Impact of Hot Mix Asphalt Properties on its
Permanent Deformation Behaviour
Author
Imran Hafeez
05-UET/PhD-CE-22
Supervisor
Dr. Mumtaz Ahmed Kamal
Professor, Department of Civil Engineering
DEPARTMENT OF CIVIL ENGINEERING
FACULTY OF CIVIL & ENVIRONMENTAL ENGINEERING
UNIVERSITY OF ENGINEERING AND TECHNOLOGY
TAXILA
October 2009
Impact of Hot Mix Asphalt Properties on its Permanent
Deformation Behaviour
Author
Imran Hafeez
05-UET/PhD-CE-22
A thesis submitted in partial fulfillment of the requirement for the degree of
PhD Civil Engineering
Thesis Supervisor
Dr. Mumtaz Ahmed Kamal
Professor, Department of Civil Engineering
Thesis Supervisor’s Signature: __________________________________________
External Examiner’s Signature:
Prof. Dr. Aziz Akbar
CED, UET, Lahore
External Examiner’s Signature
Dr. Shahab Khanzada
NHA, Islamabad
DEPARTMENT OF CIVIL ENGINEERING
FACULTY OF CIVIL & ENVIRONMENTAL ENGINEERING
UNIVERSITY OF ENGINEERING AND TECHNOLOGY, TAXILA
October 2009
II
Abstract
Over the past twenty years; road traffic (both passenger and freight) has grown significantly
and loading is progressively getting worse due to the introduction of newer and more
powerful trucks with heavier wider bodies in Pakistan. Consequently, premature rutting in
the form of shear flow in flexible pavements has been observed during high ambient
temperatures. National Highway Authority (NHA), Pakistan has been facing serious threats
like, frequent pavement failures, poor riding quality and high maintenance cost.
Modifications of Asphalt cement with polymers and rigid pavement design trends, being the
end solutions, have increased the construction cost, even more than four times than that of
conventional design. In order to cater for the growing axle load demand and to increase the
performance of asphalt concrete mixes, true prediction and accurate estimation of probable
behavior of mixes need to be investigated.
A comprehensive laboratory study was carried out using NHA aggregate gradation Class “A”
for asphaltic wearing course, which has commonly been used in the field. The main objective
of the research work was to evaluate the effects of temperature and loading on the permanent
deformation behavior of mixes, designed with the same aggregate gradation and three
commonly available asphalt cement types (A.C). Two gradations i.e. “01” and “02”, within
the envelope of the same gradation, were chosen for this study. Three asphalt cement (A.C.)
types i.e. two neat A.C with penetration grade “60/70”, “40/50” and one modified A.C. (base
“60/70” pen. grade with Elvaloy Terploymer) were selected. Six mixes ranging from finer to
coarser aggregate gradation were therefore designed at optimum filler contents, in order to
get better mix cohesion, resistant to rutting and to improve serviceability. Three percentages
of mineral fillers (i.e. 2.4%, 3.4%. & 4.4%) were trialed in order to determine the optimum
filler content for asphalt mixes.
Mix design properties i.e. optimum asphalt content, percentage air voids, voids in mineral
aggregates, voids filled with asphalt, flow and stability were determined using Marshall
Method of Mix Design. Asphalt cement consistency i.e. penetration grade, ductility,
softening point and rheological properties i.e. phase angle & complex shear modulus were
III
also measured. The temperature influence on rheological properties were determined at
constant frequency of 10 Hz using Dynamic Shear Rheometer.
Two performance tests i.e. uniaxial repeated loading strain & wheel tracking tests were
chosen to measure the permanent deformation of asphalt mixes. Relationship between rutting
and factors effecting rutting were developed. Domains of intercept and slope coefficient and
permanent deformation coefficients alpha (α) and mu (µ) were also determined. New
relationships have been developed to predict the impact of Hot Mix Asphalt properties on its
permanent deformation behaviour. It has been revealed that mixes with coarser gradation and
polymer modified asphalt showed better resistance against rutting. Permanent deformation
coefficients, such as α and µ indicate influence of mix properties on mix rut potential. It was
further observed that temperature has no effect on α but influences significantly on µ.
This study presents relationships developed to correlate temperature and rut values of mixes,
tested under wheel tracker (WT). It also presents a comparison between uni-axial repeated
load strain tests (or repeated creep test) and wheel tracker tests in terms of regression
constants, shift factors and a correlation between both the test methods. Intercept coefficient
varies in a narrow range. Based on intercept coefficient, uniaxial repeated load strain test
(repeated creep test) did not provide clear ranking of mixes at specified temperatures and
stress conditions. The slope of data line between both the tests reduces with increase in
temperature and stress levels, which means permanent deformation increases with an
increase in temperature and stress. An average value has been determined for approximate
shift of creep test data to WT test data.
Furthermore a statistical estimation of Mechanistic-Empirical Model has been formulated
that relates plastic to elastic strain ratio with different variable that influenced in the mix rut
development. Using 54 variables, comprising six mixes designed at low asphalt contents,
three temperatures (25, 40 and 550C) and three stress levels (100, 300 and 500 kPa) under
repeated load test, a mathematical model was developed to assess the magnitude of plastic to
elastic strain ratio. It was observed that plastic to elastic strain ratio is a function of number
of load repetitions, temperature, stress levels, shear complex modulus of asphalt cement and
IV
aggregate gradations respectively. Despite few limitations i.e. single source of aggregate,
single method for mix design, two test methods, certain parameters that help to estimate the
permanent deformation in the asphalt layers of flexible pavements were successfully
captured in the model.
Keywords: Flexible Pavement, Hot Mix Asphalt, Uniaxial Repeated Loading, Wheel Tracker,
Permanent Deformation
V
Undertaking
I certify that research work titled “Impact of Hot Mix Asphalt Properties on its Permanent
Deformation Behaviour” is my own work. The work has not been presented elsewhere for
assessment. Where material has been used from other sources it has been properly
acknowledged / referred.
Signature of Scholar
Imran Hafeez
05-UET/PhD-CE-22
VI
Acknowledgement
I would like to thank those individuals who were instrumental in the completion of this
research. Specially, I would like to thank my supervisor, Professor Dr. Mumtaz Ahmed
Kamal, for willingly sharing his knowledge, guidance, encouragement and patience. His
support and enthusiasm are greatly appreciated. I look forward to future endeavors.
I would also like to thank Professor Dr. Waseem Mirza for sharing his experience and for his
advice and fruitful comments. Special thanks to Professor Dr. Tanveer Iqbal Qayyum for his
help and advice. Thanks to my examining committee members, Professor Dr. Hashim Nisar
and Prof. Dr. Qaiser Uz-Zman.
My special thanks to Taxila Institute of Transportation Engineering staff for their support and
help in conducting this research and their patience and understanding. I would like to express
my appreciation to persons working in Transportation Engineering Laboratory.
I would like to give my deepest thanks to my wife, my daughter and my son for their
understanding and unlimited patience. Without their support this work would have not been
possible.
VII
Table of Contents
Abstract
III
Undertaking
VI
Acknowledgement
VII
VIII
Table of contents
List of Figures
XII
List of Tables
XV
List of Frequently used symbols and abbreviations
Chapter One:
XVII
Introduction
1.1
Background
02
1.2
Problem Statement
03
1.3
Research Objectives
03
1.4
Research Methodology
04
1.5
Organization
06
Chapter Two:
Literature Review
2.1
Introduction
09
2.2
Permanent deformation (rutting) of Asphalt Mixes
09
2.3
Influence of Asphalt Mixture Properties on Rutting
12
2.4
Influence of Aggregates Properties on Rutting
13
2.5
Influence of Mineral Filler’s Properties on Rutting
15
2.6
Influence of Asphalt Cement Properties on Rutting
16
2.7
Stress Strain Behavior of Asphalt Materials
21
2.8
Resistance to Permanent Deformation of HMA
24
2.9
Permanent Deformation Prediction Models
27
VIII
Chapter Three:
Material Characterization
3.1
Introduction
34
3.2
Aggregates for Asphalt Mixtures
34
3.2.1
Size and Grading
35
3.2.2
Particle Shape
36
3.2.3
Absorption
38
3.2.4
Clay Content (Sand Equivalent)
38
3.2.5
Toughness
39
3.2.6
Soundness
39
3.2.7
Deleterious Materials
40
3.3
3.4
Chapter Four:
Characteristics of Asphalt Cement
41
3.3.1
Consistency of Asphalt Cement
41
3.3.2
Elastic Recovery
43
3.3.3
Torsional Recovery
44
Dynamic Shear Rheometer (DSR)
45
Mix Design Methods
4.1
Introduction
50
4.2
Combined Grading of Aggregates
50
4.3
Asphalt Mixtures
52
4.4
Mix Design Properties
52
4.5
Mineral Filler Optimization
54
4.6
Asphalt Mixtures Volumetric
54
4.6.1
Specific Gravity
55
4.6.2
Voids in Mineral Aggregates, VMA
56
4.6.3
Air Voids, Va
57
4.6.4
Voids Filled with Asphalt, VFA
57
IX
Chapter Five:
Uniaxial Repeated Load Strain Test (URLST)
5.1
Introduction
60
5.2
Universal Testing Machine (UTM-5P)
60
5.3
Uniaxial Repeated Load Strain Test (URLST)
61
5.4
Test Conditions
62
5.4.1
Load Conditions
62
5.4.2
Temperature Conditions
62
5.5
Testing Methodology
62
5.6
Discussion of URLST Results
63
5.7
Resilient Strain
65
5.8
Summary of Results
70
Chapter Six:
Wheel Tracking Test
6.1
Introduction
72
6.2
Wheel Tracking Device
72
6.3
Specimen Preparation on Roller Compactor
75
6.4
Discussion of Results
78
Chapter Seven:
Results and Discussions
7.1
Introduction
83
7.2
Regression Coefficients
83
7.2.1
Intercept Coefficient (a)
83
7.2.2
Slope Coefficient (b)
85
7.3
Permanent Deformation Coefficient
86
7.3.1
Alpha (α)
86
7.3.2
Mu (µ)
88
7.4
Comparisons of Results to Other Researchers
90
7.5
Regression Analysis
92
7.6
Ranking of Mixes using Intercept Coefficient
93
7.7
Shift Factor Computations
94
7.8
Correlations Between URLST and WT Test
95
7.9
Summary
100
X
Chapter Eight:
Modeling the Permanent Deformation Properties of Hot
Mix Asphalt
8.1
Introduction
102
8.2
Data Analysis and Model Development
103
8.2.1
General Trends of Variables
103
8.2.2
Development of Prediction Model
106
8.2.3
Comparison of Computed Model with
108
MEPDG Model
8.2.4
Chapter Nine:
Sensitivity Analysis
109
Conclusions and Recommendations
9.1
Introduction
112
9.1
Conclusions
112
9.2
Recommendations for Future Study
113
REFERENCES
114
ANNEXURES
126
Annexure –A: Plots of Accumulative Strain
128
Annexure –B: Trends of Accumulated Strains in Mixes
137
Annexure –C: Plots of Resilient Strain
146
Annexure –D: Influence of Wheel Tracker Load Cycle on Rut Depth
149
XI
List of Figures
Fig. No. Description
Page No.
1.1
Flow Chart Diagram (Scope of Work)
5
2.1
Accumulated Plastic Strains in Pavement
9
2.2
Typical Repeated Load Permanent Deformation Behaviors of
11
Pavement Materials
2.3
Rutting Caused by Shear Displacement of Weak Asphalt Layer
12
2.4
Temperature Shift Behavior of Asphalt Binder
17
2.5
Microscopic View of Liquid Flow Properties
18
2.6
Stress Strain Behavior of Bituminous Material
19
2.7
Shear Loading Behavior of Aggregate
22
2.8
Contrasting Stone Skeletons
22
2.9
Permanent Strain under Load Pulse
23
2.10
Accumulated Permanent Strain Under Repeated Load
23
2.11
26
2.12
Effect of Aggregate and Mortar on the failure Behavior of
Asphalt Mixtures
Log-Log Form of Power Model
29
3.1
Aggregate Gradation ‘01’ and ‘02’ under NHA Class ‘A’
36
3.2
Briquette Specimens
43
3.3
Trimming of Specimens
43
3.4
Torsional Recovery Test Apparatus
44
3.5
Dynamic Shear Rheometer
46
3.6
Test Specimen on DSR
46
3.7
Influence of Temperature on Linear Visco-Elastic Properties
48
5.1
Universal Testing Machines (UTM-5P)
61
5.2
Influence of Temperature and Stress Levels on Accumulative Strain
64
5.3
Influence of Stress Levels on Resilient Strain, Mix 1a
67
5.4
Influence of Stress Levels on Resilient Strain, Mix 1b
67
5.5
Influence of Stress Levels on Resilient Strain, Mix 1c
68
5.6
Influence of stress levels on resilient strain, Mix 2a
68
XII
Fig. No. Description
Page No.
5.7
Influence of Stress Levels on Resilient Strain, Mix 2b
69
5.8
Influence of Stress Levels on Resilient Strain, Mix 2c
69
6.1
Wheel Tracker
72
6.2
Wheel Tracker Solid Rubber Tyre
73
6.3
Wheel Tracker out put Display
74
6.4
Slab Compaction on Roller Compacter
75
6.5
Rut Development in Mixes
77
6.6
Relationship of Load Cycle & Rut Depth, Mix 1a
78
6.7
Relationship of Load Cycle & Rut Depth, Mix 1b
79
6.8
Relationship of Load Cycle & Rut Depth, Mix 1c
79
6.9
Relationship of Load Cycle & Rut Depth, Mix 2a
80
6.10
Relationship of Load Cycle & Rut Depth, Mix 2b
80
6.11
Relationship of Load Cycle & Rut Depth, Mix 2c
81
7.1
Influence of Temperature and Stress Levels on Alpha
87
7.2
Influence of Temperature and Stress Levels on Mu
89
7.3
Domains of Alpha
91
7.4
Domains of Mu
91
7.5
Influence of Test Type on Intercept Coefficient
92
7.6
Shift of URLST Data to Wheel Tracker Data
94
7.7
Correlations between URLST and WT Test Data at 250C & 100kPa
95
7.8
Correlations between URLST and WT Test Data at 250C & 300kPa
96
7.9
Correlations between URLST and WT Test Data at 250C & 500kPa
96
7.10
Correlations between URLST and WT Test Data at 400C & 100kPa
97
7.11
Correlations between URLST and WT Test Data at 400C & 300kPa
97
0
7.12
Correlations between URLST and WT Test Data at 40 C & 500kPa
98
7.13
Correlations between URLST and WT Test Data at 550C & 100kPa
98
7.14
Correlations between URLST and WT Test Data at 550C & 300kPa
99
7.15
Correlations between URLST and WT Test Data at 550C & 500kPa
99
8.1
General Trends of Plastic Strains
104
XIII
Fig. No. Description
Page No.
8.2
General Trends of Elastic Strains
105
8.3
Measured Versus Predicted εp/εr, 10 in/in
107
8.4
MEPDG Versus Computed Model εp/εr, 106 in/in
108
8.5
Sensitivity of εp/εr Model to Predictor Variables
110
6
XIV
List of Tables
Table
Description
Page No.
No.
3.1
Adopted Aggregate Gradations & Specifications
35
3.2
Physical Properties Of Aggregates
40
3.3
Consistency of Different Types of Ac
42
3.4
Specifications of Asphalt Cement
42
3.5
Elastic Recovery Test Results
44
3.6
Torsional Recovery Test Results
45
3.7
Rheological Properties of Binders
47
4.1
Summary of Sieve Analysis Results
51
4.2
Adopted Gradations & Specifications
51
4.3
Selection of Mix Types
52
4.4
Hot Mix Asphalt Design Properties
53
4.5
Mineral Filler to Asphalt Cement Ratio
54
4.6
Hot Mix Asphalt Design Volumetric Properties
55
5.1
Specimens at Each Test Condition
63
5.2
Mean Accumulated Strain (εp) of Mixes
63
5.3
Resilient Strain of Mixes
66
6.1
Details of Specimens
75
6.2
Compaction Pressures
76
6.3
Rut depth of Mixes Measured on Wheel tracker
76
6.4
Intercept Coefficient “a” of Mixes
81
6.5
Slope Coefficient “b” of Mixes
81
7.1
Intercept Coefficient ‘a’ at 250C
84
7.2
Intercept Coefficient ‘a’ at 400C
84
7.3
Intercept Coefficient ‘a’ at 550C
84
7.4
Slope Coefficient ‘b’ at 250C
85
0
7.5
Slope Coefficient "b” at 40 C
85
7.6
Slope Coefficient "b” at 550C
85
XV
Table
Description
Page No.
No.
7.7
Permanent Deformation Parameter "α” at 250C
86
7.8
Permanent Deformation Parameter "α” at 400C
86
7.9
Permanent Deformation Parameter "α” at 550C
86
7.10
Permanent Deformation Parameter “µ” at 250C
88
7.11
Permanent Deformation Parameter “µ” at 400C
88
7.12
Permanent Deformation Parameter “µ” at 550C
88
7.13
Ranking of Mixes for Wheel Tracker
93
7.14
Ranking of Mixes for URLST
93
8.1
Influence of Test Conditions and Mix Parameters on Strain
106
XVI
List of Frequently used symbols and abbreviations
Symbols
Meaning
a
Intercept Coefficient
α
Alpha
b
Slope Coefficient
εr
Resilient Strain
εp
Plastic Strain
G
Complex Shear Modulus
µ
Mu
s
Normal Stress
T
Temperature
Gmm
Maximum Theoretical Specific Gravity
Gmb
Bulk Specific Gravity
Abbreviation
Meaning
AASHTO
American Association of State Highway and Transportation Officials
ASTM
American Society for Testing Materials
AC
Asphalt Cement
NHA
National Highway Authority
MEPDG
Mechanistic-Empirical Pavement Design Guide
PMA
Polymer Modified Asphalt
HMA
Hot Mix Asphalt
WT
Wheel Tracker
URLST
Uniaxial Repeated Load Strain Test
UTM
Universal Testing Machine
LVDT
Linear Variable Displacement Transformers
VMA
Voids in Mineral Aggregate
VFA
Voids Filled with Asphalt
DSR
Dynamic Shear Rheometer
XVII
Chapter One
1
Chapter One
Introduction
1.1
Background
Inconsistent asphalt-concrete properties together with illegal Axle loads, high tire pressures
and high temperatures are considered to be the most commonly observed causes behind
pavement frequent failures, in Pakistan. National Highway Authority (NHA), Pakistan has
continuously been modifying the aggregate gradations and penetration grade of asphalt
cements, without prior investigation of the mix behaviour under the prevailing axle load and
environmental conditions of the country. Traffic growth rate on major highways in Pakistan,
in particular the truck traffic has been increasing day by day due to its geostrategic locations
and international trade corridor for the countries like, China, Afghanistan and many states of
Russia. Indiscriminate use of materials in the construction of flexible pavements, especially in
asphalt- concrete layers, has become one of the major causes of failures, in areas, having high
temperatures and vehicle-load profile.
Premature rutting in the form of shear flow of asphalt concrete, being the consequences, has
directly been effecting the pavements service life, riding quality and their economic life cycles
cost. True prediction of asphaltic material behaviours and their precise selection on the basis of
performance can be one of the solutions towards this chaotic problem. At higher temperatures
i.e. 40oC and above, the rutting susceptibility of asphalt mixes needs to be studied in the
laboratory before its laying at site. Comprehensive laboratory investigation is required, to study
the influence of physical and mechanical properties of aggregates on rutting resistance or the
permanent deformation behaviour of asphalt mixes.
2
1.2 Problem Statement
The accumulation of permanent deformation in the asphalt surfacing layer appears to be the
major cause of rutting. To minimize this, it is necessary to pay extra attention to material
selection, mixture design and rutting measurement techniques. The questions of how to
measure rutting resistance of asphalt mixtures, what parameters to use as a measure of
resistance, and how to model and predict the development of permanent deformation need to
be addressed. In particular, measurement of rutting with simple performance test has become
the focus of current research.
A number of studies were carried out to investigate the permanent deformation behaviour of
asphalt concrete materials using various testing procedures and mainly based on uniaxial tests.
There was a need to make a more detailed study of the permanent deformation response of
asphalt concrete mixtures. An attempt has been made in this study to tackle the issues raised in
the preceding paragraphs. Based on laboratory tests that are judged to be simulative of field
loading conditions, the study attempts to provide more knowledge on the effect of volumetric
composition, loading, and temperature conditions on permanent deformation response of
asphalt concrete mixtures. In particular substantial effort has been made to correlate two
performance test procedures in terms of permanent deformation or rutting. Modeling the
permanent deformation behaviour of asphalt concrete mixes forms the other major part of this
study.
1.3
Research Objectives
The objectives of this research were:
1.
To design asphalt concrete mixes for NHA, Class-A, aggregate gradation at optimum
filler content, using three commonly used AC types and to determine the effect of asphalt
cement penetration grade on HMA properties.
3
2.
To determine the influence of stress levels, temperatures on permanent deformation
coefficients (α) and mu (µ) and their domains of asphalt mixes.
3.
To evaluate the resistance to permanent deformation of asphaltic mix using wheel tracker
4.
To define a relationship in terms of plastic to elastic strain ratio from laboratory data
using unconfined repeated creep test and to measure its validation by comparing with
AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) Model
1.4
Research Methodology
The methodology adopted to meet the objectives of this study involves a review of literature
and a laboratory investigation. The literature review is conducted to identify important
component material properties that influence the permanent deformation response of mixtures,
and available permanent deformation models and their theoretical basis. Testing methods that
are used to characterize permanent deformation property of asphalt mixtures are also reviewed.
The laboratory investigation is conducted using two testing procedures; the uniaxial repeated
load strain test (repeated creep test) and the wheel tracker. Specimens made with six mixes are
tested in both procedures.
The uniaxial repeated load strain test (repeated creep test) results are used both for modeling
purposes and evaluation of the effect of various factors on the permanent deformation
response. The wheel tracker test results are used to study the rutting potential of mixes under
different temperature conditions or to define a measure of resistance to rutting (permanent
deformation). Scope of experimental work has been shown in Fig 1.1.
4
Scope of Experimental
Work
Study of Properties of
Aggregates & AC
Hot Mix Asphalt Design
(Six Mixes)
Performance testing on
HMA Specimens
Physical &
Mechanical
Properties of
Aggregate
Shape Test
Consistency and
Rheological
Properties of
Asphalt Cement
Softening Point
Tests
Water Absorption
Test
Penetration Tests
LAA Tests
Ductility Tests
Soundness Test
Sp. Gravity Tests
Sand Equivalent
Test
Phase Angle
Deleterious Test
Uni-axial Repeated
Load Strain Test:
(At 25, 40, 550C
temperature)
&
(100, 300, 500 kPa,
stress levels)
Wheel Tracker Test
(At 25, 40, 550C
temperature)
&
Standard 720 Newton
Load
Shear Complex
Modulus
Analysis of results
Figure 1.1: Flow Chart Diagram (Scope of Work)
5
1.5
Organization
The work carried out to achieve these objectives is described in the following chapters:
Chapter Two provides a background and literature review on permanent deformation of
asphalt mixtures. The chapter entails a brief discussion of the various issues related to
permanent deformation of asphalt mixtures. The general concept is also offered in the context
of the stress-strain behaviour and resistance to permanent deformation of asphalt mixtures.
Particular attention was paid to the influence of the different components in the asphalt mixture
on the permanent deformation behaviour. Permanent strain models have also been reviewed
carefully.
Chapter Three explains material characterization, which includes physical and mechanical
properties of coarse aggregates, mineral filler, consistency of two neat binders and polymer
modified asphalt. Modification of PMA and its formulation have also been reported in this
chapter.
Results of rheological characterization of PMA, carried out on Dynamic Shear
Rheometer at different temperatures have also been provided.
Chapter Four explicates procedure adopted for the design of six HMA mixes, using two
aggregate gradations, three AC types and optimum filler content. Results of volumetric
properties, stability, stiffness index of mixes and effects of mineral filler percentages to design
properties of HMA have also been reported in this chapter.
Chapter Five elucidates the measure of resistance to permanent deformation of HMA, under
Uni-axial Repeated Load strain testing at different temperatures and stress levels. Mixes
behaviour in terms of elastic response i.e. resilient strain, resilient modulus and plastic
response i.e. accumulative strains were studied. The results were then plotted among different
variables for an in depth study of the behaviour of different mixes.
Chapter Six reports mixes behaviour under accelerated loading using wheel Tracker. This
chapter also provides relationships between load cycles and rut depth at different temperatures.
6
Chapter Seven provides and discusses summary of results in terms of mixes (with different
aggregate grading and AC types) behaviour under different conditions of loadings and
temperatures. Relationships among different variables and the outputs in terms of permanent
deformation behaviour of mixes have been presented graphically.
Chapter Eight convolutes modeling the permanent deformation behaviour of apshlt mixes
using the uniaxial load strain (creep) test and a comparison of Mechanistic-Empirical
Pavement Design Guide Model with the proposed model. It also concludes the most influential
parameter in the permanent deformation of mixes.
Chapter Nine concludes the findings of the research work and suggests recommendations.
7
Chapter Two
8
Chapter Two
Literature Review
2.1
Introduction
Rutting or permanent deformation is considered one of the major distress mechanisms in
flexible pavements, which may occur in all layers of the pavement structure and results from
lateral distortion and densification. In this chapter, rutting specifically in asphalt layers of
flexible pavements, which is mainly due to Hot Mix Asphalt (HMA), are defined and its
possible causes and distress mechanisms in particular have been discussed. The consideration
of permanent deformation at mixes design and its measurement are also discussed.
2.2
Permanent Deformation (Rutting) of Asphalt Mixes
Permanent deformation in asphalt (flexible) pavements, commonly referred to rutting, usually
consists of longitudinal depressions in the wheel paths, which are an accumulation of small
amounts of unrecoverable deformation caused by each load application as shown in Figure 2.1
(Asphalt Institute,1996). If an asphalt mixture ruts, it is normally because the mixture has
insufficient shear strength to support the stresses to which it is submitted. (Sousa et al. 1991)
Figure 2.1: Accumulated Plastic strains in Pavements (After Asphalt Institute, 1996)
9
Researches in the history showed that rutting in the Hot Mix Asphalt (HMA) layer will
generally occur within the top 3- to 5-in. 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, improving the material properties and mix characteristics will be significant in
decreasing the rut depth (Kennedy, et al, 1996).
There are several wheel path rutting classifications, one of which was provided in 1979 by the
Federal Highway Administration, which classified rutting into three levels of severity:
1.
Low, from 6 to 12.5 mm (0.25 to 0.5 inches),
2.
Medium, from 12.5 to 25 mm (0.5 to 1.0 inches), and
3.
High, over 25 mm (1 inch).
For normal cross slope values, a rut depth of 12.5 mm (0.5 inch) is typically accepted as the
maximum allowable rut depth (Huang, 1993 & Kennedy, et al, 1996).
Mechanistic-Empirical Design Guide (MEPDG) has defined three distinct stages for the
permanent deformation behaviour of pavement asphalt materials under a given set of material,
load and environmental conditions. Primary stage has high initial level of rutting, with a
decreasing rate of plastic deformations, predominantly associated with volumetric change.
