,1 u ,‘e "’11 VVVVVV
.. . ,.‘.,.;.‘;,Mhm;,p, 41 , ‘
. $1.31 "2% ‘l'mu‘b {R 9":1'“
'5ng kiwi” “"01"; ' “flma‘iii‘! “a“

 
 
 
    
       
  

  
 

r ‘,. 1
I'm; ‘ 21:1?

' . ”f?
3 “11,1 yfiifi'}.

"‘£‘

mil-Mu “3’

  
 

“he:

 
     
         
  
   
   
   
  
   
 
        
 
      
    

      
 

 
 

  

   
 

   

      
          
    
  
      

 
  
  
   

     

    

  
    
  

    
   
     
   
   
     
  

   

 
  

 

 
 
   
        

 

  
  

 

   
 

 

 

:1 ' i .1521?
‘ ~;. 1:91;; QTY ‘ ‘ (v Jfgl 15"
,v 43-13% ‘ “1,1; mm LA; «3 z
f; ‘;;=.§§;Ed'.,{{;.ai:i¥." "‘9‘33' Jivu'smil {Wm “23%;. 4‘.
L 1 '1'»... aw; ':i . 1.; .gW . fig," . Wi :;:_
: “jwjfifi;dwwflflyvmipmmbfflgfi‘WWW%“T“W"$EwE 3
»: " "w * '1'? t -Mt‘l‘il‘~~;'
, "‘33? WM" 1 - : mall "“33 film, 13 =."-'- '7 Lu: _
n“ "..,.l:;1t)£“‘l ft n 1‘!” law” nix”: . ' 32!; _
mm [1.3% 3‘, .1“ "21?.“ Li 131%": = 3:;
i "h 1:;«tizl:!:' :i h :J “I“??? “g ‘13.”; ‘ > >‘
":1.“ "“‘n‘\ :1} I" ‘; " . “r ‘I Alt '1 1.43.4 v is. .
k il;::![i:':"fl c'iigtm: "3‘ 1:21;,3'2’1 ”Iii '33,}. ’35 .~ < Rig ‘
‘::i:1!"‘f".. . ' 5 .
:.Jlu%f‘.. l;-
i .‘L "‘nHY":
~ .1‘.- a
- u.’ .‘ :. ,,;:
; ”star; .3:
z 3-2;";q‘,
; ‘w '.
;
E iikg‘:
M'
[I
J.
a "1:14 " i" V 3'. ‘.
if 1‘ ‘. it . ; Z‘K‘Zj
‘ If: -JY i; »¢f€
.' . 11w “-i' up};
£. I: .‘ r l ' , :L‘ :1 ._‘ .31 ti '
. .{ ' "-. "r. r ' n‘ .;'J "- 1" 3." '.:' ‘P' ' 2"“
S ‘ 13-» v; .m, ;: ' .. ' 424(33141‘“: :55}? 311i?
p214. 27‘3-«1 '1 [131??? “""?{"'" “ 1' am
' upbmifl'"N31 ‘kll"t;|’}r.fz“."13hHgfiEHW 3‘12“
| A], {megwh‘ [flifld ”I 1-11. ['1' l‘ft "M Hm. :II 517}
* :JWWwLWMWWJW+HWMw?iWk :+
} 1 "bn mTulip/419;? E'W’niir ‘z‘fifihi'y'flt Mg '
~ la Hum; Win w W W 1M
l ' -we‘~“rw..“.t:‘.t“ W Iz}: "m :3! 45
fi‘ifimmfih~l$fiu ”M’s ,fi? ’
a" W M, MW ~ .d‘IflzV f. .‘3'
‘ .‘E'Hln Wu : L'L. -.

 

m
o
n
7.;
7.x.

E5
u...
-4‘
'1":
,V

THESIS

Z GANHISTATE

IIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIII

31293 01707 4224

 

 

 

This is to certify that the

thesis entitled

Chemical, Physical, and Thermodynamic Properties
of Neat and Polymer Modified Asphalt Binders

presented by

Lawrence Michael France

has been accepted towards fulfillment
of the requirements for

Master of Science Chemical Engineering
degree in

 

 

Wigaxég /

Major professor

Date jig"! /Q LIL/f? a]

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DIJE DATE DUE DATE DUE

HIM—34999—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1I98 chlRC/DateDuopGS—n“

Chemical, Physical, and Thermodynamic Properties of Neat and Polymer

Modified Asphalt Binders

Lawrence M. France

A THESIS

Submitted to
Michigan State University

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

Department of Chemical Engineering

1997

ABSTRACT

Chemical, Physical, and Thermodynamic Properties of Neat and Polymer
Modified Asphalt Binders

by

Lawrence M. France

Asphalt was modified with four different polymers to investigate the benefits of
polymer modification. The modifiers studied were styrene-butadiene rubber, Elvaloy
AM, recycled crumb rubber, and ethylene vinyl acetate. A mixing procedure was
developed for each modifier based on improvement in the rheological properties of the
blend. A number of physical properties were investigated to determine the benefit of
polymer modification. Polymer modification improved high temperature properties and
did not affect low temperature properties.

Fourier Transform Infrared Spectroscopy and Gel-Permeation Chromatography
were used to fingerprint the modified binders. Phenomenological models were developed
for SBR and CRM rheological properties. The properties of SBR are a combination of
two phenomena: percolation and network formation. The properties of CRM are a result
of simple two-phase dispersed mixing. Laser scanning and environmental scanning
electron microscopy were used to verify the models. The applicability of the SHRP

performance based specifications for modified asphalts was investigated.

DEDICATION
This work is dedicated to my parents, Larry and Cheri, without whom this work would
not have been possible. Their support, both financial and emotional, has been invaluable.
Thank you for everything and God bless you.
and
In Loving Memory of A. L. “Van” VanTassel

February 8, 1926 - January 26, 1997
May the Lord bless you and keep you.

ACKNOWLEDGMENTS

The author wishes to thank Dr. Martin C. Hawley for his guidance and support
during the course of the investigation. His willingness to let the author's interests dictate
the course of the research is greatly appreciated. Thanks are also extended to Dr.
Lawrence T. Drzal and Dr. Gilbert Baladi for serving as co-principal investigators of the
MSU/MDOT project. Thank you to the staff of the Composite Materials and Structures
Center for their instruction and guidance regarding experimental techniques. Thank you
to the staff at the Michigan Department of Transportation Buituminous Lab, especially
John Barak and Doug Coleman, for their input on how to keep the research focused on
real world problems and applications.

The author wishes to thank his family for their emotional support, financial
support, and general encouragement.

The Michigan Department of Transportation supported this research.

TABLE OF CONTENTS

List of Figures .................................................................................................................... ix
List of Tables .................................................................................................................. xiii
Chapter One: Scope ............................................................................................................. 1
Chapter Two: Introduction and Background ........................................................................ 5
2.1 Asphalt Composition ..................................................................................................... 6
2.2 Asphalt Failure Mechanisms .......................................................................................... 9
2.3 Reasons for Polymer Modification .............................................................................. 10
2.4 Polymers Commonly used for Modification ................................................................ 12
2.5 Recent Results from the Literature .............................................................................. 14
2.6 Summary of SBS/SEBS Research ............................................................................... 24
Chapter Three: Testing Procedures .................................................................................... 27
3.1 SHRP Background ....................................................................................................... 28
3.2 SHRP Binder Specifications ........................................................................................ 28
3.2.1 Safety Test .................................................................................................. 29
3.2.2 Viscosity Test ............................................................................................. 30
3.2.3 Permanent Deformation (Rutting) Test ...................................................... 30
3.2.4 Excessive Aging Test ................................................................................. 31
3.2.5 Fatigue Cracking Test ................................................................................ 31
3.2.6 Thermal Cracking Test ............................................................................... 31
3.3 Detailed Test Procedures ............................................................................................. 32
3.3.1 Flash Point ................................................................................................. 32
3.3.2 Dynamic Shear Rheometry ........................................................................ 32

3.3.3 Rotational Viscometry ............................................................................... 36

3.3.4 Thin Film Oven Aging (ASTM D 1754-87) .............................................. 37

3.3.5 Pressure Vessel Aging ............................................................................... 38

3.3.6 Bending Beam Rheometry ......................................................................... 39

3.3.7 Thermal Mechanical Analysis .................................................................... 41

3.3.8 High Performance Gel Permeation Chromatography ................................ 43

3.3.9 Fourier Transform Infrared Spectroscopy .................................................. 44

3.3.10 Compatibility Index ................................................................................... 45

3.3.11 Asphaltene Fractionation ........................................................................... 45

3.3.12 Asphalt/Polymer Blending ......................................................................... 45
Chapter Four: Polymer Modification Results .................................................................... 48
4.1 Styrene-Butadiene Rubber ........................................................................................... 48
4.1.1 Mixing Procedure ....................................................................................... 49

4.1.2 Dynamic Shear Rheometry ........................................................................ 51

4.1.2.1 Optimum Polymer Content ........................................................... 51

4.1.2.2 SHRP High Temperature Performance Grade .............................. 53

4.1.3 Rotational Viscometry ............................................................................... 55

4.1.4 Bending Beam Rheometry ......................................................................... 56

4.1.5 Fourier Transform Infrared Spectroscopy .................................................. 59

4.1.6 Gel Permeation Chromatography ............................................................... 61

4.1 .7 Thermomechanical Analysis ...................................................................... 62

4.1.8 Effect of Polymer Molecular Weight on Modified Binder Properties ....... 62

4.2 Elvaloy® AM ............................................................................................................... 68

vi

4.2.1 Mixing Procedure ....................................................................................... 68
4.2.2 Mixing Scale-Up ........................................................................................ 72
4.2.3 Dynamic Shear Rheometry ........................................................................ 73
4.2.4 Rotational Viscometry ............................................................................... 75
4.2.5 Bending Beam Rheometry ......................................................................... 76
4.2.6 Fourier Transform Infrared Spectroscopy .................................................. 78
4.2.7 Gel Permeation Chromatography ............................................................... 81
4.2.8 Thermomechanical Analysis ...................................................................... 82
4.3 Recycled Crumb Rubber .............................................................................................. 84
4.3.1 Mixing Procedure ....................................................................................... 85
4.3.2 Dynamic Shear Rheometry ........................................................................ 89
4.3.3 Rotational Viscometr}I ............................................................................... 91
4.3.4 Thermomechanical Analysis ...................................................................... 93
4.4 Ethylene Vinyl Acetate ................................................................................................ 93
4.4.1 Mixing Procedure ....................................................................................... 95
4.4.2 Dynamic Shear Rheometry ........................................................................ 97
4.4.3 Rotational Viscometry ............................................................................. l 01
4.4.4 Fourier Transform Infrared Spectroscopy ................................................ 102
4.4.5 Thermal Mechanical Analysis .................................................................. 104
Chapter Five: Modeling ................................................................................................... 105
5.1 Phenomenological Modeling ..................................................................................... 105
5.1.1 Styrene Butadiene Rubber ........................................................................... 106
5.1.2 Recycled Crumb Rubber ............................................................................. 1]]

vii

5.2 Correlation Based Modeling ...................................................................................... 112

5.2.1 Materials Tested .......................................................................................... 117

5.2.2 Experiments ................................................................................................ 117

5.2.3 One Variable Correlations .......................................................................... 119

5.2.4 Multivariable Correlation ............................................................................ 119
Chapter Six: Applicability of SHRP Performance Grading to Modified Binders ........... 130
6.1 Flash Point Test .......................................................................................................... 130
6.2 Viscosity Test ............................................................................................................. 130
6.3 Rutting Resistance ..................................................................................................... 131
6.4 Excessive Aging ......................................................................................................... 131
6.5 Fatigue Cracking ........................................................................................................ 132
6.6 Low Temperature Cracking ....................................................................................... 132
Chapter Seven: Conclusions ............................................................................................ 135
7.1 Styrene-Butadiene Rubber ......................................................................................... 136
7.2 Elvaloy® AM ............................................................................................................. 136
7.3 Recycled Crumb Rubber ............................................................................................ 138
7.4 Ethylene Vinyl Acetate .............................................................................................. 139
7.5 Recommendations for Further Study ......................................................................... 140
List of References ............................................................................................................ 142

LIST OF FIGURES
Chapter Two

Figure 2.1 - The two phase asphalt model ....................................................................... 8

viii

Figure 2.2 -
Figure 2.3 -
Figure 2.4 -
Figure 2.5 -
Figure 2.6 -
Figure 2.7 -
Figure 2.8 -
Figure 2.9 -
Figure 2.10 -
Figure 2.11 -

Figure 2.12 -

Figure 3.1 -
Figure 3.2 -
Figure 3.3 -
Figure 3.4 -
Figure 3.5 -
Figure 3.6 -
Figure 3.7 -
Figure 3.8 -
Figure 3.9 -

Figure 3.10 -

Spring/shock absorber model of asphalt .................................................... 1 1

