chains.
Consequently , the lime particles are better able to intercept and deflect
micro cracks , preventing them from growing together into large cracks
that can cause pavement failure .
From the above mentioned Figures, it is obvious that an increase in the
(ITS) value can appear due to the increase in carbon fiber in the HMA
mix. Modification of the asphalt binder is one of many approaches which
can be considered to improve the pavement performance. The addition of
fibers to asphalt enhances material strength and fatigue characteristics.
Fatigue characteristics of the mixture were expected to improve with the
addition of discrete carbon fibers, and because of the high tensile strength
of carbon fibers and of their inherent compatibility with asphalt cement
and excellent mechanical properties, carbon fibers might offer an
excellent potential for asphalt modification.
Figure (4-17) shows the effect of testing temperature on the indirect
tensile strength (ITS) values corresponding to the selected mixes. It
indicates that, for each mix type, as the testing temperature increases
from (20C° to 40C° and 60C°) the tensile strength of mixture decreases
82
and this is expected, since, the increase in temperature; decrease the

cohesion of asphalt binder.
Marshall mixes
Superpave mixes
1400
1200
1000
800
600
400
200
0
20 40 60
Te m perature
Figure (4-17) Effect of Testing Temperature on the Indirect Tensile

Strength.
The creep test is performed on Marshall specimen at corresponding

optimum asphalt content (O.A.C) for various mix types under constant
stress of 0.1 Mpa , 25C° test temperature for one hour loading followed
by one hour unloading . The creep test results are reported in Appendix
(D), and presented in the form of strain – time curves for various mixes as
shown in Figures (4-18) to (4-27).
83
These Figures show that Marshall mixes with optimum of Marshall have
higher values of strain than those of Marshall mixes with optimum of
Superpave as a result of high flow and low stiffness.
This is because an increase in the asphalt content leads to high flow.
R1 and R9 Without Additives R1 marshall with opt. of marshall

R1 marshall with opt. of superpave
R9 marshall with opt. of marshall
4.0
Strain *10^-3 (mm/mm)
3.5
3.0
2.5
2.0
1.5
0 40 80 120
Time (min)
Figure (4-18) Strain –Time relationship for Marshall Mixes

(Without Additives).
84

R1 and R9 with 0.5% carbon fiber R1 marshall with opt. of superpave
4.0
3.5
3.0
2.5
2.0
1.5
0 40 80 120
Time (min)
Figure (4-19) Strain – Time relationship for Marshall Mixes(With 0.5%Carbon Fiber).

R1 and R9 with 1% carbon fiber R9 marshall with opt. of marshall
3.5
3.0
2.5
2.0
1.5
1.0
0 40 80 120
Time (min)
Figure (4-20), Strain – Time relationship for Marshall Mixes (With 1% Carbon Fiber).
85

R1 and R9 with 2% lime
3.6
3.4
3.2
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
0 40 80 120
Time (min)
Figure (4-21), Strain – Time relationship for Marshall Mixes(With 2% Lime).
R1 marsh al l wi th opt. of marsh al l

R1 and R9 with 4% lime R1 marsh al l wi th opt. of su perpave
R9 marsh al l wi th opt. of marsh al l
R9 marsh al l wi th opt. of su perpave
2.0
1.8
1.6
1.4
1.2
1.0
0 40 80 120
Time (min)
Figure (4-22), Strain – Time relationship for Marshall Mixes (With 4% Lime).
86
R1 W ith ou t addi ti ve s
R1 Marshall with Optimum of Marshall R1 wi th 0.5% carbon fi be r
R1 wi th 1% carbon fibe r
R1 wi th 2% l i me
R1 wi th 4% l i me
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0 40 80 120
Time (min)
Figure (4-23) Strain-Time relationship for R1 Gradation of Marshall Mixes with

different Additives.
R9 Without additives
R9 Marshall with Optimum of Marshall R9 with 0.5% Carbon Fiber
R9 with 1% Carbon Fiber
R9 with 2% Lime
2.8 R9 with 4% Lime
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0 40 80 120
Time (min)
Figure (4-24) Strain-Time relationship for R9 Gradation of Marshall Mixes with

different Additives.
87
Marshall with opt.Marshall

Marshall with opt. Superpave
4.0
Permanent deformation * 10^-3 (mm/mm)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
TRZ ARZ
Type of gradation
Figure (4-25) Effect of Gradation on Permanent Deformation for Marshall Mixes.
Marshall mixes with optimum of marshall

Marshall mix with optimum of superpave
4.0
Permanent deformation*10^-3 (mm/mm)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
R1 without add R1 with 0.5% R1 with 1% R1 with 2% R1 with 4%
fiber fiber lime lime
Figure (4-26) Effect of Additives on Permanent Deformation for Marshall Mixes

( for R1 Gradation).
88
Permanent deformation *10^-3 (mm/mm) 2.5 Marshall mixes with optimum of marshall
Marshall mixes with optimum of superpave
2.0
1.5
1.0
0.5
0.0
R9 w ithout R9 w ith o.5% R9 w ith 1% R9 w ith 2% R9 w ith 4%
additive s fiber fibe r lim e lim e
Figure (4-27) Effect of Additives on Permanent Deformation for Marshall Mixes

( for R9 Gradation) .
Figures (4-28) to (4-33 ) show , that Superpave mixture with optimum of Marshall
has higher values of strain than Superpave mixture with optimum of Superpave,
indicating that , the lower asphalt content for both mix design shows smaller strain
values than the higher asphalt content mixes .
It is observed that Superpave with optimum of Superpave is the strongest mix and
Marshall with optimum of Marshall is the weakest mix among the tested mixes .
To evaluate the effect of restricted zone on creep test, it is shown, that the TRZ
gradation has higher values of strain than the ARZ gradation. As a result, TRZ
mixes show lower stiffness. This is because the ARZ has lower values of flow,
higher stability, and as a result higher stiffness.
89
R1 superpave with opt. of marshall

R1 and R9 Without additives R1 superpave with opt. of superpave
R9 superpave with opt. of superpave
4.0
3.5
3.0
2.5
2.0
1.5
0 40 80 120
Time (min)
Figure (4-28) Strain – Time relationship for Superpave Mixes (Without Additives).
R1 su perpave wi th opt. of marsh al l

R1 and R9 with 0.5% Carbon Fiber R1 su perpave wi th opt. of su pe rpave
R9 su perpave wi th opt. of marsh al l
R9 su perpave wi th opt. of su pe rpave
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0 40 80 120
Time (min)
Figure (4-29) Strain- Time relationship for Superpave Mixes (with 0.5% Carbon Fiber).
90

R1 and R9 with 1% Carbon Fiber R1 superpave with opt. of superpave
R9 Superpave with opt. of superpave
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0 40 80 120
Time (min)
Figure (4-30) Strain – Time relationship for Superpave Mixes (with 1% Carbon Fiber).
R1 su perpave wi th opt. of marsh all

R1 and R9 with 2% Lime R1 su perpave wi th opt. of su pe rpave
R9 su perpave wi th opt. of marsh all
R9 su perpave wi th opt. of su pe rpave
3.4
3.2
3.0
Strain *10^-3 (mm /mm )
2.8
2.6
2.4
2.2
2.0
1.8
1.6
0 40 80 120
Ti me (min )
Figure (4-31) Strain – Time relationship for Superpave Mixes (with 2% Lime).
91

R1 and R9 with 4% Lime R1 superpave with opt. of superpave
R9 superpave with opt. of superpave
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0 40 80 120
Time (min)
Figure (4-32) Strain – Time relationship for Superpave Mixes (with 4% Lime).
Superpave with opt. Superpave

Superpave with opt.Marshall
4.0
Permanent deformation*10^-3
3.5
3.0
2.5
(mm/mm)
2.0
1.5
1.0
0.5
0.0
TRZ ARZ
Type of gradation
Figure (4-33) Effect of Gradation on Permanent Deformation for Superpave Mixes.
92
Figures (4-34) to (4-37) show the effect of adding (hydrated lime) to the mixture on
resisting creep, as a result of rutting. Rutting is permanent deformation of the
asphalt, and is caused when the elasticity of the material is exceeded. It indicate
that increasing the amount of lime (2% to 4%) leads to decrease in permanent
deformation. Hydrated lime significantly improves the performance of asphalt in
this respect . Lime is chemically active rather than inert . It reacts with the bitumen,
removing undesirable components at the same time its tiny particles disperse
throughout the mix, making it more resistant to rutting and fatigue cracking. The
stiffening that results from the addition of lime can increase the PG rating of an
asphalt cement , depending upon the amount used .
Figures (4-34) to (4-37) show , the effect of adding carbon fiber on resisting creep,
it indicates that increasing carbon fiber leads to lower strain and as a result to more
resisting to permanent deformation. Carbon fiber modified asphalt mixtures were
expected to show an increased stiffness and resistance to permanent deformation.
93
R1 without additives
R1 Superpave with opt. of Superpave R1 with 0.5% Carbon Fiber
R1 with 2% Lime
4.0 R1 with 4% Lime
3.5
3.0
2.5
2.0
1.5
1.0
0 40 80 120
Time (min)
Figure (4-34) Strain-Time relationship for R1 Gradation of Superpave Mixes with different
Additives.
R9 Without additives
R9 Superpave with optimum of Superpave R9 with 0.5% Carbon Fiber
R9 with 2% Lime
2.8 R9 with 4% Lime
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0 40 80 120
Time (min)
Figure (4-35) Strain-Time relationship for R9 Gradation of Superpave Mixes with different
Additives.
94
4.0 Superpave with opt. of Superpave
Permanent deformation*10^-3 (mm/mm)

Superpave with opt. of Marshall
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
TRZ TRZ with TRZ with TRZ with TRZ with
without 0.5% fiber 1% fiber 2% lime 4% lime
additives
Figure (4-36) Effect of Additives on Permanent Deformation for Superpave Mixes

(for R1 Gradation).
Superpave with opt. of Superpave

Superpave with opt. of Marshall
2.5
Permanent deformation *10^-3 (mm /mm )
2.0
1.5
1.0
0.5
0.0
ARZ w ithout ARZ w ith 0.5% ARZ w ith 1% ARZ w ith 2% ARZ w ith 4%
additive s fibe r fibe r lim e lim e
Figure (4-37) Effect of Additives on Permanent Deformation for Superpave Mixes