Secondary stage has small rate of rutting exhibiting a constant rate of change of rutting that is
also associated with volumetric changes; however, shear deformations increase at increasing
rate. While the tertiary stage has a high level of rutting predominantly associated with plastic
(shear) deformations under no volume change conditions as shown in Figure 2.2 (AASHTO
Design Guide, 2002).
If an asphalt material is loaded with a stress that is above the flow strength of the material, at
that temperature the material will start to deform (Stumpf, 2007). First the material will deform
rapidly, then, after some strain hardening has taken place, the material gets to a stage with a
lower creep rate as shown in Figure 2.2. This stage is known as secondary creep, or steady
state creep. In the third stage the material becomes unstable and rapid collapse is the result.
10
Permanent Strain εp
Primary
Secondary
Tertiary
Flow Point
Load Repetitions
Figure 2.2
Typical Repeated Load Permanent Deformation Behaviours of Pavement
Materials (AASHTO Design Guide, 2002)
The proportion of permanent deformation taking place as shown in Figure 2.2 in the different
creep phases is important. The critical rut depth is generally set at 10 mm, if this depth is reached
in the primary phase or in the first part of the secondary phase, the functional life of the Hot Mix
Asphalt (HMA) layer is reduced drastically. In secondary phase, the rate of deformation slows
down considerably. The dominant mode of deformation is caused by shear stress that overcomes
flow strength of the material. The flow strength consists of two components, i.e. friction and
cohesion. The deformation takes place in small iterations with each load application. Eventually
the void condition and the level of permanent strain will cause the HMA to enter the tertiary
phase and rapid unstable shear failure occurs (Carpenter, 1993).
Traffic-associated permanent deformation, rutting in particular, results from a rather complex
combination of densification and plastic flow mechanisms. Plastic flow involves essentially no
volume change, and gives rise to shear displacements in which both depression and heave are
usually manifested as shown in Figure 2.3.
11
Figure 2.3: Rutting caused by Shear Displacement of Weak Asphalt Layer (After Asphalt
Institute 1996)
Plastic flow occurs when the shear stresses imposed by traffic exceed the inherent strength
of the pavement layers mostly at high temperatures. Multiple studies have identified this
mechanism as a primary cause of rutting problems in North America (Huber and Heiman 1989,
Sousa, et al. 1991, Brown and Cross 1992).
2.3
Influence of Asphalt Mixtures Properties on Rutting
Brown and Pell (1974) concluded that a dense graded asphalt mixture exhibits less deformation
than a gap graded mixture due to less aggregate interlocking in gap graded mixture. Evidences
show that the effects of rutting can be reduced by use of dense aggregate gradations. On proper
compaction, mixtures with dense or continuous aggregate gradations are more closely spaced
than open or gap graded mixtures and therefore have fewer voids. Also at higher temperatures,
the aggregate interlocking becomes more prominent so gap graded mixtures are more susceptible
to rutting at higher temperature which was later on confirmed by test track results [Huang 1995,
White 2003, Sivasubramaniam 2004]. Surface texture of the aggregate is particularly important
12
for good rutting resistance in thicker asphalt-bound layers and hotter climates where a rough
surface texture is required. Shape of the particle is also an important factor.
Mathews and Monismith (2003) studied the effect of excessive asphalt cement content, excessive
fine grained aggregate and high percentages of natural, rounded aggregate particles and
concluded that this excessiveness can be a common material-related causes of permanent
deformation.
2.4
Influence of Aggregates properties on Rutting
Rutting resistance of asphalt concrete under traffic and environmental loads depends on the
aggregate structure in the asphalt mix. Aggregate gradation and aggregate shape properties or
morphology of aggregate materials have been recognized by the Strategic Highway Research
Program (SHRP) among the top factors that influence the stability of hot mix asphalt (HMA). In
dense graded asphalt mixes, coarse aggregate size and shape properties are believed to some
extent contribute to the rutting resistance of asphalt concrete. Previous research studies realizes
the important role, the coarse aggregate plays in the rutting behaviour of HMA related aggregate
structure stability to coarse aggregate morphologies (Klaus, 2003).
Conversely, instead of locking together, smooth, rounded aggregate particles tend to slide past
each other. If the aggregate provides a high degree of internal friction (Ø), the shear strength of
the asphalt mixture will be increased and, therefore, the resistance to rutting. This is
accomplished by selecting an aggregate that is angular, cubical, has a rough surface texture, and
is graded in a manner to develop particle to particle contact Mc Gennis et al, (1994).
Fred (1967) reported that aggregate gradation appeared to have more influence than aggregate
type (at constant asphalt content) and at longer time of loading. He also concluded that the
temperature susceptibility characteristics of the asphalt appear to have more influence at longer
time of loading.
13
Crawford (1989) concluded from a study related to tender mixtures that particle shape and the
amount of material passing No. 4 sieve (4.75-mm) were major factors contributing to the
tenderness of an asphalt concrete mixture. He also stated that rounded, uncrushed aggregates are
more likely to contribute to tender mixtures and, therefore, more rutting susceptible, especially as
the amount of uncrushed material passing No. 4 sieve increases.
Kim et al. (1992) demonstrated that aggregate type has significant effects on fatigue resistance
and permanent deformation of asphalt concrete. Gradation only had minor effects on permanent
deformation. Interactions of aggregate type with gradation, asphalt type, air voids, and
temperature were found to be significant for the permanent deformation of asphalt concrete
Krutz and Sebaaly (1993) evaluated the effects of aggregate gradation on permanent deformation
of HMA mixtures for the Nevada Department of Transportation and concluded that the best
aggregate gradation is dependent on the type and source of aggregate and the coarse aggregate
gradations (bottom of band) performed the worst and fine aggregate gradations (middle and top
band) produced better performing mixtures.
Yeggoni et al. (1994) conducted a laboratory study to evaluate the influence of coarse aggregate
shape and texture on permanent deformation characteristics of HMA mixtures. The authors
concluded that an increase in the percentage of crushed coarse aggregate resulted in increased
Hveem stability, Marshall Stability, and resistance to permanent deformation. They also found a
strong correlation between rutting potential and the shape of the coarse aggregate particles as
measured using image analysis.
Kennedy et al. (1996) stated that, in order to prevent permanent deformation of HMA
pavements, one should avoid gradations near the maximum density because, although they
theoretically produces the strongest HMA mixtures, due to their relatively low voids in the
mineral aggregate, these types of mixtures would be very sensitive to asphalt content and might
presented the risk of flushing due to inevitable variations during construction. It would be better
to use aggregates with angular particles because they exhibited greater interlock and internal
friction and, hence, resulted in greater mechanical stability than rounded particles. It would be
14
better to use aggregates with rough surface texture because they might tend to form stronger
mechanical bonds when compared to smooth-textured aggregates and provided higher VMA in a
compacted mass.
Mallick and Kandhal (2001) summarized their observations by stating that the statistical analyses
of Asphalt Pavement Analyzer (APA) rut depth data obtained on all mixtures indicated
significant differences in performance among different gradations. They observed that, for
granite and limestone, Below Restricted Zone (BRZ) generally exhibited the highest and
Through Restricted Zone (TRZ) exhibited the lowest rut depths, and above restricted zone (ARZ)
showed intermediate rut depths. For river gravel mixtures, the order from highest to lowest rut
depth was ARZ, BRZ, and TRZ. The BRZ limestone mixture yielded the highest peak shearing
strain for both wearing and binder courses.
It is commonly understood that larger, more angular aggregates, with rough surface texture
increases rutting resistance (Button et al. 1990, Sousa et al. 1991, Brown and Bassett 1990,
Kandhal and Mallick 2001). Angular rocks are considered to provide better stone on stone
interlock than rounded aggregate, there by reducing the susceptibility to rutting (Asphalt Institute
1996, Ahlrich 1996, Marks et al. 2001).
2.5
Influence of Mineral Fillers Properties on Rutting
Herrin and Goetz (1954) observed from a laboratory evaluation that the strength of the asphalt
mixture, regardless of the type of coarse aggregate, increased substantially when the fine
aggregate was changed from rounded sand to crushed fine aggregates.
Uge and Van de Loo (1974) found that at moderate or high temperature, the relative
displacements of mineral particles occurring during laying or compaction under prolonged
loading may accelerate rutting. Consequently, they recommended the use of mixtures with low
workability in order to improve the arrangement of mineral skeleton and internal friction and
hence to minimize rutting potential. They also concluded that harsh mixtures that are well
compacted after laying will be highly resistant to rutting.
15
Asphalt Institute (1995) in Marshall Method of mix design reported that mix durability has also
been related to the amount of fine “dust or dirt” particles in the mixture. Excessive fine lowered
the quality of the asphalt film on the aggregate. Depending on the size of these particles, the mix
may be stiffer or tenderer.
Shashidhar et al (1999) worked on the maximum packing fraction and generalized Einstein
coefficient to characterize the stiffening potential of mineral filler in Hot Mix Asphalt (HMA)
and proposed that those parameters had contributed towards stiffening and physio-chemical
improvement of asphalt concrete.
Ahmad et al. (2004) studied the effect of filler type especially Portland cement and lime stone on
Marshall stability and retained strength of asphalt concrete in order to compare the effect of
different types and ratios of mineral filler on the strength properties of asphalt mixes. They
concluded that cement filler resulted higher values of retained strength.
National Cooperative Highway Research Program (2005) has reported that mineral filler were
originally added to dense-graded asphalt concrete to fill the voids in the aggregate skeleton and
to reduce the voids in the mixture. The additions of fines to the asphalt binder can have three
main effects; extend the asphalt binder, or stiffness the asphalt binder, or both.
Huang et al (2007) investigated the effect of mineral filler on mix design and performance
characteristics of HMA mixtures by selecting three types of mineral fillers and four filler
contents in order to study the relationship between filler contents and rut depth. They concluded
that filler with rough texture and high percentage usually increases the stiffness and decreases the
rut depth. They further recommended that filler contents ranges would be required to be
investigated in order to ensure the performance of the mixture.
2.6
Influence of Asphalt Cement Properties on Rutting
Asphalt Cement (bitumen) AC is a viscoelastic material with suitable mechanical/rheological
properties for traditional paving and roofing applications because of their good adhesion
16
properties to aggregates (Akmal and Usmani, 1999). Bitumen is a material characterized by a
time of loading and temperature dependence of the mechanical response to loading [Burger et al,
2001]. However, the increase in traffic load requires improving the mechanical properties of
conventional asphalt mixtures. The chemical composition of the AC has a significant effect on
its viscoelastic properties and hence on its performance as road paving material in asphalts.
The behaviour of asphalt at high temperature conditions for short time spans is equivalent to its
performance at low temperature conditions for longer time durations. This concept floated by
McGennis, et al. (1995) and is called temperature shift or in other words the superposition theory
of asphalt binder, which has been explained by Asphalt Institute (2003) as in Figure 2.4.
Figure 2.4: Temperature Shift Behaviour of Asphalt Binder (After Asphalt Institute, 2003)
In hot climatic conditions or under slow moving trucks, asphalt behaves like a viscous liquid and
only aggregates are the contributing element of hot mix asphalt that bear the traffic loads. At
micro level, the contiguous layers of molecules seem sliding pass each other. This phenomenon
has been presented by Asphalt Institute (2003) as shown in Figure 2.5. Whereas in cold climatic
conditions or under fast moving trucks (rapidly applied loads), asphalt behaves like an elastic
solid and deforms when loaded, but returns to its original shape when unloaded. If it is stressed
beyond its strength, it may rupture.
17
Figure 2.5: Microscopic View of Liquid Flow Properties (After Asphalt Institute, 2003)
At intermediate temperature conditions, asphalt binder exhibits the characteristics of both
viscous liquids and elastic solids. Due to this property of asphalt, it is considered to be an
excellent adhesive material for use in paving. When heated, asphalt acts as a lubricant, allowing
the aggregate to be mixed, coated, and tightly-compacted to form a smooth and dense surface.
After cooling, it acts as a glue to hold the aggregate together in a solid matrix. In its finished
state, the behaviour of the asphalt is termed as visco-elastic i.e., it has both elastic and viscous
characteristics, which depends on the temperature and rate of loading. The elastic response or
recoverable part, viscoelastic response and plastic response or non-recoverable which appears in
the form of permanent deformation have been illustrated in Figure 2.6 (Gibb, 1996).
Rutting at high temperatures, crack initiation and propagation in the low temperature region and
other forms of pavement defects are not only due to traffic loads but also due to the capability of
the asphalt concrete to sustain temperature changes. Increased traffic factors such as heavier
loads, higher traffic volume, and higher tire pressure demand higher performance pavements.
High performance pavement requires AC that is less susceptible to high temperature rutting or
low temperature cracking, and has excellent bonding to stone aggregates (Ait-kadi et al, 1996).
18
Figure 2.6: Stress Strain Behaviour of Bituminous Material (After Gibb, 1996)
William, et al., (1967) studied the influence of rheological properties of asphalt on rate of
deformation and strength of asphalt concrete and reported a direct relation between them. The
asphalt viscosity directly affects the strength of asphalt concrete in compression (rutting) for the
practical range of temperatures. The log of pavement resistance and of cohesion varies directly
with the log of asphalt viscosity. Modulus of elasticity in compression was influenced by the
type of asphalt, temperature and amount of lateral confinement. Repeated loading produced a
marked decrease in flexural strength of asphalt concrete.
19
Brown and Snaith (1974) suggested that the increase in deformation is related to the decrease in
binder viscosity at high temperatures (400C), thereby leading to a lower interlock between the
aggregates. The contribution of the aggregate skeleton towards the behaviour of the mixture
becomes more significant at higher temperatures. Pellinen and Witczak (2002) found the
aggregate influence to be more dominant than the influence of the binder on the modulus at high
temperatures and the binder influence to be more dominant over the aggregate influence at low
temperatures. At high temperatures the effects of confining stresses play a significant role in the
permanent deformation of the mixture. Awad (1972) and Morris (1973) found that the effect of
confining stresses on elastic and permanent deformation strains respectively was more important
at high temperatures than at low temperatures.
Mahboub and Little (1988) found that mixtures containing less viscous asphalts are less stiff and
are more prone to rutting. Monismith et al. (1999) also recommended more viscous asphalt
cements in thicker pavements and hotter climates on the basis of similar observations.
Monismith and Tayebali (1988) examined the relative behaviour of mixtures with and without
modifiers. They found that mixtures containing modified asphalt cement showed better
resistance to rutting at high temperatures than the mixture containing the neat asphalt cement.
They also reported that resistance to rutting may be improved by the use of modifiers (polymers,
micro fillers, etc.) which make asphalt binder more viscous at higher temperatures without any
adverse effect at low temperature
Anderson, et al. (1995) found that asphalt binder behaves both as viscous liquids and elastic
solids at normal pavement temperatures.
Heukelom (1999) developed charts to study the effects of temperatures on the mechanical
behaviour of asphaltic bitumen. He performed the standard laboratory consistency tests and
appraised the data used for entering Van der Poel’s stiffness nomograph. He concluded that
ordinary laboratory tests at different temperatures could show relationships and stiffness
modulus of bitumen & could be obtained from Van der Poel’s nomograph.
20
Robert (2000) reported that the rutting tendency of a pavement, greatly influenced by the ratio of
the complex modulus (G*) to the phase angle ( ). High value of G*and low values of
of
bitumen are required for more rut resistance and low value of “G*” and “ ” are required for more
fatigue resistance.
Tarefder, et al. (2003) investigated the most important factors affecting rutting and performance
grade (PG) of bitumen and determined that specimen type, test temperature and moisture has
significant influence on binder performance.
Kanitong et al. (2005) compared the rutting performance of polymer modified bitumen with
unmodified and concluded that the overall performance of polymer modified binder was better
than those of unmodified.
As an alternative to larger top size and coarser mixes, polymer-modified asphalt cement and
other modified asphalt cement products are also being investigated by many agencies to increase
the resistance to permanent deformation (Ponniah and Kennepohl 1996, Prowell 2001).
2.7
Stress Strain Behaviour of Asphaltic Materials
Neat asphalt shows very complex stress-strain behaviour. It behaves as an elastic solid at low
temperatures and high frequencies of loading, and as a viscous fluid at high temperatures and
low loading frequencies. It has been mentioned that rutting occurs in mixtures with low shear
resistance or strength compared to the repeated stress it is subjected to. One can get an insight
into the effect of aggregate properties on shear strength of mixtures by considering their effect on
cohesion and angle of internal friction as illustrated in Figure 2.7. For a given level of stress,
temperature and rate of loading, the shear strength depends on the cohesion and angle of internal
friction. The cohesion is affected by the viscosity of asphalt binder and the proportion of fines.
The angle of internal friction is obtained from aggregate interlocking. Higher values of angle of
internal fraction can be obtained by using rough textured, angular and well graded aggregates.
The mechanical interlock of the aggregate particles thus plays a key role in shearing resistance.
The binder content is also known to affect angle of internal fraction because it changes the
21
degree of mechanical interlock between the particles, i.e., the higher the proportion of binder in
the mix, the further apart the aggregate particles are spread (Cheung, 1995).
Aggregate properties and aggregate gradation play a major role in the potential for rutting of an
asphalt pavement. Cubical, rough-textured aggregates are more resistant to the shearing action of
traffic than rounded, smooth-textured aggregates. Cubical aggregates also tend to interlock
better, resulting in a more shear resistant mass of material as shown in Figure 2.8.
In addition, increased compaction during construction or the use of higher percentages of coarse
aggregate fractions in the aggregate gradation provides more stone-to-stone contact in the asphalt
mix which, in turn, helps to reduce pavement rutting.
Figure 2.7: Shear Loading Behaviour of Aggregate (After Asphalt Institute, 2003)
Figure 2.8: Contrasting Stone Skeletons (After, Asphalt Institute, 2003)
22
As the response of the asphalt mixture is linear visco-elastic under the application of the load,
this element of the total strain is irrecoverable and with the repeated load application it
accumulates, leading to the formation of surface ruts (Gibb, 1996) as shown in figure 2.9 &
2.10.
Figure 2.9: Permanent Strain under Load Pulse (After Gibb, 1996)
Figure 2.10: Accumulation of Permanent Strain under Repeated Load (After Gibb, 1996)
23
2.8
Resistance to Permanent Deformation of HMA
Resistance to permanent deformation or shearing stress is a stability-related phenomenon.
HMA mixtures are designed with adequate stability to ensure adequate performance and are
considered to be the core aspect of HMA mixture design with respect to rutting. Stability is
affected by type/grade and amount of asphalt binder, aggregate properties (such as absorption,
texture, and shape of particle), gradation, compaction level, and temperature. Higher stability is
promoted by using hard cubical aggregates with rough surface textures, dense gradations,
comparatively low asphalt binder contents, harder (stiffer) asphalts, and well-compacted
mixtures as long as the air voids do not fall below a certain level.
The asphalt binder used in the mix also affects the rut resistance of an asphalt mix but to a
lesser degree than the aggregate characteristics. A mix made with a soft grade of asphalt
cement will be less resistant to rutting at high temperatures than a comparable mix that
contains a harder (more viscous) asphalt grade. Rutting in an asphalt mix normally occurs in
the early years of a pavement’s life when the asphalt binder is relatively low in viscosity.
Whereas, rutting is less likely to occur in a pavement after the asphalt binder has aged or
oxidized with exposure to the elements to a higher viscosity.
Findley, et al. (1976) studied the resistance to deformation of HMA under creep loading and
concluded that under constant static or repeated loading, a visco-elastic material undergoes
flow or ‘creep’, which includes recoverable and irrecoverable, time dependent and time
independent components of deformation. They also concluded that instantaneous elasticity,
creep under constant stress, instantaneous recovery, delayed recovery, and permanent strain
can be used to characterize viscoelastic materials including HMA.
Cooper, et al. (1985) concluded that low voids in mineral aggregates (VMA) results in good
resistance to permanent deformation. They also reported that rutting resistance of asphalt
mixture can also be improved by reducing air voids up to a certain limit & higher compaction
energy results in a low air void content in the field.
24
Huber, et al. (1987) examined 9 test sites in Saskatchewan to evaluate the mix design
characteristics and performance and concluded that higher asphalt contents and voids filled
with asphalt were the basic parameters that had increased the rutting potential of mix.
Brown (1990) reported that HMA pavements constructed at approximately 7.0 to 8.0 percent
air voids are further compacted to approximately 4.0 percent air voids under traffic loads, if the
mix is properly designed.
Molenaar (1993) reported that plasticity models such as the Mohr-Coulomb yield criterion
describe the response of a material in relation to an ultimate surface beyond which no stresses
are permitted to occur. If the state of stress is inside the ultimate surface, the deformations are
purely elastic and plastic deformation occurs at states of stress on the ultimate surface. The
Mohr-Coulomb failure criterion expressed in Equation 2.1 provides an elegant means of
demonstrating the effect of both aggregates and the mortar on the permanent deformation
behaviour (FHWA). Figure 2.11 illustrates the effect of the aggregates and the mortar on the
failure behaviour of asphalt mixtures. For simplicity purposes, the cohesive strength is
considered to be wholly influenced by the mortar while the angle of internal friction is
influenced entirely by the aggregates. If the binder has a high cohesion at high temperatures,
then it will provide better resistance to permanent deformation than a binder with a lower
cohesion. Similarly aggregates with a higher internal angle of friction can be expected to
provide a better resistance to permanent deformation than similar aggregates with a lower
angle of internal friction. However, it must be emphasized that this is a very simple idealization
of the asphalt mixture.
τ = c + σ tan φ
Where:
t = shear strength,
c = cohesion,
σ= normal stress,
Φ = angle of internal friction.
25
(2.1)
Figure 2.11: Effect of Aggregate and Mortar on the Failure Behaviour of Asphalt Mixtures
Cominsky, et al (1994) studied the resistance of HMA to rutting and reported that both the
mineral aggregate and asphalt cement contributes towards the shear strength of HMA. Collop,
et al. (1995) treated permanent deformation or rutting in bituminous pavements as a linear
visco-elastic flow phenomenon and found that permanent deformation per wheel pass is
directly proportional to the static axle load and inversely proportional to the vehicle speed. The
phenomenon is quite useful for investigating the road damaging potential of heavy vehicles and
evaluating important trends.
Ford, et al. (1988) investigated that pavements with air voids lower than 3.0 percent tend to rut
while those with higher air voids do not, as long as the aggregate quality is satisfactory. They
also concluded that pavements with air voids lower than 3.0 percent have a tendency to rut
severely.
Kamal, et al (2005) studied the insitu behaviour of asphalt concrete with and without PMA
under same temperature and loading conditions and compared resilient modulus and creep
stiffness of both type of mixes, using the indirect tensile strength test (ASTM D4123) and
repeated load uniaxial stain test. They reported a drastic reduction of about 85% in resilient
modulus was observed for an increase in temperature from 250C to 400C.
26
Ziari et al (2007) studied the effects of temperature and different percentage of bitumen on the
resistance to permanent deformation of HMA mixture and concluded that significant degree of
confidence that the mix will not fail on the roadway due to permanent deformation can be
achieved by simulating the laboratory test findings with field performance of mixes.
The distribution of stresses in the pavement depends on the stiffness, Poisson’s ratio and
thickness of the pavement layers. A typical asphalt pavement structure consists of a top layer,
base, subbase and subgrade.
Similar to granular materials, the resistance to permanent
deformation of asphalt mixtures is to a large extent dependent on the magnitude of the
confining stresses that result from the wheel load. Generally, an increase in confinement
enhances the resistance to permanent deformation. By smartly stacking the pavement layers,
significant confinement levels can be generated leading to enhanced resistance to permanent
deformation. Such a stacking mostly involves a layer with a high modulus below the asphalt
top layers (Huang 1993).
2.9
Permanent Deformation Prediction Models
The permanent deformation response of asphalt concrete mixes to loading, temperature and
other influential parameters may have to be characterized by a number of models in the
previous research work. Accordingly, recommendation and limitations have been reported. A
few models have been briefed in this section.
Garba (2002) reported the effects of material properties on permanent deformation of asphalt
mixtures, mechanisms of the permanent deformation, and methods of its prediction. The main
objectives of research work included the review and evaluation of available models for
permanent deformation of asphalt concrete mixtures, investigation of the effect of volumetric
composition, loading and temperature conditions on the permanent deformation of asphalt
concrete, and the identification and definition of simple measures of resistance to permanent
deformation. In order to meet that objectives, repeated load triaxial creep and recovery tests
were conducted at 250C and 500C under varying stress conditions. Binder content, void
content, confining stress and deviator stress were found to influence the permanent
27
deformation characteristics significantly. It was reported that most of these parameters are not
sensitive to changes in volumetric composition and therefore are not suitable for comparison of
mixtures made from the same materials but with varying proportion of the components. The
bounding surface plasticity approach was found to be a convenient method to model the
accumulation of permanent deformation. It was demonstrated that deformations calculated
using cyclic hardening model based on bounding surface plasticity fits the measured
deformation quite well. The elasto-viscoplastic model, which is based on strain decomposition
approach, provides a suitable method for analysis of creep and recovery test results.
Deformations of asphalt mixtures, calculated using this model also fit the measured
deformation quite well.
The NCHRP 1-26 study (Barenberg and Thompson, 1990) recommended the use of the
permanent strain accumulation model developed at Ohio State University (Majidzadeh et al
1981). This strain model predicts total rutting, considers the rutting rate of the pavement as
indicated by the following equation:
Єp /N=A(N)m
Where
(2.2)
Єp =
permanent strain
N=
number of load application
A=
Experimental constant (depends on material type and stress state)
m=
Experimental constant (depends on material type)
Equation 2.2 is valid for describing the progression of rutting in pavement layers, asphalt
surface and base courses, granular base and subbase courses, and subgrade soils (Majidzadeh,
et al 1981).
Several material permanent strain accumulation models have been developed so far to predict
the permanent deformation in AC pavement layers. Pavement system rutting models were also
evaluated in the NCHRP 1-26 (Barenberg and Thompson 1990). The study have revealed that
those models which are related to log of permanent strain to the log of load repetition appear to
be the most appropriate and versatile for practical use.
28
This power model is often fitted to the accumulated permanent deformation curve. It is
probably the most commonly used permanent deformation equation. The power model plots as
straight line on log-log scale. It has also been thought that the slope and intercept of this model
when plotted on log-log scale may be used as indicators of rutting resistance (Garba, 2004).
The basic permanent strain to load repetition model expressed as
Єp =aNb
(2.3)
It was initially proposed for subgrade and unbound materials by Monismith (1976); and
initially used for asphalt concrete mixes by (Khedr, 1986). Later on, various researchers used
the same model for asphaltic concrete (Diylajee and Raymond, 1982), (Vuong and Amstrong
1991), (Behzadi and Yandell, 1996). Where a & b are intercept and slope coefficients and N is
the load repetition. The curve of power model on log-log scale between load repetition and
permanent strain can be expressed in graphical form in Figure 2.12.