Polymer/asphalt phase behavior ................................................................. 13
Modifier effect on fatigue cracking ............................................................ 15
Modifier effect on rutting ........................................................................... 15
Modifier effect on thermal cracking .......................................................... l6
Modifier effect on raveling ........................................................................ 16
Modifier effect on age hardening ............................................................... 17
Chemical structure of Elvaloy® AM ......................................................... 21
Loss and storage moduli at 60°C of AC-S/SBS binders (Batch #1) ........... 25
Loss and storage moduli at 60°C of AC-S/SBS binders (Batch #2) ........... 25
Depiction of SBS modified asphalt theory at service temperatures ........... 26

Chapter Three
Variations of three viscosity-graded asphalts ............................................ 27
Viscoelastic behavior of asphalt ................................................................ 34
Sample setup for DSR ................................................................................ 35
Reproducibility of rheological data ............................................................ 35
Rotational viscosity setup .......................................................................... 36
Thin film oven test apparatus....................... .............................................. 38
Pressure aging vessel and components ...................................................... 40
Bending beam rheometer apparatus ........................................................... 40
BBR deflection and m-value ...................................................................... 42
Typical TMA curve showing softening point and melt temperature ......... 43

Figure 3.11 -

Figure 4.1 -
Figure 4.2 -
Figure 4.3 -
Figure 4.4 -
Figure 4.5 -
Figure 4.6 -
Figure 4.7 -
Figure 4.8 -
Figure 4.9 -
Figure 4.10 -
Figure 4.11 -
Figure 4.12 -
Figure 4.13 -
Figure 4.14 -
Figure 4.15 -
Figure 4.16 -
Figure 4.17 -
Figure 4.18 -
Figure 4.19 -

Figure 4.20 -

Mixing apparatus ....................................................................................... 46

Chapter Four
Storage modulus of AC-5/SBR binders with different mixing times ........ 50
Tan 6 curves of AC-5/SBR binders with different mixing times .............. 50
Loss and storage moduli of AC-S/SBR binders ......................................... 52
Tan 8 curves of AC-5/SBR binders ........................................................... 53
Dynamic shear modulus of AC-S/SBR binders ......................................... 54
Melt viscosity of AC-5/SBR binders ......................................................... 56
Creep stiffness of AC-S/SBR binders ........................................................ 57
M-value of AC-S/SBR binders .................................................................. 58
Typical FTIR spectrum for AC-2.5/SBR blends ........................................ 60

Dynamic shear modulus of A05 with three different SBR polymers ...... 65

Low temperature rheology of AC-5/SBR blends ....................................... 66
Melt viscosity of AC-S/SBR binders ......................................................... 67
Creep stiffness of AC-5/SBR binders at -12°C .......................................... 67
M-value of AC-S/SBR binders at -12°C .................................................... 68
Tan 8 curves of AC-S/EAM binders mixed for two hours (no cure) ......... 70
Tan 8 curves of AC-5/EAM binders with 23 hours cure time ................... 69
Tan 6 curves of AC-S/EAM binders with 69 hours cure time ................... 70

Dynamic shear modulus of different batch sizes of AC-S/EAM blends....73
Melt viscosity of AC-5/Elvaloy® AM modified binders ........................... 75

Melt viscosity of AC-2.5/Elvaloy® AM modified binders ........................ 76

Figure 4.21 -
Figure 4.22 -
Figure 4.23 -

Figure 4.24 -

Figure 4.25 -

Figure 4.26 -

Figure 4.27 -
Figure 4.28 -
Figure 4.29 -

Figure 4.30 -

Figure 4.31 -

Figure 4.32 -
Figure 4.33 -
Figure 4.34 -
Figure 4.35 -
Figure 4.36 -

Figure 4.37 -

Figure 5.1 -

BBR data of AC-2.5/SBR and AC-2.5/EAM binders at -1 8°C .................. 78

Typical AC-S/Elvaloy® AM FTIR spectrum ............................................ 79
Typical AC-S/Elvaloy® AM GPC curve ................................................... 82
Dynamic shear modulus of 10% (w/w) CRM modified AC-lO mixed

for one hour at various temperatures .......................................................... 87
Tan 8 curves of 10% (w/w) CRM modified AC-10 mixed for one hour

at various temperatures .............................................................................. 87
Dynamic shear modulus of 10% (w/w) CRM modified AC-10 mixed

at 300°F for various amounts of time ......................................................... 88
Dynamic shear modulus of AC-lO/CRM binders ...................................... 89
Tan 6 curves of AC-lO/CRM binders ........................................................ 90
Viscosity of AC-lO/CRM binders .............................................................. 91
Percent increase in viscosity of AC-lO/CRM binders after high

temperature storage .................................................................................... 92

Dynamic shear modulus of AC-S/EVA binders with different mixing

conditions ................................................................................................... 97
Tan 8 curves of AC-5/EVA binders ........................................................... 98
Storage modulus of AC-S/EVA binders .................................................... 99
Storage and loss modulus at 60°C of AC-5/EVA (P3104) blends ............ 99
Dynamic shear modulus of AC-5fEVA blends ........................................ 101
Viscosity of AC-5/EVA binders .............................................................. 102
FTIR calibration curves for AC-S/EVA (P8104) blends ......................... 103
Chapter Five
Rheological properties of AC-5/SBR blends ........................................... 109

xi

Figure 5.2 -
Figure 5.3 -
Figure 5.4 -
Figure 5.5 -
Figure 5.6 -
Figure 5.7 -
Figure 5.8 -
Figure 5.9 -
Figure 5.10 -
Figure 5.11 -
Figure 5.12 -
Figure 5.13 -
Figure 5.14 -
Figure 5.15 -
Figure 5.16 -
Figure 5.17 -
Figure 5.18 -
Figure 5.19 -
Figure 5.20 -
Figure 5.21 -
Figure 5.22 -

Figure 5.23 -

Laser Scanning Microscope micrographs of SBR modified binders ....... 110
Pictorial representation of SBR behavior model ...................................... 11 l
Rheological properties of AC-lO/CRM blends ........................................ l 13
Scanning electron micrograph of 20% (w/w) CRM modified AC-lO ..... 114
Softening and melt temperature versus Mn ............................................. 121
Softening and melt temperature versus Mw ............................................ 121
Softening and melt temperature versus asphaltene fraction ..................... 122
Penetration versus Mn .............................................................................. 122
Penetration versus Mw ............................................................................. 123
Penetration versus asphaltene fraction ..................................................... 123
Viscosity versus Mn ................................................................................. 124
Viscosity versus Mw ................................................................................ 124
Viscosity versus asphaltene fraction ........................................................ 125
PVN versus Mn ........................................................................................ 125
PVN versus Mw ....................................................................................... 126
PVN versus asphaltene fraction ............................................................... 126
Viscosity and penetration ratios versus Mn ............................................. 127
Viscosity and penetration ratios versus Mw ............................................ 127
Viscosity and penetration ratios versus asphaltene fraction .................... 128
VTS and aging index versus Mn .............................................................. 128
VTS and aging index versus Mw ............................................................. 129
VTS and aging index versus asphaltene fraction ..................................... 129

xii

Table 2.1 -

Table 2.2 -

Table 2.3 -

Table 2.4 -

Table 3.1 -

Table 3.2 -

Table 4.1 -

Table 4.2 -

Table 4.3 -

Table 4.4 -

Table 4.5 -

Table 4.6 -

Table 4.7 -

Table 4.8 -

Table 4.9 -

Table 4.10 -

LIST OF TABLES

Chapter Two
Elemental analysis of several asphalts ......................................................... 6
Asphaltene/Resin/Oil ratio of three asphalts ................................................ 7
Asphalt distress modes ............................................................................... 10
Most common polymers used for modification ......................................... 14
Chapter Three
SHRP performance grading worksheet ...................................................... 29
Mixing time and temperature for all modifiers .......................................... 47
Chapter Four
Optimum SBR content ............................................................................... 52
High temperature SHRP PG for SBR modified binders ............................ 55
Low temperature SHRP PG of SBR modified binders .............................. 58
FTIR oxidation ratios of AC-2.5/SBR binders .......................................... 61
FTIR oxidation ratios of AC-S/SBR binders ............................................. 61
Sofiening and melt temperatures of SBR modified binders ...................... 63
Sofiening point of AC-5 modified with three different SBR polymers ..... 64

Melt temperature of AC-5 modified with three different SBR polymers ..64
Optimum Elvaloy® AM content ................................................................ 73

SHRP high temperature PG of Elvaloy® AM modified binders ............... 74

xiii

Table 4.11 -

Table 4.12 -

Table 4.13 -

Table 4.14 -

Table 4.15 -

Table 4.16 -

Table 4.17 -

Table 4.18 -

Table 4.19 -

Low temperature SHRP PG of Elvaloy® AM modified binders ............... 77
FTIR calibration curves of Elvaloy® AM modified binders ..................... 79

FTIR ratios of different batch size Elvaloy® AM modified binders ......... 80

Softening and melt temperatures of Elvaloy® AM modified binders ....... 83
SHRP high temperature PG of AC-lO/CRM binders ................................ 90
Melt temperature and softening point of AC-lO/CRM binders ................. 94
Physical properties of EVA modifiers ....................................................... 95
High temperature SHRP PG of AC-S/EVA blends ................................. 100
Melt temperature and softening point of AC-S/EVA blends ................... 104

xiv

Chapter One

Scope

The performance of asphalt pavements depends on traffic loads, material
properties, environmental factors, and construction practices. Pavement deteriorates over
time due to the increasing number of load repetitions and the cycling through temperature
extremes in summer and winter. This deterioration manifests itself in a number of ways.
Common distresses include rutting, fatigue cracking, thermal cracking, reflective
cracking, aging, raveling, and stripping 1'3. Conventional asphalts may resist one type of
distress but are susceptible to others. For example, hard asphalts are resistant to rutting
but are susceptible to thermal and fatigue cracking. Soft asphalts, on the other hand, are
resistant to thermal and fatigue cracking but suffer from serious rutting problems. It is
therefore desirable to modify the binder to extend the temperature range of acceptable
performance. The most common method is the addition of polymer modifiers to the
asphalt. Much work has been done to show that polymer modification can improve
pavement performance 4’6. The majority of the studies that have been done have been
rather limited in either modifiers or properties tested. There is a great need for a
systematic study of polymer modified asphalts.

Michigan State University, in conjunction with the Michigan Department of
Transportation, is currently conducting a comprehensive study of modified asphalts. The
study is divided into three sections: the fundamental physical, chemical, and
thermodynamic properties of asphalts, modifiers, and blends; the basic morphology and

microstructure of modified asphalt-aggregate mixtures; and the engineering and structural

2

properties of modified asphalt-aggregate mixtures. The focus of this thesis is the
physical, chemical, and thermodynamic properties of asphalt binders.

The main objective of this study is to evaluate the use of polymer modified
asphalts as a method of improving pavement performance. To do this, the interaction
between polymer and asphalt was characterized. Possible theoretical models were
investigated, as well as the relationship between the chemical and physical properties of
the neat binders.

A second objective of this research was the development of a comprehensive
database of physical and chemical properties of neat and modified binders. This database
leads to many secondary deliverables of the study, including: determination of optimum
polymer content for each modifier; modifications to the fingerprinting protocol developed
earlier in the study 7; and a review of the SHRP and MDOT binder specifications with
suggested revision for use with polymer modifications.

The four most common asphalt grades used in Michigan roads (AC-2.5, AC-5,
AC-IO, and A020) were selected for this study. They are viscosity graded asphalts
supplied by Amoco. They are graded from low viscosity (AC-2.5) to high viscosity (AC-
20). The polymers studied were styrene-butadiene rubber (SBR) supplied in latex form
by Ultrapave, Elvaloy® AM (EAM) supplied in pellet form by DuPont, crumb rubber
particles supplied by Rouse Rubber, and ethylene vinyl acetate (EVA) supplied in powder
form by Exxon.

The following experiments were conducted to determine binder properties:

0 Dynamic Mechanical Analysis

0 Rotational Viscometry

3

0 High Performance Gel Permeation Chromatography
o Fourier Transform Infrared Spectroscopy

- Thermal Mechanical Analysis

0 Differential Scanning Calorimetry

0 and, Bending Beam Rheometry

Chapter Two contains relevant background information. It includes a discussion
of asphalt behavior and fatigue methods, a discussion of polymer modification research
found in the recent literature, and a summary of the work conducted earlier in the
MSU/MDOT project by Jeffrey Shull which was concerned with styrene-butadiene-
styrene (SBS) and styrene-ethylene-butylene-styrene (SEBS) modification.

Chapter Three contains a detailed description of the experiments used in this
study. It starts with an introduction to the Strategic Highway Research Program (SHRP)
findings, continues with a discussion of the SHRP performance based binder
specifications and test procedures, and finishes with a discussion of additional binder
tests not found in the SHRP specifications.

Chapter Four contains the physical and chemical property data for the modifiers
studied. It includes, but is not limited to, rheological properties, molecular weight
distributions, functional group determination, softening point, and SHRP performance
grade. It also contains the mixing procedures used for each modifier, the determination
of the optimum polymer content for each polymer/asphalt system, and a discussion of the
effect of polymer molecular weight on SBR and EVA modification.