(for R9 Gradation) .
95
The moisture susceptibility test is used to evaluate HMA against stripping. This test
is not a performance based test but serves two purposes . First , it identifies whether
a combination of asphalt binder and aggregate is moisture susceptible . Second , it
measures the effectiveness of anti-stripping additives . The indirect tensile strength
(ITS) test provides properties that are useful in characterizing moisture
susceptibility of hot mix asphalt (HMA). TSR tensile strength ratio is one of the
important properties to reflect the strength of asphalt materials against stripping.
The TSR value shows the susceptibility of HMA to stripping or reduction in
strength under a wet conditioning process, in addition the TSR also can be used to
evaluate the cracking potential of an asphalt mixture.
Figure (4-38), shows that Marshall with optimum of Marshall mixes have higher
tensile strength than Marshall with optimum of Superpave.
Marshall mixes
800
Superpave mixes
Tensile strength in wet condition (Kpa)
700
600
500
400
300
200
100
0
w ith opt. of Mars hall w ith opt. of Supe rpave
Figure (4-38) Effect of Asphalt content on Moisture Damage for Marshall and Superpave Mixes.
96
Evaluation of mixture moisture sensitivity is currently the final step in the

Superpave volumetric mix design process. The moisture affects asphalt mixes in
three ways: loss of cohesion, loss of adhesion and aggregate degradation. The loss
of cohesion and adhesion is important to prevent stripping. A reduction in cohesion
results in a reduction in strength and stiffness . A loss of adhesion is the physical
separation of the asphalt cement and aggregate , primary caused by the action of
moisture .
The presence of water (or moisture) often results in premature failure of asphalt
pavements in the form of isolated distress caused by debonding of the asphalt film
from the aggregate surface or early rutting and / or fatigue due to reduced mix
strength. Damage due to moisture is controlled by limiting the values of the tensile
strength ratios ( TSR ) or the percent loss in tensile strength or loss in the pavement
strength due to moisture damage which indicates that the individual tensile strength
of the mixtures after conditioning will govern the rutting and fatigue of the
mixtures ( N.Paul K. , 2005) .
Figures (4-39) to (4-40), show the evaluation of the use of the gyratory compactor
versus the Marshall hammer for mixture tensile strength ratio (TSR) values. It has
been determined that the Superpave mixture gives higher TSR values than Marshall
mixtures, the higher TSR values are most likely caused by the different aggregate
orientation of the specimen in the SGC and the difficulty of specimen saturation,
therefore, Superpave mixes are less affected by water compared to Marshall mixes.
The investigation into the effect of restricted zone, shows that TRZ gradation is less
susceptible to moisture damage than the ARZ gradation. This behavior can be
attributed to the fact , that TRZ mixture with low air voids , will be less sensitive to
97
moisture damage . After investigating the effect of additives to mixture, it can be

seen that mixes with lime have higher TSR values than the mixes without lime.
This can be attributed to the fact that when lime is added to hot mix, it reacts with
aggregates and strengthening the bond between the asphalt and the aggregate. At
the same time, it treats with the aggregate and with the asphalt itself. Lime reacts
with highly polar molecules that can otherwise react in the mix to form water-
soluble soaps that promote stripping .When those molecules react with lime , they
form insoluble salts that no longer attract water . In addition, the dispersion of the
tiny hydrated lime particles throughout the mixes makes them stiffer and tougher,
reducing the likelihood of bond between the asphalt cement and the aggregate
which will be broken mechanically, even if water is not present. In addition, it can
be seen, that the additive of a carbon fiber to HMA mix will also increase the
strength of the mixture.
98
Marshall mixes
Superpave mixes
100
90
80
70
% TSR VALUE
60
50
40
30
20
10
0
w ithout w ith 1% fibe r w ith 0.5% fibe r w ith 2% lim e w ith 4% lim e
additives
Figure (4-39) Effect of Mix Design Method on TSR Ratio.
TRZ (R1)
80
ARZ (R9)
79
78
77
% TSR value
76
75
74
73
72
71
without with 1% with 0.5% with 2% with 4%
additives fiber fiber lime lime
Figure (4-40) Effect of Gradation on TSR Ratio.
99
Marshall specimen will be taken to measure stability and flow. Stability of a HMA
pavement is its ability to resist shoving and rutting under loads (traffic). A stable
pavement maintains its shape and smoothness under repeated loading; and unstable
pavement develops ruts (channels), ripples (wash boarding or corrugation), raveling
and other signs of shifting of the HMA. Because stability for a pavement depends
on the traffic expected to use the pavement, stability should be high enough to
handle traffic adequately, but not higher than traffic conditions required. The
stability of a mix depends on internal friction and cohesion. Internal friction among
the aggregate particles (inter-particle friction)is related to aggregate characteristics
such as shape and surface texture. Cohesion results from the bonding ability of the
binder. A proper degree of both internal friction and cohesion in HMA prevents the
aggregate particles from being moved past each other by the forces exerted by
traffic. In general, the more angular the shape of the aggregate particles and the
more rough their surface texture, the higher the stability of the HMA will be. The
binding force of a HMA is called cohesion. Cohesion increases with the increase in
loading (traffic) rate. Cohesion also increases as the viscosity of the binder
increases, or as the pavement temperature decreases. Additionally, cohesion will
increase with the increase in binder content, up to a certain point. Past that point,
increasing binder content creates too thick a film on the aggregate particles,
resulting in loss of inter-particle friction. From Table (4-3), it shows that Marshall
specimens meet the criteria of specification mentioned in chapter three.
100
Stability and flow value is not the design criteria in Superpave mix design .
However, to make a comparison with Marshall mix design, specimens were
prepared at optimum of Superpave mix on gyratory compactor and tested for
stability and flow values. Comparative results are shown in Table (4-3).
Table (4-3) Stability and Flow values for each Mix .
Marshall Mixes
Type of Gradation Stability (KN) Flow (mm)
R1 11.5 11.8 12 4.4 4.9 5
R9 13 12.5 13.5 2.3 2.2 2.5
Superpave Mixes
R1 13.8 14.1 14.4 4.2 4.5 4.1
R9 15.7 15.1 16.3 2.2 2.1 2
It can be seen from Figures (4-41) to (4-42) that the stability values of the
Superpave mixes are higher than that of Marshall mixes , while the flow values of
Superpave mixes are slightly less than that of Marshall mixes . After studying the
effect of restricted zone, it can be seen that ARZ gradation has higher Marshall
stability and lower flow value than the TRZ gradation. This can be related to the
fact that the ARZ gradation has more internal friction; therefore, it has the highest
stiffness.
101
Marshall mixes
Superpave mixes
18
16
14
Stability KN
12
10
8
6
4
2
0
R1 R9
Type of gradation
Figure (4-41) Effect of Type of Gradation on Stability values of Mixes.
Marshall mixes
6
Superpave mixes
4
Flow Values mm
0
R1 R9
Type of gradation
Figure (4-42) Effect of Type of Gradation on Flow values of Mixes.
102
Recent studies have shown that asphalt mix durability is directly related to asphalt
film thickness (Kandhal, and Chakraborty, 1996). Asphalt film thickness is directly
related to durability and moisture susceptibility of HMA (Chadbourn, et al;1999).It
is generally agreed that high permeability, high air voids and thin asphalt coatings
on the aggregate particles are the primary causes of excessive aging (Kandhal, et
al;1998).
The asphalt film thickness is an indicator of the amount of binder coating the
aggregate particles. It is measured in microns and calculated by dividing the
effective volume of asphalt binder by the total estimated surface area of the
aggregate. Surface area is affected mainly by aggregate gradation. This parameter
is slightly affected by the percentage passing the larger sieves sizes significantly
affected by the percent passing small sizes. For this reason, surface area and asphalt
film thickness could be an issue for low traffic volume HMA applications with a
large percent of fines . As a consequence, it is possible to increase or decrease
surface area by increasing or decreasing the amount of fines in the mixture, and
especially by altering the amount of dust, material finer than 75 mm, present in the
HMA , (Reyes,2003).
The total surface area of an aggregate blend is then determined as the sum of
surface area factors times the percentage passing each size:
SA =∑ Sfi * Pi
Where
SA =Surface area m2 /kg
Sfi =Surface factor for sieve i
Pi =Percent passing sieve in decimal form
103
Thicker asphalt binder films produce mixes which are flexible and durable, while
thin films produce mixes which tend to crack and ravel excessively. An insufficient
coating on aggregate particles is one of the causes leading to premature aging of the
asphalt binder. Lacking of film thickness also leads to inadequate cohesion between
particles known as “dry” mixes. Also, aggregates coated by a thin asphalt film are
easily penetrated by water causing striping and brittle (Chadbourn, et al;1999).
Since the minimum asphalt content will be different for mixes with different
gradations, a more rational approach for VMA should be based on the minimum
average film thickness rather than a minimum VMA. An average film thickness of
8 microns at 4 percent air voids was used and recommended by (Kandhal, et
al;1998).
A rational approach based on a minimum asphalt film thickness has been proposed
and validated . The film thickness approach represent a more direct , equitable , and
appropriate method of ensuring asphalt mix durability and encompasses various
mix gradation . The amount of material passing the 75-mm sieve has a significant
effect on HMA properties ( Anderson , 1987) . Increasing the amount of material
passing the 75mm sieve will result in an increase in the surface area of the
aggregate blend. Consequently, the average film thickness is thinner producing a
lower VMA. Some additional effects of dust on HMA properties are presented by
(Chadbourn, et al; 1999) as follows:
 Stiffening the asphalt binder ,
 Extending the asphalt binder content
 Altering the moisture resistance of the mix
 Affecting the aging characteristics of the mix , and
 Affecting the workability and compaction characteristics of the mix .
104
Estimated asphalt film thickness for asphalt concrete mixtures and the results are
shown in Table (4-4).
The rationale behind the current Superpave VMA requirement is to incorporate a