Figure 2.12: Log-Log Form of Power Model (After Khedr, 1986)
The log relationship of Єp =aNb, can be expressed in the following form;
29
Log Єp
=
b Log N + a,
(2.4)
They further states that this laboratory relationship between permanent strain and load
application can be plotted and extrapolated for general analysis purpose. A typical value for the
exponent “b” varies between 0.1 and 0.2; while “a” is highly dependent upon the magnitude of
the repeated stress state. It was further noted that limiting the stress ratio (repeated deviator
stress/maximum deviator stress) is a good approach to minimize the permanent deformation,
and that the stress ratios (repeated stress/maximum strength) less than approximately 60 to 70
percent will generally help to limit the accumulation of permanent deformation. Permanent
deformation is also somewhat dependent on the previous stress history to which the material
has been subjected. The initial application of large stress repetitions is much more damaging
than the initial application of low stresses (Hassan, 1998). Using equation 2.3, intercept
coefficients and slope coefficients have been determined for mixes in chapter Six & Seven.
According to VESYS, 1991 progressing of rutting with load repetition can be measured using
layer elastic theory, where in all layers can be modeled using a constitutive model in the form
given in equation 2.5 (partial differentiation form).
∂ε p
∂N
= ε pn =
(
∂ aN b
∂N
)
(2.5)
ε pn = abN ( b−1)
The resilient strain (εr) is assumed to be independent of load repetition. The ratio of plastic to
resilient strain can thus be defined as:
ε pn ⎛ ab ⎞ b−1
= ⎜⎜ ⎟⎟ N = µN −α
εr ⎝ εr ⎠
(2.6)
From Equation 2.6, the rate of plastic strain (1-b) can be defined with a permanent deformation
coefficient alpha (α) & plastic to elastic strain ratio can be defined with a coefficient mu (µ) as
shown;
µ=
εr
ab
α = 1− b
(2.7)
30
Where
Єp = Permanent Strain (rut value)
N = Number of Load Application
a = Intercept Coefficient; and
b = Slope Coefficient
µ = mu (ratio of plastic to elastic response)
α = alpha (rate of change of the plastic response)
Where µ is the permanent deformation parameter representing the constant of proportionality
between permanent strain and resilient strain (i.e. plastic strain at N=1) and α is a permanent
deformation parameter indicating the rate of decrease in incremental permanent deformation as
the number of load applications increases. Thus two set of parameters a &b and µ& α are
closely related.
Rita and Matthew (1991) determined alpha and mu using log10Єp-log10N as base relationship
and proposed that these parameters could be used to assess plastic strain to elastic strain ratio.
Alpha and mu according to Sullivan (2002) are the stress and temperature dependent non linear
parameters and can be used for modelling permanent deformation of mixes.
The current Mechanistic Empirical Pavement Design Guide (MEPDG) incorporates a power
model for generating rutting predictions for asphalt concrete (Stephen et. al. 2007). Rutting
model developed from laboratory uniaxial repeated load strain tests as provided in MEPDG in
the following form has been used as basis to estimate the relationships between the predictor
variables and the permanent deformation parameters:
εp
= a1T a N a
εr
2
3
(2.8)
Where, εp, εr, are the plastic and elastic strains respectively, at N repetitions of load and ai are
the non linear regression coefficient. This model was proposed using the uniaxial repeated
31
haversine pulse loading and measuring the strains with on-sample Linear Variable
Displacement Transformers (LVDT).
An attempt is made in this study to model the permanent deformation behaviour of asphalt
concrete mixes using equation 2.8 as base guide line. In laboratory testing, off-sample LVDT’s
were used to measure the strain response under application of square shape pulse loading.
Modeling the permanent deformation behaviour of asphalt concrete mixes is thus the main part
of this study.
32
Chapter Three
33
Chapter Three
Material Characterization
3.1
Introduction
Hot Mix Asphalt generally consists of combination of different size of aggregates with mineral
fillers, uniformly mixed and coated with asphalt cement, each having its own particular
characteristics, which will be more suitable to specific design and construction purposes.
Before designing asphalt paving mixes, selection, proportioning and characterization of
individual material are imperative to obtain the desired quality and properties of finished mix.
For the current study, aggregates were obtained from the local source i.e., Margalla Crush
Quarry and asphalt cement from Attock Refinery, Rawalpindi, & National Refinery, Karachi,
Pakistan.
3.2
Aggregate for Asphalt Mixtures
The largest portion of the resistance to permanent deformation of the mixture is provided by
the aggregate structure. Aggregate is expected to provide a strong stone skeleton to resist
repeated load applications. Gradation, shape, and surface texture have a great influence on
HMA properties. Angular, rough-textured aggregates provide more shear strength than
rounded, smooth-textured aggregates. When a load is applied to the aggregate in an asphalt
mixture, the angular, cubical, rough-textured aggregate particles lock tightly together and
function as a large, single elastic mass, thus increasing the shear strength of the asphalt
mixture.
To set up a laboratory study and investigate effect of coarse aggregate properties on the HMA
aggregate structure and the stability of asphalt concrete (AC) pavements, aggregate consensus
and source properties have to be properly taken into account together with the aggregate sizes
or gradations.
34
The suitability of aggregates from Margalla Crush Quarry for use in asphalt construction was
determined by evaluating the material in terms of the followings;
̇
̇
̇
̇
̇
̇
3.2.1
Size and grading
Particle shape
Absorption
Clay Contents (Sand Equivalent)
Toughness
Soundness
Size and Grading
Maximum size and aggregate grading are directly controlled by NHA aggregate grading
“Class-A” for wearing coarse (NHA General Specification, 1998). Maximum size of aggregate
is related with the typical lift thickness used on National Highways in Pakistan. NHA in its
general specifications has specified two aggregate gradations, namely class ‘A’ and ‘B’, the
coarser and finer respectively. Two gradations i.e. “01” and “02”, within the envelope of NHA
class ‘A’ gradation (commonly used in the field) have arbitrarily been selected by the author
for this study as reported in Table 3.1.
Table 3.1: Adopted Aggregate Gradations & Specifications
Sieve Size
Combined grading (Asphaltic Wearing Course Class-A)
Gradation “1”
Inch
mm
Gradation
Gradation “2”
NHA
Targeted
Targeted Specifications
Gradation
‘Class-A’
value
value
Asphalt
Institute
Gradation
(1994)
1
25.00
100
100
100
100
100
100
3/4
19.00
90-100
90
95-100
100
90-100
90-100
1/2
12.50
-
-
-
3/8
9.50
56-69
56
59.1-69.1
69.1
56-70
56-80
#4
4.75
38-46
38
38.2-48.2
48.2
35-50
35-65
#8
2.36
25-33
25
24.3-30.3
30.3
23-35
23-49
#50
0.300
5-12
5
4.5-10.5
10.5
5-12
5-19
#200
0.075
3.4-5.3
3.4
3.3-5.3
5.3
2-8
2-8
-
35
The targeted values of aggregate gradation “01” were on the coarser side of the limits, while
that of aggregate gradation “02” were on the finer side of the gradation limits. Both the
gradations have been shown graphically in Figure 3.1.
Aggregate Gradation Limit
100
90
Gradation "02"
80
Percentage Passing
70
60
Gradation"01"
50
40
30
20
10
0
0.01
0.1
1
10
Seive Size (mm)
Gradation "02"
100
Gradation "01"
Figure 3.1: Aggregate Gradation ‘01’ and ‘02’ under NHA Class ‘A’
3.2.2
Particle Shape
Particle shape changes the workability of the paving mix as well as the compactive efforts,
necessary to obtain the required density. Particle shape has an effect on the strength of the
asphalt mix. Irregular or angular particles, such as crushed stone and gravels, and some natural
gravel and sands, tends to interlock when compacted and resist displacement. Best interlock is
generally obtained with sharp-cornered cubical shaped particles. Round particles, such as most
natural gravels and sands from stream beds, can be used successfully in asphalt paving mixes.
36
Hot Mix Asphalt can generally be made more stable and resistant to rutting by requiring that a
significant portion i.e. 50% of the coarse aggregate be angular. (Asphalt Institute, MS-4, 2003).
Coarse aggregate angularity can be defined as “The percentage by mass of the aggregates
larger than 4.75 mm with one or more fractured faces” (Asphalt Institute SP-2, 2003). High
shear strength for rutting resistance and a high degree of internal friction can be achieved by
specifying this property (ASTM D5821). The type of aggregate surface texture depends on
hardness, grain size, and pore characteristics of the parent rock as well as the extent to which
forces acting on the particle surface have smoothened or roughened the surface (Su 1996). The
type of aggregate surface texture is categorized by shape features such as angular, rounded,
smooth or rough. The aggregate surface texture affects the aggregate interlock. Durable
angular aggregates with a rough surface texture are normally considered to offer good
aggregate interlock.
Fine aggregate angularity can be defined as “The percent air voids present in loosely
compacted aggregates smaller than 2.36 mm” (Asphalt Institute SP-2, 2003). Fractured faces
are indicated by the void content measured as un-compacted void contents of fine aggregates
(AASHTO T304). Greater the void contents more will be the fractured faces. High degree of
internal friction and high shear strength for rutting resistance can be achieved by specifying
this property. Particle shape, surface texture and grading influence fine aggregate angularity.
Aggregates shapes were measured using flakiness and elongated index gauges.
Flat and elongated particles can be defined as “The percentage by mass of coarse aggregates
that have a maximum to minimum dimension ratio greater than five” (Asphalt Institute SP-2,
2003). Flakiness of aggregate particles are classified as flaky when they have thickness (small
dimension) of less than 0.6 of their mean sieve size, this size being taken as the mean of the
limiting sieve apertures used for determining the size fraction in which the particle occurs.
While aggregate particles are classified as elongated when they have a length (greatest
dimension) of more than 1.8 of their mean sieve size, this size being taken as the mean of the
limiting sieve apertures used for determining the size fraction in which the particles occurs
37
(ASTM D4791). Results of flakiness index and elongation index have been reported in Table
3.2.
3.2.3
Absorption
The porosity of an aggregate is generally indicated by the amount of water it absorbs when
soaked in water. A porous aggregate will also absorb asphalt which will tend to make asphalt
mixtures dry or less cohesive. An extra amount of asphalt must be incorporated into the paving
mix to compensate for the absorption of asphalt by the aggregate. Moreover, aggregate that are
very porous tend to require a significant amount of extra asphalt to make up for the high
absorption rate. Highly porous aggregates normally are not used unless they possess certain
other qualities or properties that make them desirable in spite of the high absorption rate
(Asphalt Institute MS-4, 2003). If absorptive aggregates that have high water content are used,
will be required in the production of HMA to ensure that the moisture in the pores can
evaporate. Otherwise, the asphalt may not be properly absorbed, leading to compaction
difficulties (Transportation Research Board, 2000). Results of absorption of aggregates have
been reported in Table 3.2.
3.2.4
Clay Content (Sand Equivalent)
Asphalt Institute defines the clay content as “clay content is the percentage of the clay material
contained in the aggregate fraction that is finer than a 4.75 mm sieve” (Asphalt Institute SP-2,
2003). Aggregate cleanliness may often be determined by visual inspection, but a washed-sieve
analysis generally provides positive proof. The sand equivalent test, developed by the
California Division of Highway and described in ASTM D2419 (AASHTO T176) is a method
of determining the relative proportion of detrimental fine dust or clay-like materials in the
portion of the aggregate passing the 4.75mm sieve(Asphalt Institute MS-4, 2003).
equivalent of fine aggregate have been reported in Table 3.2.
38
Sand
3.2.5
Toughness
Aggregates are subjected to additional crushing and abrasive wear during manufacture, placing
and compaction of asphalt paving mixes. Aggregates are also subjected to abrasion under
traffic loads. They must exhibit, to a certain degree, ability to resist crushing, degradation, and
disintegration; Asphalt Institute’s defines the toughness as “toughness is the percent loss of
material from an aggregate blend during the Los Angeles Abrasion Test (ASTM C131 or
AASHTO T96)”. The toughness property test estimates the coarse aggregate resistance to
abrasion and mechanical degradation that occurs during handling, construction and service. To
perform the test, coarse aggregates larger than 2.36 mm are subjected to impact and grinding
by steel spheres. Due to this mechanical degradation, the mass percentage of the coarse
material lost gives the toughness. Typically the maximum loss values range from 35 – 45
percent.
Aggregates at or near the pavement surface require greater toughness than aggregate in the
lower layers where loads have dissipated or are not as concentrated. Relatively high resistance
to wear, as indicted by a low percent of abrasion loss, is desirable characteristics of aggregates
to be used in asphalt pavement surface layers. Aggregates having higher abrasion losses, within
limits, may generally be used in lower pavement layers where they will not be subjected to the
high stresses caused by traffic. Los Angles Abrasion (LAA) values determined in the
laboratory for Margalla Quarry aggregates have been reported in Table 3.2.
3.2.6
Soundness
Aggregates must be resistant to breakdown and disintegration from weathering (wetting/drying
and freezing/thawing) or they may break apart and cause premature pavement distress.
Durability and soundness are terms, typically given to an aggregate’s weathering resistance
characteristic. Aggregates used in HMA are dried in the production process and therefore
should contain almost no water. Thus, for aggregate used in HMA, freezing/thawing should
not be a significant problem. According to Asphalt Institute’s Superpave Series No. 2,
soundness is defined as “The percent loss of material from an aggregate blend during the
39
sodium or magnesium sulfate soundness test (AASHTO T104)”. The resistance of aggregate to
in-service deterioration is determined by using this test. The test is applicable to both coarse
and fine aggregates. Results of soundness of aggregates have been reported in Table 3.2.
3.2.7
Deleterious Materials
Asphalt Institute’s Superpave Series No. 2 defines the deleterious materials as “The mass
percentage of contaminants such as clay lumps, shale, wood, mica, and coal in the blended
aggregates (AASHTO T112)”. The test is applicable to both coarse and fine aggregates. Wet
sieving aggregate size fractions over the specified sieves were performed and the percentage by
mass of the material lost as a result of it yielded the percent of the clay lumps and the friable
particles. Maximum allowable percentage of these materials ranges from as low as 0.2 percent
to as high as 10 percent and depends upon the exact composition of the contaminant (Asphalt
Institute, SP-2, 2003). Results of deleterious materials have been reported in Table 3.2.
Table 3.2: Physical Properties of Aggregates
Sr.
Test Description
No.
Specification
Results
Recommended Values
(NHA General Specification)
Reference
1
Flakiness Index (FI)
BS 812, Part 1
4.75%
15(Max)
2
Elongation Index (EI)
BS 812, Part 1
15(Max)
3
Aggregate Absorption
AASHTO T 166
4.4 %
0.88%
4
Sand Equivalent
AASHTO T 176
72
45(Min)
23%
30 (Max)
3.32%
12 (Max)
1
0.2-10
5
6
7
Loss Angles Abrasion
value (LAA)
AASHTO T96
ASTM C 131
Sodium Sulphate
AASHTO T104,
Soundness value
ASTM C88
Deleterious materials
Asphalt Institute,
SP-2,
40
3.3
Characteristics of Asphalt Cement
For engineering and construction purposes, three properties or characteristics of asphalt are
important i.e. consistency, purity and safety (Asphalt Institute MS-4, 2003). Asphalts are
thermoplastic materials, since they gradually liquefy when heated. They are characterized by
their consistency or ability to flow at different temperatures. Consistency is the term used to
describe the viscosity or degree of fluidity of asphalt at any particular temperature. The
consistency of asphalt cement varies with temperature; therefore, it is necessary to use a standard
temperature when comparing the consistency of one asphalt cement with another. Asphalt
cements are graded based on ranges of consistency at a standard temperature. Consistency of
paving asphalt is commonly specified and measured by a viscosity test or a penetration test
(Asphalt Institute MS-4, 2003). Other tests i.e. ductility, softening point provides additional
information and confidence towards consistency.
3.3.1 Consistency of Asphalt Cement
The penetration test determines the penetration of semi-solid and solid asphaltic materials. The
needles, containers and other conditions described in this test method provide for the
determinations of penetrations up to 500. Higher values of penetration indicate softer
consistency. Where the conditions are not specifically mentioned, the temperature, load, and time
are understood to be 250C, 100 g, and 5s, respectively (AASHTO T 49-03 or ASTM D5). Results
of penetration tests for three AC types have been reported in Table 3.3.
Softening point of asphalt cement was performed by using ring and ball apparatus confirming to
AASHTO T53. This test method covers the determination of the softening point of bitumen in
the range from 30 to 1570C (86 to 3150F) using the ring-and-ball apparatus immersed in distilled
water (30 to 800C), glycerin (above 80 to 1570C), or ethylene glycol (30 to 1100C). Bitumen’s
are visco-elastic material without sharply defined melting points; they gradually become softer
and less viscous as the temperature rises. Due to this property, softening points must be
determined by an arbitrary and closely defined method for the reproducibility of the test. Results
of softening point test for three AC types have been reported in Table 3.3.
41
The ductility of a asphalt cement can be defined as the “distance to which it will elongate before
breaking when two ends of a briquette specimen of the material, are pulled apart at a specified
speed (5cm/min ±5.0%) and at a specified temperature (25±0.50C) (ASTM 113-99 or AASHTO
T 51-00). This test method provides measure of tensile properties of bituminous materials and
may be used to measure ductility for specification requirements. Ductility is an indicator of how
flexible behaviour of bitumen under various temperatures. Results of ductility test for three AC
types have been reported in Table 3.3.
Specific gravity of two neat bitumen with penetration grade “60/70” & “modified bitumen”,
obtained from Attock Refinery Limited, Rawalpindi, and penetration grade “40/50” from
National Refinery Limited, Karachi, Pakistan were measured as per AASHTO T 228-2, and
reported in Table 3.3. Table 3.4 shows the specifications of these grades.
Table 3.3: Consistency of Different Types of AC
Sr. No.
Description
Type
1
2
Ring & Ball
Softening Point
Penetration
0
3
Ductility @ 25 C
4
Specific gravity
PMA
(60/70 base)
Modified
penetration grade
“60/70”
neat
penetration grade
“40/50”
neat
58
49
56
46
65
44
45
100
67
1.023
1.03
1.032
Table 3.4: Specifications of AC
Test Description
Test Methods
cm
ASTM T113
100
C (0F)
AASHTO T92
232 (450)
232 (450)
Min
1/10 mm
ASTM D 5
60-70
40-50
-
AASHTO T 53
50
60
Min
Ductility @ 250C
Flash Point
Penetration @ 250C
Softening Point
AASHTO M-20
Units
0
0
C
42
P.G 60/70
P.G 40/50
Max/Min
Min
Polymer modified bitumen used for this research work was obtained from Attock Refinery. The
modification of bitumen was carried out by Mathy Technology, USA (Gerald, 2001). It was
based on the climatic as well as the traffic conditions of one of the flexible pavement trial section
of NHA. The base bitumen i.e., 60-70 penetration AC was modified with 1.6% Elvaloy® 4170 in
the presence of superphsophoric acid. The addition of the 1.6% Elvaloy® 4170 caused the tensile
strength of the modified binder to increase significantly relative to the unmodified binder
(Gerald, 2001). The only reason for selecting the PMA is to compare the resistance to rut of
modified bitumen with neat asphalt. The properties of PMA determined in the laboratory have
been discussed in the proceeding sections.
3.3.2
Elastic Recovery
It is defined as “the recovered strain, as a percentage of the original strain attained at the break
point (end of loading phase)” (AASHTO T 301). Elastic recovery test was performed on AC by
means of a ductilometer with briquette specimens as shown in Figure 3.2. The specimens were
pulled apart at 5cm/min and held after a specified elongation as shown in Figure 3.3. This
procedure was used to evaluate the elastic recovery of PMA by the percent of recoverable strain
measured after the elongation of a sample.
Figure 3.2 : Briquette Specimens
Figure 3.3: Trimming of Specimens
(After Asphalt Institute, 2003)
(After Asphalt Institute, 2003)
43
Elastic recovery was calculated using the following equation and has been reported in Table 3.5.
Elastic Recovery (%) = 100 ×
a
b
(3.1)
Where “a” is the recovered displacement (or strain), in mm & “b” is original displacement (or
strain), in mm.
Table 3.5 Elastic Recovery Test Results at 25 oC
Types
PMA
60/70
40/50
Loading Rate
5 cm/min
5 cm/min
5 cm/min
Elongation Recovery
(%)
3.3.3
71% 77.5%
74%
66.5% 70%
68%
72.5% 75%
71%
Torsional Recovery
Austroads defines torsional recovery as the measure of the extent of recovery of the originally
applied rotation which is usually 180 degrees, in percentage. A flat-bottomed, cylindrical,
seamless tin 55mm in diameter and 35 mm in depth, Disc made of aluminum and spindle
assembly, bolt, and pointer are made of steel. The recommended device provides a scale, of 80
mm radius and graduated in degrees around at least half of its circumference (Austroads T-122).
Figure 3.4 shows the assembly of torsional recovery apparatus.
Graduation scale up to
1800
Pointer
Spindle
Figure 3.4 : Torsional Recovery Test Apparatus
44
Torsional recovery in %age was calculated for each type of asphalt cement at 250C, using the
following equation;
Torsional Recovery (%) = 100 ×
A
180
(3.2)
Where, A is the recovered angle, in degrees. The results have been reported in Table 3.6.
Table 3.6 Torsional Recovery Test Results.
Type of AC
40-50 Pen. grade
PMA
60-70 Pen. grade
Recovery Time
30 sec
30 sec
30 sec
Recovery
(degrees)
90
Recovery (%)
5%
3.4
140
120
240
300
220
200
180
170
7.7% 6.6% 13.3% 16.6% 12.2% 11.11% 10% 9.44%
Dynamic Shear Rheometer (DSR)
This test method covers the determination of the dynamic shear modulus and phase angle of AC
binders when tested in dynamic (oscillatory) shear using parallel plate geometry. It is applicable
to AC binders having dynamic shear modulus values in the range from 100 Pa to 10 mPa
(AASHTO T 315-05). This test method is intended for determining the linear viscoelastic
properties of AC binders as required for specification testing and is not intended as a
comprehensive procedure for the full characterization of the viscoelastic properties of AC
binders.
A dynamic shear rheometer test system consists of parallel metal plates, an
environmental chamber, a loading device, and a control and data acquisition system and shown
in Figure 3.5.
45
Figure 3.5: Dynamic Shear Rheometer
Metal test plates of 8.00 ±5.00 mm in diameter smooth polished surface. A chamber for
controlling the test specimen temperature, by heating (in steps or ramps), or cooling (in steps or
ramps), to maintain a constant specimen environment. A temperature controlling device
capable of maintaining the specimen temperature from 5 to 120 0C ± 0.1 0C test temperature.
The loading device which applies a sinusoidal oscillatory load to the specimen is shown in
Figure 3.6, to the frequency of 10.0 + 0.1 rad/sec.
Specimen
Figure 3.6: Test Specimen on DSR
46
Rheology is the study of deformation and flow of bitumen [Thomase, 2002] that explains the
elastic and viscous behaviour of bitumen, when subjected to a stress [Barnes et al 1989].
Complex modulus (G*) and phase angle ( ) are considered to be the principal rheoligcal
parameters, normally measured from a device known as Dynamic Shear Rheometer (DSR)
[Huang, and Zeng, 2007]. The complex modulus (G*) is the peak-to-peak shear stress to
absolute value of peak-to-peak shear strain and the phase angle ( ) is the angle in radian between
a simultaneously applied stress and the resulted sinusoidal stress in a controlled strain testing
mode (Asphalt Institute, SP-2, 2003). Following rheological properties can be determined using
DSR;
•
•
•
Strain amplitude, nearest 0.01 percent.
Complex modulus (G*) for the 10 measurements, kPa to three significant figures.
Phase angle ( ) for the second 10 cycles, nearest 0.1 degrees.
Table 3.7 shows the results of the tests being performed on DSR. The relationships between
different rheological properties are shown graphically in Figure 3.7.
Table 3.7: Rheological Properties of Binders
Sample
Type.
40-50
60-70
PMA
Frequency
Temperature
Phase Angle
(ºC)
( )
(rad/s)
(Hz)
10
1.59
25
10
1.59
40
10
1.59
55
10
1.59
25
10
1.59
40
10
1.59
55
10
1.59
25
10
1.59
40
10
1.59
55
47
Mean Complex
Modulus (G*)
G*/sin( )
(kPa)
72.5
462
484
80.8
37
37.8
86.6
3
3.22
74.7
995
1031.56
83.8
75
75.4
87.7
6
5.56
66.9
875
951
69.8
60
63.72
70.7
7
7.97
(Phase Angle & Complex Shear Modulus)
G* (kPa) 2000
40/50
60/70
100 δ [ ο ]
1800
90
1600
80
G* for 60/70
1400
Phase
Angle
PMA
70
60
1200
1000
800
G* for 40/50
G*
G* for PMA
PMA
50
40/50
40
Phase Angle for 40/50
30
600
60/70
400
20
200
10
Phase Angle for 60/70
Phase Angle for PMA
0
0
25
40
55
0
Temperature ( C)
Figure 3.7: Influence of Temperature on Linear Visco-Elastic Properties
Rheological characteristics like phase angle, complex modulus and G*/sin( ) of three asphalt
cement types i.e. 60/70 & 40/50 penetration grade AC and PMA are studied in dynamic shear
rheometer at three temperatures i.e. 25, 40 and 550C. Study revealed that temperature has
significant effect on binder rheology, specifically at intermediate to high temperatures.
Complex modulus and G*/sin ( ) reduces significantly at higher temperature (550C). Phase
angle in case of 60/70 and 40/50 penetration grade bitumen was observed to be more effected
than PMA.
48
Chapter Four
49
Chapter Four
Mix Design Method
4.1
Introduction
A mix design method is used for determining the gradation of course and fine aggregates to be
combined to achieve a predetermined percentage of air void volume and voids in mineral
aggregate for a given quantity of asphalt cement. Mixes are designed for heavy-duty asphalt
pavements, keeping in view the specifications and requirements of National Highway
Authority, Pakistan in terms of stability and durability. The aggregate gradations meet NHA
gradation requirements. The materials used in mix design practice conformed to specifications
and testing procedures as per Asphalt Institute (MS 2), American Association of State
Highway and Transportation Officials (AASHTO) and American Society for Testing and
Material (ASTM). This chapter presents HMA design using Marshall Method of Mix Design
for the preparation of six mixes.
4.2
Combined Grading of Aggregate
Local aggregates sources and asphalt cement as described in section 3.2, 3.3, & 3.4 were
chosen for mix design preparation. Three aggregate fractions at source were sieved and mean
passing size have been reported in Table 4.1. Two aggregate gradations with one-inch (1”)
maximum size as shown in Table 4.2 were developed from mean sieve sizes. Coarse aggregate,
which is the material, retained on an AASHTO NO.4 Sieve, consist of 100% crushed rock,
having at least two faces mechanically fractured. The filler consisted of final divided mineral
matter such as rock dust.