Chapter Five contains the physical and chemical property data for the neat binders

studied. It shows the correlation between molecular weight and various physical

4

properties of asphalt. It also investigates correlations between asphaltene content and
physical properties.

Chapter Six contains an evaluation of the applicability of the SHRP performance
grading system for modified binders. Chapter Seven concludes the thesis with a summary

of the findings and recommendations for future research.

Chapter Two

Introduction and Background

A large majority of roads in the United States are paved with asphalt concrete.
These roads are showing signs of premature stress, especially in states like Michigan
whose climate causes both high and low temperature extremes. Roadways can be
designed using conventional asphalt to perform extremely well in areas with only high
temperature extremes and reasonably well in areas with only low temperature extremes.
Unmodified asphalts do not hold up well in areas with both extremes. There are a
number of ways to modify asphalts: careful selection of crudes and refining processes can
be used to give a better asphalt; asphalt constituent composition can be engineered to
give improvements in certain properties; the asphalt can be aged by heating which results
in better high temperature properties due to oxidation; or a modifier such as polymers or
fibers can be added to improve the properties. The majority of modification being done
involves polymer addition. Michigan State University, in conjunction with the Michigan
Department of Transportation, is currently finishing a three year study of polymer
modified asphalt. The goal of this study is to provide a systematic protocol for selecting
asphalt modifiers and for establishing mixture processing conditions to provide
pavements with longer service life. The focus of the research in this thesis is a greater
understanding of the underlying chemical properties which govern asphalt physical

properties and how modification can effect these properties.

2.1. Asphalt Composition

Asphalt is a complex mixture of organic molecules. An elemental analysis of a
generic asphalt shows the following composition: 84 percent carbon, 10 percent
hydrogen, 1 percent oxygen, and the remainder consisting of several trace elements
including nitrogen, sulfur, vanadium, nickel, and iron 8. The molecular structure is much
more important with regard to properties than the actual elemental components. Asphalt
consists mainly of linear and organic ring structures with an average molecular weight of
500 to 5000 8. Typical elemental analysis and molecular weight data for three different

asphalt grades studied are presented in Table 2.1 8.

Table 2.1 - Elemental analysis of several asphalts

 

4 away—u...— .a mum-mu»- ---—_—.

Acgs A6510 A020

no...» ~‘W‘M ,_. n.»

Carbon, % 85.7 82.3 84.5
Hydrogen, % 10.6 10.6 10.4
Oxygen, % ------ 0.8 1.1

Nitrogen, % 0.54 0.54 0.55
Sulfur, % 5.4 4.7 3.4
Vanadium, PPM 163 220 87

Nickel, PPM 36 56 35

Iron, PPM ------ 16 100
Aromatic C, % 32.5 31.9 32.8
Aromatic H, % 7.24 7.12 8.66

MolecularWeigbt- 570890 3.1.9230. . 8591399--

 

Asphalt components have traditionally been divided into three categories: oils,
resins, and asphaltenes. Asphaltenes, which govern many of the properties of asphalts,
are the heaviest molecules. They have an average molecular weight greater than 2000.
They contain the trace elements such as nickel, iron, and vanadium and are soluble in

carbon tetrachloride. Resins are intermediate molecular weight (800-2000) polar

7

molecules. Resins often contain sulfur and nitrogen and are soluble in light petroleum
naphtha 9. Oils are low molecular weight (25-800) molecules which have a large number
of side chains and few ring structures. They have a carbon/hydrogen ratio of less than 0.6
and are soluble in hexane 9. An typical asphalt sample has an asphaltene/resin/oil ratio of
23/27/50 with harder asphalts having a higher concentration of asphaltenes. Table 2.2

lists the ratios for three different asphalts '0.

Table 2.2 -Asphaltene/Resin/0il ratio of three asphalts

 

Asphaltene. Resin. Oil ,

Average Sample 23 27 50
Mexican Asphalt 18.6 39.7 41.22
(170 pen, AC-20)

Mexican Asphalt 28 37.7 44

-.-Qéfiné.€:49).--..-_.,..

..~._..

This three component classification leads to a general two phase asphalt model 9.
The model consists of an assembly phase (asphaltenes and resins) dispersed in an oil
phase. This can be seen in figure 2.1. The resins serve to stabilize the asphaltenes in the
oil phase.

The assembly phase can vary between 40 and 60 angstroms in diameter. The size
of the assembly phase increases with increasing asphaltene content. This adds structure
to the asphalt and makes the asphalt stiffer. This translates to better high temperature
properties. The asphaltenes and oil combination acts as a colloid, with the polar
resinsstabilizing the colloid. The asphaltenes can exist as discrete particles or in a micelle
structure. Depending on whether the polar groups are oriented inward or outward

determines whether it is an oil-extemal or oil-intemal micelle. The presence of these

8

micelles is affected by temperature, resin content, and any other modifiers present in the

asphalt. These micelles greatly affect engineering properties of the asphalt.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.1 - The two phase asphalt model

Asphalt molecules contain many different organic functional groups. Asphalt is
known to contain carboxylic acids, ketones, phenols, sulfoxides, acid anhydrides,
pyrroles, and quinones 8. These functional groups can react with a reactive polymer

modifier, such as an epoxy. They are also important in age hardening, which is

essentially oxidation of these functional groups which causes increased stiffness.

2.2 - Asphalt Failure Mechanism

9

Asphalt manifests failure in a number of mechanisms. The most common forms
of distress are thermal cracking, block cracking, fatigue cracking, rutting, and raveling or
stripping. Table 2.3 summarizes the most common forms of pavement distress.

Thermal cracking can be caused by three mechanisms. First, fracture can be
caused by the aggregate and binder having different coefficients of thermal expansion.
As the temperature changes, the aggregate and binder expand at different rates causing
cracking. Second, fracture can be caused by the presence of water in existing cracks in
the asphalt. The freeze/thaw cycle found in climates like Michigan cause crack
propagation. Third, fracture can be caused by increased brittleness in cold temperatures.
This is usually a result of oxidative aging. When this brittle pavement is exposed to a
load it is susceptible to fracture initiation.

Block cracking is similar to thermal cracking. It is caused at all temperatures by
embrittlement of the pavements. Again, this embrittlement is usually caused by age
hardening.

Fatigue cracking is the least understood pavement distress mode. This is because
the results of the fatigue tests are dependent on the test methods. It is difficult to
determine the period of time in which damage grows from an initial state to a critical
failure level. This fatigue cracking is the result of repeated cyclic and static loads and the
tensile strains that they produce in the bottom of the asphaltic layers.

Rutting is the major high temperature failure mode. It results from the asphalt
flowing under loads at high temperature. The asphalt suffers permanent plastic
deformation. A harder asphalt can be used to prevent rutting but will be more susceptible

to low temperature distress.

10

Raveling or shipping occurs when the adhesion between aggregate and binder is

poor. The fracture is at the aggregate/binder interface.

Table 2.3 - Asphalt distress modes

 

o
l‘

 

...DistreSS.Mode.. _ ,_ a H . . Cause w _ . V . . . ., +_ V
Thermal Cracking 0 thermal expansion differences between aggregate and binder
0 water freeze/thaw cycle
0 low temperature embrittlement
Block Cracking o embrittlement of asphalt binder
Fatigue Cracking a tensile failure
Rutting o microstructural rearrangement due to asphalt plasticity under
load
__BaYelin§/Stfippin_g__-..__,-_.-. interfesiel-£za.9t2:e99¢t9.19?!binéerhggtsgeteesbesi9r}--.

 

w- .. .- nhl—um -—-. «—

2.3 - Reasons for Polymer Modification

Asphalt is a Viscoelastic material whose properties are a very strong function of
temperature. For optimum performance, asphalt must be soft at low temperatures to
prevent cracking and hard at high temperatures to prevent rutting. This is the exact
opposite of the way most materials behave. One way to improve the properties of asphalt
is to add a modifier, such as polymer.

Goodrich proposed a model of asphalt that consists of a shock absorber and a
spring as shown in Figure 2.2 H. The spring represents the elastic properties and the
shock absorber represents the viscous properties. At high temperatures, the asphalt shows
good viscous flow behavior and behaves mostly like a shock absorber. Rutting occurs
because of the lack of elastic behavior. At low temperatures, the asphalt becomes a brittle
elastic solid with little or no viscous character. In Goodrich's analogy, the spring

becomes overloaded and snaps resulting in low temperature cracking. To reduce the

A - v. ova-.m-v- .3. drama..-» - -

_u. M» ”v

11

rutting and cracking potentials, various modifiers were added to asphalt. The goal is to
increase the range of both the elastic and viscous properties of the asphalt binders.
Convenient measures of these properties are G', the elastic storage modulus, and G", the
viscous loss modulus. The ideal case will have high temperature 0' increasing and low

temperature G" increasing with modification.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.2 - Spring/shock absorber model of asphalt

12

2.4 - Polymers Commonly Used for Modification

The basic criteria for selecting a polymer modifier are performance, ease of
processing, and economics. According to the functions and behaviors of various
modifiers in asphalt, modifiers can be classified into five main types: dispersed
thermoplastics, network thermoplastics, reacting polymers, fibers, and crumb rubber
(CRM) particles. Dispersed thermoplastics behave like asphaltenes and normally require
peptizing agents like resins to stabilize the modified system. Without peptizing agents,
these polymers tend to phase separate from the asphalt binder. Usually it requires a
considerable amount of dispersed thermoplastic material before formation of a
macrostructural network occurs. Network thermoplastics behave somewhat like resins
and will form a network of themselves inside asphalts. Network thermoplastics generally
form a macrostructural network upon addition of 3% (w/w) to 7% (w/w) polymer.
Reacting polymers bond chemically to the asphaltenes and form either asphalt/polymer
networks or larger asphaltene groups. Fibers increase the available wetting surface area
and behave as binder thickeners. Crumb rubber particles behave as aggregates if they are
large (> 100 pm) or as dispersed themioplastics if they are small (< 100 um). Surface
modification of CRM particles, such as sulfonation, can create functional groups to react
with the asphalt. Figure 2.3 is a pictorial representation of four different phase behaviors
seen in modified asphalt systems '3. Table 2.4 lists many polymers that have been studied
as additives. Reactive polymers are the most expensive, followed by network polymers
and dispersed polymers. However, the amount of modifier needed for a given level of

improvment generally evens out the costs.

13

 

 

Asphalt Dispersed in Polymer

 

 

 

   

= asphaltene
Polymer Network in Asphalt Polymer grafted onto Asphalt

 

 

 

Figure 2.3 - Polymer/asphalt phase behavior

One or two of the most promising modifiers of each type were selected for this
study. The polymers used were styrene-butadiene-styrene (SBS), styrene-ethylene-
butylene-styrene (SEBS), styrene—butadiene rubber (SBR), Elvaloy® AM (EAM), ethyl
vinyl acetate (EVA), polyethylene (PE), and crumb rubber particles (CRM). Work on
SBS and SEBS was conducted prior to the author joining the research team and is
summarized in section 2.6. This thesis focuses on SBR, EAM, EVA, and CRM

modification.

14

Table 2.4 - Most common polymers used for modification

 

 

 

 

 

 

 

 

 

 

 

 

Polymer Abbreviation Cost ($llb.) Tg 1°C!
DISPERSED THERMOPLASTICS
Ethylene-vinyl acetate EVA * 23 to 39
Polypropylene PP 0.3 -30 to 20
Atactic Polypropylene APP
Polyethylene PE 0.4 -130 to -15
Low Density Polyethylene LDPE
Polystyrene PS 0.55 80 to 100
Polybutadiene PBDt "' -107 to -83
Hydroxyl Terminated Polybutadiene HTPB
NETWORK COPOLYMERS
Styrene Butadiene Rubber SBR 0.75 -64 to -59
Styrene-Butadiene-Styrene SBS
Styrene-Butadiene SB 0.75 -60 to -90
Styrene-Ethylene-Butylene-Sgyrene SEBS
REACTIVE POLYMERS
Elvaloy® AM EAM 1.50 *

 

* Not measured at this time

2.5 - Recent Results from the Literature

A number of studies have been conducted investigating the effects of polymer
modification on asphalt performance. While the results have varied from study to study,
the general trend shows that polymer modification improves asphalt pavement
performance. Figures 2.4 through 2.8 show the results of the literature survey conducted
by MSU for the Polymers in Bituminous Mixtures, Phase 1 project '2. All of these figures
show the number of studies which reported positive and negative results of modification
on various properties used to measure performance. These figures show that the majority

of the studies have found that polymer modification does enhance pavement performance.

15

 

 

N
\I

N
A
Y'vv—r'v

Good Cases
G a

“

IIIIIILIJIAIAIIIIIJJ

 

 

 

Bad Cases

p
P
F

D
I

P
b

 

12 SBS Reacting
EVA PE PO R Polymer QRM C
Dispersed Network Particle

 

 

lllllLll

 

 

Figure 2.4 - Modifier eflect on fatigue cracking

 

 

Good Cases

 

 

 

 

 

 

*2 gm 9: gos'ggEEEé Beasts? 33% f
isperse e wor 0 ice

 

 

 

Bad Cases
3..