minimum asphalt content into the mix to ensure its durability (Kandhal, et
al;1998).
Because the Superpave mix design often suggests a lower optimum asphalt content
than that of the Marshall mix design, the durability of the Superpave mix is
questionable and needs to be evaluated.
Thicker asphalt binder films produce mixes that are flexible and durable, whereas
thin films produce brittle mixes which result in a reduction of pavement service
life. The asphalt film thickness of the mixtures is calculated to ensure the adequacy
of the estimated binder contents. The average asphalt film thickness as shown in
Figure (4-43), generally recommends ranges from six to eight microns (Campen, et
al;1959). Kandhal also suggested an optimum film thickness value of 8 microns
(Kandhal, et al;1998).
Figure (4-43) Asphalt Film Thickness
105
The film thickness of the Superpave at the optimum asphalt content is determined
to be as shown in Table (4-4).
Gradation, VTM and dust content affect film thickness, with dust content having
the greatest impact on asphalt film thickness.
In this study, the film thickness is mainly affected by the compaction methods and
the asphalt contents since the Superpave and Marshall mix designs use the same
aggregate gradation. The Marshall mixes yield a higher asphalt film thickness than
the Superpave mixes. Therefore, they are more durable to oxidation and
polymerization than the Superpave mixtures.
Figure (4-44) shows values of asphalt film thickness for both Superpave and
Marshall mixtures. In general, Marshall mixtures show values higher than that of
Superpave mixture. This is expected since Marshall mixtures have higher optimum
asphalt contents, in addition, there is more amount of fines passing No.200, then
increasing surface area as a result of decreasing asphalt film thickness for
Superpave mixes.
106
Table (4-4) Asphalt Film Thickness for both Mixes.
Asphalt Film Thickness (µ)
Marshall Mixes Superpave Mixes
TRZ (R1) ARZ (R9) TRZ (R1) ARZ (R9)
6.22119E-06 5.9491E-06 5.8334E-06 5.82796E-06
0.0000063 Marshall Mixes

Superpave Mixes
0.0000062
Asphalt Film Thickness
0.0000061
0.000006
0.0000059
0.0000058
0.0000057
0.0000056
TRZ (R1) ARZ (R9)
Type of Gradation
Figure (4-44) Effect of Mix Design Method on Asphalt Film Thickness.
107
Chapter F ive Effect of M ix D esign M ethod on the Pavement Structural Performance
One of the primary goals in pavement engineering is the prediction of pavement

performance. Such predictions, until recently, have been exclusively based on
empirical procedures based on road tests carried out some 50 years ago. Such
procedures are only applicable to similar loading (i.e., tire pressures and tire types),
pavement layer materials, and environmental conditions that were present at the road
test sites. On the other hand, mechanistic procedures enable pavement engineers to
undertake design and analysis at a site that has different conditions (loading,
materials and environmental) than the road test site. Such procedures require
pavement response (e.g. stress, strain, deflection) induced by traffic loads to predict
pavement performance (Raj V.S ,et,al,2003).
Recent pavement design procedures, including the SHRP Superpave procedure use
the pavement response as a critical input.
Keeping in mind that flexible pavement design deals primarily with structural aspects
(i.e., the selection of appropriate materials, characterization of strength or load-
carrying properties, layer thickness determination), it can be said that the state of the
art in flexible Pavement design is manifested in mechanistic, or mechanistic-
empirical (M-E), based design. (Stephen B.S,).
The mechanistic-empirical approach would more realistically characterize in-service
pavements and improve the reliability of designs. Much of the expected improvement
is a result of characterizing paving materials through the application of engineering
mechanics rather than empirical.
١٠٨
M-E design methods are based on the mechanics of materials that relate an input,
such as wheel load, to an output of pavement response, such as stress or strain. In the
ME Design Guide procedure, the pavement is regarded as a multi-layered elastic
system. The materials in each of these layers are characterized by modulus of
elasticity (E) and Poisson’s ratio (ν). This method requires the determination of
critical stress, strain, or deflection in the pavement by some mechanistic method and
the prediction of resulting damages by some empirical failure criteria. Prior to the
thickness design, remaining life of the existing pavement must be evaluated. In the
ME design process, the multi-layer structure is analyzed mechanistically to estimate
the critical strains developed within the structure. These strain values are used to
estimate the structural capacity in terms of repeated traffic loading by using the
empirically derived transfer functions. The results are compared with the results
obtained from a field test section to validate the mechanistic component.
A mechanistic- procedure, has the following advantages:
 It provides more reliable and realistic analysis.
 It has an ability to predict the type of distress.

 It can be used for both existing pavement rehabilitation and new pavement
construction.
 It accommodates changing load types, environmental, and aging conditions.
 It uses material properties which relate better to actual pavement response.
 It can better characterize materials.
١٠٩
Figure ( 5-1 ) Mechanistic and Empirical design and analysis

Mechanistic techniques for asphalt pavement analysis have been around since the
1960s, although wider development and implementation started in the 1980s and
1990s. Mechanistic design is much the same as other engineering approaches used
for bridges, buildings, and dams. Essentially, the principles of physics are used to
determine a pavement's reaction to loading. As the critical points in the pavement
structure are known, one can design against certain types of failure or distress by
choosing the right materials and layer thicknesses.
In the case of the perpetual pavement, it would consist of providing enough stiffness
in the upper pavement layers to preclude rutting and enough total pavement thickness
and flexibility in the lowest layer to avoid fatigue cracking from the bottom of the
pavement structure. Since the HMA pavement is tailored to resist specific distresses
in each layer, the materials selection, mix design, and performance testing need to be
specialized for each material layer.
١١٠
The mixtures stiffness need to be optimized to resist rutting or fatigue cracking,

depending upon which layer is being considered. Durability is a primary concern for
all layers.
One of the benefits of the M-E methodologies (ideally) is that they rely primarily on
a fundamental engineering property of the individual pavement and soil layers to
determine the state of stress and predict pavement performance. That property is the
elastic modulus, and its benefit over other index properties such as Poisson’ ratio,
and other material properties that have a direct effect on the analytical models used to
predict the state of stress.
The dynamic stiffness of asphalt concrete mixtures is one of the key factors to
control pavement performance. Powell and Leech show that the dynamic stiffness of
the mixture increases with 30 % if the void content of the material is reduced by 3 %.
Linear elastic analysis of the construction as a whole shows that by reducing void
content, the thickness of the construction can be reduced by 8%. Moreover a higher
compaction level will increase the fatigue resistance of the material. The third
advantage of adequate compaction is the increase in the resistance to the permanent
deformation. An increase of 3% in compaction leads to a reduction of the permanent
deformation of about 50% after 1000 passes, measured with a pneumatic tired wheel
tracking machine (Henny H., &André M.,2000).
Hot Mix Asphalt concrete (HMA) is a complex composite made up of aggregates,

binder and air voids. In hot mix asphalt, binder, and aggregate are blended together
in precise proportions. The relative proportions of these materials determine the
physical properties of the HMA and ultimately how the HMA will perform as a
finished pavement. Theoretically, in this study, we have to studied the effect of two
mixes design method on the performance of pavement structure to determine which
mixes performs better.
١١١
In constructing an asphalt layer the compaction is probably one of the most crucial
stages. The level of compaction largely governs the structural performance of the
entire pavement construction (Henny H., &André M.,2000).
To increase the knowledge about the compaction of asphalt concrete mixtures several
approaches can be taken. In this section simulation of the process with a 2D FEM
calculation, based on elastic-viscoelastic material behavior is described.
Improving compaction results in a significant improvement in load spreading,
resistance to fatigue cracking and resistance for deformation of asphalt concrete
mixtures. These improvements undoubtedly result in extended pavement life. In the
present research, two different methods of HMA specimen preparation (Marshall
method and Superpave system), are used to predict the performance of pavement
structure.
A comprehensive software named the Michpave has been developed at Michigan
State University (MSU). This software analyzes a pavement using nonlinear finite
element program. MICHPAVE has been enhanced to use a distant lateral boundary
and many more finite elements, and the nonlinear model for granular material is
implemented. This instills confidence in the proposed M-E method, since adopting
the M-E method will yield pavement cross- sections with uniformly consistent
performance.
Furthermore, MICHPAVE Software is selected to be used for the comparison
between the structural performance of Superpave and traditional Marshall mixes. The
pavement volumetric and mechanical properties were used as an input data for
MICH-PAVE Software. These data can be seen in Table (5-1). Input data for other
layers are assumed and are presented in Table (5-2).
١١٢
Table (5-1) Data input to MICHPAVE Program for Wearing Course
Input Data for Marshall Mixes
Type of Asphalt Stiffness Density Air void

gradation Content % (Psi) Pcf %
R1 4.7 300000 146.86 4
R9 4.63 500000 145.74 4
Input Data for Superpave Mixes
Type of Asphalt Stiffness Density Air void

gradation Content % (Psi) Pcf %
R1 4.42 350000 147.42 4
R9 4.54 550000 146.62 4
Table (5-2) Assumed Input Data of layers of pavement for MICHPAVE Program.
Base Course
Type of Marshall Mixes Superpave Mixes

Gradation Stiffness Density Stiffness Density
(Psi) Pcf (Psi) Pcf
R1 295000 140 295000 140
R9 495000 140 495000 140
Sub base Course
R1 15000 120 15000 120