50
Table 4.1:
Summary of Sieve Analysis Results
Sieve size
Percentage Passing (%)
(Inches)
(mm)
0---5mm Size
5---13mm size
13---25mm size
1”
25
100
100
100
3/4
19
100
100
74.10
1/2
12.50
-
-
-
3/8
9.50
100
99.20
1.40
#4
4.75
99.50
25.20
0.60
#8
2.36
75.70
2.20
0.30
#50
0.300
23.20
1.60
0.20
#200
0.075
10.60
1.10
0.10
Table 4.2: Adopted Gradations & Specifications
Sieve Size
Combined grading (Asphaltic Wearing Course Class-A)
Gradation “1”
Inch
mm
Gradation
Gradation “2”
NHA
Targeted
Targeted Specifications
Gradation
‘Class-A’
value
value
1
25.00
100
100
100
100
3/4
19.00
90-100
90
95-100
100
1/2
12.50
-
3/8
9.50
56-69
56
59.1-69.1
69.1
#4
4.75
38-46
38
38.2-48.2
48.2
#8
2.36
25-33
25
24.3-30.3
30.3
#50
0.300
5-12
5
4.5-10.5
10.5
#200
0.075
3.4-5.3
3.4
3.3-5.3
5.3
-
51
Asphalt
Institute
Gradation
(1994)
100
100
90-100
90-100
-
-
56-70
56-80
35-50
35-65
23-35
23-49
5-12
5-19
2-8
2-8
4.3 Asphalt Mixtures
The Job Mix Formula (JMF) identifies the gradation of the combined aggregate materials and
the selected asphalt cement (Asphalt Institute, 2003). Six mixes were designed using both the
aggregate gradations and three AC types. Mixes designated as 1a, 1b, 1c and 2a, 2b, 2c with
gradation “1” and gradation “2” respectively and reported in Table 4.3.
Table 4.3: Selection of Mix Types
Gradation “1”
1a
PMA
(60/70+Elvaloy)
4.4
1b
60/70
Gradation “2”
1c
40/50
2a
PMA
(60/70+Elvaloy)
2b
60/70
2c
40/50
Mix Design Properties
An Asphalt Institute Marshall Method of Mix Design has adopted for the preparation of mixes
(Asphalt Institute, 2003). The Marshall method as presented here is applicable only to hot-mix
asphalt cements and containing aggregates with maximum size of 25mm (1 in.) or less. The
method is intended for laboratory design of asphalt hot-mix paving. The Marshall method uses
standard cylindrical test specimens of 64mm (2 ½ in.) height x 102 mm (4 in.) diameter. These
were prepared using a specified procedure for heating, mixing, and compacting the asphalt
aggregate mixtures. The two principal features of the Marshall method of mix design are
stability-flow test and density-voids analysis of the compacted test specimens. Following
criteria was adopted for the design of mixes (NHA General Specification, 1998);
Compaction, number of blows
75
Stability (kg) (minimum)
1000
Flow, 0.25mm
8-14
Percent air voids in mix (Va)
4-7
Loss in Stability (%),
20 (Max)
52
The stability of the test specimens is the maximum load resistance in Newton’s (lb.) that the
standard test specimen will develop at 600C (1400F). The flow value is the total movement or
strain, in units of 0.25-millimeter (1/100 in.) occurring in, the specimen between no load and
maximum load during the stability test (Asphalt Institute, 2003).
Stability by immersion of specimen in water at sixty (600C) for twenty four (24) hours as
compared with the stability measured after immersion in water at sixty degree (600C) for
twenty (20) minutes is the loss of stability value. Stiffness index is an empirical relationship
which is the ratio of stability to flow of mixes at 600C (NHA General Specification, 1998).
Optimum asphalt contents, stability, loss of stability, flow and stiffness index of mixes has
been reported in Table 4.4.
Table 4.4: Hot Mix Asphalt Design Properties
Mix
Types
1a
(PMA)
1b
(60/70)
1c
(40/50)
2a
(PMA)
2b
(60/70)
2c
(40/50)
Optimum
AC
Contents
(%)
3.83
Stability
(Kg)
Loss of
Stability
(%)
Flow
(0.25mm)
Stiffness index
(Stability/flow)
1378
11.00
10.80
128
3.87
1305
14.30
11.00
119
3.95
1356
12.70
10.50
129
4.29
1335
8.90
9.80
136
4.31
1298
11.00
11.30
115
4.33
1314
10.50
10.50
125
Stiffness index of mix ‘1c’ and ‘1a’has the same value. Mix ‘2a’ showed higher value; while
mixes ‘1b’ & ‘2b’ showed the lowest.
53
4.5
Mineral Filler Optimization
Mineral fillers are the part of mineral aggregates, they fill interstices and provide contact points
between larger aggregates particles and thereby strengthen the mixture. Utmost efforts were
made to ensure less organic material passing sieve no. 200 (75µm) having plasticity index (PI)
less than 4. Three percentages of filler i.e. 2.4%, 3.4%. & 4.4% in Marshall Method of mix
design were used in order to determine the optimum filler content for asphalt mixes. Mineral
filler to AC ratio determined through trials have been reported in Table 4.5, which shows that
the filler to asphalt content ratio ranges from 0.9 to 1.0 in asphalt mixes.
Table 4.5: Mineral Filler to Asphalt Cement Ratio
Sr.
No.
Description
Mixes with aggregate
Mixes with aggregate
gradation “01”
gradation “02”
Mix
Mix
Mix
Mix
Mix
Mix
“1a”
“1b”
“1c”
“2a”
“2b”
“2c”
1
Optimum Asphalt Content (%)
3.83
3.87
3.95
4.29
4.31
4.33
2
Filler Contents (%)
3.45
3.48
3.56
4.29
4.31
4.33
3
Filler to AC ratio (%)
0.9
0.9
0.9
1.0
1.0
1.0
4.6
Asphalt Mixture Volumetric
Mixes probable durability and in-service performance were determined by analyzing a
compacted paving mixture for air voids (Va), voids in mineral aggregates (VMA) and voids
filled with asphalt (VFA), commonly known as the volumetric properties of a compacted
paving mixture and represented in Table 4.6. The consequences of a change in any of the
volumetric factors or any detour in the procedure that offsets the total process will be a loss of
performance or service life. Mixtures that ultimately consolidate to less than three percent air
voids can be expected to rut and shove if placed at heavy traffic locations (Asphalt Institute,
2003).
54
Table 4.6: Hot Mix Asphalt Design Volumetric Properties
Gsb
Gmm
Gmb
1a (PMA)
Optimum AC
Contents (%)
3.83
2.650
2.522
1b (60/70)
3.87
2.65
1c (40/50)
3.95
2a (PMA)
2.373
Va
(%)
5.90
VMA
(%)
13.90
VFA
(%)
58.00
2.515
2.371
5.70
13.99
59.00
2.65
2.514
2.370
5.70
14.10
60.00
4.29
2.662
2.520
2.395
4.96
13.90
64.00
2b (60/70)
4.31
2.662
2.516
2.386
5.17
14.23
63.70
2c (40/50)
4.33
2.662
2.515
2.384
5.20
14.32
63.70
Mix Types
4.6.1
Specific Gravity
According to Asphalt Institute (2003) “The ratio of the mass in air of a unit volume of a
permeable material (including both permeable and impermeable voids in the aggregates) at a
stated temperature to the mass in air of equal density of an equal volume of gas-free distilled
water at a stated temperature is known as bulk specific gravity (Gsb) for the total aggregates”.
The ratio of the mass in air of a unit volume of an impermeable material at a stated temperature
to the mass in air of equal density of an equal volume of gas-free distilled water at a stated
temperature is known as apparent specific gravity (Gsa). While, the ratio of the mass in air of a
unit volume of a permeable material (excluding voids permeable to asphalt) at a stated
temperature to the mass in air of equal density of an equal volume of gas-free distilled water at
a stated temperature is the effective specific gravity (Gse) of aggregates.
The effective specific gravity of the aggregate is assumed constant because absorption does not
vary appreciably with change in asphalt contents. The maximum specific gravity (Gmm) at
different asphalt contents was measured to calculate air voids as shown in Table 4.6. The
volume of asphalt binder absorbed by an aggregate would be less than the volume of water
absorbed. Consequently, the value for the effective specific gravity of an aggregate is between
its bulk and apparent specific gravities. (Asphalt Institute, 2003).
55
4.6.2
Voids in Mineral Aggregates, VMA
According to Asphalt Institute (2003), the voids in the mineral aggregates, are defined as the
intergranular void space between the aggregates particles in a compacted paving mixture that
includes the air voids and the effective asphalt content, expressed as a percent of the total
volume of the sample. The VMA are calculated based on the bulk specified gravity of the
aggregates and is expressed as a percentage of the bulk volume of the compacted paving
mixture, given as follow and reported in Table 4.6.
VMA
= 100
−
G
mb
G
sb
× Ps
(4.1)
Where
VMA
= voids in the mineral aggregates, percent of bulk volume.
Gsb
= bulk specific gravity of total aggregates
Gmb
= bulk specific gravity of compacted mixture
Ps
= aggregates content, percent by total mass of mixture.
The most difficult mix design property to achieve is a minimum amount of voids in the mineral
aggregates. The goal was to furnish enough space for the asphalt cement so it could provide
adequate adhesion to bind the aggregates particles, but without bleeding when temperatures
rise and the asphalt expands. Normally, the VMA decreases to a minimum value with increase
in asphalt content. At some point as the asphalt content increases, the VMA begins to increase
because relatively more dense material (aggregates) is displaced and pushed apart by the less
dense material (asphalt content). The asphalt content on the “wet” side of VMA curve was
avoided, even the minimum air void and VMA criteria met. Design asphalt contents in this
range have a tendency to bleed and /or exhibit plastic flow when placed in the field. Any
amount of additional compaction from traffic leads to inadequate room for asphalt expansion,
loss of aggregates-to-aggregates contact, and eventually, rutting and shoving in high traffic
areas Asphalt Institute (2003).
56
4.6.3 Air Voids, Va.
According to Asphalt Institute (2003), the total volume of the small pockets of air between the
coated aggregates particles throughout a compacted paving mixture, expressed as percent of
the bulk volume of the compacted paving mixture. The air voids, Va, in the total compacted
paving mixture consists of the small air spaces between the coated aggregates particles. The
mathematical relationship has shown as;
V a = 100
×
G mm − G
mb
G mm
(4.2)
Where,
Va
=
air voids in compacted mixture, percent of total volume.
Gmm
=
maximum specific gravity of paving mixture (as determined in
previous article or as determined directly for a paving mixture by
ASTM D2041/AASHTO T209)
Gmb
=
bulk specific gravity of compacted mixture
The design range of air voids of mixes were kept from 4 to 7 percent for heavy traffic. The
overall objective was to limit adjustments of the design asphalt content to less than 0.5 percent
air voids from the median of the design criteria (four percent), especially on the low side of the
range to minimize chances of rutting in the field (Asphalt Institute, 2003).
4.6.4
Voids Filled with Asphalt, VFA.
According to Asphalt Institute (2003), the percentage portion of the volume of intergranular
void space between the aggregates particles that is occupied by the effective asphalt. It is
expressed as the ratio of (VMA-Va) to VMA. The voids filled asphalt, VFA is the percentage
of the integral void space between the aggregates particles (VMA) that are filled with asphalt.
The mathematical relationship has shown as;
57
VFA
= 100
×
VMA
− V a
VMA
(4.3)
Where,
VFA
= voids filled with asphalt, percent of VMA
VMA = voids in the mineral aggregates, percent of bulk volume
Va
= air voids in compacted mixture, percent of total volume.
Volumetric analysis i.e. specific gravity, maximum theoretical specific gravity, air voids, voids
in mineral aggregates and voids filled with asphalt, determined from each mix type have been
reported in Table 4.6. due to its high fluidity, PMA produced relatively high-density mixes
than neat AC for the same gradation.
The amount of effective asphalt cement, which ultimately affects the amount of air voids,
utilized within the asphaltic mixture has closely been controlled to produce desired results.
Void volume control in the mixture is critical. Insignificant variations in the effective
proportion of asphalt cement resulting from variations in aggregate absorption characteristics
may substantially affect the life and quality of the pavement.
Additionally, because aggregate from different rock quarries and pit locations have varying
absorptive qualities, representative samples during the testing process were carefully selected
and analyzed in order to produce an even uniform pavement mix. After successful design of
HMA mixtures, performance testing was carried out to investigate their resistance to
permanent deformation under repeated loading.
58
Chapter Five
59
Chapter Five
Uniaxial Repeated Load Strain Test
5.1
Introduction
Uniaxial Repeated Load Strain Test (URLST) conforms to the requirements of the design draft
issued by the British Standard Institute (BSI) as a method of measurement of resistance of
asphalt mixes to permanent deformation behaviour, subjected to unconfined uniaxial cyclic
loading on Universal Testing Machine (UTM-5P). A comprehensive laboratory investigation
was carried out on six specified mixes to study the permanent deformation of asphalt concrete
mixes at different temperatures and stress levels. The main objective of this chapter is to
investigate the performance of HMA material in terms of resilient & permanent strains under
cyclic loading conditions.
5.2
Universal Testing Machine (UTM-5P)
UTM as shown in Figure 5.1, is a series of closed loop servo control material testing machine
designed to transmit energy to the HMA specimens using high pressure air acting on double
sided piston (actuator). The actuator is mechanically coupled to the specimen through a
reaction loading frame. Air pressure is controlled by a servo valve where in small electric
currents are used to open and close the control spool of the valve. These transducers convert
mechanical movement into standard electronic signals and via Control and Data Acquisition
System (CDAS) displays output on the personal computer. The transducers signals are also
used to control the system.
60
Figure 5.1: Universal Testing Machines (UTM-5P)
5.3
Uniaxial Repeated Load Strain Test (URLST)
In UTM-5P, Controlled strain test applied 1800 block (square) repeated load pulses, with a
pulse width and pulse period of 500 milli second and 2000 milli second respectively, to the
specimens to make maximum damage to the specimens. With these stress levels (100 kPa, 300
kPa, 500 kPa) and temperatures (250C, 400C, & 550C), shear mode of failure for some
specimens were observed beyond 1800 cycles (K.B. De Vose and Feeley, 2002). A static
conditioning time of 100 sec and conditioning stress of 10kPa was also applied prior to
commencement of actual test. Following the conditioning period, a fixed twenty seconds rest
period was programmed where the applied stress was set to zero.
Data is collected from the loading pulses at a linear (i.e. equidistant) time interval and then
stored in a buffer. Digital filtering is applied to the data at various stages of its capture and
processing to ensure smoothness in the reported and displayed data. As test proceeds, plotted
61
data is displayed with linear vertical and horizontal axis in terms of accumulated strain,
derivation of accumulated strain (slope), resilient strain, resilient modulus and creep stiffness.
5.4 Test Conditions
5.4.1
Load Conditions
Specimens were subjected to repeated pulse loading of 1800 cycles at 100, 300 and 500 kPa
stress levels. Pulse width and pulse period were kept 500 milli second and 2000 milli second
respectively. A range of stresses from 100 kPa to 500 kPa were selected for this study to cover
almost all the domains of loads plying on the road.
The magnitude of the loading stress and the timing width of the pulse applied to the specimen
during a test were directly controlled by the “test loading stress” and “pulse width” parameters
in the test set up and control edit screen. A minimum force of 20N ensured that the loading
actuator would not be lifted off the specimen between pulse applications. As pulse loading
continued, the permanent deformation in terms of accumulated strain was measured using two
Linear Variable Displacement Transformers (K.B. De Vose and Feeley, 2002).
5.4.2
Temperature Conditions
The typical test temperatures i.e. 250C, 400C, & 550C, are adopted in the laboratory for
research study to accommodate the influence of temperature conditions in summer expected on
National Highways. It is the representative of typical temperature conditions observed on
flexible pavements in Pakistan
5.5
Testing Methodology
For each stress level, three specimens were tested at 250C, 400C and 550C. Samples are
evaluated for 54 test conditions (27 for each type of gradation) and total 162 specimens are
prepared. Test matrix for different conditions, used in the experimental study has been shown
in Table 5.1.
62
Table 5.1: Specimens at each Test Condition
Stress
(kPa)
Temperature
(0C)
Mixes with Gradation 01
1a
1b
1c
100
25
300
500
100
300
500
100
300
500
40
55
5.6
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Mixes with Gradation 02
2a
2b
2c
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Discussion of URLST Results
Influence of pulse counts (1800 cycles) on accumulated strains for six types of mixes (1a, 1b,
1c, 2a, 2b, 2c) at three temperatures (250C, 400C, & 550C) and three stress levels (100 kPa, 300
kPa, & 500 kPa) have been shown in Annexure-A and summarized in Table 5.2.
Table 5.2: Mean Accumulated Strain (εp) of Mixes
Sr.
Temperature
Stress
Percentage Accumulated Strain (εp) of Mixes
No.
(0C)
(kPa)
1a
1b
1c
2a
2b
2c
1
25
100
0.193
0.281
0.183
0.286
0.403
0.315
2
25
300
0.399
0.564
0.375
0.516
0.572
0.493
3
25
500
0.616
0.686
0.5926
0.907
0.958
0.926
4
40
100
0.332
0.424
0.389
0.540
0.590
0.536
5
40
300
0.547
0.742
0.666
0.609
0.774
0.676
6
40
500
0.881
0.990
0.946
0.989
1.057
0.995
7
55
100
0.438
0.881
0.526
0.833
0.941
0.747
8
55
300
1.114
1.32
1.124
1.052
1.487
1.058
9
55
500
1.242
1.692
1.247
1.320
1.951
1.375
63
Table 5.2 show that permanent strain increases with the increase in stress levels and
temperature. The effect of temperature variation is significant as compared to stress levels.
Results of Table 5.2 have been shown graphically in Figure 5.2.
Mixture with Gradation "01" & PMA
(1a)
1.30
Mixture with Gradation "02" & PMA
(2a)
1.50
1.30
1.10
500kPa
1.10
0.90
500kPa
0.90
Єp (% )
Єp (%)
0.70
300kPa
0.70
300kPa
0.50
0.50
100kPa
100kPa
0.30
0.30
0.10
0.10
10
25
0
Temperature ( C)
40
100kPa
55
300kPa
10
500kPa
Mixture with Gradation "01" & 60/70 Pen. Grade
(1b)
1.90
25
0
Temperature ( C)
1.90
1.50
1.70
55
500kPa
300kPa
Mixture with Gradation "02" & 60/70 Pen. Grade
(2b)
2.10
1.70
40
100kPa
1.50
1.30
500kPa
0.90
300kPa
0.70
0.50
500kPa
1.30
Єp (%)
Єp (%)
1.10
100kPa
1.10
0.90
300kPa
0.70
100kPa
0.50
0.30
0.30
0.10
10
25
40
0.10
55
10
0
Temperature ( C)
300kPa
500kPa
Mixture with Gradation "01" & 40/50 Pen. Grade
(1c)
1.50
25
0
Temperature ( C)
1.30
1.30
1.10
1.10
40
100kPa
300kPa
55
500kPa
Mixture with Gradation "02" & 40/50 Pen. Grade
(2c)
1.50
500kPa
0.90
0.90
500kPa
Єp (%)
Єp (%)
100kPa
0.70
300kPa
0.50
300kPa
0.70
100kPa
0.50
100kPa
0.30
0.30
0.10
10
25
0
Temperature ( C)
40
100kPa
0.10
55
300kPa
10
500kPa
25
0
Temperature ( C)
40
100kPa
300kPa
Figure 5.2: Influence of Temperature and Stress Levels on Accumulated Strain
64
55
500kPa
Figure 5.2 shows that permanent strain (εp) in all mixes increases from low stress level (100
kPa) to high stress level (500kPa) at low temperature (250C). At low temperature and high
stress level, densification of mixes may lead to a decreasing trend. At medium temperature
(400C), εp has a constant rate of increase with the increase in stress. The stiffness offered by the
mix at this specific temperature may be combined effects of both the gradation and the binder.
At higher temperature (550C), the trends are likely the same as at low temperatures i.e. εp
increases at the decreasing rate, but higher than low temperature. The reasons may be the shift
of resistance offered by binder to aggregate skeleton.
Results of percentage accumulated strain drawn against pulse counts were multiplied by one
million to obtain positive values on logarithmic scale according to Figure 2.12 in section 2.9.
Straight line trends were developed as shown in Annexure-B. Equations for each stress level
have been shown on the respective figures. Where in percentage accumulated strain is denoted
“y”, pulse counts with “x” and regression coefficients with “a” and “b”, which are namely
intercept coefficients and slope coefficients respectively. Detail discussion of regression
coefficients have been given in chapter Seven.
5.7
Resilient strain
Resilient strain is the recoverable strain phase after a loading event. Resilient strain is mainly
affected by the density of mix, moisture content, temperature, particle size and shape,
aggregate gradations, lateral confinement, loading condition and bulk stress. Resilient strain
can be calculated using the following relation;
Єr
=
∆r / havg.
Єr
=
resilient strain
∆r
=
resilient deformation
=
∆r,h - ∆r,I
havg.
=
average height of test specimen
∆r,h
=
high deformation reading for the LVDT and
∆r,I
=
low deformation reading for the LVDT
(5.1)
Where
65
The results of resilient strain of asphalt mixes under applied uniaxial loads have been plotted
graphically and shown in Annexure-C, in order to make a comparison among gradations,
asphalt types, temperatures and stress levels on resilient strain values.
Resilience is the property of a material to absorb energy when it is deformed elastically and
then, upon loading to have this energy recovered. In other words, it is the maximum energy per
volume that can be elastically stored. Results of resilient strain have been shown in Table 5.3,
which shows strong correlations among all the four variables. Resilient strain increases with
the increase in stress levels and temperatures similar to accumulative strain as shown in Figure
5.3 through 5.8. It was observed that PMA and 40-50 penetration grade asphalt mixes has
almost equal resilient strain with gradation “01”.
Table 5.3: Resilient Strain of Mixes
Sr.
Temp.
Stress
No.
(0C)
(kPa)
1
25
100
2
25
300
3
25
500
4
40
100
5
40
300
6
40
500
7
55
100
8
55
300
9
55
500
Resilient Strain of Mixes
1a
1b
1c
2a
2b
2c
0.0364
0.0574
0.0346
0.0231
0.0454
0.0217
0.0450
0.0728
0.0423
0.0553
0.0750
0.0601
0.0698
0.0768
0.0635
0.0884
0.1096
0.0928
0.0461
0.0620
0.0509
0.0308
0.0680
0.0254
0.0487
0.0792
0.0620
0.0647
0.0826
0.0672
0.0757
0.0809
0.0732
0.1076
0.1312
0.1109
0.0568
0.0629
0.0580
0.0394
0.0826
0.0524
0.0622
0.1185
0.0734
0.0885
0.1287
0.1018
0.1011
0.1177
0.1044
0.1157
0.1677
0.1339
66
Table 5.3 shows that at low stress level (100 kPa) mixes with coarser gradation exhibits more
resilient strain and at high stress level (500 kPa) mixes with finer gradation expressed more
strain. The main reason behind is the total strain developed under same stress conditions.
Mixes with fine gradation has more value of total strain, hence the amount of recoverable
strain has also higher value than coarse mixes. It has also been observed in Table 5.3 that at
higher temperatures (400C & 550C) the percentage difference in rate of recoverable strain or
Resilient Strain( %)
resilient strain reduces. These results have further been elaborated in Figure 5.3 through 5.8.
0
Mixture with Gradation "01" & PMA
(1a)
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0
100
200
300
400
500
55 C
0
40 C
0
25 C
600
Stress (kPa)
Figure 5.3: Influence of Stress Levels on Resilient Strain, Mix 1a
Resilient Strain( %)
Mixture with Gradation "01" & 60/70 Pen. Grade
(1b)
0.13
0
55 C
0.12
0.11
0.10
0
0.09
40 C
0.08
0
25 C
0.07
0.06
0.05
0.04
0
100
200
300
400
Stress (kPa)
500
Figure 5.4: Influence of Stress Levels on Resilient Strain, Mix 1b
67
600
Resilient Strain( %)
Mixture with Gradation "01" & 40/50 Pen. Grade
(1c)
0.11
0
55 C
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0
40 C
0
25 C
0
100
200
300
400
500
600
Stress (kPa)
Figure 5.5: Influence of Stress Levels on Resilient Strain, Mix 1c
Mixture with Gradation "02" & PMA
(2a)
0.14
0
55 C
Resilient Strain( %)
0.12
0
40 C
0.1
0.08
0
25 C
0.06
0.04
0.02
0
0
100
200
300
400
500
Stress (kPa)
Figure 5.6: Influence of Stress Levels on Resilient Strain, Mix 2a
68
600
Mixture with Gradation "02" & 60/70 Pen. Grade
(2b)
0.18
0
55 C
Resilient Strain( %)
0.16
0
40 C
0.14
0.12
0
25 C
0.1
0.08
`
0.06
0.04
0.02
0
0
100
200
300
400
500
600
Stress (kPa)
Figure 5.7: Influence of Stress Levels on Resilient Strain, Mix 2b
Resilient Strain( %)
Mixture with Gradation "02" & 40/50 Pen. Grade
(2c)
0.16
0
55 C
0.14
0
0.12
40 C
0.10
0
25 C
0.08
0.06
0.04
0.02
0.00
0
100
200
300
400
500
Stress (kPa)
Figure 5.8: Influence of Stress Levels on Resilient Strain, Mix 2c
69
600
5.8
Summary of Results
A comprehensive laboratory investigation was carried out on six mixes ranging from finer to
coarser. Specimens were subjected to cyclic loading on UTM-5P to study the resistance against
permanent deformation of the mixes at 250C, 400C and 550C. At low temperatures and stress
levels, both coarse and fine graded mixes showed less accumulated strain, whereas at higher
temperatures and stress levels, coarse graded mix with PMA showed good resistance to
permanent deformation.
Mixes with coarser gradations offered more resistance to rutting than mixes with finer
gradations under URLST. Overall performance of mixes with penetration grade “60/70”
bitumen were lower as compared to mixes with penetration grade “40/50” and PMA.
70
Chapter Six
71
Chapter Six
Wheel Tracking Test
6.1
Introduction
Wheel Tracker typically measures the rut, created by repeated passage of a wheel over
prismatic asphalt concrete samples. It was used to assess the resistance to rutting of the
asphaltic material, under standard defined conditions of load and temperatures i.e. 25, 40 and
55 0C. This chapter presents the relationship established between rutting and factors effecting
rutting. Ranges of intercept and slope coefficient were also determined.
6.2
Wheel Tracking Device
A loaded wheel shown in Figure 6.1 tracks a sample under specified conditions of speed and
temperature while the development of the rut is monitored continuously during the test.
Figure 6.1: Wheel Tracker (After Cooper, 2006)
72
The rut resistance can be quantified as the rate of rutting during the test or the rut depth at the
conclusion of the test, measured with Linear Variable Displacement Transformers (LVDT)
25mm (min). Slab specimens were prepared in the laboratory for research study. The
susceptibility of an asphaltic material to deform is assessed by measuring the rut formed by
repeated passes of a loaded wheel at specified temperatures. The wheel tracking apparatus
consists of loaded wheel which bears on a sample held on a moving table as shown in Figure
6.2. The moving table reciprocate with simple harmonic motion with a frequency of 26.5
passes per minutes (European Standard- PrEN 13108/12697-22, 2002).
Figure 6.2: Wheel Tracker Solid Rubber Tyre (After Cooper, 2006)
The wheel is fitted with solid rubber tyre of outside diameter 200 mm. The tyre is a rectangular
section 50 ± 1 mm wide and 10mm to 13 mm thick. The wheel tracker is fitted with a
temperature controlled cabinet with a maximum temperature up to 650C ± 10C. Square slab
specimens (305x305mm) of asphalt mixes with typical asphalt wearing course thickness of
50mm thick, fitted with wheel tracker (WT) table, clamps for securing specimen holders.