 

 

 

Figure 2.5 - Modifier effect on rutting

16

 

 

 

 

 

 

 

 

 

 

 

30 - (:2: Binder 3
27'- "I '1 Mix

- — Field ;

m 24: .

o 21 - 1

:6; ‘3: 1

15 - j

'8 12- 1 1

8 9: 1

6 r 1

3 ' ‘ 2

m - 2

5 -3: .

s :3: L- L L. I

m 12E sass Reacting :
EVA PE E0 Polymer QRM Q
Dispersed Network Particle

 

 

 

Figure 2.6 - Modifier eflect on thermal cracking

 

 

Good Cases
8

OH
my]
1

 

 

 

Bad Cases
3»

-12 S Reacting 1
m PE P9 RS Polymer QRM Q

Dispersed Network Particle

 

 

 

Figure 2. 7 - Modifier eflect on raveling

17

 

:21 Binder
/'/ ’ /« Mix
I— _ Field

 

U U'U'V‘UT'V‘T'I'U"

 

3.
0' D
_l __I
~12 SBS Reactin
W assess Pol ymerg Lu.

Dispersed Network aPrticle

 

 

 

 

 

 

 

 

Bad Cases Good Cases
0

I'V'Ifr

 

 

 

Figure 2.8 - Modifier eflect on age hardening

For dispersed thermoplastics, modifiers that improved the low temperature
susceptibility of asphalts included ethylene acrylic copolymer '4, hydroxylterminated
polybutadiene (HTPB)'5, and polypropylene wax (PPW)'5. Modifiers that improved the
fatigue cracking resistance included HTPB'S, and PPW'S. When polyethylene (PE) was
used to modify AC-20 asphalt, the modified binder had a higher resistance to permanent
deformation and thermal cracking15 "7.

Reviews of previous studies show that PE is the most promising candidate in the
dispersed thermoplastics family. The glass transition temperature of PE ranges from -l30
to -15°C. Polyethylene, however, is insoluble in asphalts (tends to phase separate in the

asphalt/polymer blend). For example, PE will coagulate and separate from the asphalt

phase in half an hour at 160°C after mixing if no stabilization technique is introduced.

18

The properties of a phase separated asphalt/polyethylene blend are less than those of
straight asphalt. Therefore, phase separation needs to be prevented.

There are two ways to prevent phase separation. One is to reduce the storage
time, such as done by Novophalt. This technique requires a huge mixing truck on the
construction site so that asphalt, aggregate and PE can be mixed just prior to being laid
down. Usually, the smaller the PE particle size, the better the performance of the PE
modified pavement. Another way to prevent PE phase separation is through stabilization
of the asphalt/PE blend.

Many methods can be used for stabilizing an asphalt/PE blend. They are,
structural stabilization by gelling agents, specific polymer/asphalt fimctional reactivity,
specific asphalt selection/modification, partial dissolution by polymer oxidation, and
steric stabilization by a block or graft copolymer.

For structural stabilization by gelling agents, both inorganic and organic agents
have been studied. When inorganic fillers are added to a liquid, a three dimensional
network is formed '6. This has a thickening effect which tends to increase the brittle
nature of the composite. Therefore, when inorganic agents are added to asphalt, the
agents may introduce cracks at low temperatures. Organic gelling agents may give the
blend macro scale stability but micro scale phase separation can still occur.

Through the introduction of certain functional groups to the polymer backbone,
such as ester forming groups, polyethylene can be made partially soluble and stabilized in
liquid bitumen. These methods include phosphonization, chlorosulfonation, maleic
anhydride grafting, and acrylic acid grafting. If an asphalt only contains a small amount

of polar aromatics (<5%). PE is a good modifier and the asphalt/PE blend will be stable.

19

Polyethylene chlorinated to less than 15 weight percent and polyethylene maleated to less
than four weight percent were used to modify asphalts and resulted in improved
properties '8.

Partial dissolution by polymer oxidation is another method that can be used for
asphalt/PE stabilization. The oxidation of polymer makes PE more compatible with
asphalt. Oxidation techniques include hot air, high shear, and catalytic oxidation.
However, the method causes deterioration of the low temperature properties of asphalt.

In steric stabilization, PE coalescence is prevented using a form of steric barrier to
keep the particles apart far enough so that Van der Waals interactions can be overcome by
thermal forces. The steric barrier should be soluble in asphalt and reactive with
polyethylene. Stabilizers studied include styrene-butadiene rubber, Kraton Gl652, and
styrene hydrogenated butadiene-styrene tri-block copolymer.

This study used poly ethylene vinyl acetate in place of polyethylene. EVA results
in performance similar to polyethylene without the need to stabalize the asphalt/polymer
blend.

For network thermoplastics, modifiers that increased binder resistance to rutting
included styrene-butadiene-styrene (SBS)'5"9'22, styrene-ethylene-butylene-styrene
(SEBS)20’2', and styrene-butadiene random copolymer (SBR)5‘6"9‘22. However, they only
improved binder low temperature properties slightly. SBR modified asphalts also had
less fatigue cracks 5"9‘22. However, ethylene-vinyl acetate (EVA) modified AC-20
showed more rutting and fatigue cracking as well as a tendency of stripping 5"5‘20‘22.

SBS/SEBS and SBR are the most promising candidates in the network

thermoplastics family. SBR modifier is usually used in the latex form, a polymer/water

20

mixture. SBR latex is manufactured at temperatures of 100-110°F with a resulting
polymer size of 0.1 micron and an overall solid weight percentage of 31%. Mechanical
agglomeration is used to concentrate the latex and the final commercial product is 70 %
(w/w) solids with a polymer size of half a micron. Typically, 3 to 5 % (w/w) of dry SBR
is added to asphalt. Generally, a homogeneous blend is desired with asphalt as the
continuous phase. In SBR modified asphalt, asphalt remains the continuous phase if the
SBR concentration is below approximately 7 % (w/w). Molecular weight distribution
and average polymer size in the latex are two variables that can be adjusted to increase
compatibility between asphalts and polymers in the network thermoplastic family.

For reacting polymers, epoxy 24 and Elvaloy®AM modified asphalts showed less
rutting, thermal cracking, and temperature susceptibility. Furfural 23‘” modified asphalts
had lower temperature susceptibility, higher resistance to rutting and low temperature
cracking, higher freeze-thaw resistance, and better adhesion, but lower cohesion. Maleic
anhydride (MAH)26 modified asphalts had lower temperature susceptibility, higher
resistance to rutting and low temperature cracking, and better adhesion, but lower
cohesion. For the reacting polymers, previous studies are very limited.

Polymers with epoxide and hydroxyl groups are recommended for the reacting
polymer family. Elvaloy®AM is a modified ethylene copolymer with a reacting epoxide
group developed by DuPont/Chevron for the modification of asphalt cement. Most
polymers, when added to asphalt cement, become dispersed and upon addition of more
polymer, form a network which gives the desired properties. Elvaloy®AM, on the other
hand, chemically reacts with asphalt cement creating a new material. Because of this

reaction, DuPont claims that only 1-2 percent by weight of their material needs to be

21

added to asphalt cement versus 3-5 weight percent of most other polymers to obtain the
same desired properties. Elvaloy®AM contains an epoxy functional group that DuPont
believes reacts with aromatic carboxylic acid functional groups found in the asphaltene

portion of asphalt cement. Figure 2.9 is a chemical schematic of Elvaloy®AM.

 

W\c ———(EmyI-no m—c V\/\

l |
O O
l |
cm. (Int,
CH
/ \
OHg—O

 

 

 

Figure 2.9 - Chemical structure of Elvaloy® AM

The reaction of Elvaloy®AM with asphalt cement was tested by looking at
infrared measurements that showed the epoxy functional groups disappeared after being
reacted with asphalt cement. Too much Elvaloy®AM will result in gelling the asphalt
mixture due to the extensive crosslinking structures caused by the reacted epoxy
functional groups. The amount of polymer needed for asphalts from different sources
also varies.

DuPont tested Elvaloy®AM modified asphalt cements using dynamic mechanical

analysis and found that the material remained elastic at high temperatures which would

22

help prevent rutting. The binder properties remained the same at low temperatures as
those of straight asphalts. In order to improve both high and low temperature properties,
a softer asphalt should be used with Elvaloy®AM.

DuPont and Chevron tested their product using the Chevron creep test, bending
beam test, split tensile measurements, and resilient modulus measurements. All tests
showed the DuPont product to be superior to SBS and SBR. It should be noted that all of
these tests were conducted by the manufacturer of the superior polymer.

Mixing Elvaloy®AM with asphalt cement requires a low horse power stirrer, a
heated tank, and an elevator assembly that will transfer the polymer to the tank. The
polymer must be allowed to react with the asphalt cement for 2-48 hours at 350°F. The
polymer is mixed in a solid form and DuPont claims to have never gelled a tank.
Aging/oxidation was found not to be a problem when processing at these high
temperatures because the tanks are sealed and no air is allowed to circulate in the system.
The manufacturer stated a case where Elvaloy®AM modified asphalt cement was stored
for 100 days at 350°F, and no viscosity change was recorded with only a one hour per day
mixing period.

Elvaloy®AM is handled at temperatures similar to those of other polymer
modified asphalts, but has been found not to be stringy and sticky like SBS and SBR
modified asphalts. No recyclability experiments have been conducted on Elvaloy®AM
modified asphalts. At the present time, Elvaloy®AM costs between $1.20 and $2.00 per
pound.

For crumb rubber modifiers (CRM), both good and bad field results were

27.28

reported. Good field performance was reported for low temperature susceptibility.

23

29'3' included severe aging, lower tensile and shear

The reported bad field performances
strength, raveling, reflective cracking, and stripping.

Though the benefits of crumb rubber modification to asphalts are still being
debated, CRM has to be studied due to environmental regulations. The principle source
of CRM is scrap tire rubber in forms of whole tire, cut tire, shredded tire, or retread
buffing waste. Tire rubber is primarily a composite of a number of blends of natural and
synthetic rubbers and carbon black. There are three processing methods to produce
CRM: granulator, crackmill, and micromill. The granulator process shears apart the scrap
tire rubber and cuts the rubber with revolving steel plates that pass at a close tolerance.
The product of this process is called granulated CRM which has a low surface area and
particle size range of 9.5 mm to 2.0 m (3/8” - No.10 sieve). The crackmill process, the
most common process, tears apart scrap tire rubber and reduces the size of the rubber by
passing the material between rotating corrugated steel drums. The product of this process
is called ground CRM which has a large surface area and particle size range of 4.75 mm
to 425 pm (No.4 - 40 sieve). The micro-mill process further reduces the crumb rubber to
very fine ground particles. The product of this process is called fine ground CRM which
has a large surface area and a particle size range of 425 pm to 75 um (No.40 - 200 sieve).
The average cost of commercial CRM ranges form 20 to 35 cents per kilogram for coarse
and medium crumb and up to 55 cents per kilogram for fine ground crumb.

Among the five types of modifiers, the order of importance in Michigan is
network thermoplastics, reacting polymers, dispersed thermoplastics, fibers, and CRM

particles. The CRM particles could become more important if congressional legislation is

implemented requiring their use.

24

2.6 - Summary of SBS/SEBS Results

Year 1 of the MSU/MDOT study focused on SBS and SEBS as modifiers.32
Laboratory results indicate that there is measurable improvement in pavement
performance for SBS and SEBS modified binders. The rheological properties of the
binders were improved with increasing polymer content up to a certain "optimum"
content, after which improvement was minimal. SBS showed an unexpected reduction in
properties at around 3.5% - 4.5% (w/w) polymer. Figure 2.10 shows the storage and loss
moduli as a function of SBS content for modified AC-S binders. At first glance this
appears to be an experimental artifact. However, a new series of AC5/SBS binders was
mixed and tested resulting in similar phenomena. Figure 2.11 shows the storage and loss
moduli for the second set of binders. It was theorized that at low polymer concentrations,
SBS is slowly destroying the natural structure of the asphalt by absorbing the oils. At
some intermediate polymer content, the asphalt suffers a sharp reduction in properties due
to the disruption of the polar resin-asphaltene network by the SBS. At a slightly higher
polymer content, the SBS forms a network structure and dominates the properties of the

blend. Figure 2.12 is a graphical representation of this rheological theory.