R9 15000 120 15000 120
Subgrade Course
R1 5000 120 5000 120

R9 5000 120 5000 120
١١٣
The following cross-section which represents the pavement structure is proposed to

be used during the application of MICH-PAVE Software.
Wearing course 5cm
Base course 15 cm
Subbase course 30 cm
Subgrade course 90 cm
Figure (5-2) assumed pavement structure used in software.
Output data from this program will eventually be used to make detailed prediction of
pavement performance. In other words, the output will allow to estimate the
performance life of HMA to achieve a certain level of rutting and fatigue cracking.
١١٤
There are some factors which affect the performance of pavement structure. These
factors can be listed as follows:
 Mix Design Information
Hot Mix Asphalt pavement mixtures are expected to perform over extended periods
under a variety of traffic and environmental conditions. HMA properties are very
important in resisting permanent deformation and cracking under traffic loads.
 Pavement Layers
The thickness of the asphalt concrete surface course plays a crucial role in bearing
load repetitions, because a given percentage of increase in the expected loads can be
accommodated by a much smaller percent increase in pavement thickness.
 Material Properties of Pavement Components
According to the multi-layered elastic theory, the material properties of each layer
such as resilient modulus and Poisson’s ratio will contribute to the magnitudes of
stress and strain in and between each layer and thus can directly reflect the fatigue
characteristics and permanent deformation behavior of pavements.
 Traffic Loading and Volume
Traffic including loading and volume is one of the most important criteria in
pavement design. The consideration of traffic should include both the loading and the
number of load repetitions. The current AASHTO design method is based on the
total number of passes of the standard equivalent single axle load (18-kip ESAL)
during the design period. The important fact is that most pavement distresses are
load-associated even for the non-load-associated distresses; the load repetitions
certainly exacerbate the deterioration,(Shiou-San K., Hesham S.M, et al ,2003 ) .
 Temperature and Precipitation
Mix temperature is considered as the most important factor in achieving proper
pavement compaction. The mix temperature at the time of compaction is affected by
conditions at the hot-mix plant, the paving process, thermal properties of the hot-mix,
١١٥
thickness and density of the pavement layer and environmental conditions (air
temperature, base temperature, wind velocity and solar radiation).If the temperature
is too low, the mix will be understressed; if it is too high, the mix will be
overstressed. This underscores the importance of determining and maintaining an
optimal temperature at which maximum densification can take place (Bruce A.C. et
al ,1998).
Temperature change may affect the existing insitu resilient modulus of HMA. When
the pavement surface cools, the asphalt binder will slowly transform from a ductile
into a brittle material. Inherent in the pavement structure are a large number of flaws
that are unable to transmit loads and will therefore act as stress concentrators. At
crack tips in the binder and , at the binder – aggregate interface or within broken
aggregate , thermal induced stresses will concentrate which may allow cracks to
initiate and / or propagate , ( Namir .G.A.,2002 ).
A computer program developed by Dr. Namir G.A.(2002), for the Prediction of

Critical Pavement Temperature (PCPT) was used for the determination of critical
pavement temperature in asphalt concrete mixture. In this program, the temperature
at which the accumulated thermal stress exceeds the tensile strength is defined as
critical pavement temperature. This definition is applied as the main concept in
performing the program that predicts the critical pavement temperature. The required
data for the experimental approach obtained from the test results are presented in
chapter four. The input and output of the PCPT program performed for this purpose
can be seen in Figures (5-4) to (5-11), the output of the PCPT program is shown in
Table (5-3).
١١٦
Table (5-3) Output Results of the PCPT Program.
Critical Pavement Temperature
Type of Gradation Marshall Mixes Superpave Mixes
TRZ (R1) -31.75 -31.25
ARZ (R9) -31.95 -31.55
Figure (5-3) shows the low temperature crack in HMA.
Figure (5-3) Low Temperature Crack in HMA.
١١٧
Figure (5-4) Input Data PCPT Program for Marshall Mixes (for R1 Gradation).
Figure (5-5) Output Results of PCPT Program for Marshall Mixes(for R1 Gradation).
١١٨
Figure (5-6) Input Data PCPT Program for Superpave Mixes (for R1 Gradation).
Figure (5-7) Output Results PCPT Program for Superpave Mixes (for R1 Gradation).
١١٩
Figure (5-8) Input Data PCPT Program for Marshall Mixes (for R9 Gradation).
Figure (5-9) Output Results of PCPT Program for Marshall Mixes (for R9 Gradation).
١٢٠
Figure (5-10) Input Data PCPT Program for Superpave Mixes (for R9 Gradation).
Figure (5-11) Output Results PCPT Program for Superpave Mixes (for R9 Gradation).
١٢١
MICHPAVE is a user-friendly, non-linear finite element program for the analysis of

flexible pavements. The program computes displacements, stresses and strains within
the pavement due to a single circular wheel load. Useful design information such as
fatigue life and rut depth are also estimated through empirical equations. Most of
MICHPAVE is written in FORTRAN 77.
This section gives a summary of the modeling and analysis so that the user becomes
aware of the capabilities and limitations of the MICHPAVE program.
 Modeling of the Pavement
Each layer in a pavement cross section is assumed to extend infinitely in the
horizontal directions, and the last layer is assumed to be infinitely deep. All the
pavement layers are assumed to be fully bonded so that no slip occurs due to the
applied load. Displacements, stresses and strains due to a single circular wheel load
are computed. Due to the assumptions used, the problem is reduced to an
axisymmetric one.
 Granular and Cohesive Material Models
The so-called K-model is used to characterize the resilient moduli of granular
materials.
This model is of the form in which = 1 + 2 + 3 = bulk stress and MR = resilient
modulus, and K1 and K2 are material properties. For this model, log MR varies
linearly with log as shown in Figure (5-12).
The resilient modulus for cohesive soils is specified in terms of the deviatoric stress
through the bilinear model:
MR={ k2+k3(k1-(δ1-δ3)) , when (δ1-δ3)≤k1
MR={K2+K4{( δ1- δ3)-K1} , when (δ1- δ3)>k1
This model is illustrated in Figure (5-13).
١٢٢
 Gravity and Lateral Stresses

The MICHPAVE program includes the effect of gravity and lateral stresses arising
from the weight of the materials. At any location within the pavements, the vertical
gravity stress (g) is computed as the accumulation of the layer thicknesses
multiplied by the appropriate unit weights.
Figure (5-12) Resilient Modulus Model for Granular Soils.
The lateral stress is taken as σh= Ko σ g, where Ko = coefficient of earth pressure at

rest. For granular soils Ko =1 sin and for cohesive soils Ko = 1 0.95 sin ,
where = angle of internal friction.
To approximately account for “locked-in” stresses caused by compaction, the user
can input a value for Ko higher than the coefficient of earth pressure at rest.
 Finite Element Analysis
The pavement response model uses output from the material property and
environmental effects model to predict stresses and strains using a two –
dimensional, axisymmetric finite element approach.
The performance testing and performance prediction models represent an important
new tool for engineers in designing and managing pavements.
١٢٣
In this software, rectangular four-noded axisymmetric finite elements with linear

interpolation functions are used in all upper layers and through the depth specified by
the user for the last layer (the roadbed).
A lateral boundary is placed at a radial distance of 10a from the center of the loaded
area, where a = radius of the loaded area.
A default mesh is initially generated, but this may be modified by the user. The
default mesh has the following characteristics:
Figure (5-13) Resilient Modulus Model for Cohesive Soils
 In the radial direction, the total width of 10 radii is divided into four regions.
Within any region, all elements have the same horizontal dimension. The first
region, between 0 and 1 radius, is equally divided into four elements; the second
region, between 1 radius and 3 radius, is equally divided into four elements; the
third region, between 3 radii and 6 radii, is equally divided into three elements;
and the fourth region, between 6 radii and 10 radii, is equally divided into two
elements.
 Within any layer, all elements have the same vertical dimension. The number of
elements in each layer in the vertical direction is dependent on the layer thickness,
but at least four elements are used at the top (AC) layer, and at least two elements
١٢٤
are used at all other layers. A typical default finite element mesh is shown in
Figure (5-14).
Figure (5-14) Typical Finite Element Mesh
Displacements, stresses and strains are computed only within the region modeled by
finite elements. In order to increase accuracy, and to reduce the memory and
computation time required by the program, the infinite extent of the last layer is
modeled by using a flexible bottom boundary (Harichandran and Yeh 1989).
The half-space below the bottom boundary is assumed to be homogeneous and
linear elastic. The modulus of the half-space is taken as the average moduli of the
finite elements immediately above the bottom boundary.
The non-linear analysis consists of several iterations. A linear analysis is performed
in each iteration, after which the resilient modulus of each finite element is revised if
necessary. If the Mohr-Coulomb failure criterion is violated in any granular or
cohesive soil element, the principal stresses are modified to reflect the failure
١٢٥
condition, and the resilient moduli are determined from the modified stresses (Raad
and Figueroa 1980). The iteration is repeated until the resilient moduli of all the
elements stabilize.
At the end of the analysis, MICHPAVE outputs an equivalent resilient modulus for
each pavement layer. These equivalent moduli may be useful if further analyses are
to be performed using other programs that assume linear elastic materials. The
equivalent moduli for each layer is computed as the average of the moduli of the
finite elements in that layer that lies within an assumed 2:1 load distribution zone
(Harichandran et. al. 1990). Results from the non-linear mechanistic analysis,
together with other parameters, are used as input to the performance models derived
on the basis of field data (Baladi 1989), to predict the fatigue life and rut depth.
These performance models are currently restricted to three-layer pavements with
asphalt concrete (AC) surface, base and roadbed soil, and four-layer pavements with
AC surface, base, subbase and roadbed soil. Fatigue life and rut depth estimated for
other types of sections may be meaningless. The models relate the fatigue life and rut
depth to the number of equivalent 18-kip single-axle loads, surface deflection,
moduli and thicknesses of the layers, percent air voids in the asphalt, tensile strain at
the bottom of the asphalt layer, average compressive strain in the asphalt layer,
kinematics viscosity of the asphalt binder and average annual air temperature
(Ronald S.H. , Gilbert Y.B., 2000 ).
The data setting of the MICHPAVE program performed for this purpose can be seen
in Figures (5-15) to (5-19).
١٢٦
Figure (5-15), describe the current job Figure (5-16), Input the loading
for identification purposes. and design thresholds
Figure (5-17), Specify pavement cross-section Figure (5-18), Input to output sections,
and material type which displacements, stresses and strains
were computed
١٢٧
Figure (5-19), Input of boundary conditions
Using the mechanical properties of the HMA and the performance prediction models,
mix design engineers will be able to estimate the combined effect of asphalt binders,
aggregates, and mixture proportions. The analysis will take into account the
structure, condition, and properties of the existing pavement (if applicable) and the
amount of traffic to which the proposed mixture will be subjected over its
performance life. The output of the analysis will be millimeters of rutting, life time of
fatigue cracking, and strains in other layers. By using this approach, the HMA mix
design system will become the ultimate design procedure by linking material
properties with pavement structural properties to predict actual pavement
performance. When the pavement analysis is completed, the benefit (or detriment) of
new materials, different mix designs, asphalt modifiers, and other products can be
quantified in terms of cost versus predicted performance. This capability would
reduce the dependency on field test sections for relative comparisons.
١٢٨
This integrated mixture and structural analysis system will allow the designer to
evaluate and compare the costs associated with using various materials and
applications.
When high quality materials are used, distresses are typically due to traffic loading,
resulting in rutting and fatigue cracking. Environmental conditions such as
temperature and water can have highly significant affect the performance of asphalt
concrete pavement as well.
When a wheel load is applied to a pavement, two stresses are transmitted to the
HMA: vertical compressive stress within the asphalt layer, and horizontal tensile
stress at the bottom of the asphalt layer. The HMA must be internally strong and
resilient to resist the compressive stresses and prevent permanent deformation within
the mixture. In the same manner, the material must also have enough tensile strength
to withstand the tensile stresses at the base of the asphalt layer, and also be resilient
to withstand many load applications without fatigue cracking.
Figures (5-20) and (5-21) illustrate the initiation of cracks and stress, strain within
the pavement layers.
Figure (5-20) Illustration of Propagation Cracks.
١٢٩
Figure (5-21) Illustration of Stress and Strain within Pavement Layers.
The asphalt mixture must also resist the stresses imparted by rapidly decreasing
temperatures and extremely cold temperatures. While the individual properties of
HMA components are important, asphalt mixture behavior is best explained by
considering asphalt cement and mineral aggregate acting together. One way to
understand asphalt mixture behavior is to consider the primary asphalt pavement
distress types that engineers try to avoid such as: permanent deformation, fatigue
cracking and low temperature cracking. These are the distresses analyzed in HMA.
The performance of the mixtures in fatigue and rutting is not affected by tensile
strength alone. A large set of mixture properties influences the performance of the
mixtures. Fatigue life of a mixture is influenced by the percent voids filled with
asphalt (VFA), asphalt content, nominal maximum size of the aggregate, air void etc,
and on the tensile strength and the stiffness of the mixtures. Similarly , the rutting
characteristic of a mixture are influenced by shear strength , air void , percent in the
mineral aggregate , asphalt content , percent aggregate fines than No.200 . (N.Paul K.
,2005-14).
By observing the output results of the MICHPAVE software in Appendix (G), it can
be seen that Superpave is better than traditional mixes (Marshall Mixes) for service
١٣٠
life of pavement structure. So, Superpave offers several advantages over the
traditional design process. These benefits include longer pavement life with less
rutting and have the approximately the same resistance to fatigue cracking. These
important results will support the recommendation of using Superpave mixes instead
of traditional Marshall mixes especially from maintenance, rehabilitation and the
economic point of view. In addition, and depending on the previous mentioned
results, applying Superpave system in the evaluation process of the HMA is strongly
recognized to produce better quality of mixes and enhance the pavement
performance.
The summary of output results of application of MICHPAVE Program is as follows:
Design Summary of Marshall Mixes for R1 Gradation:
Maximum tensile (radial) strain in the asphalt layer = 3.083E-05
Average compressive (vertical) strain in the asphalt layer = 8.205E-05
Maximum compressive (vertical) strain at top of subgrade = 3.848E-04
Fatigue life of asphalt pavement = 3.178E+07 ESAL applications.
Total expected rut depth of the pavement = 2.13187E-01 inches
Expected rut depth in the asphalt course = 5.83811E-02 inches
Expected rut depth in the base and/or subbase course = 7.13512E-02 inches
Expected rut depth in the roadbed soil = 8.34548E-02 inches
Design Summary of Marshall Mixes for R9 Gradation:

١٣١
Design Summary of Superpave Mixes for R1 Gradation:

Design Summary of Superpave Mixes for R9 Gradation:

١٣٢
Chapter Six Conclusions and Recommendations
Based on the study findings, the following conclusions are appropriate:
1. Superpave achieves lower air void content than Marshall mixes; this prevents
additional compaction under traffic, which could result in the wheel path.
2. Superpave mixes yield lower asphalt content than Marshall Mixes. As a
result, Superpave mixes are better from the economical point of views than
Marshall Mixes.
3. Superpave mixes show better moisture susceptibility than Marshall Mixes.
4. The mixes prepared under the Superpave method pass the Marshall criteria,
and the mixes prepared under the Marshall method pass the Superpave
criteria, this indicates that using Superpave method to design and construct
pavement should not face unusual difficulties with Superpave mixes.
5. The TRZ gradation blends meet all the Superpave mix design requirements
and may be expected to perform adequately.
6. It is concluded that the asphalt mix durability is directly related to asphalt
film thickness. Therefore, the minimum VMA should be based on the
minimum desirable asphalt film thickness rather than minimum asphalt
content because the latter will be different for mixes with different gradation.
7. It is concluded that using (carbon fiber and lime) as additives, results; in
increase in tensile strength, more resistance to water action and reduce in the
133
permanent deformation for both types of mixes. Accordingly, this will

enhance the performance of asphalt concrete mixtures.
8. Mixes with TRZ gradation have higher values of tensile strength and creep as
compared with mixes of ARZ gradation.
9. Regarding the MICHPAVE–Finite Element Software results, the Superpave
mixes offer several advantages over the traditional Marshall design process.
The advantages include longer pavement life with less permanent
deformation, and less fatigue cracking. These results reflect a significant
reduction in maintenance and rehabilitation cost.
10. The results of PCPT Software show that both mixes have the approximately
same resistance to thermal cracking.
The main recommendations based on this work are summarized as follows:

1. Superpave mixes can be used instead of traditional Marshall mixes, since
Superpave mix design yields a more economical cost, it prescribes lower
asphalt content and longer life of pavements.
2. It is recommended to select (Filler / Asphalt) ratio carefully to be in an
adequate range within the specification to achieve easy and good
performance.
3. Using additives (Lime and Carbon fiber) to achieve performance of asphalt
paving mixtures, since they improve fatigue resistance, decrease permanent
deformation and have better moisture susceptibility of HMA.
134
The following can be listed as the recommendations for the future researches;
1. Similar studies should be conducted on a larger variety of aggregate
gradation and binder types to establish more robust confidence in Superpave
mix design criteria.
2. Further evaluation of the Superpave gyratory compactor, SGC, mixes
designed for different values of million ESALs could be performed to
evaluate the suitability of the current Superpave compactive effort
requirements for this design traffic mixes.
3. The gradation limits used in this study have produced acceptable mixes.
However, more extensive research is needed to verify the gradation control
point limits recommended from this research.
4. Field verification of different applications should be conducted to monitor
the performance of these Superpave mixtures.
5. A greater variety of mixes containing natural sand should be used to compare
Marshall and Superpave 12.5 mm NMAS mixes in order to clearly establish
the role of natural sand in mixture performance.
6. Since aggregate properties vary from source to source, the effect of
aggregate properties and sources on the performance of Superpave mixes is
recommended to be studied.
135
References
1. Abdul Haqh. A.AL.H., (January / 2000), " A suggested model for stress &
deflection analysis of multi-layer elastic flexible pavement systems using
finite element method " , thesis , M.Sc. University of Technology.
2. Amirkhanian S.N., (June,2001), "Development of a Gyratory Design for
Conventional SCDOT Hot Mixed Asphalt Mixtures" , FHWA-SC-01-
05.South Carolina Department of Transportation. Columbia S.C.
3. Anderson, D. ,(1987), "Guidelines for the Use of Dust in Hot-Mix-Asphalt
Concrete Mixtures" ,Proceedings of the Association of Asphalt Paving
Technologists, Vol. 56,pp.492-516.
4. Asphalt Institute, (1996), " Superpave mix design " , Superpave series
No.2 ( SP -2 ) .
5. ASTM, (1989), Annual Book of American Society for Testing and
Materials , Standard Section 4 , Vol. 04.03.
6. Baladi, G.Y.,(1989).“Fatigue life and permanent deformation
characteristics of asphalt concrete mixes,” Transportation Research
Record,1227,75–86.
7. Brown E.R., and M.S. Buchanan, (January ,2001), "Consolidation of the
Ndesign Compaction Matrix and Evaluation of Gyratory Compaction
Requirements", NCHRP Web Document 34 Project D9 -9 Contractor’s
Final Report, National Cooperative Highway Research Program.
8. Bruce A.C. et al , (June 1998), "An Asphalt Paving Tool For A adverse
Conditions " , University of Minnesota.
١٣٦
References
9. Campen W. H. , J.R. Smith , L. G .Erickson ,and L .R.Mertz, (1959),