Mixes were evaluated under a loaded wheel (700 ± 20 N) tracked with simple harmonic motion
73
through a distance of 305mm on specimens under specified conditions i.e. 53 passes per
minute at temperatures 250C, 400C and 550C (European Standard- PrEN 13108/12697-22,
2002).
The operational software run under Windows 95 to start and stop the WT, control speed and
acquire deformation and temperature data. An on-screen display provides a continuously
updated graph of time versus deformation as shown in Figure 6.3. The test data are stored in a
text file for subsequent analysis using a spreadsheet.
Figure 6.3: Wheel Tracker out put Display (After Cooper, 2006)
74
6.3
Specimen Preparation on Roller Compactor
Specimens of each type of asphalt concrete mix were prepared with the help of roller
compacter at specific temperatures and number of specimen for each condition is given in
Table 6.1. Weight of each specimen was taken 12.50 kg (approximately). Mixing of aggregate
with binder was carried out at a working temperature of 1550C + 50C.
Table 6.1: Details of Specimens
Number of Specimens for Each Mix Type
Gradation 01
Gradation 02
1b
1c
2a
2b
Temperature
(0C)
1a
25
3
3
3
3
3
3
40
3
3
3
3
3
3
55
3
3
3
3
3
3
2c
Roller Compactor compacts the specimen uniformly with its specially designed press system to
a specified pressure as shown in Figure 6.4.
Figure 6.4: Slab Compaction on Roller Compacter (After Cooper, 2006)
75
For each temperature, compaction of slab specimens at roller compactor was undertaken in
four stages. Pressure at each stage and number of passes corresponding to each pressure have
been reported in Table 6.2.
Table 6.2: Compaction Pressures
1
Compaction Stages
2
3
4
Pressure (bar)
2
5
4
3
No. of Passes
10
10
5
5
Wheel tracker tracked for 18000 pulses or 25mm rut depth, which ever happens first, at
standard conditions of applied load (720N), frequency (26.5rpm) on compacted confined slab
specimens for six mixes. Rut development of mixes under wheel tracking test was analyzed
and summary of rut depth, measured at three temperatures have been reported in Table 6.3.
Table 6.3: Rut Depth of Mixes Measured on Wheel Tracker
Temperature
Sr.
(0C)
No.
Measured Rut Depth (mm) of Mixes
“1a”
“1b”
“1c”
“2a”
“2b”
“2c”
1
25
2.74
3.90
2.82
4.53
5.99
4.60
2
40
6.20
9.99
6.62
10.86
14.60
12.08
3
55
8.53
15.20
11.61
17.80
23.40
19.00
Table 6.3 shows that mixes with coarser aggregate gradation has lesser value of rut depth as
achieved by mixes with finer gradation, irrespective of AC type. At 250C & 400C temperatures,
mixes with PMA and 40/50 penetration grade AC have shown almost similar rut potential, but
at 550C mixes with PMA has the least value of rut depth among all the mixes. Percentage
difference in rut potential of mixes with aggregate gradation “01” with three AC types is more
than mixes with gradation “02”. The development of rut in asphalt concrete specimens under
the wheel pass of the Wheel Tracker has been shown in Figure 6.5.
76
Figure 6.5: Rut Development in Mixes
77
6.4
Discussion of Results
Influence of load cycles on rut depth was measured and plotted using the data out put file and
shown graphically in Annexure-D. Results of WT show that permanent deformation is a
function of number of load repetitions. At high temperature (550C), PMA mix shows better
resistance to rutting as total rut depth achieved is only 8.5 and 17.8 mm in mixes with
gradation “01” and “02” respectively. Similarly, rut depth of 60/70 pen. grade & 40/50 pen
grade asphalt mixes was 15.2mm & 11.61mm with aggregate gradation “01”, while 23.4mm &
19.00mm with aggregate gradation “02” respectively.
Rut development on WT is further plotted on log-log scale as discussed earlier in section 2.9
and shown in Figure 2.12 and the results are expressed in Figure 6.6 through 6.11. Straight line
trends have shown good agreement (R2). Relationships between load repetitions and rut depth
can be measured at any value of N using these plots.
Mixture with gradation "01" & PMA
(1a)
7.5
y 25 = 0.2694Ln(x) + 3.7261
R2 = 0.9524
6
Rut Depth (mmx10 )
7.0
6.5
6.0
5.5
5.0
y 55 = 0.2231Ln(x) + 4.7724
R2 = 0.9944
y 40 = 0.2037Ln(x) + 4.8079
R2 = 0.9965
4.5
4.0
10
1a, 25C
100
1a, 40C
1000
Load Cycle (LOG N)
1a, 55C
Log. (1a, 25C)
10000
Log. (1a, 40C)
Figure 6.6: Relationship of Load Cycle & Rut Depth, Mix 1a
78
100000
Log. (1a, 55C)
Mixture with gradation "01" & 60/70 Pen. Grade
(1b)
7.5
y 25= 0.2737Ln(x) + 3.9131
R2 = 0.9838
6
Rut Depth (mmx10 )
7.0
6.5
6.0
5.5
5.0
y 55 = 0.2174Ln(x) + 5.1163
R2 = 0.9591
y 40 = 0.2295Ln(x) + 4.8107
R2 = 0.9452
4.5
4.0
10
100
1b, 25C
1b, 40C
1b, 55C
1000
Load Cycle (LOG N)
10000
Log. (1b, 25C)
Log. (1b, 40C)
100000
Log. (1b, 55C)
Figure 6.7: Relationship of Load Cycle & Rut Depth, Mix 1b
Mixture with gradation "01" & 40/50 Pen. Grade
(1c)
7.5
y 25 = 0.2957Ln(x) + 3.4551
R2 = 0.3927
7.0
6
Rut Depth (mmx10 )
6.5
6.0
5.5
5.0
4.5
y 55 = 0.2844Ln(x) + 4.3368
R2 = 0.9809
y 40 = 0.2111Ln(x) + 4.7834
R2 = 0.9909
4.0
10
1c 25C
100
1c, 40C
1c, 55C
1000
Load Cycle (LOG N)
Log. (1c 25C)
Log. (1c, 40C)
Figure 6.8: Relationship of Load Cycle & Rut Depth, Mix 1c
79
10000
100000
Log. (1c, 55C)
Mixture with gradation "02" & PMA
(2a)
7.5
y 25= 0.1317Ln(x) + 5.3749
R2 = 0.9926
6
Rut Depth (mmx10 )
7.0
6.5
6.0
5.5
y 40 = 0.2161Ln(x) + 4.9767
R2 = 0.9676
5.0
10
2a, 25C
100
2a, 40C
y 55 = 0.2217Ln(x) + 5.1177
R2 = 0.9875
1000
Load Cycle (LOG N)
2a, 55C
Log. (2a, 25C)
10000
Log. (2a, 40C)
100000
Log. (2a, 55C)
Figure 6.9: Relationship of Load Cycle & Rut Depth, Mix 2a
Mixture with gradation "01" & 60/70 Pen. Grade
(2b)
8.0
y 25 = 0.1633Ln(x) + 5.1857
R2 = 0.9869
6
Rut Depth (mmx10 )
7.5
7.0
6.5
6.0
5.5
y 55 = 0.2118Ln(x) + 5.3468
R2 = 0.9691
y 40 = 0.1968Ln(x) + 5.2741
R2 = 0.9801
5.0
10
2b, 25C
100
2b, 40C
1000
Load Cycle (LOG N)
2b, 55C
Log. (2b, 25C)
10000
Log. (2b, 40C)
Figure 6.10: Relationship of Load Cycle & Rut Depth, Mix 2b
80
100000
Log. (2b, 55C)
Mixture with gradation "01" & 40/50 Pen. Grade
(2c)
8.0
7.5
y 25 = 0.2606Ln(x) + 4.1251
R2 = 0.9942
6
Rut Depth (mmx10 )
7.0
6.5
6.0
5.5
5.0
y 55 = 0.239Ln(x) + 5.0268
R2 = 0.9165
y 40 = 0.199Ln(x) + 5.1743
R2 = 0.9824
4.5
4.0
10
2c, 25C
100
2c, 40C
1000
Load Cycle (LOG N)
2c, 55C
Log. (2c, 25C)
10000
100000
Log. (2c, 40C)
Log. (2c, 55C)
Figure 6.11: Relationship of Load Cycle & Rut Depth, Mix 2c
For quick review, summary of intercept coefficients and slope coefficients from Figure 6.6
through 6.11 has been reported in Table 6.4 & 6.5 respectively. It shows that ranges of
intercept coefficient “a” and slope coefficient “b” for all the six mixes were observed from
3.46 to 5.38 and 0.16 to 0.30 respectively.
Table 6.4: Intercept Coefficient “a” of Mixes
Sr.
No.
Temp.
(0C)
1
25
2
3
40
55
Intercept Coefficient “a” of Mixes
“1a”
“1b”
“1c”
“2a”
“2b”
“2c”
3.73
4.81
4.77
3.91
4.81
5.12
3.46
4.78
4.34
5.38
4.98
5.12
5.19
5.27
5.35
4.13
5.17
5.03
Table 6.5: Slope Coefficient “b” of Mixes
Sr.
No.
1
2
3
Temp.
(0C)
25
40
55
“1a”
0.27
0.21
0.22
“1b”
0.27
0.23
0.22
Slope Coefficient “b” of Mixes
“1c”
“2a”
“2b”
0.30
0.13
0.16
0.21
0.22
0.20
0.28
0.22
0.21
81
“2c”
0.26
0.20
0.24
Chapter Seven
82
Chapter Seven
Results and Discussions
7.1
Introduction
Comprehensive performance based testing on several specimens were carried out after
complete characterization of materials and determinations of HMA properties as described in
the previous chapters. One of the objectives of the testing was to investigate the effect of HMA
properties on its permanent deformation behaviour. The evaluation of the resistance to
permanent deformation of six mixes was carried out using both the test methods. This chapter
comprised of calculating the regression coefficients i.e. intercept and slope coefficient and
Permanent Deformation Coefficients alpha and mu. A brief introduction of these coefficients
has already been provided in chapter Two. Also, computation of domains of “α” & “µ” and its
comparison with other researcher has been reported in this chapter. In addition, a comparison
of both the test have also been made using the shift factors techniques, ranking of mixes (based
on intercept coefficients) and a correlation between both the test methods have also been made
a part of this study.
7.2
Regression Coefficients
The derivation of regression coefficients i.e. intercept coefficients “a” & slope coefficients “b”
are based on log-log scale between load repetition and permanent strain as already explained in
section 2.9. Relationships have been plotted for six mixes under each temperature and stress
conditions as reported in Annexure-D. This section provides a comparison of results obtained
for each coefficient.
7.2.1 Intercept Coefficient (a)
Intercept coefficient (a) of mixes for each temperature and stress condition have given in
83
Table 7.1, 7.2, & 7.3. One can observe the accuracy of computed data from the Standard
Deviation (SD) & Coefficient of Variance (CV)
Table 7.1: Intercept Coefficient ‘a’ at 250C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
4.07
4.74
5.13
4.41
4.91
5.29
4.55
4.91
5.52
4.06
4.66
5.08
4.53
4.92
5.25
4.86
5.21
5.25
0.06
0.06
0.02
Coefficient
of
Variance
1.3
1.2
0.4
Table 7.2: Intercept Coefficient ‘a’ at 400C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
4.77
5.15
5.25
4.73
4.89
5.07
4.93
5.15
5.25
5.06
5.14
5.36
5.15
5.13
5.31
4.91
5.07
5.10
0.06
0.04
0.04
Coefficient
of
Variance
1.22
0.74
0.78
Table 7.3: Intercept Coefficient ‘a’ at 550C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
4.73
4.82
4.98
4.82
5.06
5.32
4.82
4.97
5.09
4.78
4.83
5.05
5.02
5.07
5.48
4.83
4.99
5.32
0.04
0.04
0.07
Coefficient
of
Variance
0.74
0.83
1.39
From Table 7.1, 7.2 and 7.3, it can be observed the intercept coefficient increases with an
increase in stress level, irrespective of aggregate gradations, A.C types, mix types, and test
temperatures. Domains of ‘a’ for temperature 250C, 400C and 550C are 4.07 to 5.52, 4.73 to
5.36, and 4.733 to 5.481 respectively. Irrespective of test temperature and stress levels, domain
of mixes with aggregate gradation “01”was observed to be 4.06 to 5.52 and for mixes with
gradation “02” was 4.06 to 5.481. Standard deviation measured for ‘a’ has a range from 0.02 to
0.0723.
84
7.2.2
Slope Coefficient (b)
Slope coefficient shows a reverse trend unlike intercept coefficient as tabulated in Table 7.4, 7.5
and 7.6.
Table 7.4: Slope Coefficient ‘b’ at 250C
Mixes with gradation "01"
Stress
Level
1a
1b
1c
100
300
500
0.20
0.12
0.10
0.15
0.10
0.08
0.15
0.09
0.07
Mixes with gradation "02"
2a
2b
2c
Standard
Deviation
0.20
0.14
0.12
0.15
0.11
0.10
0.10
0.08
0.08
0.02
0.01
0.01
Coefficient
of
Variance
9.33
7.52
7.41
Table 7.5: Slope Coefficient ‘b’ at 400C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
0.12
0.08
0.08
0.13
0.10
0.09
0.18
0.12
0.10
0.10
0.09
0.09
0.11
0.10
0.09
0.17
0.12
0.11
0.01
0.01
0.01
Coefficient
of
Variance
8.76
6.18
4.20
Table 7.6: Slope Coefficient ‘b’ at 550C
Stress
Level
100
300
500
Mixes with gradation "01" Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
0.16
0.15
0.12
0.16
0.14
0.10
0.19
0.16
0.13
0.16
0.13
0.11
0.13
0.11
0.10
0.16
0.12
0.11
0.004
0.007
0.005
Coefficient
of
Variance
8.76
6.18
4.61
It decreases with the increase in stress level, irrespective of aggregate gradations, A.C types, mix
types, and test temperature. For all the mixes, the domain is higher and percentage difference of
‘b’ is more at low temperature (250C) and stress level (100 kPa) than higher temperature (550C)
and stress levels (500 kPa). Mixes with “PMA” and “40/50” penetration grade have higher value
of ‘b’ as compared to “60/70” penetration grade. Overall domain of ‘b’ for mixes with gradation
“01” and “02” is 0.07 to 0.20 and 0.08 to 0.20 respectively. Standard deviation observed for “b”
has a range from 0.004 to 0.02.
85
7.3
Permanent Deformation Coefficients
Two very use full coefficients i.e. Alpha (α) & Mu (µ) have been computed in this section from
straight line trends reported in Annexure-D, by using the derivation of base log-log scale of
power model as explained earlier in section 2.9. This section provides a comparison of results
obtained for each coefficient.
7.3.1
Alpha (α)
Permanent deformation coefficient (α) as given in Tables 7.7, 7.8 and 7.9 and graphically shown
in Figure 7.1 has similar trends as intercept coefficient i.e. with an increase in stress level, alpha
increases.
Table 7.7: Permanent Deformation Parameter ‘α’ at 250C
Stress
Level
100
300
500
Mixes with
gradation "01"
1a
1b
1c
0.80 0.85 0.85
0.88 0.90 0.91
0.90 0.92 0.93
Mixes with gradation "02"
2a
0.80
0.86
0.88
2b
0.85
0.89
0.90
2c
0.90
0.92
0.92
Table 7.8: Permanent Deformation Parameter ‘α’ at 400C
Mixes with gradation
Mixes with gradation "02"
Stress
"01"
Level
1a
1b
1c
2a
2b
2c
100
0.88 0.87
0.82
0.91
0.89
0.85
300
0.92 0.90
0.88
0.91
0.90
0.88
500
0.92 0.91
0.90
0.92
0.91
0.89
Standard
Deviation
0.02
0.01
0.01
Coefficient
of
Variance
1.79
0.86
0.74
0.01
0.01
0.01
Coefficient
of
Variance
1.32
0.70
0.42
Standard
Deviation
Coefficient
of Variance
0.004
0.007
0.005
0.45
0.85
0.57
Standard
Deviation
Table 7.9: Permanent Deformation Parameter ‘α’ at 550C
Stress
Level
100
300
500
Mixes with gradation
"01"
1a
1b
1c
0.84
0.84
0.84
0.85
0.87
0.85
0.88
0.91
0.87
Mixes with gradation
"02"
2a
2b
2c
0.84
0.87
0.84
0.87
0.90
0.88
0.89
0.90
0.90
Overall value of ‘α’ for coarse as well as fine graded mixes has a range from 0.80 to 0.92, thus
showing minor effect of temperature and stress levels. However trends show that ‘α’ decreases
86
with increase in temperature and increases with increase in stress levels. Standard deviation
measured for ‘α’ has a range from 0.004 to 0.02.
Mixture with gradation "01" & PMA
(1a)
0.92
0.92
0.90
0.90
500 kPa
0.88
0.86
300 kPa
0.84
0.82
300 kPa
0.86
0.84
100 kPa
0.82
100 kPa
0.80
0.78
0.78
20
30
0
Temperature ( C)
40
100 kPa
50
60
300 kPa
20
0
0.92
500 kPa
300 kPa
0.86
0.84
A lpha
0.90
0.88
100 kPa
0.82
20
30
0
Temperature ( C)
40
100 kPa
50
300 kPa
60
0.86
300 kPa
0.84
0.82
100 kPa
0.80
20
30
0
Temperature ( C)
40
100 kPa
50
60
300 kPa
500 kPa
30
40
50
60
100 kPa
300 kPa
500 kPa
500 kPa
300 kPa
100 kPa
`
20
30
0
Temperature ( C)
Figure 7.1: Influence of Temperature and Stress Levels on Alpha
87
500 kPa
Mixture with gradation "02" & 40/50
(2c)
0.93
0.92
0.91
0.90
0.89
0.88
0.87
0.86
0.85
0.84
0.83
Alpha
0.88
300 kPa
100
Temperature ( C)
500 kPa
0.90
100 kPa
60
300
0
0.92
50
500
20
500 kPa
40
Mixture with gradation "02" & 60/70
(2b)
0.92
0.91
0.90
0.89
0.88
0.87
0.86
0.85
0.84
Mixture with gradation "01" & 40/50
(1c)
0.94
30
Temperature ( C)
500 kPa
Mixture with gradation "01" & 60/70
(1b)
0.94
A lp h a
500 kPa
0.88
0.80
Alpha
Mixture with gradation "02" & PMA
(2a)
0.94
Alpha
A lpha
0.94
40
100 kPa
50
300 kPa
60
500 kPa
7.3.2
Mu (µ)
Mu, as given in Table 7.10, 7.11 and 7.12 and graphically shown in Figure 7.2 has similar trends
as slope coefficient.
Table 7.10: Permanent Deformation Parameter ‘µ’ at 250C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
20.061
13.534
8.037
20.45
9.793
6.315
19.854
11.066
6.017
17.953
9.091
6.796
16.171
10.423
6.382
18.813
9.432
5.678
0.601
0.707
0.307
Coefficient
of
Variance
3.12
6.87
4.70
Table 7.11: Permanent Deformation Parameter ‘µ’ at 400C
Stress
Level
100
300
500
Mixes with gradation "01"
Mixes with gradation "02"
1a
1b
1c
2a
2b
2c
Standard
Deviation
13.865
9.190
6.393
12.276
9.283
5.820
21.235
11.636
7.083
17.078
8.250
5.817
10.499
6.731
4.603
20.369
10.45
6.003
1.640
0.634
0.304
Coefficient
of
Variance
10.29
6.85
5.11
Table 7.12: Permanent Deformation Parameter ‘µ’ at 550C
Stress
Level
100
300
500
Mixes with gradation "01"
1a
1b
1c
15.406 12.412 19.862
13.124
7.868 12.401
7.145
4.491 6.897
Mixes with gradation "02"
2a
2b
2c
21.573 8.828 17.670
9.185 4.572
6.902
5.135 3.316
4.627
Standard
Deviation
1.579
1.224
0.553
Coefficient
of Variance
10.55
13.58
10.50
It decreases with an increase in stress levels and temperature, irrespective of aggregate
gradations, binder types and mix types. Overall domain of ‘µ’ for all the mixes was observed to
be is 21.57 to 3.31. Standard deviation measured for ‘µ’ has a range from 0.304 to 1.64.
88
Mixture with gradation "01" & PMA
(1a)
25.00
Mixture with gradation "02" & PMA
(2a)
25.00
100 kPa
20.00
20.00
15.00
100 kPa
300 kPa
10.00
Mu
Mu
15.00
5.00
300 kPa
10.00
5.00
500 kPa
500 kPa
0.00
0.00
20
30
40
0
Temperature ( C)
50
100 kPa
60
300 kPa
20
35
0
100 kPa
40
45
100 kPa
50
55
300 kPa
500 kPa
15.00
100 kPa
300 kPa
Mu
Mu
15.00
10.00
60
Mixture with gradation "02" & 60/70
(2b)
20.00
20.00
30
Temperature ( C)
500 kPa
Mixture with gradation "01" & 60/70
(1b)
25.00
25
10.00
300 kPa
5.00
5.00
500 kPa
0.00
20
25
30
35
40
45
50
55
500 kPa
0.00
60
20
Temperature ( C)
100 kPa
300 kPa
35
40
45
100 kPa
50
55
300 kPa
Mixture with gradation "02" & 40/50
(2c)
25.00
60
500 kPa
100 kPa
20.00
20.00
100 kPa
15.00
15.00
300 kPa
Mu
Mu
30
Temperature ( C)
500 kPa
Mixture with gradation "01" & 40/50
(1c)
25.00
25
0
0
300 kPa
10.00
10.00
5.00
5.00
500 kPa
500 kPa
0.00
0.00
20
25
30
35
40
45
50
55
60
20
30
0
0
Temperature ( C)
100 kPa
300 kPa
500 kPa
Temperature ( C)
Figure 7.2: Influence of Temperature and Stress Levels on Mu
89
40
100 kPa
50
300 kPa
60
500 kPa
7.4
Comparisons of Results to Other Researchers
Alpha and mu reported in this research ranges from 0.8 to 0.92 and 3.31 to 21.57 respectively.
Average values of alpha and mu from all the results are 0.881 and 10.812 respectively.
Specimens, prepared at 5 to 7 % air voids, were subjected to 1800 load repetitions at constant
load frequency 60 cycles/minutes (0.1 sec load duration and 0.9 sec. rest period), under
unconfined conditions, at different stress levels (100, 300, 500 kPa) and temperatures (25, 40,
550C). Domains of alpha and mu with respect to temperature and stress level have been shown in
Figure 7.3 & 7.4 respectively.
Aalpha and mu reported by Rita et al. (1991) ranges from 0.254 to 0.81 and 0.006 to 3.10 and the
average values for the same from all the results were 0.56 and 0.58 respectively. In their study, a
constant load frequency of 60 cycles/minutes (0.1 sec load duration and 0.9 sec. rest period) was
used for all test specimens, prepared at 3.5%, 6.5% and 9.5% air voids. Specimens were
subjected upto 30,000 load repetitions with a deviator stress range from 68 kPa to 200 kPa and
three temperatures i.e. 65, 80, and 95 0F.
Results of alpha and mu, derived by Monismith and Rauhut are 0.56, and 0.28 to 0.35
(Monismith et al 1976), 0.45 to 0.9 and 0.1 to 0.48 (Rauhut et al 1978) respectively.
Slope parameter report by Garba (2002) are in the range of 0.1617 to 0.1806 for HMA specimens
with 4%, 4.7% and 5.4% asphalt cement contents and from 0.135 to 0.1836 for specimen with
3%, 5% and 8% air voids. This parameter had been found within narrow range and suggested not
a good measure of resistance to measure the permanent deformation of mixes prepared with
same materials. Intercept parameter was concluded to be a sensitive parameter in the power
model that showed variation with both the percentage asphalt contents and air voids. Intercept
parameter varies from 2.031 to 4.52 for specimens with 4%, 4.7% and 5.4% asphalt contents and
from 1.649 to 6.91 for specimens with 3%, 5% and 8% air voids.
Permanent Deformation Coefficients computed in this study shows similar domains as in
previous researchers has reported.
90
Domains of Alpha
0.95
Alpha
0.90
0.85
0.80
0.75
0
200
400
Stress level
600
25C,1a
25C,1b
25C,1c
25C,2a
25C,2b
25C, 2c
40C,1a
40C,1b
40c, 1c
40C,2a
40c,2b
40C,2c
55C,1a
55C,1b
55C,1c
55C,2a
55C,2b
55C,2c
Figure 7.3: Domains of alpha
Domains of Mu
25.00
Mu
20.00
15.00
10.00
5.00
0.00
0
100
200
300
400
500
600
Stress levels
25C, 1a
25C,1b
25C,1c
25C,2a
25C,2b
25C,2c
40C,1a
40C,1b
40C,1c
40C,2a
40C,2b
40C,2c
55C,1a
55C,1b
55C,1c
55C,2a
55C,2b
55C,2c
Figure 7.4: Domains of Mu
91
7.5
Regression Analysis
Further, mixes were ranked using intercept coefficient ‘a’ in descending order, which is a
representative of permanent strain at N=1, on a log-log scale, obtained from the base power
model as explained earlier in section 2.9. The results of Intercept coefficient for uniaxial load
strain test and WT have been reported in Table 6.4, 7.1, 7.2 & 7.3. It can be observed that ‘a’
increases with an increase in stress level, irrespective of aggregate gradations, bitumen types,
mix types, and test temperatures. The intercept coefficient ‘a’ of mixes under all temperature and
stress conditions in creep and WT test ranges from 4.73 - 5.48 and 4.34 - 5.35 respectively.
Results of these Tables have been shown in Figure 7.5, which clearly shows that ‘a’ varies in a
narrow range of 4.34 to 5.48 for both the tests.
40C,100kpa
Comparison of WT & URLST
40C,300kpa
8
1a
Intercept Coefficient (a)
7
1b
1c
2a
2b
40C,500kpa
2c
55C,100kpa
6
5
55C,300kpa
4
55,500kpa
3
40C,WT
2
55C,WT
1
25C, 100kpa
0
0
1
2
3
4
5
6
7
25C, 300kpa
25C, 500kpa
Mixes(1a, 1b, 1c, 2a, 2b, 2c)
Figure 7.5: Influence of test type on intercept coefficient
Abscissa in Figure 7.5 shows six mixes (1a, 1b, 1c, 2a, 2b, 2c) ascending order.
92
25C, WT
7.6
Ranking of Mixes using Intercept Coefficient
Further, mixes were ranked based on ‘a’ value to observe the best possible options of mix
performance under given temperatures for both the tests and tabulated in Table 7.13 and 7.14.
Table 7.13: Ranking of Mixes for Wheel Tracker
Sr.
Temperature
No.
(0C)
1st
2nd
3rd
4th
5th
6th
1
25
1c
1a
1b
2c
2b
2a
2
40
1c
1a
1b
2a
2c
2b
3
55
1c
1a
2c
2a
1b
2b
Ranking of Mixes based on Intercept Coefficient “a”
Table 7.13 shows that increase in temperature from 40-550C has affected only the position of
mix ‘2c’ and ‘1b’ from rank 5 to 3 respectively. The reasons may be that 40/50 pen. grade is
more harder grade than 60/70 pen. grade and it showed lower intercept value at 550C. Intercept
coefficients of mixes with coarser gradation (1a, 1b & 1c) have lower value than finer mixes (2a,
2b &2c) in wheel tracker test.