25

 

1.00E+05 -

 

1,005+“ I
‘3
$ .
% 1.00E+03 '
3
I .
I I G"
1.00E+02
100E+01 ' ‘ t I. ’f - . o - . .
0 1 2 3 4 5 6 7 8 9 10

Polymer Content (96 (whim

 

 

 

Figure 2.10 - Loss and storage moduli at 60°C of A C -5/SBS binders (Batch #1 )

 

LIKE-KS:
O
O
‘ o 8 .".."‘
1005mm ° ° 8 ° 8 ° "
g . ‘3’....\ . p.’
‘5' o a". ' ‘9." \o'
2 r‘ O
3 ‘ I
'8 I” "'G
5 1.ooe+oe,’
1 Gt
0 1 2 3 4 5 6 7
PolyrrierComeitPMvdM)

 

 

 

Figure 2.11 - Loss and storage moduli at 60C of A C -5/SBS binders (Batch #2)

26

Rheological Property

 

. u'l’ .‘
_yg‘VJ '-
w»“““"'“ \' Asphalt
9 a.
M“

M I
'.-m'ms~u.l -.. u . w . —

 

 

 

Increaslng Polymer Content —D

Figure 2.12 - Depiction of SBS modified asphalt theory at service temperatures

Chapter Three

Testing Procedures

Penetration and viscosity have traditionally been used to differentiate between
different grades of asphalt. Both methods leave a lot to be desired. There are many
chemical and physical differences between asphalts with the same penetration grade or
viscosity grade. Figure 3.1 shows three asphalts that have the same viscosity grade 33. It
is apparent that these materials do not have the same physical properties and may exhibit
extreme performance differences. State highway agencies recognized the deficiency in
the current system. In 1987, the Strategic Highway Research Program (SHRP) was

formed to develop a new method for grading asphalt cement.

 

 

Consistency
(pen or vis)
A
hard pen
vis
a A
. B
soft vie. C
-15 25 60 135

Temperature. C

Figure 3.1 - Variations of three viscosity-graded asphalts

27

28

3.1 - SHRP Background

The major result of this initiative was the development of performance based
binder specifications. Physical properties were selected that correlate well with pavement
performance. A minimum specification value was determined that differentiated a good
performing asphalt from a poor performing binder. The temperature at which the asphalt
meets this specification determines its performance grade. The value of the specification
remains constant because good pavement performance is expected in both Arctic and
desert pavements. The SHRP tests simulate the three critical stages of the binder's
service life: neat binders, processing and compaction, and long term aging. It is
important to remember that these tests assume that the compaction and construction

methods are sound. If not, even the best binder will perform poorly in an actual road.

3.2 - SHRP Binder Specifications

This section explains how the SHRP tests relate to pavement performance. Table
3.1 shows the SHRP binder specification sheet 33. The PG contains two numbers, for
example 52-40. The first number represents the average 7-day maximum pavement
temperature that this asphalt can sustain, in this case 52°C. The second number
represents the minimum instantaneous temperature that the pavement can withstand, in
this case -40°C. By looking at climate data for the specific construction site a contractor
can choose the appropriate asphalt grade. For example, Juneau, Alaska calls for a PG 46-
26 while Phoenix, Arizona calls for a PG 70-10 33. Clearly different asphalts are needed

for these two locations.

29

Table 3.1 - SHRP performance grading worksheet

 

 

 

 

 

 

 

 

 

 

Performance Grade PG-76 PG-82
-16l -22 I -28 -10 I -16 [-22
7-Day Maxrmum Pavement <76 <82
Temperature °C
Minimum Pavement Temperature °C >-16]>-22l>-28 >-10[>-16I>—22
Original Binder
Flash Point Temp, Min. °C 230
Viscosity, ASTM D 4402
Maximum 3 Pas 135
Test Temp, °C
Dynamic Shear, TP5
G*lsin 5, minimum 1 kPa 76 82
Test Temp @ 10 rad/s, °C

 

 

 

Rolligg Thin Film Oven or Thin Film Oven Residue

 

Mass Loss, maximum %

1.00

 

Dynamic Shear, TP5
G*lsin 8, minimum 2.2 kPa
Test Temp @ 10 rad/s, °C

76

 

82

 

 

Pressure A in

Vessel Residue

 

 

PAV Aging Temperature °C

100

100

 

Dynamic Shear, TP5
G*sin 5, maximum 5000 kPa
Test Temp Q 10 rad/s, °C

34

 

31 [28

40

 

37

 

34

 

Physical Hardening

Re ort

 

Creep Stiffness, TP1
S, Maximum 300 Mpa
m-value, minimum 0.300
Test Temp @ 60 sec, °C

-18

 

 

Direct Tension, TP3
Failure Strain, minimum 1%
Test Tenm @ 1 mm/min, °C

 

-12

 

 

 

 

 

 

 

30

3. 2.1 - Safety Test
The asphalt must have a minimum flash point of 230°C to be graded by the SHRP

system. All binders in this study passed this test easily.

3. 2.2 - Viscosity Test for Pumping
This test ensures that the asphalt binders can be processed using the currently
employed equipment. To pass, the binder must not have a viscosity greater than 3 Pas at

a temperature of 335°C. Any binder which fails this test cannot be graded.

3.2.3 - Permanent Deformation (Rutting) Test

Rutting is the major cause of pavement distress at high temperatures. Rutting
resistance is measured using a Dynamic Shear Rheometer (DSR). The response of
asphalt to load consists of two parts, elastic and viscous. Rutting is the accumulation of
the non-recoverable viscous response. The high temperature stiffness, or rutting
resistance, is characterized by the quantity G*/sin 5. G* is the complex shear modulus
and 5 is the phase angle. To minimize rutting, this value must be at least 1.00 kPa for
neat binders and 2.20 kPa for RTFO aged binders. The specification promotes the use of
stiff, elastic binders to prevent rutting 33. The temperature at which the binder meets both

of these specifications determines the high temperature PG of the material.

3. 2.4 - Excessive Aging Test
Excessive aging susceptibility is determined by measuring the mass loss during

aging. The mass loss can be no greater than one percent of the original mass to meet the

31

specification. This prevents the use of an asphalt that would lose an excessive amount of

volatiles during processing.

3. 2.5 - Fatigue Cracking Test

Fatigue cracking occurs at low to moderate temperatures in pavements that have
been in service for a while. G* and 5 are important in determining fatigue cracking
resistance. Unlike rutting resistance, these values are determined at lower temperatures
on binders which have been aged using RTFO and PAV. They are measured using a
DSR, however the important value in fatigue cracking is G*sin 6. The specification
requires a maximum value of 5000 kPa. Low values of G* and 6 are desirable to prevent

fatigue cracking 33.

3. 2.6 - Thermal Cracking Test

The most important form of low temperature distress is thermal cracking. When
the pavement temperature decreases, the asphalt shrinks and causes tensile stress build-
up. When these stresses exceed the tensile strength of the binder, a crack forms 33. A
Bending Beam Rheometer (BBR) is used to apply a creep load to the aged sample and
measure the creep stiffness at low temperatures. To prevent cracking, the creep stiffness
must be less than 300 MPa. The rate at which the binder stiffness changes with time is
also important and is represented by the m-value. The m-value represents the slope of the
stress-strain curve. A high m-value is desirable because it results in lower stresses and,
consequently, less opportunity for cracking. The m-value must be at least 0.300 after 60

seconds of loading. A third factor in thermal cracking is direct tension strain. This

32

specification is only used if the stiffness is between 300 and 600 MPa and the m-value is
over 0.300. This occurs rarely and therefore the direct tension specification is of less
importance. The temperature at which the binder passes the fatigue cracking and thermal

cracking tests determines the low temperature PG.

3.3 - Detailed Test Procedures
This section offers detailed procedures for all of the SHRP tests as well as all
additional tests used to characterize neat and modified binders. Mixing procedures for

each asphalt/polymer system are included in Chapter 4.

3.3.1 - Flash Point
The flash point test is conducted at the MDOT laboratory. The sample is heated

to 230°C in accordance with the standard test method (AASHTO T 48).

3. 3.2 - Dynamic Shear Rheometry

Dynamic Shear Rheometry (DSR) is used to measure the rheological properties of
aged and unaged binders. The experiment is conducted by sandwiching a thin (1.6 mm)
asphalt sample between two parallel plates. One of the plates oscillates at a user-
deterrnined frequency and the other plate is fixed. The response of the material as the
plate oscillates is measured by a transducer connected to the opposite plate.

The DSR is used to characterize both elastic and viscous behavior by measuring
the phase angle (8) and the complex shear modulus (G*). G* contains an elastic

component (0') and a viscous component (G"). G' is often referred to as the storage

33

modulus and G" is called the loss modulus. The SHRP specification uses the dynamic
shear modulus for the rheological properties. This modulus, defined as G*/sin 6,
incorporates both the phase angle and complex modulus into one number to give a simple
measure of the rheological characteristics. The rheological prOperties are related by the

following equations:

0* = Jana"2 and tan6 = 9—

G'
The values of the rheological properties are highly dependent on temperature. At high
temperatures, asphalt behaves as a completely viscous fluid. In this case, G' would be
small and 6 would be 90°. At low temperatures asphalt behaves like an elastic solid with
a large G', small G" and 6 equal to 0°. Under normal operating temperatures asphalt
behaves as a combination of the two. This behavior is called Viscoelastic. Figure 3.2
shows a plot of G* and 6 for two asphalts with the same G* 33. Even though both
asphalts have the same value of G*, Asphalt 2 is more elastic than Asphalt 1 because its 6
is smaller. Asphalt 2 will recover much more deformation than Asphalt 1. It is important
to consider both G* and 6 when describing asphalt rheology.

G’ is a measure of the elastic response of the binder. At high temperatures, elastic
response is an important factor in rutting resistance. A large value of G’ indicates that the
binder is better at preventing rutting, assuming all other variables are the same. G” is a
measure of the viscous response of the binder. At low temperatures, viscous response is
an important factor in thermal cracking. A large value of G” indicates less resistance to
flow under load which corresponds to lower stiffness. This lower stiffness causes an

increase in the cracking resistance of the binder.

 

Viscous Behavior

A

at both viscous and
v, 1 elastic behavior

GI

V2 /2

‘62
1
b
51 52
Elastic Behavior

 

 

 

 

 

 

Figure 3.2 - Viscoelastic behavior of asphalt

In this study, a Rheometrics RMS—800 apparatus was used for measuring
rheological properties. Experiments on unaged binders were conducted using 25 mm
diameter plates and those on aged binders used 8 mm diameter plates. A gap width of 1.6
mm was used for both aged and unaged binders. Sample handling procedures were
consistent with those used for the SHRP Bohlin instrument in which the sample is poured
hot directly on the plates and allowed to cool to room temperature. Figure 3.3 shows the
sample setup. Temperature sweeps were conducted from 25°C to 80°C with
measurements taken at five degree intervals. The equilibration period between
temperatures was two minutes. A frequency of 10 rad/sec was used in compliance with
the SHRP specification. Strain levels were controlled to ensure that the testing was
conducted in the Viscoelastic range. Tests were conducted multiple times to ensure
reproducibility. Figure 3.4 shows the reproducibility of storage modulus versus

temperature for a triplicate. This reproducibility is excellent.

35

 

Oscillating
Plate
\. :3 . _,. Proper Amount

_ /0f Asphalt

Fixed
Plate

 

 

 

 

Figure 3.3 - Sample setup for DSR

 

 

 
   
 

 

 

 

1.00E+06
_- :SEFpIe1
g 1.00E+05 +Sample2
3 f§mple3
'U
0
E 1.00E+04 --
U!
8
2
U)
1.00E+03 --
1.005+02 I ,
20 30 40 50 60
Temperature (°C)

 

70

 

Figure 3.4 - Reproducibility of rheological data

 

36

3.3.3 - Rotational Viscometry

Rotational Viscometry is conducted to ensure that the binder will be able to be
pumped at the hot mix facility. The viscosity of the sample is determined by measuring
the torque required to keep the spindle suspended in the asphalt turning at a given
rotational velocity. The torque is directly related to the material viscosity at the test
temperature. Figure 3.5 shows a schematic of the test setup.

This study used a Brookfield viscometer for these measurements. Samples were
heated at ~135°C until they flowed well. Ten grams of the asphalt was poured into a
preheated sample container. A No. 27 spindle was suspended in the sample and the
temperature was allowed to equilibrate for 15 minutes. The spindle was rotated at 5 to
100 rpm depending on the viscosity. The rotation rate was set so that the strain was as
close to 10% as possible without being less. This ensured that the viscosity was being
measured in a newtonian region. Each sample was tested at 30°F intervals from 220°F to
370°F. The temperature was allowed to equilibrate for 15 minutes after each temperature

change.

 

sample

  
 

sample ' ' spindle

chamber

\

 

 

 

Figure 3.5 - Rotational viscosity setup

37

3.3.4 - Thin Film Oven Aging (ASTM D 1754-87)

Aging occurs by two main mechanisms, loss of volatiles due to high temperatures
and oxidation of the binder at moderate temperatures and high oxygen exposure. During
processing, both mechanisms are important. The Thin Film Oven Test (TFOT) was used
to simulate aging that occurs during processing. The SHRP program calls for use of the
Rolling Thin Film Oven Test for simulation of processing aging. However, this
equipment was unavailable during this study and TFOT was used as a substitute. Figure
3.6 is a schematic of the TFOT apparatus. It consists of a convection oven with a
suspended, horizontal, rotating shelf suspended from the center. The test calls for a
temperature of 163°C for a period of five hours. The shelf rotates at a rate of 5.5 :t 1.0
RPM.