“The Relationships Between Voids, Surface Area, Film Thickness and
Stability in Bituminous Paving Mixtures." Proceedings, Association of
Asphalt Paving Technologists, Vol. 28.
10. Chadbourn B.A. , et al, (1999), " The Effect of Voids in Mineral
Aggregate (VMA) on Hot-Mix Asphalt Pavements", Minnesota
Department of Transportation, St. Paul, Minnesota.
11. Chad W. Hawkins , P.E. (December , 2004)," Evaluation of the use of
gyratory compacted asphalt specimens for tensile strength ratio (TSR)
determination" ,South Carolina Department of Transportation .
12. Collins, R. et al., (January 1997)," Evaluation of the effect of aggregate
degradation on the specimens compacted by Superpave gyratory
compactor". Preprint for the Transportation Research Board 76th Annual
Meeting. Washington, D.C.
13. Cominsky R.,R. Leahy B. , and E.T. Harrigan , (1994) . " Level one mix
design : materials selection , compaction , and conditioning " ( SHRP-A-
408) ,Transportation Research Board , National Research Council,
Washington , DC.
14. D’Angelo, J.A., C. Paught, T.P.Harman, and J. Bukowski, (1995),
"Comparison of the Superpave Gyratory Compactor to the Marshall for
Field Quality Control", Journal, Association of Asphalt Paving
Technologist,Vol.64,pp.611-635.
15. Dallas Little and Jon Epps., (May 2001) ," The Benefits Of Hydrated
Lime In Hot Mix Asphalt", National Lime Association, www.lime.org.
١٣٧
References
16. " General Specification for Roads & Bridges " , (2003), Iraq State
Commission of Roads & Bridges , Department of planning & Studies.
17. Goode , J.F. and L.A. Lufsey , (1962) , " A new graphical chart for
evaluating aggregate gradations " proceedings , Association of Asphalt
paving technologists , Vol.31.
18. Habib A. , et al., (1998), “Comparison of Superpave and Marshall
Mixtures for Low-Volume Roads and Shoulders”, Transportation
Research Record 1609:Paper 98-0590,Washington,DC.
19. Harichandran, R.S. ,Baladi, G.Y., and Yeh, M-S. ,(1989), “Development
of a computer program for design of pavement systems consisting of
layers of bound and unbound materials,” Report No.FHWA-MI-RD-89-
02, Michigan Department of Transportation, Lansing, Michigan.
20. Harichandran, R.S .and Yeh, M-S. ,(1989,. “Flexible boundary in finite
element analysis of pavements” ,Transportation Research Record,
1207,50–60.
21. Harichandran, R.S. ,Yeh ,M-S., and Baladi , G.Y. ,(1990),
“MICHPAVE:A nonlinear finite element program for the analysis of
flexible pavements.” Transportation Research Record,1286,123–131.
22. Henny H., &André M., (2000),"Design and evaluation of a simulation
tool for the compaction process of asphalt pavements", Transportation
Research Board , , Washington , DC.
23. Heukelom, W., (1965), “The Role of Filler in Bituminous Mixtures,”
Asphalt Paving Technology, Proceedings: Association of Asphalt Paving
Technologists Technical Sessions, Vol.34,pp.396-429.
24. Johannson, L., (1998), “Bitumen Aging and Hydrated Lime,” Ph.D.
,Dissertation, Kungl Tekniska Högkolan, Royal Institute of Technology.
١٣٨
References
25. Jones, G.M., (May 22,1997), “The Effect of Hydrated Lime on Asphalt in
Bituminous Pavements,” NLA Meeting, Utah DOT.
26. John A . D' Angelo , " Superpave mix design tests methods and
requirements ", U.S. Federal Highway Administration.
27. John P.Zaniewski , David Diazgranados Diaz ,(April 2004), " Evaluation
Of 4.75 mm Superpave Mix Criteria For West Virginia" ,Asphalt
Technology Program, Department of Civil and Environmental
Engineering Morgantown, West Virginia .
28. John P.Zaniewski , Jason Nelson , (June , 2003)," Comparison of 9.5 mm
Superpave and Marshall wearing I mixes in west Virginia", Asphalt
Technology Program .
29. Kandhal P.S. and S. Chakraborty, (January 1996,) "Effect of Asphalt Film
Thickness on Short and Long Aging of Asphalt Paving Mixtures", NCAT
Report No.96-1, National Center for Asphalt Technology, Auburn, Al.
30. Kandhal, P.S,K.Y.Foo, and R. B. Mallick, (January 1998), "A Critical
Review of VMA Requirements in Superpave. " NCAT Report No.98-1,
National Center for Asphalt Technology, Auburn, Al.
31. Khaled Ksaibati and Jason Stephen ,(July 1998), " A preliminary
evaluation of Superpave one mix design procedure " University of
Wyoming.
32. Lesueur, D., Little, D.N. and Epps, J.A., (June 1998), “Effect of Hydrated
Lime on the Rheology , Fracture and Aging of Bitumen and Asphalt
Mixtures,” Paper Presented at Lhoist HMA Symposium, Dusseldorf,
Germany,.
33. Mansour S. , et al, (January 1999), " Evaluation of Superpave gyratory
compactors " , Center for Transportation Research .
١٣٩
References
34. M. Aren cleven , (2000), " Investigation of the properties of carbon fiber
modified asphalt mixtures ", Thesis, M.SC, Michigan Technological
University.
35. Maupin, G.W., (January 1998), "Comparison of Several Asphalt Design
Methods" ,Virginia Transportation Research Council, Charlottesville,
VA,.
36. Namir G.A. , (December , 2002), " The development of models for the
prediction of thermal cracking in flexible pavements", Ph.D, thesis ,
University of Baghdad.
37. Nijboer ,L.W,(1948), " Plasticity as a factor in the design of dense
bituminous road carpets . " Elsevier publishing company , Inc.
38. N.paul knosla and sachiyo kawaguchi, (2000) , " Comparative evaluation
of design and performance of Marshall and Superpave mixes "
Transportation Research Board annual meeting , Washington ,DC .
39. N. Paul K. " A design and evaluation tool for Superpave mixtures " ,2005-
14)., North Carolina State University .
40. Petersen, J.C., Plancher, H., and Harnsberger, P.M., (1987), “Lime
Treatment of Asphalt to Reduce Age Hardening and Improve Flow
Properties,” AAPT, Volume 56.
41. Powell, W.D .and Leech, D., (1983), "Compaction of bituminous road
materials using vibratory rollers", TRRL Laboratory Report 1102.
42. Prithvi S. Kandhal , L.Allen Cooley , (2002 ), "Coarse versus fine –
graded Superpave mixtures : comparative evaluation of resistance to
rutting ", National Center for Asphalt Technology .
43. Raj V.S. , N.Krishnamenon , and Peter E.Sebaaly, (2003), "Pavement
Response Evaluation Using Finite-Layer Approach ".
١٤٠
References
44. Randy C.West , Jingna Zhang and Allen Cooley.Jr. ,( July 2004) ,
"Evaluation Of The Asphalt Pavement Analyzer For Moisture
Sensitivity Testing " , National Center for Asphalt Technology, NCAT
Report 04-04.
45. "Resistance to plastic flow of bituminous mixtures using Marshall
apparatus" ,( 1993), AASHTO T245 .
46. Reyes C.H., (2003), "Evaluation of the Effect of Fines on Asphalt
Concrete", Masters , Thesis, West Virginia University, Morgantown,
WV.
47. Roberts , et al , T.W, (1996), " Hot mix asphalt materials , mixture design
and construction" , NAPA Research and Education Foundation , Lanham
, MD .
48. Rommel N. Y.,(2004), " Influence of avoiding the Superpave restricted
zone on asphalt concrete performance " ,M.Sc , Thesis , University of
Baghdad .
49. Ronald S.H. , Gilbert Y.B., (2001) ," Flexible Pavement Design in
Michigan: Transition from Empirical to Mechanistic Methods",
Transportation Research Board , , Washington ,DC .
50. Ronald S. Harichandran , Gilbert Y.Baladi , (January 2000),
"MICHPAVE User's Manual", Michigan State University,.
51. Santucci, L. E., and Schmidt, R.J.,(1962), “Setting Rate of Asphalt
Concrete,”, Highway Research Board,Washington,D.C.,1962,pp.1 9.
52. Serfass, J.P.; J. Samanos, (1996), “Fiber-Modified Asphalt Concrete
Characteristics, Applications and Behavior.” Journal of the Association
of Asphalt Paving Technologists, Vol.65,p 193-230.
١٤١
References
53. Shiou-San K., Hesham S.M, et al , (2003), "Development Of Flexible
Pavement Performance Prediction Model Based On Pavement Data",

TRB, Annual Meeting .
54. Stephen A. Cross, et al , (1999), , "Effects Of Gradation On
Performance Of Asphalt Mixtures ", University of Kansas, Transportation
Research Board 78th Annual Meeting January 10-14, Washington, D.C.
55. Stephen B.S, Nichols Consulting Engineers, Reno, Nevada , "Flexible
Pavement Design Summary of the State of the Art", A2B03:Committee on
Flexible Pavement Design Chairman .
56. Tarrer, Ray, (1996), “Use of Hydrated Lime to Reduce Hardening and
Stripping in Asphalt Mixes,” 4th Annual ICAR Symposium, Atlanta,
Georgia.
57. Teruhisa M., et al, (February 2004), "Material Properties For
Implementation Of Mechanistic –Empirical (M-E) Pavement Design
Procedures ", Ohio Research Institute for Transportation &the
Environment (ORITE).
58. Vasavi Kanneganti , (2002), " Comparison of 19mm Superpave and
Marshall base II mixes in west Virginia" ,Thesis , M.SC, department of
civil and environmental engineering.
59. Yoder, and Witczak , (1975), Principles of pavement design , 2nd
edition, John Wiley & Sons, Inc.
١٤٢
APPENDICES:
APPENDIX A: Data Analysis for Marshall Mixes Design
APPENDIX B: Data Analysis for Superpave Mixes Design
APPENDIX C: Marshall Test Results and Superpave
Mixes Analysis
APPENDIX D: Creep Test Results for Mixes Design
APPENDIX E: Indirect Tensile Strength Results for Mixes
Design
APPENDIX F: Moisture Damage Results for Mixes Design
APPENDIX G: Output Results of Application of
MICHPAVE Program
The measurements and calculations needed for a void analysis are:
The bulk specific gravity for the total aggregate is calculated using:
P 1  P 2  ......  PN
Gsb 
  ...... 
P1 P2 PN
G1 G 2 GN
where Gsb = bulk specific gravity for the total aggregate
P1, P2, PN = individual percentages by mass of aggregate
G1, G2, GN = individual bulk specific gravities of aggregate
The effective specific gravity of the aggregate, Gse, includes all void
spaces in the aggregate particles except those that absorb asphalt.
Gse is determined using :
Pmm  Pb
Gse 

Pmm Pb
Gmm Gb
where Gse = effective specific gravity of aggregate
Gmm = maximum specific gravity (ASTM D 2041 / AASHTO T 209)
Of paving mixture (no air void).
Pmm = percent by mass of total loose mixture = 100
Pb = asphalt content at which ASTM D 2041 / AASHTO T 209 test was
performed , percent by total mass of mixture .
Gb = specific gravity of asphalt
A1
After calculating the effective specific gravity of the aggregate from each
measured maximum specific gravity and averaging the G se results, the
maximum specific gravity for any other asphalt content can be obtained
using the equation shown below.
Gmm 
Pmm