Table 7.14: Ranking of mixes for URLST
Sr.
Temperature
Stress
No.
(0C)
(kPa)
1st
2nd
3rd
4th
5th
6th
1
25
100
2a
1a
1b
1c
2b
2c
2
25
300
2a
1a
1b
1c
2b
2c
3
25
500
2a
1a
2b
2c
1b
1c
4
40
100
1b
1a
2c
1c
2a
2b
5
40
300
1b
2c
1a
2a
1a
1c
6
40
500
1b
2c
1c
1c
2b
2a
7
55
100
1a
2a
1c
1b
2c
2b
8
55
300
1a
2a
1c
2c
1b
2b
9
55
500
1a
2a
1c
2c
1b
2b
Ranking of Mixes
93
It is difficult to conclude mixes ranks using intercept coefficient in the repeated creep test.
However, one can draw a conclusion from Table 7.13 & 7.14 that mixes with PMA can be
ranked at high temperature (550C). Increase in stress levels has relatively minor effects on
permanent strains and hence ‘a’ value. However, significant influence of temperature on the
regression constant has been observed. Although, confinement of slab on WT may has some
more stiffness effects on the mix, which ultimately reduces the rate of rut depth.
7.7
Shift Factor Computations
Permanent strain obtained from repeated creep test can be converted to rut depth using layer
strain method expressed as;
[
RutDepth = ∑ ε p (i )× hi
N
i =1
]
(7.1)
Where i remain as one, N is the total number of load repetitions; εp is the permanent strain and hi,
the thickness of Marshall Specimens (63mm). Rut depth obtained from the above method, was
plotted on log-log scale after multiplying εp with one million in order to obtain straight line
trends. Data obtained from both the tests were plotted graphically in Figure 7.5, and shift factors
were determined.
w t-A
Wheel Tracker versus Uniaxial Repeated Load Test
w t-B
w t-C
A-500kpa
B-500kpa
9
C-500kpa
D-500kpa
8
Log (Rut Depthx10) (mm)
E-500kpa
7
F-500kpa
A-300kpa
6
6
B-300kpa
5
C-300kpa
D-300kpa
4
E-300kpa
F-300kpa
3
Creep Test
Data
2
A-100kpa
WT Test
Data
B-100kpa
C-100kpa
1
D-100kpa
-
E-100kpa
1
10
100
1000
Load Cycles LOG (N)
Figure 7.6: Shift of URLST Data to Wheel Tracker Data
94
10000
100000
F-100kpa
Figure 7.6 shows that plots of both the tests don’t have the same trend lines, but to compare
repeated creep test with WT test an average shift factor of 0.48 have been used.
7.8
Correlations between URLST and WT Test
A comparison of rut on log-log scale between both types of test data has been made to develop
relationships at each temperature and stress levels as shown in Figure 7.6. Rut data of URLST
have been taken on coordinate and that of WT test on abscissa.
It can be observed from Figure 7.6 through 7.15 that the slope of plots gradually reduces by
increasing stress levels and temperatures. At the same time, increase in temperature and stress
levels has also reduces the range of data variation. Wheel tracker specimens being confined in
the test mould showed less rate of increase of permanent strain (rut depth) than unconfined
repeated creep test. One can draw a conclusion from the preceding discussion that rutting can be
prediction from any of the method and can be compared reasonably with one another.
0
Comparison at 25 C & 100 kPa
8.50
WT (Rut Depth)mm
8.00
7.50
7.00
y1a = 85.785x2 - 397.21x + 465.17
R2 = 0.976
y2a = 15.825x2 - 69.623x + 82.453
R2 = 0.99
y1b = 125.29x2 - 585.55x + 689.66
R2 = 0.996
6.50
y2b = 81.2x2 - 380.72x + 452.4
R2 = 0.992
6.00
5.50
5.00
4.50
4.00
2.30
y2c = 272.96x2 - 1295.1x + 1541.8
R2 = 0.996
y1c = 280.42x2 - 1335.1x + 1594.8
R2 = 0.963
2.32
2.34
2.36
2.38
2.40
UTM (Rut De pth)mm
2.42
2.44
2.46
1a
1b
1c
2a
2b
2c
Poly. (1c)
Poly. (2c)
Poly. (1b)
Poly. (1a)
Poly. (2b)
Poly. (2a)
Figure 7.7: Correlations between URLST and WT Test Data at 250C & 100kPa
95
0
Comparison at 25 C & 300 kPa
8.50
y1a = 298.39x2 - 1418.6x + 1691.4
R2 = 0.9792
WT (Rut Depth)mm
8.00
y2a = 39.932x2 - 184.87x + 219.99
R2 = 0.9977
7.50
7.00
6.50
y1b = 331.11x2 - 1577.9x + 1885.3
R2 = 0.9936
y2b = 171.4x2 - 816.93x + 979.53
R2 = 0.9896
6.00
5.50
5.00
4.50
y2c = 703.71x2 - 3390x + 4088.3
R2 = 0.9981
y 1c= 443.38x2 - 2127.4x + 2557.4
R2 = 0.9205
4.00
2.30
2.35
2.40
2.45
2.50
UTM (Rut Depth)mm
1a
Poly. (2c)
1b
Poly. (1c)
1c
Poly. (1b)
2a
Poly. (1a)
2b
Poly. (2b)
2c
Poly. (2a)
Figure 7.8: Correlations between URLST and WT Test Data at 250C & 300kPa
0
Comparison at 25 C & 500 kPa
8.50
8.00
y2a = 62.036x2 - 293.36x + 352.84
R2 = 0.9963
y1a = 431.89x2 - 2081.2x + 2512.7
R2 = 0.9707
y2b = 267.36x2 - 1290.6x + 1563.7
R2 = 0.9941
WT (Rut Depth)mm
7.50
7.00
6.50
y1b = 659.79x2 - 3189.4x + 3859.9
R2 = 0.9965
6.00
5.50
5.00
4.50
y2c = 362.89x2 - 1747.4x + 2109
R2 = 0.9984
y1c = 8017x2 - 39289x + 48141
R2 = 0.9663
4.00
2.30
2.35
2.40
2.45
2.50
UTM (Rut Depth)mm
1a
Poly. (1c)
1b
Poly. (1a)
1c
Poly. (2c)
2a
Poly. (1b)
Series5
Poly. (2b)
2b
Poly. (2a)
Figure 7.9: Correlations between URLST and WT Test Data at 250C & 500kPa
96
2c
0
Comparison at 40 C & 100 kPa
8.50
y2a = 0.7355x2 - 6.4169x + 19.854
R2 = 0.9905
y1a = 0.6168x2 - 5.0671x + 16.073
R2 = 0.9977
8.00
WT (Rut depth)mm
7.50
y1b = 0.57x2 - 4.7054x + 15.449
R2 = 0.9979
7.00
6.50
y2b = 0.6085x2 - 4.9251x + 15.582
R2 = 0.9973
6.00
5.50
5.00
y2c = -0.2671x2 + 3.5816x - 4.4757
R2 = 0.9986
y1c = 1.1523x2 - 11.566x + 34.706
R2 = 0.9988
4.50
4.00
4.00
4.50
5.00
5.50
6.00
6.50
UTM-5P (Rut De pth) mm
1a
Poly. (2c)
1b
Poly. (2b)
1c
Poly. (2a)
2a
Poly. (1b)
2b
Poly. (1a)
2c
Poly. (1c)
Figure 7.10: Correlations between URLST and WT Test Data at 400C & 100kPa
0
Comparison at 40 C & 300 kPa
8.50
8.00
y1a = 1.1835x2 - 11.121x + 31.842
R2 = 0.9982
y2a = 0.6856x2 - 5.5859x + 16.72
R2 = 0.989
WT (Rut depth) mm
7.50
7.00
6.50
6.00
5.50
5.00
4.50
y2b = 0.5542x2 - 4.6522x + 15.612
R2 = 0.9949
y1b = 0.466x2 - 3.7412x + 13.028
R2 = 0.9982
y2c = 0.2145x2 - 1.0869x + 6.2694
R2 = 0.9977
y1c = 0.4675x2 - 4.1183x + 14.55
R2 = 0.9978
4.00
4.50
5.00
5.50
6.00
6.50
UTM-5P (Rut Depth) mm
1a
1b
1c
2a
2b
Poly. (2c)
Poly. (1a)
Poly. (2a)
Poly. (1b)
Poly. (2b)
2c
Figure 7.11: Correlations between URLST and WT Test Data at 400C & 300kPa
97
Poly. (1c)
0
Comparison at 40 C & 500 kPa
8.50
WT (Rut Depth)mm
8.00
7.50
y2a = 1.2742x2 - 12.696x + 37.66
R2 = 0.9943
y1a = 0.7785x2 - 6.9721x + 20.988
R2 = 0.9978
7.00
6.50
6.00
5.50
5.00
4.50
4.00
5.00
y2b = 0.0168x2 + 1.5601x - 2.6272
R2 = 0.9903
y1b = 0.6082x2 - 5.2359x + 16.726
R2 = 0.9974
y2c = -0.0175x2 + 1.6625x - 1.9362
R2 = 0.9963
y1c = 0.7005x2 - 6.2399x + 19.11
R2 = 0.9952
5.20
5.40
5.60
5.80
6.00
6.20
UTM-5P (Rut Depth)mm
1a
Poly. (2c)
1b
Poly. (2b)
1c
Poly. (2a)
2a
Poly. (1b)
2b
Poly. (1a)
2c
Poly. (1c)
Figure 7.12: Correlations between URLST and WT Test Data at 400C & 500kPa
0
Comparison at 55 C & 100 kPa
8.50
WT (Rut Depth)mm
8.00
7.50
7.00
y2a = 0.5813x2 - 4.9708x + 16.739
R2 = 0.9897
y1a = 0.6854x2 - 5.7827x + 18.043
R2 = 0.9975
y1b = 0.3647x2 - 2.4325x + 9.3175
R2 = 0.9937
y2b = -0.1539x2 + 3.4111x - 6.9468
R2 = 0.9926
6.50
6.00
5.50
5.00
4.50
4.00
5.00
1a
Poly. (1c)
y2c = -1.6635x2 + 19.995x - 52.588
R2 = 0.9796
y1c = 2.0859x2 - 20.177x + 54.608
R2 = 0.9893
5.20
1b
Poly. (2c)
5.40
5.60
UTM (Rut Depth)mm
1c
Poly. (2b)
2a
Poly. (2a)
5.80
6.00
2b
Poly. (1a)
Figure 7.13: Correlations between URLST and WT Test Data at 550C & 100 kPa
98
2c
Poly. (1b)
0
Comparison at 55 C & 300 kPa
8.50
WT (Rut Depth)mm
8.00
y2a = 0.3999x2 - 3.224x + 12.556
R2 = 0.9931
y1a = 0.3732x2 - 2.939x + 11.467
R2 = 0.9969
7.50
7.00
6.50
y2b = -0.5394x2 + 7.5384x - 18.263
R2 = 0.993
6.00
y1b = -0.0623x2 + 1.7455x - 0.7971
R2 = 0.989
5.50
5.00
y2c = -0.6765x2 + 8.6389x - 20.027
R2 = 0.9802
y1c = 0.5508x2 - 4.5584x + 15.06
R2 = 0.986
4.50
4.00
5.00
5.20
1a
Poly. (2c)
5.40
1b
Poly. (2a)
5.60
5.80
UTM (Rut Depth)mm
1c
Poly. (2b)
2a
Poly. (1b)
6.00
2b
Poly. (1c)
6.20
2c
Poly. (1a)
Figure 7.14: Correlations between URLST and WT Test Data at 550C & 300 kPa
0
Comparison at 55 C & 500 kPa
8.00
WT (Rut Depth)mm
7.50
y2a = 0.794x2 - 6.4726x + 17.549
R2 = 0.9884
y1a = 0.4169x2 - 3.4149x + 12.603
R2 = 0.9955
7.00
6.50
y2b = 0.4236x2 - 3.7442x + 14.44
R2 = 0.9981
6.00
5.50
y1b = -3.8074x2 + 46.259x - 133.23
R2 = 0.9931
5.00
4.50
4.00
5.20
y2c = -5.9797x2 + 71.54x - 206.66
R2 = 0.9878
y1c = 0.6175x2 - 5.3107x + 16.976
R2 = 0.9827
5.40
5.60
5.80
6.00
6.20
6.40
UTM (Rut Depth)mm
1a
Poly. (2c)
1b
Poly. (1b)
1c
Poly. (2b)
2a
Poly. (1c)
2b
Poly. (1a)
Figure 7.15: Correlations between URLST and WT Test Data at 550C & 500 kPa
99
2c
Poly. (2a)
7.9
Summary
Straight line trends of URLST data files were plotted between Cycles (N) and Rut depth, using
log techniques, for the determinations of regression constants i.e. ‘intercept coefficients’ and
‘slope coefficients’. Intercept coefficient ‘a’ increases with an increase in stress levels, but
slope coefficient ‘b’ decreases under the same conditions. Within the same class of aggregate
grading (NHA Class “A”), it is difficult to study the influence on slope coefficient. Ranges of
intercept coefficient “a” and slope coefficient “b” for all the six mixes were observed from
3.46 to 5.38 and 0.16 to 0.30 respectively. Mixes initially designed for URLST test were also
tested under wheel tracker using the same three temperatures and measurements of rut depth
against number of load cycles were plotted. Mixes with PMA & 40/50 penetration grade
bitumen have almost same rut potential at lower temperature (250C).PMA performed better
than that of neat binders at higher temperatures (550C).
Intercept coefficient (a) and slope coefficient (b) were also computed using the log-log scale on
the test data files. It was observed that with increase in temperature, magnitude of “a” increase,
indicating rut susceptibility of mixes. Similarly, alpha (α) increases with the increase in stress
levels, but mu (µ) decreases under same conditions. Effect of temperature observed on value of
alpha was minor, while mu varies significantly. Domains of alpha and mu observed for low
asphalt content mixes are 0.80 to 0.92 and 3.32 to 21.57 respectively.
Results of both the test methods as described in the previous chapters were further correlated
using shift factor, ranking of mixes and correlation developments. It was observed that rutting
can be predicted from any of the method and can be compared reasonably with one another.
Uniaxial load strain (creep) test did not provide a clear indication of mixes ranks than wheel
tracker test.
Shift factor is useful number in order to compare the results and plots of repeated creep tests
and it can be shifted to that of WT test with an approximate average of 0.48. However, it
requires further investigation for precision. The plots drawn between both the tests, clearly
indicates a reduction in the slope of line with the increase in temperature and stress levels.
100
Chapter Eight
101
Chapter Eight
Modeling the Permanent Deformation Properties of Hot Mix Asphalt
8.1
Introduction
This chapter mainly focused on the statistical estimation of Mechanistic-Empirical Model that
relates plastic to elastic strain ratio with different variable that influenced in the mix rut
development. Using 54 variables, comprising six asphalt mixes, three temperatures (250C,
400C and 550C) and three stress levels (100, 300 and 500 kPa) under repeated load test,
mathematical model have been developed to assess the magnitude of plastic to elastic strain
ratio. Despite certain limitations, critical parameters have been captured in the model that helps
estimating the permanent deformation in the asphalt layers of flexible pavements. This chapter
provides an alternative test methodology with similar accuracy of AASHTO Mechanistic
Empirical Pavement Design Guide (MEPDG) model for the prediction of permanent
deformation.
The current MEPDG incorporates a power model for generating rutting predictions for asphalt
concrete. Rutting model developed from laboratory uniaxial repeated load strain tests as
provided in MEPDG in the following form has been used as basis to estimate the relationships
between the predictor variables and the permanent deformation parameters (Stephen et. al.
2007) :
εp
= a1T a N a
εr
2
3
(8.1)
Where, εp, εr, are the plastic and elastic strains respectively, at N repetitions of load and ai are
the non linear regression coefficient.
102
8.2 Data Analysis and Model Development
8.2.1 General Trends of Variables
Graphical study of characteristics of the data sample help to observed general trends [
McCuen, 1985]. The analysis of data initiates with sorting individual effects of stresses,
temperatures and mix types on the plastic and elastic properties. For example, the effects
of temperatures and stress levels are evaluated by calculating the mean values of elastic and
plastic parameters of asphalt mixes as given in Table 5.2 & 5.3. The effects of binder type
can be seen in the form of mix type in alphabetical order of mix designations i.e. ‘a’, ‘b’
and ‘c’.
The general trends of ‘εp,’& ‘εr’ have been shown graphically in Figure 8.1. On the abscissa
are the stress levels and on the ordinate are the plastic & elastic strain components. One can
observe the effects of different variables on plastic and elastic strains. Plots for each
temperature (250C, 400C, 550C) have been drawn. Figure 8.1 shows that accumulative
plastic strain or the permanent strain increases with increase in stress as well as the
temperature each asphalt mix.
Figure 8.2 shows an increase in resilient strain with
increases in the temperature.
103
1.50
0
55 C
0
40 C
1.00
0
25 C
0.50
Mixture with Gradation "02" & PMA
(2a)
2.00
0
55 C
1.50
0
40 C
1.00
0
25 C
0.50
0.00
0
0.00
0
100
200
300
400
500
55 Deg. C
0
40 C
1.00
0
25 C
0.50
Mixture with Gradation "02" & 60/70 Pen. Grade
(2b)
2.00
0
55 C
0.00
0
55 C
1.50
0
40 C
1.00
0.50
0
25 C
0.00
0
100
200
300
400
500
600
0
100
Stress (kPa)
25Deg. C
40 Deg. C
Mixture with Gradation "01" & 40/50 Pen. Grade
(1c)
1.50
0
55 C
0
40 C
1.00
0
25 C
0.50
200
300
400
Stress (kPa)
25 Deg. C
55 Deg. C
0.00
500
600
40 Deg. C
55 Deg. C
Mixture with Gradation "02" & 40/50 Pen. Grade
(2c)
2.00
Accumulated Plastic
Strain (εp) %
Accumulated Plastic Strain
(εp) %
40 Deg. C
55Deg. C
1.50
2.00
600
600
Mixture with Gradation "01" & 60/70 Pen. Grade
(1b)
2.00
200
400
Stress (kPa)
25 Deg. C
Stress (kPa)
25 Deg.C
40 Deg.C
Accumulated Plastic
Strain (εp ) %
Accumulated Plastic Strain
(εp) (%)
2.00
Accumulated Plastic
Strain (εp) %
Accumulated Plastic Strain
(εp) (%)
Mixture with Gradation "01" & PMA
(1a)
0
55 C
1.50
0
40 C
1.00
0
25 C
0.50
0.00
0
100
200
300
400
Stress (kPa)
25 Deg. C
500
40 Deg. C
600
0
55 Deg. C
100
200
300
400
Stress (kPa)
25 Deg. C
Figure 8.1: General trends of Plastic Strains
104
40 Deg. C
500
600
55 Deg. C
Mixture with Gradation "01" & PMA
(1a)
0
0.14
55 C
0.10
40 C
0
0
25 C
0.06
0.02
25 Deg. C
400
600
0
40 C
0.1
0
0.06
25 C
55 Deg. C
40 Deg. C
0
0.18
Mixture with Gradation "01" & 60/70 Pen. Grade
(1b)
0.14
55 C
0.10
40 C
0
0
0
25 C
0.06
200
0.02
400
25 Deg. C
Stress (kPa)
Resilient Strain( %)
200
Stress (kPa)
Resilient Strain( %)
0
55 C
0.14
0.02
0
600
55 Deg. C
40 Deg. C
0.18
Mixture with Gradation "02" & 60/70 Pen. Grade
(2b)
0
55 C
0.14
40 C
0.1
25 C
0
0
0.06
0.02
200
Stress (kPa)
25 Deg. C
400
55 Deg. C
600
40 Deg. C
0.18 Mixture with Gradation "01" & 40/50 Pen. Grade
(1c)
0.14
0
55 C
0.10
0
40 C
0
0.06
0
25 C
0.02
100
Stress (kPa)
0.18
Resilient Strain( % )
0
Resilient Strain( %)
Mixture with Gradation "02" & PMA
(2a)
0.18
Resilient Strain( %)
Resilient Strain( %)
0.18
200
300
400
500
600
25 Deg. C
55 Deg. C
40 Deg. C
Mixture with Gradation "02" & 40/50 Pen. Grade
(2c)
0
55 C
0.14
0
40 C
0.10
0
25 C
0.06
0.02
0
Stress (kPa)
100
200
300
400
500
600
25 Deg. C
55 Deg. C
40 Deg. C
105
0
Stress (kPa)
100
200
300
25 Deg. C
400
500
55 Deg. C
600
40 Deg. C
Figure 8.2: General Trends of Elastic Strains
To conclude the influence of data trends as shown in Fig. 8.1 and 8.2, Table 8.1 provides a
qualitative view of the effects of test conditions on the elastic and plastic strains.
Table 8.1: Influence of Test Conditions and Mix Parameters on Strain
Predictor Variables
Criterion Variables
Plastic Strain (εp)
Elastic Strain (εr)
Temperature
Very Strong
Strong
Asphalt Type (G*)
Strong
Strong
Stress
Strong to Moderate
Very Strong
Aggregate Type (Gradation)
Low
Moderate
As shown in Table 8.1, the elastic and plastic stains are highly dependent on the temperature,
and stress conditions and moderately dependent on the mean shear complex modulus of asphalt
cement. Very little effect has been observed with the change in the gradation. Also, plastic
strain component is more sensitive as compared to elastic strain, especially at 400C and 550C.
8.2.2 Development of Prediction Model
The MEPDG power model was initially adopted in this study to observe the agreement
between the predicted and the measured plastic to elastic ratio under each test condition. The
simplest technique is to combine all the variables and perform ordinary least-square regression
on the entire data set. Predictor variables as given in Table 6 were then incorporated in step
wise step iteration process using the Solver tool. The Microsoft Excel Solver (MES) can find
an optimal value of coefficients for a formula using “what-if analysis” tools. Non liner
regression coefficients (a1, a2, a3, a4) were then obtained at least-square of difference between
the predicted and measured ‘εp/εr’ ratio. The equation that provided the minimum least square
and at the same time highest coefficient of determination (R2) was selected as final model. The
plots of all predictor variables versus measured variables using the same model at 54 test
conditions and 162 data files have been shown in Figure 8.3
106
Agreement of Model with Measured data
Predicted εp/εr
20.0
15.0
10.0
5.0
nonlinear least square
0.0
0.0
5.0
10.0
15.0
Measured εp/εr
20.0
Figure 8.3: Measured versus Predicted εp/εr, 106 in/in
The plot in Figure 8.3 shows a variation in the trends, especially at higher numbers, but there is
no bias observed in the model. One observes from Figure 8.3 that the procedure adopted for
mathematical development of model, yields a reasonable agreement between the predicted and
measured values of strain ratio (R2 = 0.573). The mathematical form of the model is presented
here;
εp σ a Ta Na
=
εr
G*
1
2
3
a4
Where:
εp
= Plastic Strain
εr
= Elastic Strain
σ
= Stress Levels (kPa),
T
= Temperature (0C)
N
*
= Load Repetitions
G
= Mean Complex Shear Modulus of Asphalt Cement and,
a1, a2, a3, a4
= Non linear regression coefficient.
107
(8.2)
The above model can be used to determine the permanent deformation of mixes under repeated
loading in terms of plastic to elastic strain ratio. Taking the logarithm of above model and
placing values of non liner regression coefficients, the above model can be shown as;
⎛ε ⎞
log⎜⎜ p ⎟⎟ = 0.018 log σ + 0.121log T + 0.277 log N − 0.031log G *
⎝ εr ⎠
(8.3)
The above model is an implicit form to predict permanent deformation as a function of stress,
temperature, load cycle and shear modulus of binder.
8.2.3 Comparison of Computed Model with MEPDG Model
Computed model was then compared with the MEPDG model using test data to check the
relevant accuracy. The trends plotted between both the models have been shown graphically in
Figure 8.4, which produced an excellent agreement (R2 =0.99). This analysis enhanced the
confidence in creating the alternatives for estimating the rutting predictions.
MEPDG Model Versus Computed Model
16
R2 = 0.9935
14
Computed Model
12
10
8
6
4
2
0
0
5
10
MEPDG MODEL
Figure 8.4: MEPDG Versus Computed Model εp/εr, 106 in/in
108
15
8.2.4 Sensitivity Analysis
Following the model development, a sensitivity analysis was conducted to ascertain the
comparative importance of predictor variables in the recommended model. Figure 8.5 provides
a summary of the relative sensitivity of εp/εr model to predictor variables and a picture that
temperature is the most sensitive parameter, followed by stress levels and complex shear
modulus of asphalt (descending order). However, regression constant against number of load
repetition has been observed higher than others, which indicates its relative importance in any
of the test condition.
In summary, the above research work presents an assessment of influence of the mix
parameters and test conditions on the permanent deformation characteristics of asphalt concrete
in terms of εp/εr. Based on the comprehensive laboratory study on six mixes, it was observed
that plastic to elastic strain ratio is a function of number of load repetitions, temperature, stress
levels, shear complex modulus of asphalt cement and aggregate gradations respectively. Also,
the sensitivity analysis indicates that the temperature has influential effects on permanent
deformation than other variables as considered in this experimental study. Plastic to elastic
strain can be computed mathematically, using the above model with a relatively good
reliability. The proposed model has a best compatibility with the MEPDG model.
109
Mixture with Gradation "01" & PMA
(1a)
Temperature (T)
Mixture with Gradation "02" & PMA
(2a)
Stress (σ)
Temperature (T)
Stress (σ)
Percentage change in strain ratio
Percentage change in strain ratio
260
260
160
60
-40 0
100
200
300
400
500
-140
Percentage Change in temp & stress
60
-40 0
Stress (σ)
T(1a,100kPa)
σ (1a,40C)
σ (1a,55C)
T (1a,300kPa)
T (1a,500kPa)
Temperature (T)
Stress (σ)
160
60
-40 0
100
200
300
400
500
60
-40 0
100
200
300
400
500
-140
-140
Percentage Change in temp & stress
Percentage Change in temp & stress
-240
-240
σ(1a,25C)
T(1a,100kPa)
σ (1a,40C)
σ(1a,55C)
T(1a,300kPa)
T(1a,500kPa)
σ (1a,25C)
T(1a,100kPa)
σ(1a,40C)
σ (1a,55C)
T(1a,300kPa)
T (1a,500kPa)
Mixture with Gradation "02" & 40/50 Pen. Grade
(2c)
Percentage change in strain ratio
Mixture with Gradation "01" & 40/50 Pen. Grade
(1c)
Temperature (T)
Stress (σ)
260
Percentage change in
strain ratio
500
σ (1a,25C)
260
160
160
60
-40
400
Mixture with Gradation "02" & 60/70 Pen. Grade
(2b)
Percentage change in
strain ratio
Percentage change in strain ratio
Temperature (T)
300
Percentage Change in temp & stress
Mixture with Gradation "01" & 60/70 Pen. Grade
(1b)
260
200
-140
σ(1a,40C)
T(1a,500kPa)
T (1a,100kPa)
T (1a,300kPa)
100
-240
-240
σ (1a,25C)
σ (1a,55C)
160
0
100
200
`
300
400
500
-140
-240
Percentage Change in temp & stress
260
Temperature (T)
Stress (σ)
160
60
-40 0
100
200
300
400
500
-140
-240
Percentage Change in temp & stress
σ(1a,25C)
T(1a,100kPa)
σ (1a,40C)
σ (1a,25C)
T(1a,100kPa)
σ (1a,40C)
σ (1a,55C)
T(1a,300kPa)
T(1a,500kPa)
σ(1a,55C)
T(1a,300kPa)
Temp. (1a,500)
Figure 8.5: Sensitivity of εp/εr Model to Predictor Variables
110
Chapter Nine
111
Chapter Nine
CONCLUSIONS AND RECOMMENDATIONS
9.1
Introduction
This chapter presents the findings of study carried out on six asphalt mixes by using two test
methods. The main objectives of the research work were to assess the influence of the mix
parameters and test conditions on the permanent deformation characteristics of asphalt
concrete. Two test procedures were conducted on six mixes in this research study, comprising
two aggregate gradations and three binders. The effects of variables on the mix performance
were studied using regression constants (intercept and slope coefficients), permanent
deformation coefficients (alpha and mu) and plastic to elastic strain ratios (εp/εr). Relationship
has also been developed between two commonly known tests (uniaxial creep and wheel tracker
test), to predict resistance of permanent deformation or rutting of mixes at elevated
temperatures. Relationship between HMA properties to its permanent deformation behaviour in
the form of a model has been proposed as the key findings of the study.