Samples are prepared by heating the asphalt at 135°C until a good melt is
achieved. Approximately 50 grams of material are poured into the sample containers.
The sample containers are 140 mm deep in diameter and 9.5 mm deep. They are made of
aluminum and have a flat bottom. The pans are allowed to cool to room temperature and
then are weighed to the nearest 0.001 grams. The pans are put into the preheated TFOT
oven and allowed to heat for five hours after the system has regained the equilibrium
temperature. After heating, the pans are removed and allowed to cool to room
temperature. They are again weighed to the nearest 0.001 grams to determine the percent
mass loss. If it is more than one percent the sample does not meet the excessive aging
specification. The material is then left in the pans for further aging in the Pressure Aging

Vessel.

38

 

Motor

 

 

  
   

Thermometer

Sample Tins

 

 

 

 

Figure 3.6 - Thin film oven test apparatus

3.3.5 - Pressure Vessel Aging

Prior to Superpave, long term in-service aging effects were not included in any
specifications. This was a major oversight because the properties of the asphalt during its
service life can be significantly different than the properties of the asphalt during
construction. The Pressure Aging Vessel (PAV) was used to simulate in-service aging.
The PAV exposes TFOT aged binder to high pressure air and elevated temperatures for
20 hours 33.

The PAV consists of the high pressure vessel and a forced draft oven. Oxygen is
supplied as clean, dry, compressed air in a regulated cylinder. The PAV operates at 2070

kPa and 100°C. The oven is able to control the temperature to i0.5°C.

39

The pressure vessel is preheated to the desired test temperature prior to loading.
After the samples are loaded, the temperature is allowed to reach the test temperature
before pressurization. After 20 hours at high pressure, the pressure is slowly reduced
over about 10 minutes to prevent foaming. The samples are removed and degassed in a

163°C oven for 30 minutes. Figure 3.7 is a schematic of the PAV apparatus.

3.3.6 - Bending Beam Rheometer

The Bending Beam Rheometer (BBR) is used as a predictor of low temperature
cracking of asphalt pavements. The SHRP specification calls for the test to be done on
TFOT and PAV aged binders to gauge low temperature performance during service life.
The BBR is a standard three-point bending apparatus submersed in a bath of ethylene
glycol, methanol, and water. Figure 3.8 shows the BBR apparatus. Test temperatures are
in the range of -6°C to -36°C. The BBR is used to measure how much a binder deflects
under a constant load at a constant temperature. This data is converted to a binder
stiffness (a measure of how fast stress builds up for a given strain) and m-value (a
measure of the binders ability to dissipate stresses). High stiffness and low m-value are
indicative of a large low temperature cracking potential.

The asphalt beams have dimensions of 6.25 mm X 125 mm X 12.5 mm. They are
prepared by pouring hot asphalt into metal molds from one end to the other in a
continuous motion. They are allowed to cool to room temperature for an hour, placed in
a fi'eezer for 15 minutes, and then allowed to equilibrate at the test temperature for an

additional hour before testing.

40

 

air

pressure

temperature
probe

/ asphalt

/
.1
z
x

6

pressure vessel sample rack sample pan

   

 

 

 

Figure 3. 7 - Pressure aging vessel and components

 

I Deflection
/ Transducer
Control and (Air Bearing

Data Acqulsmon 1 Load Ce“
Asphalt Beam / / Fluid Bath
. l: \ a 3‘1 W
”is \ i ,

 

 

    

 

 

 

 

Figure 3.8 - Bending beam rheometer apparatus

The sample is tested under load for a period of 240 seconds. Using the time-
temperature superposition principle, the stiffness at 60 seconds correlates to the stifliless
afier two hours at a temperature 10°C lower than the test temperature. Figure 3.9 shows a

graphical representation of the stiffness and m-value measurements. The m-value is

41

simply the slope of the stiffness vs. time curve at 60 seconds on a log-log plot. There is
some debate as to whether the SHRP specification values are correct for use with

modified asphalts. This topic is addressed in detail in Chapter Six.

3.3. 7 - Thermal Mechanical Analysis

Thermal Mechanical Analysis (TMA) was used to measure the softening point
and melt temperature of asphalt binders. The samples were heated at a specified rate
under a probe loaded with one gram. The pressure exerted by this probe is similar to that
exerted by a truck tire on pavement. The change in height of the probe was measured and
used to determine the sofiening point and melt temperature.

TMA tests were conducted using a DuPont Instruments 943 Thermal Mechanical
Analyzer. The samples were placed in a 7 mm diameter by 5 mm deep aluminum holder
at room temperature and packed firmly forming a flat top surface. The sample was then
quenched to a temperature of approximately -100°C and allowed to equilibrate. The
sample was heated at a rate of 5°C per minute under a one gram load and its height
change was measured until a total melt (corresponding to a 1 mm depression of the
sample) was achieved. A typical TMA curve is shown as Figure 3.10. The softening
point was taken as the intersection of the extrapolation of the two straight portions of the
TMA curve. The melt temperature corresponds to the temperature at which the sample

shows a 1 mm depression.

42

 

 

 

6(1)

Log Creep
Stiffness, S(t)

15

 

30

Wei-ass
*eruoursdloc
lonerhtrp

Time, sec

slope - m-velue

/

 

60 120

Log Loading Time. t (sec)

240

 

Figure 3.9 - BBR deflection and m-value

 

 

 

Mel Tenperdue

 

 

 

Taverna

 

 

 

Figure 3.10 - Typical TMA curve showing soflening point and melt temperature

3.3.8 - High Performance Gel Permeation Chromatography

High Performance Gel Permeation Chromatography (HP-GPC) is used to
determine size and molecular weight distributions in an asphalt sample. A sample is
injected into a solvent stream which flows through columns packed with a porous silica
gel. The larger particles encounter fewer pores and are eluted first while the small
particles which encounter many pores are eluted last. The molecular weight of the
material can be calibrated to molecular size to determine molecular weight distributions.
This assumes that the molecules are similar in shape to the standards used for calibration.
The molecular weight range of the equipment is about 50 - 600,000 g/mol.

For this study, a Waters' HP-GPC system was used to analyze modified and
unmodified binders as well as asphaltene fi'actions of unmodified asphalts. The system

was equipped with a refractive index detector to measure molecular size, a photo diode

44

array to measure UV absorption spectra, a dual reciprocating pump, and a manual
injection port. A PC with Millennium software was used to control the apparatus and to
collect and analyze data. The system uses four Waters' Styragel columns that gave an
effective molecular weight range of 50 - 600,000 g/mol. The mobile phase used for this
study was tetrahydrofuran (T HF ) at a flow rate of 1.0 ml/min. The sample concentration

was kept constant at 10 grams per liter and the injection volume was 250 pl.

3.3.9 - Fourier Transform Infrared Spectroscopy

Fourier Transform Infiared Spectroscopy (F TIR) is used to identify chemical
functional groups. All molecules possess natural resonating frequencies due to bond
stretching, bending, or twisting that can be detected by examining the infrared absorbence
spectrum. Wavelengths that are the same as those of the molecule are absorbed and this
absorbence is converted to an infrared spectrum.

FTIR was used to examine aged, unaged, and modified binders. The samples
were prepared by evaporating a THF/binder solution on a potassium bromide pellet
leaving a thin film of material. The solution was mixed at a concentration of 0.5 g of
sample per 10 ml of THF. The pellets were allowed to dry overnight at room temperature
followed by 15 minutes at 135°C in a vacuum oven.

This study used a Perkin Elmer Model 1800 spectrometer. Each sample was
scanned 16 times at a resolution of 4 cm". The system is controlled using a PC with

Perkin Elmer Spectrum software.

45

3.3.10 - Compatibility Index

The compatibility index was first reported by Heithaus in 1960”. It is a measure
of how well the asphaltenes are held in solution by the resins and the oils in the asphalt.
A modified procedure used by Glover, et al. 35 was used in this study. The compatibility
index (15) was determined fiom the amount of heptane required to initiate precipitation of
asphaltenes in an asphalt-toluene solution. Starting with 0.1 gram of asphalt in 1 ml
toluene, the Ic is calculated as

Ic = (volume of toluene) + (volume of heptane)
volume of toluene

The more heptane required to initiate precipitation indicates better compatibility. A large
Ic indicates better compatibility than a small Ic. Because of the qualitative nature of

determining onset of flocculation, these tests were done three times to ensure accuracy.

3.3.11 - Asphalt Fractionation

Asphalt was separated into four fractions (asphaltenes, naphthene aromatics, polar
aromatics and saturates) using ASTM D4124-86 Method B”. Asphalt is dissolved in
heptane (100 ml/ gram asphalt) and stirred with a magnetic stirrer for approximately two
hours. The precipitated asphaltenes are vacuum filtered and air dried for at least 24

hours.

3.3.12 - Asphalt/Polymer Blending
Industry ofien recommends a mixing time based on visual homogeneity of the

asphalt and polymer blend. It was found that this is inadequate for laboratory practice. A

46

mixing procedure for each polymer system was developed based on an improvement in
the high temperature rheological properties of the blends. The experiments used to make
this determination are described in detail in Chapter Four. Table 3.2 lists the mixing time
and temperature for all of the polymers studied.

The actual mixing was done using a Fischer-Scientific Jumbo 115V low shear
mixer equipped with a four blade, 5 cm diameter impeller. The stirring speed was
determined for each polymer according to experiments and manufacturer suggestions.
The asphalt was heated during mixing in an oil bath held at constant temperature. The

temperature of the bath was different for each polymer system.

 

 

Ad'peteble Stand

  
 

Removable sample bullet whemple in oil bath
..__..__‘ ‘

 

 

Temperature Controls

 

To Power Supply -

 

 

 

Figure 3.11 - Mixing apparatus

47

Table 3.2 - Mixing time and temperature for all modifiers

 

 

 

 

 

 

 

Polymer Temperature (”F) Mixing Time
SBS/SEBS 350 2 hours
SBR 325 30 minutes
EVA 350 2 hours
Crumb Rubber 350 30 minutes
Elvaloy® AM 375 2 hours

 

 

 

 

Chapter Four

Experimental Results and Discussion

Four different modifiers were considered in this study. This chapter will detail
and discuss the experimental results for each of the modifiers. The general outline of
each section will be an introduction to the modifier followed by a presentation of the

physical and chemical properties of the binders.

4.1 - Styrene Butadiene Rubber

Styrene-butadiene rubber (SBR) is one of the most promising polymers in the
family of network thermoplastics. An SBR latex was obtained from the Ultrapave
division of the Textile Rubber and Chemical Company. The water-based UP®70 SBR
latex solution contains approximately 70% (w/w) polymer confirmed by
thermogravimetric analysis. The polymer is a high molecular weight, random block
copolymer consisting of 24 mole % styrene and 76 mole % butadiene. The material
maintains a stable latex at room temperature with pH of 10 and has a viscosity of about
1.65 Pa-s. The consistency of the polymer is similar to that of latex paint. Two other
Ultrapave polymers were studied briefly. UP®7576 is a medium molecular weight SBR
latex while UP®7289 is a low molecular weight SBR latex. The effect of the polymer

molecular weight on binder properties was investigated.

48

49

4.1.1 - Mixing Procedure

The SBR polymer was mixed with asphalt binders by using a preparation method
developed in the laboratory. It is believed that this material behaves as dispersed
thermoplastics at low SBR content and then forms an entangled polymer network within
the asphalt as the SBR content is increased. This network structure manifests itself in an
increase in the rheological properties of the modified binder. The mixing procedure for
SBR modification was determined to maximize its rheological properties and to minimize
asphalt degradation. Approximately 150 g of neat asphalt was heated in an oven to 135°C
to obtain a good melt. The asphalt in a beaker was placed in a pre-heated (180°C) oil bath
and SBR was added slowly while stirring at low-shear. It was important to add the latex
slowly to allow the water to evaporate without gelling the SBR. Each asphalt/polymer
blend was mixed for only 30 minutes at 180°C. The stirring speed varied as a function of
the consistency and concentration of the blend. The current industry practice is to flash
evaporate the latex into the hot mix. This may provide sufficient mixing because of the
violent nature of the flash evaporation. The mixing procedure is consistent with that
recommended by Ultrapave. Figure 4.1 shows the storage modulus as a fimction of
polymer content for various mixing times for modified AC-S binders. It is apparent that
the majority of the improvement is seen with only 30 minutes of mixing, especially for
temperatures below 60°C. Figure 4.2 shows the Tan 8 curves for the same binders. A flat
Tan 5 is desired because it represents a consistency in the rheological properties with
temperature changes. The figure shows that at a mixing time of 30 minutes the Tan 6
curve is essentially flat and that further mixing does not significantly effect the flatness of

the curve.

5O

 

 

 

 

 

 

 

 

 

 

 

 

 

120000
100000
a: 80000
3 +0 Min l
g + 30 Min
2 60000 +60 Min:
:- + 9994‘".
40000
3
20000
0
20 30 4o 50 60 70 80
Temperature (°C)
Figure 4.1 - Storage modulus of A C -5/SBR binders with diflerent mixing times
a
7
6
5 _ ° ° -<>-0M’in ‘
«o . ‘ +30 Min
r: 4 .
.3 . ° ‘-I-—60 Mun
3 ' ‘ .— ‘. . /’ , 90M"!
2
1
0 . .
20 3O 40 50 60 70 80
Temperature (°C)

 

 

 

Figure 4.2 - Tan 6 curves of A C -5/SBR binders with diflerent mixing times

51

4.1.2 - Dynamic Shear Rheometry

Dynamic Shear Rheometry (DSR) testing served two purposes. First, it is the
initial test in the SHRP performance grading specifications. Secondly, it was used to
determine the optimum polymer content. The optimmn content is defined as the content
at which further increases in polymer amount do not significantly improve the high

temperature rheological properties as measured by DSR.