Ps Pb
Gse Gb
where Gmm = maximum specific gravity of paving mixture ( no air
voids)
Pmm = percent by mass of total loose mixture = 100
Ps = aggregate content, percent by total mass of mixture
Pb = asphalt content, percent by total mass of mixture
Gse = effective specific gravity of aggregate
Absorption is expressed as percentage by mass of aggregate rather than as

a percentage by total mass of mixture.
Asphalt absorption, Pba is determined using:
Gse  Gsb
Pba  100 * * Gb
GsbGse
where
Pba = absorbed asphalt , percent by mass of aggregate
Gsb = bulk specific gravity of aggregate
A2
The effective asphalt content, Pbe of a paving mixture is the total asphalt
content minus the quantity of asphalt lost by absorption into the aggregate
particles.
The formula is:
Pbe  Pb 
Pba
* Ps
100
where
Pbe = effective asphalt content, percent by total mass of mixture
Pba = absorbed asphalt, percent by mass of aggregate
The VMA is calculated on the basis of the bulk specific gravity of the
aggregate and is expressed as a percentage of the bulk volume of the
compacted paving mixture. Therefore, the VMA can be calculated as:
VMA  100 
Gmb * Ps
Gsb
where
VMA = voids in mineral aggregate (percent of bulk volume)
Gsb = bulk specific gravity of total aggregate
Gmb = bulk specific gravity of compacted mixture (ASTM D 1188 or D
2726 / AASHTO T 166)
A3
The volume percentage of air voids in a compacted mixture can be
determined using:
Gmm  Gmb
Va  100 *
Gmm
where
Va = air voids in compacted mixture, percent of total volume
Gmm = maximum specific gravity of paving mixture
Gmb = bulk specific gravity of compacted mixture
The percentage of the voids in the mineral aggregate that are filled with
asphalt , VFA , not including the absorbed asphalt , is determined using :
VMA  Va
Vfa  100 *
VMA
where
VFA = voids filled with asphalt, percent of VMA
VMA = voids in mineral aggregate, percent of bulk volume
Va = air voids in compacted mixture, percent of total volume
A4
Superpave gyratory compaction is analyzed by computing the estimated
bulk specific gravity, corrected bulk specific gravity, and corrected
percentage of maximum theoretical specific gravity each desired
gyration. During compaction, the height is measured and recorded after
each gyration. Gmb of the compacted specimen and Gmm the loose
mixture are measured. An estimate of Gmb at any value of gyration is
made by dividing the mass of the mixture by the volume of the
compaction mold:
Gmb ( est ) 
w
Wm / Vmx
where
Gmb(est) = estimated bulk specific gravity of specimen during
compaction
Wm = mass of specimen, grams
γw = density of water = 1 g/cm3
Vmx = volume of compaction mold (cm3) calculated using the equation
Vmx = 17.6715 hx
where
hx = height of specimen in mold during compaction (mm).
The estimated Gmb is corrected by a ratio of the measured to estimated
bulk specific gravity:
C
Gmb ( measured )
Gmb ( estimated )
where
C = correction factor
Gmb (measured) = measured bulk specific gravity after Ndes
Gmb (estimated) = estimated bulk specific gravity at Ndes
B1
The estimated Gmb at any other gyration level is then determined using:
Gmb (corrected) = C * Gmb (estimated)
where
Gmb corrected = corrected bulk specific gravity of the specimen at Ndes
C = correction factor
Gmb (estimated) = estimated bulk specific gravity at Ndes
Then calculate the percent Gmm at Ndes as the ratio of Gmb (corrected)
to Gmm (measured).
The percentage of air voids at Ndes is determined from the equation:
Va = 100 - %Gmm@Ndes
where
Va = air voids @ Ndes, percent of total volume
%Gmm @Ndes = maximum theoretical specific gravity @ Ndes, percent
The percent voids in the mineral aggregate is calculated using:
%VMA 100 (
%Gmm@ Ndes* Gmm* Ps
)
Gsb
where
VMA = voids in mineral aggregate, percent of bulk volume
%Gmm@Ndes = maximum theoretical specific gravity @ Ndes percent
Gmm = maximum theoretical specific gravity
Gsb = bulk specific gravity of total aggregate
Ps = aggregate content, cm3/ cm3, by total mass of mixture.
If the percentage of air voids is equal to four percent, then this data is
compared to the volumetric criteria and an analysis of this specimen is
completed. However , if the air void content at Ndes varies from four
percent ( and this will typically be the case ) , an estimated design asphalt
content to achieve 4 percent air voids at Ndes is determined , and the
B2
estimated design properties at this estimated design asphalt content are
calculated .
The estimated asphalt content at Ndes = four percent air voids is
calculated using this equation:
Pb, estimated = Pbi – (0.4 * (4-VA))
Where Pb, estimated = estimated asphalt content, percent by mass of
mixture
Pbi = initial (trial) asphalt content, percent by mass of mixture
Va = percent air voids at Ndes (trial)
The volumetric (VMA and VFA) at Ndes and mixture density at Nini and
Nmax are then estimated at this asphalt binder content using the
equations that follow.
For VMA:
%VMA estimated = %VMA initial + C *(4-VA)
where
%VMA initial = %VMA from trial asphalt binder content
C = constant = 0.1 if Va is less than 4 percent
= 0.2 if Va is greater than 4 percent
For VFA :
(% VMAest  4 )
% VFA ( est )  100 *
% VMA
For %Gmm at Nini:
%Gmm estimated @ Nini = %Gmm trial @ Nini – ( 4-Va )
The maximum allowable mixture density at Nini is 89 percent
For % Gmm at Nmax :
%Gmm estimated @ Nmax = % Gmm trial @ Nmax – (4-Va)
The maximum allowable mixture density at Nmax is 98 percent
B3
The effective asphalt content is calculated using:
(Gse  Gsb)
Pbe  (Ps * Gb) *  Pbest
Gse* Gsb
Where
Pbe = effective asphalt content, percent by total mass of mixture
Gsb = bulk specific gravity of aggregate
Dust proportion is calculated using:
DP 
P 0 . 075
Pbe
where
P0.075 = aggregate content passing the 0.075 mm sieve , percent by mass
of aggregate
Pbe = effective asphalt content , percent by total mass of mixture An
acceptable dust proportion ranges from 0.6 to 1.2 for all mixtures .
B4
Table (C-1) Data Analysis for Marshall Specimens and Superpave Specimens.
Physical properties of gradation R1 Physical properties of gradation R9
Ga)bulk 2.623 Ga)bulk 2.6153
Ga)apparent 2.683 Ga)apparent 2.689
G filler 3.12 G filler 3.12
Gse 2.624 Gse 2.616
Gmm 2.449 Gmm 2.445
Gc)bulk 2.518 Gc)bulk 2.5189
Gf)bulk 2.615 Gf)bulk 2.622
Gc)apparent 2.553 Gc)apparent 2.554
Gf)apparent 2.662 Gf)apparent 2.689
MARSHALL SPECIMENS (TRZ) MARSHALL SPECIMENS (ARZ)
A.C% 4.70 A.C% 4.63
VMA% 14.5 VMA% 14.85
VTM% 3.9 VTM% 4.49
VFA% 73.1 VFA% 69.76
density 2.353 density 2.335
SUPERPAVE SPECIMENS (TRZ) SUPERPAVE SPECIMENS (ARZ)
A.C% 4.42 A.C% 4.54
VMA% 14 VMA% 14.32
VFA% 71.42 VFA% 72
density 2.362 density 2.349
C1
Table (C-2), Marshall Test for Mix Design.
Marshall Test for Marshall Specimens
Gradation Stability KN Flow mm
R1 11.5 11.8 12 4.4 4.9 5
R9 13 12.5 13.5 2.3 2.2 2.5
Marshall Test for Superpave Specimens
Gradation Stability KN Flow mm
R1 13.8 14.16 14.4 4.2 4.5 4.1
R9 15.7 15.1 16.3 2.2 2.1 2
C2
Table (C-3), Asphalt Film Thickness for Marshall Mixes.
R1 R9 Surface Area Surface Surface
SIEVE PASSING PASSING Factor area R1 area R9
19 100 100 0.41 0.41 0.41
12.5 92 89.5
10 83.1 77.8
4.74 66.9 55.3 0.41 0.27429 0.22673
2 41.5 40.2 0.82 0.3403 0.32964
1.18 28.2 32 1.64 0.46248 0.5248
0.6 21.4 25 2.87 0.61418 0.7175
0.25 14.4 15.1 6.14 0.88416 0.92714
0.15 11.6 12.2 12.29 1.42564 1.49938
0.075 9.8 9.8 32.77 3.21146 3.21146
7.62251 7.84665
MARSHALL SPECIMENS
Gradation R1 Gradation R9
A.C 4.7 A.C 4.63
VMA 14.5 VMA 14.85
VFA 73.1 VFA 69.76
VTM 3.9 VTM 4.49
Vol. AC 10.6 Vol. AC 10.36
Weight of
AC 110.24 Weight of AC 107.744
Weight of weight of
aggregate 2235.29191 aggregate 2219.340233
Weight of Weight of AC/
AC/aggregate 0.04931794 aggregate 0.048547761
Asphalt film 6.22119E-06 Asphalt film 5.9491E-06
Thickness µ Thickness µ
C3
Table (C-4) Asphalt Film Thickness for Superpave Mixes.
R1 R9 Surface Surface Surface
SIEVE PASSING PASSING Area Factor area R1 area R9
19 100 100 0.41 0.41 0.41
12.5 92 89.5
10 83.1 77.8
4.74 66.9 55.3 0.41 0.27429 0.22673
2 41.5 40.2 0.82 0.3403 0.32964
1.18 28.2 32 1.64 0.46248 0.5248
0.6 21.4 25 2.87 0.61418 0.7175
0.25 14.4 15.1 6.14 0.88416 0.92714
0.15 11.6 12.2 12.29 1.42564 1.49938
0.075 9.8 9.8 32.77 3.21146 3.21146
7.62251 7.84665
SUPERPAVE SPECIMENS
Gradation R1 Gradation R9
A.C 4.42 A.C 4.54
VMA 14 VMA 14.32
VFA 71.42 VFA 72
VTM 4 VTM 4
Vol. AC 10 Vol. AC 10.32
Weight of
AC 104 Weight of AC 107.328
Weight of Weight of
aggregate 2248.94118 aggregate 2256.724863
Weight of Weight of AC
AC/aggregate 0.04624398 /aggregate 0.047559187
Asphalt Film 5.8334E-06 Asphalt Film 5.82796E-06
Thickness µ Thickness µ
C4
Table (C-5), Data analysis for Superpave specimens at initial asphalt
content for gradation R1.
Gradation R1
Data Analysis of Superpave Specimens
Gse 2.624
Gsb 2.623
ASPHALT INITAIL 4.7
Gmax 2.449
DENSITY 2.3613
WEIGHT 4478
HEIGHT@ N=9 129.2
HEIGHT@ N=135 116
HEIGHT@ N=220 114.8
VMX @Ndes 2049.9
VMX@ Nini 2283.2
GMB Ees.@ Nini 1.9613
GMB Ees.@ Ndes 2.1845
C FACTOR 1.0809
GMB CORR.@ Ndes 2.3613
GMB CORR.@ Nini 2.1201
%GMM @ Ndes 0.9642
%GMM@ Nini 0.8657
VOID AIR 3.5803
%VMA 14.208
PB Ees. 4.5321
%VMA Ees. 14.25
%VFA Ees. 71.929
%GMM Ees.@ Nini 86.149
PBE Effective 4.5177
density@ Nmax 2.386
air void @ Nmax 2.5725
Gmm@ Nmax 97.428
C5
Table (C-6), Data analysis for Superpave specimens at initial asphalt content for
gradation R9.
Gradation R9
Data Analysis of Superpave Specimens
Gse 2.6165
Gsb 2.6153
ASPHALT INITAIL 4.63
Gmax 2.445
DENSITY 2.3496
WEIGHT 4582
HEIGHT@N=9 129.2
HEIGHT@N=135 115.6
HEIGHT@N=220 114.8
VMX @Ndes 2042.8
VMX @Nini 2283.2
GMB EST @Nini 2.0069
GMB EST @Ndes 2.243
C FACTOR 1.0476
GMB CORR.@ Ndes 2.3496
GMB CORR.@ Nini 2.1023
%GMM @Ndes 0.961
%GMM@ Nini 0.8598
VOID AIR 3.9
%VMA 14.317
PB EST 4.59
%VMA EST 14.327
%VFA EST 72.082
%GMM EST @Nini 85.884
Pbe Effective 4.5726
density@ Nmax 2.375
air void Nmax 2.863
Gmm Nmax 97.137
C6
Table (C-7), Data Analysis for Superpave Specimens at ±0.5% and +1% for {(4.532) estimated asphalt content}
SUPERPAVE DATA ANALYSIS FOR R1