9.1 Conclusions
Based on the literature review, testing and analysis of test results, and modeling efforts under
taken in this research work, following conclusions have been drawn;
•
Plastic to elastic strain ratio is a function of number of load repetitions, temperature, stress
levels, shear complex modulus of asphalt cement and aggregate gradations. However,
AASHTO Mechanistic Empirical Pavement Design Guide (MEPDG) model which has
only two parameters (load repetitions and temperature) results in a similar accuracy with
different sensitivities of the independent variables on permanent deformation prediction.
112
•
Repeated creep testing showed equal potential of predicting permanent deformation
compared to repeated load testing. Haversine repeated load testing carried out for the
MEPDG requires more complicated system as compared to running simple repeated creep
testing.
The proposed model provides an alternative test methodology with similar
accuracy of MEPDG model for the prediction of permanent deformation. MEPDG model
works for the mixes used in this study.
•
Mixes with coarser gradation offered more resistance to rutting than mixes with finer
gradations under uniaxial repeated load strain test.
•
Intercept coefficient ‘a’ increases with an increase in stress levels, but slope coefficient ‘b’
decreases under the same conditions. Ranges of ‘a’ & ‘b’ for mixes in study were observed
from 3.46 -5.38 & 0.16-0.3 respectively. Permanent Deformation Coefficient alpha ‘α’ has
similar trends as ‘a’ and mu ‘µ’ as ‘b’. Domains of α & µ for mixes in study were 0.8-0.92
& 3.32-21.57 respectively.
•
Master curve can be plotted between Uniaxial Repeated Load Strain Test & Wheel Tracker
test by an average shift factor of 0.48.
9.3
Recommendations for Future Study
•
Comprehensive experimental study may be conducted on the moisture sensitivity as
well as on the fatigue limits of the six mixes and the aging effects of binder on mixes
behaviour. Similarly, uses of hydraulic cement, lime, crumb rubber and other locally
available materials to enhance the performance of mixes can be used to accommodate
their effects in a complete model development. At the same time, full scale accelerated
testing should be carried out to investigate the performance of these mixes in the field
•
under local environmental conditions.
Limits of permanent deformation in terms of plastic to elastic strain ratio should be
specified under the current flexible pavement design, material properties, axle loading
and temperature conditions of Pakistan.
113
References
114
REFERENCES
Ahlrich, R.C. 1996. Influence of Aggregate Properties on Performance of Heavy-Duty HotMix Asphalt Pavements. Transportation Research Record No. 1547, pp. 7-14.
Ahmad, H. Alijassar, Sayed Metwali, Mohammad A.Ali, 2004. “Effect of Filler Types on
Marshall Stability and Retained Strength of Asphalt Concrete” International Journal of
Pavement Engineering, 5:1 pp 47-51.
Ait-kadi A., B. Brahimi, M. Bousmina, 1996. Polymer Engineering Sciences, 36 (12), 17241733.
Akmal N., and Usmani A. M., 1999. “Polymers”, News, 24, 136-140.
American Association of State Highway and Transportation Officials (AASHTO). 2004.
Determining the Soundness of Aggregates by Sodium Sulphate or Magnesium Sulphate,
AASHTO Designation T104.
American Association of State Highway and Transportation Officials (AASHTO). 2004.
“Determining the Percentage Deleterious Materials present in Aggregates Blend”, AASHTO
Designation T112.
American Association of State Highway and Transportation Officials (AASHTO). 2004.
Plastic Fines in Graded Aggregate and Soils by Use of the Sand Equivalent Test, AASHTO
Designation T176.
American Association of State Highway and Transportation Officials (AASHTO). 2004.
Elastic Recovery Test of Bituminous Materials by Means of a Ductilometer, AASHTO
Designation T 301.
American Association of State Highway and Transportation Officials (AASHTO). 2004.
Uncompacted Void Content of Fine Aggregate, AASHTO Designation T304.
115
American Association of State Highway and Transportation Officials (AASHTO). 2004.
Determining the Rheological Properties of Asphalt Binder Using a Dynamic Shear Rheometer
(DSR), AASHTO Designation T315.
American Association of State Highway and Transportation Officials (AASHTO). 2002. Standard
Test Method for Density of Semi-Solid Bituminous Materials, AASHTO Designation T228.
American Association of State Highway and Transportation Officials (AASHTO). 2002. “Design
of New and Reconstructed Flexible Pavements” Chapter 3, Design Guide.
American Association of State Highway and Transportation Officials (AASHTO). 2000. Softening
Point of Bitumen (Ring and Ball Apparatus), AASHTO Designation T53.
American Society for Testing and Materials (ASTM). 2004. Standard Test Method for Penetration
of Bituminous Materials. ASTM Designation D5.
American Society for Testing and Materials (ASTM). 2004. Test Method for Theoretical
Maximum Specific Gravity and Density of Bituminous Paving Mixtures. ASTM Designation
D2041
American Society for Testing and Materials (ASTM). 2004. Test Method for determining the
Percentage of Plastic Fines in Graded Aggregate and Soils by Use of the Sand Equivalent Test.
ASTM Designation D2419.
American Society for Testing and Materials (ASTM). 2004. Flat or Elongated Particles in Coarse
Aggregates, ASTM Designation D4791.
American Society for Testing and Materials (ASTM). 2004. Test Method for determining the
Percentage of Fractured Faces in Coarse Aggregate. ASTM Designation D5821
American Society for Testing and Materials (ASTM). 2004. Test Method for Indirect Tension Test
for Resilient Modulus of Bituminous Mixtures. ASTM Designation D4123
American Society for Testing and Materials (ASTM). 2004. Standard Test Method for Soundness
of Aggregates by Use of Sodium Sulfate or Magnesium Sulfate. ASTM Designation C88
116
American Society for Testing and Materials (ASTM). 2004. Standard Test Method for
Resistance to Degradation of Small-Size Coarse Aggregate by Abrasion and Impact in the Los
Angeles Machine. ASTM Designation C131
American Society for Testing and Materials (ASTM). 2004. Standard Test Method for
Ductility of Bituminous Materials. ASTM Designation D113-99
Anderson, R. M., McGennis, R.B., Kennedy, T. W. and Solaimanian, M. 1995. Background of
Superpave Asphalt Mixture Design & Analysis. Publication No. FHWA-SA-95-003. National
Asphalt Training Center Demonstration Project 10.
Asphalt Institute. 1996. “Superpave TM Mix Design” Superpave Series SP-2, Lexington,
Kentucky, USA.
Asphalt Institute, 2003. “Mix Design Method for Asphalt Concrete and other Hot-Mix Types,”
Manual series No. 02, Six Edition, Kentucky, USA.
Asphalt Institute, 2003. “Asphalt Hand Book,” Manual series No. 04, Six Edition, Kentucky,
USA.
Asphalt Institute. 2003. “Superpave Mix Design” Superpave Series SP-2, Third Edition,
2001.Printing. Revised 2003, Lexington, Kentucky, USA.
Australian Roads Standards, 2006. “Torsional Recovery of Polymer Modified Binders”
AUSTROAD Designation T-122.
Awad, I.S., 1972. Characterisation of the Stress-Strain Relationships of Asphalt Treated Base
Material, Graduate Report, University of Washington.
Barengberg, E.J. and M.R. Thompson. 1990. Calibrated Mechanistic Structural Analysis
Procedures for Pavements Phase 1 & II of NCHRP Project 1-26. National Cooperative
Highway Research Program. Transportation Research Board. Washington, DC.
117
Barnes, H.A., J.F. Hulton and K. Walters, 1989. “Introduction to rheology”. Elsvier, Barking.
Behzadi, G. and Yandell, W.O., 1996. “Determination of elastic and plastic subgrade soil
parameters for asphalt cracking and rutting prediction.” Transp. Res. Rec. 1540, Transportation
Research Board, Washington, D.C., 97-104.
Brown, S. F. and Pell, P. S. 1974. Repeated Loading of Bituminous Materials. CAPSA 74.
National Institute for Road Research, Republic of South Africa, Durban.
Brown, S.F. and Snaith M.S., 1974. The Permanent Deformation of Dense Bitumen Macadam
Subjected to Repeated Loading. Proceedings of the Association of Asphalt Paving
Technologists Technical Sessions, vol. 43, Virginia, United States of America.
Brown, E. R., Cross, Stephen A. 1990. A Study of In-Place Rutting of Asphalt Pavements. The
Association of Asphalt Paving Technologists v. 59.
Brown, E. R. 1990. Density of Asphalt Concrete – How much is needed? Transportation
Research Record 1282, Transportation Research Board, Washington D. C
Brown, R.R., and Bassett, C.E. 1990. Effects of Maximum Aggregate Size on Rutting Potential
and Other Properties of Asphalt –Aggregate Mixtures. Transportation Research Record 1259,
pp. 107-119.
Brown, E.R., and Cross, S.A. 1992. A National Study of Rutting in Hot Mix Asphalt (HMA)
Pavements. NCAT Report 92-5, National Centre for Asphalt Technology.
Burger, A.F., Van de Van, M.F.C, J. Muller, K.J. Jenkins., “Rheology of Polymer Modified
Bitumen: A Comparison Study of Three Binders and Three Binder/Filler Systems”, 20th South
African Transport Conference, Meeting the Transport Challenges in Southern Africa, 16-20
July, 2001.
Button, J.W., Perdomo, D., Lytton, R.L. 1990. Influence of Aggregate on Rutting in Asphalt
Concrete Pavements. Transportation Research Record No. 1259, pp.141-152.
118
Carpenter, S. H., 1993. Permanent deformation: Field evaluation. Transport research record
1417 Materials and Construction, Asphalt concrete mixtures. Transportation Research Board,
National Academy Press, Washington
Cheung, C.Y. 1995. Mechanical behaviour of bitumen and bituminous mixes. Ph.D. Thesis,
University of Cambridge.
Collop, A. C., Cebon, D. and Hardy, S. A. 1995. Viscoelastic Approach to Rutting in Flexible
Pavements. Journal of Transportation Engineering v. 121, Issue 1, p. 82-93.
Cominsky, R., R. B. Leahy, and E. T. Harrigan. 1994. Level One Mix Design: Material
Selection, Compaction, and Conditioning. Report SHRP-A-408. Strategic Highway Research
Program, National Research Council.
Cooper, K. E., Brown, S. F., and Pooley, G. R. 1985. The Design of Aggregate Gradings for
Asphalt Base courses. The Association of Asphalt Paving Technologists v. 54.
Cooper Technology, 2006. “ Wheel Tracker Small Device”, Product Catalogue, Issue 1, CRTWTEN1& CRT-WTEN2 in conformance to E13108.
Crawford, C. 1989. Tender Mixes: Probable Causes, Possible Remedies. Quality Improvement
Series 108-3186. National Asphalt Pavement Association, Lanham, MD.
Diylajee, V.A., and Raymond, G.P., 1982. “Repetitive load deformation of cohesionless soil,”
J.Geotech. Engrg. Div., ASCE, 108(10), 1215-1229.
European Standard; prEN 12697-22, 2002. “Test methods for hot mix asphalt” part-22, Wheel
Tracking.
Findley, W. N., Lai, J. S., and Onaran, K. 1976. Creep and Relaxation of Non linear Viscoelastic Materials. North Holland Publishing, Amsterdam.
119
Ford, Miller C. 1988. Pavement Densification Related to Asphalt Mix Characteristics. Flexible
Pavement Construction, Transportation Research Record 1178. Transportation Research
Board, Washington D. C.
Fred, N. Finn., 1967. “Factors involved in the design of Asphaltic Pavement Surfaces,”
National Cooperative Highway Research Program Report 39, Highway Research Board,
Washington, D.C.
Garba, R., 2002. “Permanent deformation properties of asphalt concrete mixtures, PhD Thesis,
Norwegian University of Science and Technology”.
Gerald, Reinke; 2001. “Elvaloy Formulation with Attock Asphalt” Mathy Technology &
Engineering Services, Inc;
Gibb, J.M., 1996.Evaluation of resistance to permanent deformation in the design of
bituminous paving mixtures. Ph.D. Thesis, University of Nottingham.
Hassan, S. M., 1998. “Performance Prediction Models for Asphalt Concrete Pavements”,
Treatise on Flexible Pavement Design, pp, 15-28.
Herrin, M., and W. H. Goetz. 1954. Effect of Aggregate Shape on Stability of Bituminous
Mixes. In Proceedings, Vol. 33, Highway Research Board.
Hertzberg, R.W., 1996. Deformation and fracture mechanics of engineering materials. 4th
edition, John Wiley and sons, New York.
Heukelom, W. 1999. “A bitumen test data chart for showing the effect of temperature on the
mechanical behaviour of asphaltic bitumen”, Journal of the institute of Petroleum, vol 55.
Huang, Y. H. 1993. Pavement Analysis and Design. Prentice-Hall, Englewood Cliffs, NJ.
Huang, Y.H., 1993. Pavement Analysis and Design, Prentice Hall, New Jersey, United States
of America.
120
Huang, Shin-Che and Zeng Menglan, 2007. “Characterization of aging effect on rheological
properties of asphalt-filler systems”, International Journal of Pavement Engineering, 8:3, 213223.
Huang, H., 1995. “Analysis of Accelerated Pavement Tests and Finite element Modeling of
Rutting Phenomenon”, Ph.D. Thesis, Purdue University.
Huang, B., Xiang Shu, Xinwei Chen. 2007. “Effects of mineral fillers on hot-mix asphalt
laboratory measured properties”, International Journal of Pavement Engineering 8:1, pp1-9.
Huber, G. A. and Heiman, G. H. 1987. Effects of Asphalt Concrete Parameter on Rutting
Performance: A Field Investigation. Association of Asphalt Paving Technologists v. 56.
Huber, G.A., and Heiman, G.H. 1989. Effect of Asphalt Concrete Parameters on Rutting
Performance: A Field Investigation. Pavement Rutting Seminar- Roads and Transportation
Association of Canada.
K. B. De Vose. and A.J. Feeley. 2002. “Feed back controlled repeated uniaxial loading strain
test” Hardware Reference. Manual, Universal Testing Machine (UTM-5P), pp11-128, 2002.
Kamal, M.A., Shahzeb F., Baber Y., 2005. “Resilient Behaviour of Asphalt Concrete under
repeated loading & effects of temperature” Journal of the Eastern Asia Society for
Transportation Studies, Vol. 6, pp 1329-1343.
Kandhal, S. P., Mallick B. R. and Brown R. E., 1998. ”Hot Mix Asphalt For Intersections in
Hot Climates”, National Center for Asphalt Technology, NCAT Report No. 98-06, Auburn,
Alabama.
Kandhal, S. P. and Mallick R.B., 2001. “Effects of Mix Gradation on Rutting Potential of
Dense-Graded Asphalt Mixtures”, Transportation Research Board, National Research Council,
Transportation Research Record No. 1767, Washington, D.C., pp. 146-151.
Kanitong, K., Bahia, H., 2005. Relating adhesion and cohesion of asphalts to the effect of
moisture on laboratory performance of asphalt mixtures, Transportation Research Record,
1901, 33-43.
121
Kennedy, T.W., P. S. Kandhal, S. F. Brown, D. Y. Lee, and F. L. Roberts. 1996, Hot Mix
Asphalt Materials, Mixture Design and Procedure. NAPA Research and Education Foundation,
Lanham, MD.
Khedr, S.A., 1986. “Deformation mechanism in asphaltic concrete.” J. Transp. Engrg. , ASCE,
112(1), 29-45
Kim, Y. R., N. Yim, and N. P. Khosla. 1992. Effects of Aggregate Type and Gradation on
Fatigue and Permanent Deformation of Asphalt Concrete. In American Society for Testing and
Materials, STP 1147, West Conshohocken, PA,
Klaus D., 2003. Asphalt flow improvers – A new technology for reducing mixing temperature
of asphalt concrete mixes with high-resistance against permanent deformation, Proceedings of
the 22nd PIARC World Road Congress, pp7.
Krutz, N. C., and P. E. Sebaaly. 1993. The Effects of Aggregate Gradation on Permanent
Deformation of Asphalt Concrete. In Proceedings, Vol. 62, Association of Asphalt Paving
Technologists.
Mahboub, K. and Little, D. N. 1988. Improved Asphalt Concrete Design Procedure, Research
Report 474-1F. Texas Transportation Institute
Majidzadeh, K, S. Khedr, and H. Guirguis 1981. Implementation of a pavement design system,
volume 1 and 2 of the final report. Research project EES 579. Ohio State Univeristy.
Columbus, OH.
Mallick, R. B. and P. S. Kandhal.2001. Effect of Mix Gradation on Rutting Potential of Dense
Graded Asphalt Mixtures. A Paper Submitted for Presentation and Publication at the 80
th
Annual Meeting of the Transportation Research Board, Washington, D. C.
Marks, V. J., Monroe, R. W. and Adam, J. F. 2001. Relating Creep Testing to Rutting of
Asphalt Concrete Mixes. For presentation at the Transportation Research Board 7-th Annual
Meeting January 13-17. Washington, D.C.
122
Matthews, J. M. and Monismith, C. L. 2003. Effects of Aggregate Gradation on the Creep
Response of Asphalt Mixtures and Pavement Rutting Estimates. ASTM Special Technical
Publication, No 1147, ASTM, Philadelphia, PA., p p. 329–343.
McCuen, R., “Statistical Methods for Engineers,” Prentice-Hall, Inc., (1985).
McGennis, R. B., R. M. Anderson, T. W. Kennedy, and M. Solaimanian. 1994, Background of
Superpave Asphalt Mixture Design and Analysis. Report SA-95-003. Federal Highway
Administration, U.S. Department of Transportation.
Monismith, C. L., and Tayebali, A. A. 1988. Permanent Deformation (Rutting) Considerations
in Asphalt Concrete Pavement Sections. The Association of Asphalt Paving Technologists v.
57, Williamsburg, Virginia.
Monismith, L. C., Brown F. S., 1999. “Developments in the Structural Design and
Rehabilitation of Asphalt Pavement over Three Quarters of a Century”, Asphalt Paving
Technology, 75th Anniversary Historical Review and Index of Journals 1975-1999, Volume
68A.
Monismith, C.L., 1976. “Rutting prediction in asphalt concrete pavements.” Transp. Res. Rec.
616, Transportation Research Board, Washington, D.C., 2-8.
Molenaar, A.A.A., 1993. Road materials Part III-Asphaltic Materials, Lecture Notes, Delft
University of Technology, The Netherlands.
Morris, J., 1973. The Prediction of Permanent Deformation in Asphalt Concrete Pavements.
Ph.D. Thesis, University of Waterloo, Ontario, Canada.
National Highway Authority, NHA., 1998, “General Specifications”, Surface Courses, Item
No. 305, Pakistan.
National Cooperative Highway Research Program; 2005. “Aggregate properties and the
performance of superpave Designed Hot Mix Asphalt”, NCHRP Report No. 539.
123
Parker, Frazier, and Brown, E. R. 1992. Effects of Aggregate Properties on Flexible Pavement
Rutting in Albama. Effects of Aggregate and Mineral Fillers on Asphalt Performance. ASTM
STP 1147. Philadelphia.
Pellinen, T.K. and Witczak, M.W., 2002. Stress Dependent Master Curve Construction for
Dynamic (Complex) Modulus. Proceedings of the Association of Asphalt Paving
Technologists Technical Sessions, vol. 71, CD Version.
Ponniah, J., and Kennepohl, G. 1996. Polymer-Modified Asphalt Pavements in Ontario:
Performance and Cost Effectiveness. Transportation Research Record No. 1545, pp. 151-160.
Prowell, B.D. 2001. Design, Construction and early Performance of Virginia’s Hot-Mix
Asphalt Stabilizer and Modifier Test Sections. Transportation Research Record No. 1767, pp.
7-14.
Rauhut, J. and Jordhal J. 1978. “Effects on Flexible Highways of Increased Legal Vehicles
Weights Using VESYS IIM, FHWA-RD-77-116.
Rita, B. Leahy and Matthew W. Witczak, 1991. The Influence of Test Conditions and Asphalt
Concrete Mix Parameters on Permanent Deformation Coefficient Alpha and Mu. Journal of the
Association of Asphalt Paving Technologies, vol1, pp 333-363.
Robert, N. Hunter 2000. “Asphalt in roads construction” American Society of Civil
Engineering, Thomas Telford Publications, pp 75, London.
Shashidhar, N., Needham, S.P., Chollar, B.H. and Romero, P. 1999. “Prediction of the
performance of mineral fillers in stone matrix asphalt” J. Assoc. Asphalt Paving Technologists,
68, pp 222–251.
Sivasubramaniam, S., Galal K. A., Noureldin, A. S., White T. D. and Haddock J. E., 2004.
“Modeling Permanent Deformation Using Laboratory, Proto-type, and In-Service Accelerated
Pavement Testing”, Transportation Research Board Annual Meeting, Washington, D.C.
124
Sousa, J.B., Craus J., Monismith, C.L. 1991. Summary Report on Permanent Deformation in
Asphalt Concrete. Strategic Highways Research Program IR-91-104.
Stephen, P., Yusu M., and Leslie, M. McCarthy., “ Evaluation and Modeling of Repeated Load
Test Data of Asphalt Concrete for Mechanistic-Empirical Pavement Design”, Journal of
Materials in Civil Engineering, ASCE, pp 993-999, Vol. 19, No. 11, November 2007.
Stumpf, W., Van Rooyen, G.T., 2007. Mechanical Metallurgy NMM 700, Course notes,
University of Pretoria, Pretoria.
Su. Z., 1996. Mineral Aggregates; Their Classifications and Properties, ESHA Research
Report, RT010-96-02, Department of Research and Technology, Smid & Holland Services
B.V.
Sullivan, B. Development of Flow Number and Flow time candidate simple performance test
for asphalt mixtures. MS Thesis. Department of Civil and Environmental Engineering, Arizona
State University, Tempe 2002.
Tarefder, R.A., Zaman, M., Hobson, K., 2003. A Laboratory and statistical evaluation of
factors affecting rutting, International Journal of Pavement Engineering. Volume 4 Number 1,
59-68.
Thomase, G. Mezger, 2002. “The Rheology Handbook” pp13.
Transporation Research Board, (TRB) 2000. Hot-Mix Asphalt Paving Hand Book, pp 18,
USA.
Uge, P., and Van de Loo, P. J. 1974. Permanent Deformation of Asphalt Mixes,
Koninklijke/Shell-Laboratorium. Amsterdam.
Vuong, B., and Amstrong, P., 1991. Repeated load Triaxial testing on the Subgrade from
Mulgrace ALF site, Australian Road Research Board
125
White, T., Hua J. and Galal K. 2003. “Analysis of Accelerated Pavement Tests”,
Transportation Research Board Annual Meeting, Paper CS8-7, Washington, D.C.
William, L. Hewitt and Floyed, O. Slate, 1967. “The Effects of the Rheological Properties of
Asphalt on Strength Characteristics of Asphalt Concrete”, Proceedings of 2nd International
Conference on the structural design of asphalt pavements; University of Michigan, USA
pp757-768.
Yeggoni, M., J. W. Button, and D. G. Zollinger. 1994. Influence of Coarse Aggregate Size,
Shape, and Surface Texture on Rutting of Hot Mix Asphalt Concrete. Report FHWATW94/1244-6. Federal Highway Administration and Texas Department of Transportation.
Ziari, H., Mahmud Ameri˛ and Mohammad Mahdi Khabiri; 2007. “Resilient Behaviour of Hot
Mixed and Crack Sealed Asphalt Concrete under Repeated Loading” Iran Science and
Technology University, ISSN 1392-8619 print/ISSN 1822-3613 online, Vol XIII, pp 56-60.