4.1.2.1 - Optimum Polymer Content

Table 4.1 summarizes the optimum contents for the SBR modified binders. These
optimum contents were chosen by examining both the modulus curves at 60°C and the
Tan 8 curves. Sixty degrees was chosen because it is a likely maximum pavement
temperature for Michigan roads. Figure 4.3 shows the loss and storage modulus versus
polymer content of AC-5/SBR binders. It shows that the increase in moduli slows after a
content of 3% (w/w). Figure 4.4 shows the tan 8 curves of the AC-S/SBR binders. It is
desirable to have a flat tan 8 vs. temperature curve which indicates constant relative loss
modulus and storage modulus with varying temperature. It can be seen in Figure 4.4 that
no appreciable increase in the "flatness" of the Tan 8 curve is realized with polymer
contents greater than 3% (w/w). The optimum contents for the other asphalt grades were

determined using the same procedure.

52

Table 4.1 - Optimum SBR content

 

-- Asphalt Grade Optimum SBR Content

 

 

 

 

 

 

 

AC-2.5 4% (w/w)
AC-S 3% (w/w)
ACID 3% (w/w)
AC-20 3% (MW)
1.00905
1.00904
is
1:
5 1.00902
(9 A
+G'
1.00901 +6"
1.00900 .
0 1 2 3 4 5 6 7
SBR Content

 

 

Figure 4.3 - Loss and storage moduli of A C -5/SBR binders

53

 

1CD n

d

10.»

Ten Delta (G”Ki')

HHHH

 

 

 

 

 

 

 

 

Temperature (°C)

Figure 4.4 - Tan (3 curves of A C—5/SBR binders

4. 1.2.2 - SHRP High Temperature Performance Grade

The SHRP performance grading system, as discussed in Chapter Three, was used
to characterize SBR modified binders. The high temperature performance grade is
determined by computing the dynamic shear modulus, G*/sin 6. The Performance Grade
is then the temperature at which the dynamic shear modulus of the unaged binder is equal
to 1000 Pa. To complete the SHRP high temperature grading, the RTFO residue must be
analyzed with DSR. The specification calls for a minimum dynamic shear modulus of
2200 Pa. In all cases, the testing of the aged binder did not affect the performance grade
determined by the original binder. Figure 4.5 shows a typical curve of dynamic shear vs.

temperature. The temperature scale has been divided into the gradations specified in the

54

SHRP system. The graph shows that polymer contents of 2% (w/w) to 5% (w/w) result in
an increase of one PG. Polymer contents of greater than 5% (w/w) did not pass the
viscosity specification and were therefore ungradable by the SHRP system. Table 4.2

summarizes the SHRP high temperature performance grades for all four asphalts.

 

 

+052}
.‘0—174
+24
+3’/
+47
'—a—-5’/]

  

11

 

G*ISIN delta (Pa)

. ----- 704

 

 

 

 

 

 

Figure 4.5 - Dynamic shear modulus of A C-5/SBR binders

55

Table 4.2 - High temperature SHRP PG of SBR modified binders

 

SBR Content AC-2.5 AC-5 A9119. AC-20

 

0% (w/w) 52 58 64 70
1% (w/w) 58 58 70 70
2% (w/w) 58 64 70 70
3% (w/w) 58 64 76 70
4% (w/w) 58 64 76 76
5% (w/w) 64 7O 76 82

 

4.1.3 - Rotational Viscometry

The melt viscosity of SBR modified binders was determined using rotational
Viscometry. All binders with polymer contents of 5% (w/w) or less passed the SHRP
specification for melt viscosity. Figure 4.6 shows the melt viscosity curves for AC-
5/SBR binders. Viscosity increases at all temperatures with increasing polymer content
with the exception of 1% (w/w). There is no increase in viscosity seen upon the addition
of 1% (w/w) SBR. This is likely because the polymer is too dispersed to have any affect
on the viscosity. Significant increases are seen between 1% (w/w) and 2% (w/w) and
between 4% (w/w) and 5% (w/w). These two facts suggest that SBR is forming a
localized network structure at about 2% (w/w) SBR and forming a global network at 5%
(w/w) SBR. The formation of the SBR network at 5% (w/w) can bee seen using
fluorescence microscopy. At contents less than 5% (w/w) the SBR can be seen as
discrete particles or micro-networks. At 5% (w/w) and greater, the SBR forms a fibril

macro-network within the asphalt.

56

 

 

 

 

 

1.00905
e
. I 0%
‘5 n 1%
9, 1.00904 A . 2%
E
g ‘ o 3%
3 A . a 4%
z B ‘ ‘ A 5%
g 1.00903 .\ 8 A E
\ l5 ‘ ‘ e 6%
o .
O
\D
1.00902
200 220 240 260 280 300
Temperature ( F)

 

 

 

Figure 4.6 - Melt viscosity of A C -5/SBR binders

4.1.4 - Bending Beam Rheometry

Bending beam rheometry was conducted to determine the low temperature SHRP
PG of SBR modified AC-2.5 and AC-S binders. In all cases, the m-value specification
was the limiting factor in determining the performance grade. Table 4.3 summarizes the
low temperature performance grade for the SBR modified asphalts. All of the samples
tested had a low temperature PG of -28. There was no improvement in performance
grade upon addition of SBR. There was some improvement in the creep stiffness of the
binders with SBR modification. Some of the binders did meet the specification for creep
stiffness at -24°C (PG of -34) but they failed to meet the m-value specification at that
temperature. Figure 4.7 shows the creep stiffness curves for AC-5/SBR at various

temperatures and polymer contents. Figure 4.8 shows the m-value curves for the same

57

binders. It can be seen that there is no trend for creep stiffness except at —24°C. At this
low temperature, the creep stiffness improves significantly and linearly with increasing
SBR content. This improvement did not manifest itself in the performance grade because
the m-value specification was not met at -24°C. There were no trends evident in m-value
at any temperature. SHRP assumed that the m-value would increase with the increase of
polymer content 37. It appears that the m-value is not a firnction of polymer content, at

least for SBR, and is instead determined by the properties of the base asphalt used for

 

 

 

 

 

 

 

modification.
400
350
300 4
E
5 250 +420 .
: +480 J
g 200 N +-24C
a: 150 \
100
50 M
0
0 1 2 3 4 5 6 7 8
SBR Content(%)

 

 

 

Figure 4. 7 - Creep stiffness of A C -5/SBR binders

58

 

 

0.4 ‘W’
0.3 M #4!

g

V_12C

 

m-value

 

 

 

0.2 ’ +-18c
+-24C
0.1 -
0 1 2 3 4 5 6 7

SBR Content (°/o)

 

 

 

Figure 4.8 - M-value of A C -5/SBR binders

Table 4.3 - Low temperature SHRP PG of SBR modified binders

 

__-__SBR Content 69:2._-§____-_6C:5___

 

0% (w/w) ”-28 -28
1% (w/w) -28 -28
2% (w/w) -28 -28
3% (w/w) -28 -28
4% (w/w) -28 -28

5% (w/w) -28 -28

 

59

4. 1.5 - Fourier Transform Infrared Spectroscopy

F ourier-Transform Infrared Spectroscopy (F TIR) was used to characterize various
functional groups in the asphalt and the polymer. FTIR has been used to analyze the
unaged SBR modified asphalt binders. A typical FTIR spectrum for SBR/AC-2.5 blends
is shown in Figure 4.9. The aliphatic C-CH3 peak at wavelength of 1375 cm‘1 in asphalt
and the trans-1,4 contribution of butadiene (C-C=C-C) at 965 cm'1 in SBR were used as
the content indicator. These peaks were chosen because they are unique to the asphalt
and the polymer respectively. The ratio of the two characteristic absorption peaks is
supposed to be linearly proportional to the SBR content in the asphalt. A calibration
curve is constructed by preparing samples that contain known concentrations of SBR
polymer in the asphalt binders and plotting this information against the respective
absolute absorbence band ratios. It is important to note that the calibration curves are
valid only for the given asphalt/polymer blends. A new calibration curve must be
constructed for each new asphalt. The calibration curves for all SBR systems tested are ;

Polymer content (% (w/w)) = (Ratio - 0.2189) / 0.0626 for SBR/AC-2.5 ;

Polymer content (% (w/w)) = (Ratio - 0.0628) / 0.0539 for SBR/AC-S ;
Polymer content (% (w/w)) = (Ratio - 0.0858) / 0.0438 for SBR/AC-lO ;

Polymer content (% (w/w)) = (Ratio - 0.3007) / 0.0463 for SBR/AC-20.

All of the above calibration curves had an R2 value of 0.95 or greater.

60

 

 

 

 

 

 

 

 

100 1
l
80 1
1699
1602 9 7
6
60 4
%T
40 <
1377
20 '
1458
ll . . - a
4000 3000 2000 crrl 1 1500 1000 400

 

Figure 4.9 - Typical FT IR spectrum of A C -2. 5/SBR binders

Aging studies with F TIR used the absorption peak at 1700 cm" corresponding to
carboxyl groups in the asphalt as an indicator of aging. Since the number of C=O bonds
in the asphalt increases during aging due to oxidation of the binder, while the numbers of
aryl C-C and C-CH3 bonds are not affected by oxidation, the change in the ratio of the
C=O peak (1700 cm!) to the C-CH3 peak (1375 cm") due to aging is used as a measure
of aging.

PAV aged AC-2.5 and A05 binders modified with SBR have been investigated
to determine extent of aging. Aging is characterized by the ratio of the 1700 cm'l C=O
peak which indicates oxidation, to the 1375 cm" C-CH3 peak. Aging has also been

characterized by looking at the 1700 cm'I peak to the aryl C-C peak at 1600 cm". The

61

ratios for the aged binders are shown in Table 4.4 and Table 4.5. There does not appear

to be any effect of polymer modification on the oxidation ratios determined by F TIR.

Table 4.4 - FT 1R oxidation ratios of A C -2. 5/SBR binders

 

 

 

 

 

 

AC-2.5 Aged Binders
1700/1600 1700/1375

0% SBR 0.73 0.38

1% SBR 0.72 0.39

3% SBR 0.76 0.35

5% SBR 0.55 0.25

 

Table 4.5 - F TIR oxidation ratios of A C—5/SBR binders

 

 

 

 

 

 

 

AC-S Aged Binders
1700/ 1600 1700/1375

0% SBR 0.76 0.35

1% SBR 0.70 0.40

3% SBR 0.63 0.36

5% SBR 0.58 0.31

7% SBR 0.66 0.30

 

4. I . 6 - Gel Permeation Chromatography

Gel Permeation Chromatography did not prove to be a useful tool in analyzing
SBR modified binders. SBR itself could not be characterized in our system because it
was insoluble in THF due to its latex form in water. GPC was also used to examine the
SBR modified asphalt binders. SBR modified asphalt binders were not characterized
well even though the water in the blends was believed to be evaporated during the
mixing. The most likely reason is that the low molecular weight range of the SBR

overlapped with the high molecular weight range of the asphalt making much of the SBR

62

indistinguishable from the asphalt. It is also possible that the SBR, being insoluble in

THF, was filtered from the solution before injection.

4.1. 7 - T hermomechanical Analysis

Thermomechanical analysis (TMA) was used to characterize both the softening point
and melt temperature of SBR modified binders. These temperatures are important in
predicting rutting potential by determining the temperature at which the asphalt begins to
flow. Table 4.6 summarizes the softening and melt temperatures for the four asphalt
grades modified with SBR. With addition of more than 4 % (w/w), there is an increase of
approximately 4°C in the final softening point for the AC-S/SBR blends with respect to
the unmodified binders. It is important to note that the 0% SBR binders were subjected
to the same mixing condition as the SBR modified binders. For comparison purposes,
Table 4.6 also shows the softening points and melt temperatures of neat (unprocessed)

asphalt.

4. 1.8 - Effect of Polymer Molecular Weight on Modified Binder Properties

Three different molecular weight SBR polymers were studied to determine the
effect of the polymer molecular weight on the modified binder properties. All three
binders were provided by Ultrapave. The polymers in increasing order of molecular
weight are UP® 7289, UP® 7576, and UP® 70. UP® 70 was the polymer used for all of

the SBR modification presented in sections 4.1.1 - 4.1.7. There was significant variation

63

Table 4.6 - Softening and melt temperatures of SBR modified binders

AC-2.5/SBR Blends

 

 

 

 

 

 

Polymer Content (%) Softening Point (°C) Melt Temperature (°C)
neat 22 31
0 24 33
l 26 34
3 27 39
4 25 38
AC-S/SBR Blends
Polymer Content (%) Softening Point (°C) Melt Temperature (°C)
neat 26 33
0 28 38
1 27 36
2 30 42
3 27 39
4 27 42
5 31 46
6 30 46
AC-lO/SBR Blends
Polymer Content (%) Softening Point (°C) Melt Temperature (°C)
neat 31 40
0 32 41
1 35 48
2 36 49
3 35 49
4 36 46
5 37 49
AC-20/SBR Blends
Polymer Content (%) Softening Point (°C) Melt Temperature (°C)
neat 35 47
0 39 48
1 40 48
3 38 50

 

54

in the physical properties of the binders mixed with different molecular weight polymers.
All three SBR modifiers were blended with the asphalt using the mixing procedure
described in Section 4.1.1. AC-5 grade asphalt was used for all of these tests.