ASPHALT GMM Gmm dust effective Gmm
%VMA %VFA CONTENT air void @Ndes GMM density @Nmax proportion asphalt @Nini
14.54945 62.19788 4.032 5.5 94.5 2.471467 2.335537 0.9646364 2.439328 4.017499 0.857702
13.81365 74.66275 4.532 3.5 96.5 2.45387 2.367985 0.9750871 2.169306 4.517575 0.867752
14.43009 79.21011 5.032 3 97 2.436522 2.363426 0.9827074 1.953105 5.01765 0.879063
15.04248 83.3804 5.532 2.5 97.5 2.419417 2.358932 0.9844161 1.776094 5.517725 0.879412
Table (C-8), Data Analysis for Superpave Specimen at ±0.5% and +1% for {(4.59) estimated asphalt content}
SUPERPAVE DATA ANALYSIS FOR R9

ASPHALT air Gmm GMM absorbed effective dust Gmm
%VMA %VFA CONTENT void Ndes @Nmax GMM density asphalt asphalt proportion @Nini
14.92381 61.13592 4.09 5.8 94.2 0.961574 2.462718 2.319881 0.0182378 4.072508 2.406379 0.860944
14.18248 73.20638 4.59 3.8 96.2 0.9720557 2.44528 2.352359 0.0182378 4.572599 2.143201 0.871813
14.61565 78.78986 5.09 3.1 96.9 0.9816943 2.428086 2.352815 0.0182378 5.07269 1.931914 0.878843
15.04923 84.05234 5.59 2.4 97.6 0.9854258 2.411132 2.353265 0.0182378 5.572782 1.758547 0.880314
C7
‫وزارة اﻟﺘﻌﻠﯿﻢ اﻟﻌﺎﻟﻲ واﻟﺒﺤﺚ اﻟﻌﻠﻤﻲ‬
‫اﻟﺠﺎﻣﻌﺔ اﻟﻤﺴﺘﻨﺼﺮﯾﺔ‬
‫ﻛﻠﯿﺔ اﻟﮭﻨﺪﺳﺔ‬
‫ﻗﺴﻢ ھﻨﺪﺳﺔ اﻟﻄﺮق واﻟﻨﻘﻞ‬
‫ﻣﻘﺎرﻧﮫ ﺗﻘﯿﻤﯿﮫ ﻟﻤﺴﺘﻮى أداء ﻣﻮاد اﻟﺘﺒﻠﯿﻂ ﺑﺎﺳﺘﺨﺪام‬

‫طﺮق اﻟﺮص ﻣﺎرﺷﺎل و اﻟﺘﺒﻠﯿﻂ اﻟﻔﺎﺋﻖ‬
‫رﺳﺎﻟﺔ ﻣﻘﺪﻣﮫ إﻟﻰ‬

‫ﻗﺴﻢ ھﻨﺪﺳﺔ اﻟﻄﺮق واﻟﻨﻘﻞ‬
‫ﻛﻠﯿﺔ اﻟﮭﻨﺪﺳﺔ‪ /‬اﻟﺠﺎﻣﻌﺔ اﻟﻤﺴﺘﻨﺼﺮﯾﺔ‬
‫ﻛﺠﺰء ﻣﻦ أﻛﻤﺎل ﻣﺘﻄﻠﺒﺎت درﺟﮫ اﻟﻤﺎﺟﺴﺘﯿﺮ‬
‫ﻓﻲ ﻋﻠﻮم ھﻨﺪﺳﺔ اﻟﻄﺮق واﻟﻨﻘﻞ‬
‫ﻣﻦ ﻗﺒﻞ‬
‫ﻧﻮر ﻣﻌﺘﺰ إﺳﻤﺎﻋﯿﻞ اﻟﻌﺰاوي‬
‫ﺑﻜﺎﻟﻮرﯾﻮس ھﻨﺪﺳﺔ اﻟﺒﻨﺎء واﻹﻧﺸﺎءات‬
‫‪2003‬‬
‫ﺑﺈﺷﺮاف‬
‫اﻟﺪﻛﺘﻮر ﻧﻤﯿﺮ ﻏﻨﻲ اﺣﻤﺪ‬
‫ﺷﺒﺎط ‪2006 /‬‬ ‫ﻣﺤﺮم‪1426/‬‬

Comparative Evaluation For The Performan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparative Evaluation For The Performan

Uploaded by

Copyright:

Available Formats

See

COMPARATIVE EVALUATION FOR THE

Thesis · July 2006

The user has requested enhancement of the downloaded file.

COMPARATIVE EVALUATION FOR THE

MOHARRAM, 1426 FEBRUARY, 2006

‫) اﻵﻳﺔ ‪ ،109‬ﺳﻮرة اﻟﻜﻬﻒ(‬

‫‪ ‬أ‪ ฀‬ﻳﻨﺒﻮع اﻟﺤﻨﺎن واﻟﻌﻄﻒ ‪ ...............‬واﻟﺪﺗﻲ اﻟﻜﺮ ﻳﻤﻪ‪...‬‬

‫اﻟﺘﻲ راﻓﻘﺖ درﺑﻲ اﻟﻄﻮﻳﻞ ‪......‬إﺟﻼﻻ وﻋﺮﻓـﺎﻧﺎ ﺑﺎﻟﺠﻤﻴﻞ‬

‫‪ ‬أ‪ ฀‬ﻣﻦ ﻏﻤﺮوﻧﻲ ﺑﺎﺣﺘﺮاﻣﻬﻢ وﺣﺒﻬﻢ ‪ ...........‬أﺧﻮاﺗﻲ ‪...‬‬

‫إﻛﺮاﻣﺎ واﻋﺘﺰازا ﺑﻬﻢ‪.........‬‬

I certify that the preparation of this thesis entitled " Comparative

In view of the available recommendations, I forward this thesis for debate

Approved by the Dean the College of Engineering

Chapter One : Introduction

Chapter Two: Literature Review

Chapter Three: Materials and Methods of Testing

Chapter Four: Results and Discussion

Chapter Six: Conclusions and Recommendations

(2.2) Superpave Coarse Aggregate Angularity Requirements 21

This technology has a tremendous potential to be implemented in Iraq, which will

Superpave technology as a new design methodology can be rigorously used under

Appendix A presents a detailed information about data analysis for Marshall

Appendix B presents a detailed information about data analysis for Superpave

Appendix D contains the Creep test results for mixes design.

Appendix F presents the moisture damage results for mixes design.

Appendix G presents the results of MICHPAVE program application.

Aggregate gradation HMA is graded by the percentages of different – size aggregate

Marshall Stability is defined as the maximum load carried by a compacted specimen

Superpave is an acronym for Superior Performing Asphalt Pavements. It is the

The Superpave gyratory compactor (SGC), as shown in Figure (2-1) is used in

Figure (2-1) Superpave Gyratory Compactor

The SGC consists of the following components:

Figure (2-2) Mold Configuration

Figure (2-3) Densification plot

excessive rutting. N maximum is calculated using N design in the following

Table 2-1 Superpave Design Gyratory Compactive Effort

Design Average Design High Air Temperature

1-3 8 105 167

3-10 9 119 192

10-30 9 135 220

30-100 10 153 253

>100 10 172 288

Figure (2-4) Superpave Mix Design

Coarse aggregate angularity is defined as the percent by weight of aggregates larger

Superpave Coarse Aggregate Angularity Requirements

< 100 mm > 100 mm

<0.3 55/- -/-

0.3 to < 3 75/- 50/-

3 to < 10 85/80 60/-

10 to < 30 95/90 80/75

Table (2-3) Superpave Fine Aggregate Angularity Requirements

Superpave Fine Aggregate Angularity Requirements

< 100 mm > 100 mm

Note: Criteria are presented as percent air voids in loosely

Superpave Clay Content Requirements

Traffic , million ESALs Sand equivalent ,

Superpave Flat Elongated Particle Requirements

Traffic , million ESALs Percent , maximum

Note: Criteria are presented as maximum percent by weight of flat

To specify gradation, Superpave uses a modification of an approach already used by

Figure (2-5) 0.45 power gradation

The maximum density gradation represents a gradation in which the aggregate