126
Annexure
127
Annexure -A
Plots of accumulative strain: at 250C
o
Permanent Strain~ Load Repetition at 25 C
1
500kPa
0.9
0.8
εp (%)
0.7
0.6
300kPa
0.5
0.4
0.3
100kPa
0.2
0.1
0
0
500
1000
1a,100kPa
Load Repetition (N)
1500
1a,300kPa
2000
1a,500kPa
Figure A.1: Influence of Pulse count on accumulative strain at 250C, Mix 1a
o
εp (%)
Permanent Strain~ Load Repetition at 25 C
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
500 kPa
300 kPa
`
100 kPa
0
500
Load Repetition (N)
1000
1b,100kPa
1500
1b,300kPa
2000
1b,500kPa
Figure A.2: Influence of Pulse count on accumulative strain at 250C, Mix 1b
128
o
Permanent Strain~ Load Repetition at 25 C
1
0.9
500 kPa
0.8
εp (%)
0.7
0.6
300 kPa
0.5
0.4
0.3
100 kPa
0.2
0.1
0
0
500
Load Repetition (N)
1000
1500
1c,100kPa
1c,300kPa
2000
1c,500kPa
Figure A.3: Influence of Pulse count on accumulative strain at 250C, Mix 1c
o
Permanent Strain~ Load Repetition at 25 C
500 kPa
1
0.9
0.8
εp (%)
0.7
300 kPa
0.6
0.5
0.4
0.3
100 kPa
0.2
0.1
0
0
500
1000
2a,100kPa
Load Repetition (N)
1500
2a,300kPa
Figure A.4: Influence of Pulse count on accumulative strain at 250C, Mix 2a
129
2000
2a,500kPa
o
Permanent Strain~ Load Repetition at 25 C
500 kPa
1
0.9
0.8
εp (%)
0.7
300 kPa
0.6
0.5
0.4
0.3
100 kPa
0.2
0.1
0
0
500
1000
2b,100kPa
Load Repetition (N)
1500
2000
2b,300kPa
2b,500kPa
Figure A.5: Influence of Pulse count on accumulative strain at 250C, Mix 2b
o
εp (%)
Permanent Strain~ Load Repetition at 25 C
500 kPa
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
300 kPa
100 kPa
0
Load Repetition (N)
500
1000
2c,100kPa
1500
2c,300kPa
Figure A.6: Influence of Pulse count on accumulative strain at 250C, Mix 2c
130
2000
2c,500kPa
Plots of accumulative strain: at 400C
0
Permanent Strain~ Load Repetition at 40 C
500 kPa
1
εp (%)
0.8
300 kPa
0.6
0.4
100 kPa
0.2
0
0
Load Repetition (N)
500
1000
1a,100kPa
1500
2000
1a,300kPa
1a,500kPa
Figure A.7: Influence of Pulse count on accumulative strain at 400C, Mix 1a
0
Permanent Strain~ Load Repetition at 40 C
1
500 kPa
εp (%)
0.8
300 kPa
0.6
0.4
0.2
100 kPa
0
0
Load Repetition (N)
500
1000
1b,100kPa
1500
1b,300kPa
2000
1b,500kPa
Figure A.8: Influence of Pulse count on accumulative strain at 400C, Mix 1b
131
0
Permanent Strain~ Load Repetition at 40 C
1
0.9
εp (%)
0.8
0.7
500 kPa
0.6
300 kPa
0.5
0.4
0.3
0.2
100 kPa
0.1
0
0
500
1000
1c,100kPa
Load Repetition (N)
1500
2000
1c,300kPa
1c,500kPa
Figure A.9: Influence of Pulse count on accumulative strain at 400C, Mix 1c
0
Permanent Strain~ Load Repetition at 40 C
500 kPa
1
0.9
0.8
300 kPa
εp (%)
0.7
0.6
0.5
0.4
100 kPa
0.3
0.2
0.1
0
0
Load Repetition (N)
500
1000
2a,100kPa
1500
2a,300kPa
Figure A.10: Influence of Pulse count on accumulative strain at 400C, Mix 2a
132
2000
2a,500kPa
0
Permanent Strain~ Load Repetition at 40 C
500 kPa
1.4
1.2
εp (%)
1
300 kPa
0.8
0.6
0.4
100 kPa
0.2
0
0
500
Load Repetition (N)
1000
2b,100kPa
1500
2b,300kPa
2000
2b,500kPa
Figure A.11: Influence of Pulse count on accumulative strain at 400C, Mix 2b
0
Permanent Strain~ Load Repetition at 40 C
1
0.9
500 kPa
0.8
εp (%)
0.7
300 kPa
0.6
0.5
100 kPa
0.4
0.3
0.2
0.1
0
0
500
1000
2c,100kPa
Load Repetition (N)
1500
2c,300kPa
Figure A.12: Influence of Pulse count on accumulative strain at 400C, Mix 2c
133
2000
2c,500kPa
Plots of accumulative strain development: at 550C
o
Permanent Strain~ Load Repetition at 55 C
1.4
1.2
500 kPa
εp (%)
1
0.8
300 kPa
0.6
0.4
100 kPa
0.2
0
0
500
Load Repetition (N)
1000
1a,100kPa
1500
2000
1a,300kPa
1a,500kPa
Figure A.13: Influence of Pulse count on accumulative strain at 550C, Mix 1a
0
Permane nt Strain~ Load Repetition at 55 C
1.4
1.2
500 kPa
εp (%)
1
300 kPa
0.8
0.6
0.4
100 kPa
0.2
0
0
500
Load Repetition (N)
1000
1b,100kPa
1500
1b,300kPa
2000
1b,500kPa
Figure A.14: Influence of Pulse count on accumulative strain at 550C, Mix 1b
134
o
Permanent Strain~ Load Repetition at 55 C
1.4
1.2
500 kPa
εp (%)
1
0.8
300 kPa
0.6
0.4
100 kPa
0.2
0
0
500
Load Repe tition (N)
1000
1c,100kPa
1500
1c,300kPa
2000
1c,500kPa
Figure A.15: Influence of Pulse count on accumulative strain at 550C, Mix 1c
0
Permanent Strain~ Load Repetition at 55 C
500 kPa
1.40
1.20
300 kPa
εp (%)
1.00
0.80
0.60
100 kPa
0.40
0.20
0.00
0
Load Repetition (N)
500
1000
2a,100kPa
1500
2a,300kPa
Figure A.16: Influence of Pulse count on accumulative strain at 550C, Mix 2a
135
2000
2a,500kPa
0
Permanent Strain~ Load Repetition 55 C
1.6
1.4
500 kPa
1.2
εp (%)
1
300 kPa
0.8
0.6
0.4
100 kPa
0.2
0
0
500
Load Repetition (N)
1000
2b,100kPa
1500
2000
2b,300kPa
2b,500kPa
Figure A.17: Influence of Pulse count on accumulative strain at 550C, Mix 2b
0
Permanent Strain~ Load Repetition at 55 C
1.4
500 kPa
1.2
εp (%)
1
300 kPa
0.8
0.6
100 kPa
0.4
0.2
0
0
500
Load Repetition (N)
1000
2c,100kPa
1500
2c,300kPa
Figure A.18: Influence of Pulse count on accumulative strain at 550C, Mix 2c
136
2000
2c,500kPa
Annexure -B
Trends of Accumulated Strains in Mixes: at 250C
( 250C)
% Acc. Strain x10
6
10.00
y100 = 0.202Ln(x) + 4.0717
R2 = 0.9582
y300 = 0.12Ln(x) + 4.7369
R2 = 0.966
y500= 0.1018Ln(x) + 5.1317
R2 = 0.9563
1.00
1
10
100
1000
10000
Pulse Count
1a, 25, 100
1a, 25,300
1a, 25, 500
Log. (1a, 25, 100)
Log. (1a, 25,300)
Log. (1a, 25, 500)
Figure B.1: Trends of accumulated strain at 250C, Mix 1a
o
(25 C)
6
% Accumulative StrainX10
10.00
y 500 = 0.0764Ln(x) + 5.2903
R2 = 0.9734
y 100 = 0.1525Ln(x) + 4.4119
R2 = 0.9769
y 300 = 0.0977Ln(x) + 4.9116
R2 = 0.9424
1.00
1
10
100
1000
10000
Pulse Counts
1b,100
1b,300
1b,500
Log. (1b,500)
Figure B.2: Trends of accumulated strain at 250C, Mix 1b
137
Log. (1b,300)
Log. (1b,100)
0
(25 C)
% Acc. Strain x10
6
10.00
y100 = 0.1463Ln(x) + 4.5462
R2 = 0.9779
y500 = 0.073Ln(x) + 5.522
R2 = 0.9791
y300 = 0.0938Ln(x) + 4.9078
R2 = 0.9463
1.00
1
10
1c,100
1c,300
100
Pulse Counts
1c,500
1000
Log. (1c,500)
Log. (1c,300)
10000
Log. (1c,100)
Figure B.3: Trends of accumulated strain at 250C, Mix 1c
0
(25 C)
6
% Accumulative Strainx10
10.00
y 100 = 0.2021Ln(x) + 4.0597
R2 = 0.977
y 500 = 0.119Ln(x) + 5.083
R2 = 0.8922
y 300 = 0.1384Ln(x) + 4.6635
R2 = 0.9104
1.00
1
10
100
1000
10000
Pulse Counts
2a,100
2a,300
2a,500
Log. (2a,100)
Figure B.4: Trends of accumulated strain at 250C, Mix 2a
138
Log. (2a,300)
Log. (2a,500)
o
(25 C)
6
% Accumulative Strainx10
10.00
y100 = 0.1503Ln(x) + 4.5297
R2 = 0.9662
y500 = 0.0973Ln(x) + 5.2469
R2 = 0.9881
y300 = 0.1083Ln(x) + 4.918
R2 = 0.9562
1.00
1
10
100
1000
10000
Pulse Counts
2b,100
2b,300
2b,500
Log. (2b,100)
Log. (2b,500)
Log. (2b,300)
Figure B.5: Trends of accumulated strain at 250C, Mix 2b
o
(25 C)
6
% Accu mu lativ e Strain x 1 0
10.00
y 500 = 0.0792Ln(x) + 5.247
R2 = 0.9881
y 100 = 0.0967Ln(x) + 4.8638
R2 = 0.9431
y 300 = 0.0796Ln(x) + 5.2134
R2 = 0.9794
1.00
1
10
100
1000
10000
Pulse Counts
2c,100
2c,300
2c,500
Log. (2c,100)
Log. (2c,300)
Figure B.6: Trends of accumulated strain at 250C, Mix 2c
139
Log. (2c,500)
Trends of Accumulated Strains in Mixes: at 400C
(400C)
% A ccumulative S train X 10 6
10.00
y 300 = 0.0808Ln(x) + 5.1522
R2 = 0.9387
y 100 = 0.12Ln(x) + 4.7371
R2 = 0.966
y 500 = 0.0792Ln(x) + 5.247
R2 = 0.9881
1.00
1
10
100
1000
10000
Pulse Counts
1a,100
1a,300
1a,500
Log. (1a,100)
Log. (1a,300)
Log. (1a,500)
Figure B.7: Trends of accumulated strain at 400C, Mix 1a
o
(40 C)
6
% Accumulative Strain x 10
10.00
y 300 = 0.1233Ln(x) + 4.8935
R2 = 0.9786
y 100 = 0.1289Ln(x) + 4.7331
R2 = 0.9494
y 500 = 0.1217Ln(x) + 5.0696
R2 = 0.9934
1.00
1
1b,100
10
1b,300
1b,500
100
Pulse Counts
Log. (1b,100)
1000
Log. (1b,300)
Figure B.8: Trends of accumulated strain at 400C, Mix 1b
140
10000
Log. (1b,500)
0
(40 C)
6
% Accumulative Strain X10
10.00
y 100 = 0.1808Ln(x) + 4.9329
R2 = 0.9387
y 300 = 0.1191Ln(x) + 5.1522
R2 = 0.9618
y 500 = 0.0972Ln(x) + 5.247
R2 = 0.9881
1.00
1
10
100
1000
10000
Pulse Counts
1c,100
1c,300
1c,500
Log. (1c,100)
Log. (1c,300)
Log. (1c,500)
Figure B.9: Trends of accumulated strain at 400C, Mix 1c
O
(40 C)
6
% Accumulative Strain x 10
10.00
y100 = 0.0945Ln(x) + 5.0601
R2 = 0.9028
y300 = 0.0858Ln(x) + 5.1443
R2 = 0.9591
y500 = 0.0847Ln(x) + 5.3571
R2 = 0.882
1.00
1
10
100
1000
10000
Pulse Counts
2a,100
2a,300
2a,500
Log. (2a,100)
Figure B.10: Trends of accumulated strain at 400C, Mix 2a
141
Log. (2a,300)
Log. (2a,500)
0
(40 C)
% Accumulative Strain X10
6
10.00
y 100 = 0.108Ln(x) + 5.1522
R2 = 0.9387
y 300 = 0.101Ln(x) + 5.1318
R2 = 0.9563
y 500 = 0.0919Ln(x) + 5.3092
R2 = 0.995
1.00
1
10
100
1000
10000
Pulse Counts
2b,100
2b,300
2b,500
Log. (2b,300)
Log. (2b,100)
Log. (2b,500)
Figure B.11: Trends of accumulated strain at 400C, Mix 2b
0
(40 C)
6
% Accumulative Strain X 10
10.00
y100 = 0.146Ln(x) + 4.911
R2 = 0.9984
y300 = 0.1217Ln(x) + 5.067
R2 = 0.9331
y500 =0.1088 Ln(x) + 5.096
R2 = 0.9934
1.00
1
2c,100
2c,300
10
100
Pulse Counts
2c,500
Log. (2c,100)
Figure B.12: Trends of accumulated strain at 400C, Mix 2c
142
1000
Log. (2c,500)
10000
Log. (2c,300)
5.6.6
Trends of Accumulated Strains in Mixes: at 550C
o
(55 C)
% Accumulative Strain X 10
6
10.00
y300 = 0.1525Ln(x) + 4.8194
R2 = 0.9748
y100 = 0.1595Ln(x) + 4.7329
R2 = 0.9496
y500 = 0.1234Ln(x) + 4.9792
R2 = 0.9824
1.00
1
10
100
1000
10000
Pulse Counts
1a,100
1a,300
1a,500
Log. (1a,500)
Log. (1a,300)
Log. (1a,100)
Figure B.13: Trends of accumulated strain at 550C, Mix 1a
o
(55 C)
6
% Accumulative Strain X 10
10.00
y 300 =0.1353 Ln(x) +5.0591
R2 = 0.9751
y 100 = 0.1597Ln(x) + 4.8185
R2 = 0.9037
y 500 = 0.0946Ln(x) + 5.317
R2 = 0.9923
1.00
1
10
1b,100
1b,300
100
Pulse Counts
1b,500
Log. (1b,300)
Figure B.14: Trends of accumulated strain at 550C, Mix 1b
143
1000
Log. (1b,500)
10000
Log. (1b,100)
o
(55 C)
% Accu mu lativ e Strain X 1 0
6
10.00
y 300 = 0.1548Ln(x) + 4.9669
R2 = 0.975
y 100 = 0.1896Ln(x) +4.8189
R2 = 0.9448
y 500 = 0.1287Ln(x) +5.0907
R2 = 0.9865
1.00
1
10
100
1000
10000
Pulse Counts
1c,100
1c,300
1c,500
Log. (1c,500)
Log. (1c,300)
Log. (1c,100)
Figure B.15: Trends of accumulated strain at 550C, Mix 1c
o
(55 C)
% Accu m u lativ e Strain X 1 0 6
10.00
y300 =0.1256Ln(x) + 4.8263
R2 = 0.9781
y 100 = 0.1584Ln(x) + 4.7803
R2 = 0.9933
y 500 = 0.1082Ln(x) + 5.0544
R2 = 0.9994
1.00
1
10
100
1000
10000
Pulse Counts
2a,100
2a,300
2a,500
Log. (2a,500)
Log. (2a,300)
Figure B.16: Trends of accumulated strain at 550C, Mix 2a
144
Log. (2a,100)
o
(55 C)
% Accumulativ e Strain X 10
6
10.00
y 300 = 0.1046Ln(x) + 5.0699
R2 = 0.998
y 100 = 0.1343Ln(x) + 5.0154
R2 = 0.9896
y 500 = 0.098Ln(x) + 5.4813
R2 = 0.7784
1.00
1
10
100
1000
10000
Pulse Counts
2b,100
2b,300
2b,500
Log. (2b,100)
Log. (2b,300)
Log. (2b,500)
Figure B.17: Trends of accumulated strain at 550C, Mix 2b
o
(55 C)
% Accumulative Strain X 10
6
10.00
y 300 = 0.1164Ln(x) +4.9869
R2 = 0.9793
y 100 = 0.1576Ln(x) + 4.8324
R2 = 0.9954
y 500 = 0.1053Ln(x) + 5.3167
R2 = 0.9921
1.00
1
10
100
1000
10000
Pulse Counts
2c,100
2c,300
2c,500
Log. (2c,500)
Log. (2c,100)
Figure B.18: Trends of accumulated strain at 550C, Mix 2c
145
Log. (2c,300)
Annexure-C
Plots of Resilient Strain: at 250C
o
o
Resilient S train ~ Load Repetition at 25 C
Resilient Strain~ Load Repetition at 25 c
0.14
0.18
0.12
0.16
0.14
500kPa
500 kPa
0.12
% Resilient Strain
% Resilient Stra in
0.1
0.08
0.06
`
300kPa
0.04
0.02
0.1
300 kPa
0.08
0.06
0.04
100 kPa
0.02
100kPa
0
0
0
500
1000
1a,100kPa
Load Repetition (N)
1500
1a,300 kPa
-0.02
2000
1a,500 kPa
0
500
2a,300kPa
2a,500kPa
Resilient Strain ~ Load Repetition at 25 C
0.18
0.120
500 kPa
0.080
300 kPa
0.060
`
0.040
100 kPa
0.020
% Resilient Strain
0.16
0.100
0.14
0.12
500 kPa
0.1
0.08
300 kPa
0.06
0.04
100 kPa
0.02
0.000
0
0
500
Load Repetition (N)
1000
1b,100kPa
1500
1b,300kPa
2000
0
1b,500kPa
500
Load Repetition (N)
1000
1500
2000
2b,100kPa
2b,300kPa
2b,500kPa
o
o
Resilient Strain ~ Load Repetition at 25 C
Resilient Strain~ Load Repetition at 25 C
0.12
0.140
0.1
0.120
500 kPa
% Resilient Strain
% Resilient Strain
2000
o
Resilient Strain ~ Load Repetition at 25 C
% Resilient Strain
1500
2a,100kPa
Load Repetition (N)
o
1000
0.08
0.06
300 kPa
0.04
0.02
100 kPa
0
500
1000
1c,100kPa
1500
1c,300kPa
0.080
0.060
300 kPa
0.040
0.020
0
Load Repetition (N)
500 kPa
0.100
100 kPa
0.000
2000
0
1c,500kPa
Load Repetition (N)
Figure C.1: Influence of Pulse count on resilient strain at 250C
146
500
1000
2c,100kPa
1500
2c,300kPa
2000
2c,500kPa
Plots of Resilient Strain at 400C
o
0
Resilient Strain ~ Load Repetition at 40 C
Resilient Strain ~ Load Repetition at 40 C
0.100
500 kPa
0.080
0.060
300 kPa
0.040
100 kPa
0.020
% R esilien t S tra in
% R esilient Strain
0.100
300 kPa
0.050
100 kPa
0.000
0.000
0
500
1000
Load Repetition (N)
1a,100kPa
1500
0
2000
1a,300kPa
1a,500kPa
o
0.1
300 kPa
0.05
100 kPa
0
Load Repetition (N)
1b,100kPa
1500
% R esilien t S tra in
% R esilient Strain
500 kPa
1000
2a,300kPa
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
2000
1b,300kPa
2a,100kPa
1500
2000
2a,500kPa
Resilient Strain~ Load Repetition at 40 C
0.15
500
1000
o
Resilient Strain~ Load Repetition at 40 C
0
500
Load Repetition (N)
0.2
500 kPa
300 kPa
100 kPa
0
1b,500kPa
500
Load Repetition (N)
1000
1500
2000
2b,100kPa
2b,300kPa
2b,500kPa
o
0
Resilient Strain~ Load Repetition at 40 C
Resilient Strain~ Load Repetition at 40 C
0.18
0.12
0.16
% R esilien t S tra in
500 kPa
% R esilient Stra in
500 kPa
0.08
300 kPa
0.04
100 kPa
0.14
500 kPa
0.12
0.1
0.08
300 kPa
0.06
0.04
0.02
0
100 kPa
0
0
500
Load Repetition (N)
1000
1c,100kPa
1500
1c,300kPa
0
2000
1c,500kPa
Load Repetition (N)
Figure C.2: Influence of Pulse count on resilient strain at 400C
147
500
1000
2c,100kPa
1500
2c,300kPa
2000
2c,500kPa
Plots of Resilient Strain: at 550C
0
0
Resilient Strain ~ Load Repetition at 55 C
Resilient Strain ~ Load Repetition at 55 C
0.150
500 kPa
0.100
300 kPa
0.050
100 kPa
% R esilient Strain
% R esilient Strain
0.200
0.000
0
500
Load Repetition (N)
1000
1a,100kPa
1500
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
2000
1a,300kPa
1a,500kPa
500 kPa
300 kPa
100 kPa
0
500
Load Repetition (N)
1000
2a,100kPa
1500
2a,300kPa
0
Resilient Strain ~ Load Repetition at 55 C
0
Resilient Strain ~ Load Repetition 55 C
0.35
0.160
% Resilient Strain
0.120
0.100
300 kPa
0.080
0.060
0.040
100 kPa
0.3
% R esilient Strain
500 kPa
0.140
0.25
500 kPa
0.2
0.15
300 kPa
0.1
0.05
0.020
100 kPa
0
0.000
0
500
Load Repetition (N)
1000
1b,100kPa
1500
0
2000
1b,300kPa
1b,500kPa
500
Load Repetition (N)
1000
2b,100kPa
0
Resilient Strain ~ Load Repetition at 55 C
300 kPa
0.060
0.040
100 kPa
0.020
0
Load Repetition (N)
500
1000
1c,100kPa
1500
2000
1c,300kPa
1c,500kPa
% R esilient Strain
500 kPa
0.080
2000
2b,500kPa
0
0.120
0.100
1500
2b,300kPa
Resilient Strain ~ Load Repetition at 55 C
0.140
% Resilient Stra in
2000
2a,500kPa
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
0
500
Load Repetition (N)
Figure C.3: Influence of Pulse count on resilient strain at 550C
148
500 kPa
300 kPa
100 kPa
1000
2c,100kPa
1500
2c,300kPa
2000
2c,500kPa
Annexure-D
Influence of Wheel Tracker Load Cycle on Rut Depth
Mixture with gradation "01" & PMA
(1a)
55 C
400C
5.000
4.000
3.000
250 C
2.000
0.000
40C,1a
Load Cycles (N)
100
1300
2500
3700
4900
6100
7300
8500
9700
10900
12100
13300
14500
15700
16900
20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
0.000
55C, 1a
25C,1a
Mixture with gradation "01" & 60/70 pen grade
(1b)
16.00
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
25C, 2a 40C,2A 55C,2A
Load Cycles (N)
Mixture with gradation "02" & 60/70 pen grade
(2b)
25.000
0
55 C
R ut D epth (m m )
0
0
40 C
0
25 C
55 C
20.000
0
15.000
40 C
10.000
0
25 C
5.000
Load Cycles (N)
Mixture with gradation "01" & 40/50 pen grade
(1c)
25C, 2b
16900
15700
14500
13300
12100
9700
10900
8500
7300
6100
4900
3700
2500
100
25C, 1b 40C,1b 55C,1b
Load Cycles (N)
14.000
1300
0.000
100
1500
2900
4300
5700
7100
8500
9900
11300
12700
14100
15500
16900
R u t D e p th (m m )
400 C
10.000
8.000
6.000
4.000
250C
2.000
1.000
550 C
16.000
14.000
12.000
7.000
6.000
Mixture with gradation "02" & PMA
(2a)
20.000
18.000
0
R u t D e p th (m m )
R ut D epth (mm)
9.000
8.000
40C,2b
Mixture with gradation "02" & 40/50 pen. grade
(2c)
20.000
12.000
0
55 C
0
55 C
18.000
16.000
10.000
R u t D ep th (m m )
R ut D epth (m m )
55C, 2b
8.000
0
40 C
6.000
4.000
0
25 C
14.000
0
40 C
12.000
10.000
8.000
6.000
0
25 C
4.000
2.000
2.000
0.000
Load Cycles (N)
40C,1c
55C, 1c
25C,1c
Figure D.1: Influence of Load Cycle on Rut Depth
149
Load Cycles (N)
13100
14100
15100
16100
17100
4100
5100
6100
7100
8100
9100
10100
11100
12100
100
1100
2100
3100
20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
0.000
25C, 2c
40C,2c
55C, 2c
Trends of Rut Depth Development in WT
Mixture with gradation "01" & PMA
(1a)
7.500
y25 = 0.2694Ln(x) + 3.7261
R2 = 0.9524
6
Rut Depth (mmx10 )
7.000
6.500
6.000
5.500
5.000
y40 = 0.2037Ln(x) + 4.8079
R2 = 0.9965
4.500
y55 = 0.2231Ln(x) + 4.7724
R2 = 0.9944
4.000
10
100
1000
10000
100000
Load Cycle (LOG N)
1A, 25C
Log. (1A, 25C)
1a, 40C
Log. (1a, 40C)
1a, 55C
Log. (1a, 55C)
Figure D.2: Trends of Rut Depth Development, Mix 1a
Mixture with gradation "01" & 60/70 pen. Grade
(1b)
7.5
y25 = 0.2737Ln(x) + 3.9131
R2 = 0.9838
6
Rut Depth (mmx10 )
7.0
6.5
6.0
5.5
y55 = 0.2174Ln(x) + 5.1163
R2 = 0.9591
5.0
y40 = 0.2295Ln(x) + 4.8107
R2 = 0.9452
4.5
4.0
10
100
1000
10000
100000
Load Cycle (LOG N)
1b, 25C
Log. (1b, 25C)
1b, 40C
Log. (1b, 40C)
Figure D.3: Trends of Rut Depth Development, Mix 1b
150
1b, 55C
Log. (1b, 55C)
Mixture with gradation "01" & 40/50 pen. Grade
(1c)
y25 = 0.2957Ln(x) + 3.4551
R2 = 0.3927
7.0
6
Rut Depth (mm x 10 )
7.5
6.5
6.0
5.5
5.0
y40 = 0.2111Ln(x) + 4.7834
R2 = 0.9909
4.5
4.0
10
100
1c 25C
Log. (1c 25C)
y55 = 0.2844Ln(x) + 4.3368
R2 = 0.9809
1000
10000
Load Cycle (LOG N)
1c, 40C
Log. (1c, 40C)
100000
1c, 55C
Log. (1c, 55C)
Figure D.4: Trends of Rut Depth Development, Mix 1c
Mixture with gradation "02" & PMA
(2a)
7.5
6
Rut Depth (mmx10 )
7.0
6.5
6.0
y 25 = 0.1317Ln(x) + 5.3749
R2 = 0.9926
5.5
5.0
`
y 40 = 0.2161Ln(x) + 4.9767
R2 = 0.9676
4.5
y 55 = 0.2217Ln(x) + 5.1177
R2 = 0.9875
4.0
10
100
1000
10000
100000
Load Cycle (LOG N)
2a, 25C
Log. (2a, 25C)
2a, 40C
Log. (2a, 40C)
Figure D.5: Trends of Rut Depth Development, Mix 2a
151
2a, 55C
Log. (2a, 55C)
Mixture with gradation "02" & 60/70 pen. Grade
(2b)
8.0
y25 = 0.1633Ln(x) + 5.1857
R2 = 0.9869
6
Rut Depth (mmx10 )
7.5
7.0
6.5
6.0
5.5
5.0
y55 = 0.2118Ln(x) + 5.3468
R2 = 0.9691
4.5
y40 = 0.1968Ln(x) + 5.2741
R2 = 0.9801
4.0
10
100
2b, 25C
Log. (2b, 25C)
1000
10000
Load Cycle (LOG N)
2b, 40C
Log. (2b, 40C)
100000
2b, 55C
Log. (2b, 55C)
Figure D.6: Trends of Rut Depth Development, Mix 2b
Mixture with gradation "02" & 40/50 pen. Grade
(2c)
8.0
6
Rut Depth (mmx10 )
7.5
y 25 = 0.2606Ln(x) + 4.1251
R2 = 0.9942
7.0
6.5
6.0
5.5
5.0
y 55 = 0.239Ln(x) + 5.0268
R2 = 0.9165
y 40 = 0.199Ln(x) + 5.1743
R2 = 0.9824
4.5
4.0
10
2c, 25C
Log. (2c, 25C)
100
1000
Load Cycle (LOG N)
10000
2c, 40C
Log. (2c, 40C)
Figure D.7: Trends of Rut Depth Development, Mix 2c
152
100000
2c, 55C
Log. (2c, 55C
Imran Hafeez