The softening point and melt temperature of the SBR modified binders was
greatly affected by SBR molecular weight. Table 4.7 and 4.8 show respectively the
softening and melt temperatures for these binders. The differences in initial softening
point are small (3°) while the difference in final melt temperature is rather large (9-10°C).
This is because the initial softening point depends primarily on the base asphalt while the

SBR network helps support the binder and prevent flow necessary for a total melt.

Table 4. 7 - Softening point of A C -5 modified with three different SBR polymers

 

 

 

 

 

0 % (w/w) 3% (w/w) 5% (w/w)
UP® 70 29 33 33
UP® 7576 29 32 33
UP® 7289 29 30 32

 

 

 

 

 

Table 4.8 - Melt temperature of A C -5 modified with three different SBR polymers

 

 

 

 

 

0 % (w/w) 3% (w/w) 5% (w/w)
UP® 70 37 47 52
UP® 7576 37 42 46
UP® 7289 37 38 42

 

 

 

 

 

The high temperature dynamic shear modulus of AC-S binders modified with 0%
(w/w), 3% (w/w), and 5% (w/w) of each SBR polymer was measured. These results are
presented in Figure 4.10. The figure shows that there is significant differences in the

modulus value as well as in the SHRP performance grade among the samples with the

65

same content but different molecular weight polymer. For example, the temperature at
which the samples meet the SHRP specification for 3% (w/w) polymer ranges from 63°C
for the low MW UP®7289 to 69°C for the high MW UP®70. The 5% (w/w) samples
show a similar range of 68°C to 78°C. There is some overlap in the two groups. The 3%

(w/w) UP®70 has virtually identical modulus characteristics as the 5% (w/w) UP®7289.

 

 

 

 

 

 

1.00905
A +Aceexmuno
0° 1
.E 2
1’
13
9.
in 1.009044»
2
3
1:
o
2
h
N
2
a) “IE-003'-
.9.
E
N
5%
1.00902 : : : 5 1 1 : : 4,
28 34 4o 46 52 58 64 70 7e 82 as
Tameratrlem

 

 

 

Figure 4.10 - Dynamic shear modulus of A C -5 with three different SBR polymers

The same test was run at low temperatures (0°C - 25°C) to determine low
temperature rheology. Figure 4.11 shows the results of the low temperature dynamic
shear modulus. There is very little difference between any of the samples. This indicates

that the base asphalt is controlling the rheological properties at low temperatures.

66

 

 

 

  
 

 

 

 

1.00900 ~~~——~—-m -. . e_ a?
1.00907 ;
i
75 s
c ,
07: 1.00900 +13% SBR
.x ' +316 unzae
9 I +11% UP7510 .
—x— 3% UPTO
1.00905 + 5% umee '1
+ 5% 091570 1
! —+—5°/. UP70 . I
1.005?“ -. 1 1
0 s 10 15 20 25 30 35
Temperature (°C)

 

 

 

Figure 4.11 - Low temperature rheology of A C-5/SBR blends

Melt viscosity of the AC-S/SBR binders is plotted in Figure 4.12. The viscosity is
a strong function of polymer molecular weight. The effect is so significant that the 3%
(w/w) UP®70 and the 5% (w/w) UP®7289 have virtually identical viscosities.

The effect of polymer molecular weight on low temperature properties was
investigated. Figure 4.13 shows that there is little effect of either MW or polymer content
on creep stiffness. Figure 4.14 shows that the m-value deteriorates with increasing

polymer content, however the effect of molecular weight is minimal.

67

 

Ki 1000
O

   

l +316 sen 091200
+31 an unsn

‘ +316 sen uno
+516 one owner l
+016 sen umn
+616 see um

 

100 220 240 200 no 200
Tom peretu re (°F)

 

10000 p". .. , ., . ssh.” , ”H W , _. . , , ,_,,,,, ,,,,,

 

Figure 4.12 - Melt viscosity of A C ~5/SBR binders.

 

 

60

Stiffness (MPa)

 

    

0% SBR h 73% Lime 3% UP7576 3% up7289' 7757070 up7576 5% up7289"

 

 

 

Figure 4.13 - Creep stifiess of AC—5/SBR binders at -12"C.

 

68

 

 

0.40<

 

 

 

 

0.341

 

 

00/0 SBR 3°/o UP70 3% UP7576 3°/o UP7289 5% UP7S76 5% UP7289

 

 

Figure 4.14 - M-value of A C -5/SBR binders at 42°C.

4.2 - Elvaloy® AM

Elvaloy® AM is a modified ethylene copolymer with a reacting epoxide group.
When added to the asphalt, it reacts with the asphalt creating a new material. This
reaction takes place at high temperature causing extensive oxidation of the asphalt binder.
The chemical structure of Elvaloy® AM is proprietary, however a possible structure was

presented in Figure 2.9.

4. 2.1 - Mixing Procedure
Elvaloy® AM was mixed with asphalt using the procedure recommended by the

manufacturer. The three main variables to consider in the mixing procedure are time,

 

69

temperature, and shear rate. Because Elvaloy® AM reacts with asphalt, a temperature of
195°C was required to overcome the reaction activation energy. Shear rate was chosen to
be high shear to be consistent with the other polymer systems studied. Therefore, mixing
time was the only variable to be investigated. The manufacturer suggested a mixing time
of two hours with additional curing at 195°C for 23 or 69 hours. Typical Elvaloy® AM
content is one to three percent by weight. AC-5 with three polymer contents (1% (w/w),
2% (w/w) , and 3% (w/w)) as well as controlled AC-S (0% (w/w)) were evaluated at
different mixing times. All of the samples were mixed according to the following
procedure:

0 About 150 g of the asphalt was heated in an oven until it melted. The temperature of
the oven was set at 135 °C and the asphalt was allowed to heat for 40-60 minutes.

0 After a good melt of the asphalt was obtained, the container was placed in a hot oil
bath which was preheated to 195 °C. The asphalt was stirred slowly (~800 rpm) for
10 minutes before adding the polymer.

0 The desired amount of Elvaloy® AM was slowly added to the asphalt and the stirring
speed was gradually increased to ~1600 rpm. The blend was stirred for an additional
two hours at the same temperature.

0 After two hours of mixing, a part of blend was poured into a separate container which
was stored in a freezer for testing at a later date. The sample was recorded as the ‘2
hours mixing’ and ‘0 hour cure time’ sample.

0 The container of the remaining blend was covered in order to not be exposed to air
and then placed in the hot oil bath (~3 80 °F) for the desired cure time.

0 After 23 hours of curing, some of the sample was removed for later testing. The rest
of the material was allowed to cure for 69 hours.

Figure 4.15 shows the tan 6 curves for AC-S/EAM binders with no additional cure
time. It is evident that addition of Elvaloy® AM improves the tan 6 behavior of the

asphalt. There is a big increase in tan 6 flatness with addition of 2% (w/w) polymer.

70

Figure 4.16 and Figure 4.17 show tan 6 curves for samples cured for 23 hours and 69
hours, respectively. These figures also include the 3% (w/w) Elvaloy® AM binder with
no additional cure time for comparison. It is clear from these two figures that mixing for
only 2 hours enhances the rheological properties of the blend and that additional curing
time is not beneficial. This phenomenon was also seen during the SBS and SBR mixing

studies and is attributed to either thermal degradation of the base asphalt or shearing of

the polymer.

 

12

 

  

10
+0°/0 2 hrs

’—I—1% 2 hrs
3 .+2%2hrs
—x—S%2hrs

Ten delta
0’

 

 

 

 

 

 

 

25 35 45 55 65 75 85

Tem pereture (°C)

 

 

 

Figure 4.15 - Tan 5 curves of A C -5/EAM binders mixed for two hours (no cure)

71

 

 

 

12 ..---. __._ .. ...__..__.___~__. . .-. _ ._ .. .-....._ .___ ._ _. _____ _

 
  
 
    

10 +016 23 hrs'
+196 23 hrs‘
.+2% 23 hrs
8 +31% 23 hrs
—-D—3% 2 hrs ‘

 

 

 

 

 

 

 

 

  
 

 

 

 

 

 

£
8 e
5
...
4
2 % ““k fig” M
0
25 35 45 55 65 75 85
Temperature (°C)
Figure 4.16 - Tan 5 curves of A C -5/EA M binders with 23 hours cure time
12 Ft... .~.. ..-__.___ ....._........... -..-............---...“...a___.._--_..__.-.-..-..-.._........__.- .-.- . .-.. .— _ .. _ - - __ -....- _.
10
+096 69hrs
+196 69 hrs
8 ‘+2% 69 hrs
-D—3% 2 hrs
6
4
2 Mic ‘n TM
0
25 35 45 55 65 75 85

 

 

Figure 4.1 7 - Tan 6 curves of A C -5/EA M binders with 69 hours cure time

 

 

72

4. 2.2 - Mixing Scale- Up

The mixing procedure developed for Elvaloy® AM is only valid for mixing 150
gram batches. Direct scale-up to 300 gram batches gave vastly different, as well as less
desirable, properties. The other polymers studied to date have scaled up perfectly to large
batches. The network formation in these polymers is much less dependent on perfectly
replicating the small batch mixing in the large batch. The reaction of Elvaloy® AM with
the asphalt can be affected by shear rate, asphalt turnover, access to oxygen, and any
number of other factors affected by the mixing process. To get identical properties in the
large batch would require duplicating the three dimensional flows and thermal gradients
produced in the small batch during mixing. This is beyond the scope of this project.
Figure 4.13 illustrates the differences between the two batches. It is a graph of the high
temperature dynamic shear modulus for two large batches and one small batch of 1.5%
(w/w) Elvaloy® AM modified AC-5. Dynamic shear modulus for small batches was
reproducible to within ten percent. Figure 4.13 shows not only the inferior properties of

the large batch but also the irreproducibility.

73

 

 

.3

Am

a

a.

warm - _ - 0 4,1
g 44211114310109);
3 10000 I+2|-E1.R$T2(1fln)|
z +21-{1REI'3CIDQ
0':

r

0

 

 

 

1m+ 4
$3406$$647076Q$

Tarpaauem

 

 

 

 

Figure 4.18 - Dynamic shear modulus of diflerent batch sizes of A C -5/EA M blends

4. 2.3 - Dynamic Shear Rheometry

Dynamic shear rheometry was used to determine the optimum Elvaloy® AM
content for each asphalt. The manufacturer claims that half as much Elvaloy® AM is
needed to get properties similar to those of SBS modified asphalts. Because of this, the
Elvaloy® AM is priced at twice the price of SBS. Our data confirm this claim. Table 4.9
shows the optimum Elvaloy® AM content for the four asphalt grades. The optimum

contents were determined using the same method as outlined in Section 2.2.1.

Table 4. 9- Optimum Elvaloy® AM content

 

.__A§phe_1t§.r2§§ OMimum Elvaloy® AM Contest---
AC-2.5 2.5% (w/w)
AC-5 1.5% (w/w)
AC-10 1% WW)

 

 

74

DSR was also used to determine the SHRP high temperature performance grade. Table
4.10 shows the high temperature performance grade for Elvaloy® AM modified binders.
There is a large increase in SHRP grade due to oxidation as seen by the high PG of the
0% (w/w) Elvaloy® AM samples. These samples have a PG three or four grades higher
than the neat asphalt used for the mixtures. There is little increase in the performance
grade upon addition of Elvaloy® AM. While the SHRP PG for Elvaloy® AM samples is
very high, the performance may be degraded in a real road application because much of

the improvement is from oxidation in the mixing process and not the polymer itself.

Table 4.10 - SHRP high temperature PG of Elvaloy® AM modified binders

 

Asphalt —> AC-2.5 AC-S AC-10 AC-5 AC-5
Cure Time —1 2 hr. 2 hr. 2 hr. 23 hr. 69 hr.

 

0% (w/w) 70 76 76 70 70
1% (w/w) 70 76 82 76 70
2% (w/w) 70 82 82 82 82
3% (w/w) 82 82 82 82 N/A

4% (w/w) N/A 82 N/A N/A N/A

75

4. 2.4 - Rotational Viscometry

The viscosity of Elvaloy® AM modified binders was much higher than SBR
modified binders. This is mainly due to the oxidative hardening that occurs during
processing. Figures 4.19 and 4.20 are the viscosity versus temperature curves for A05
and AC—2.5 blends respectively. Of the AC-S blends, only samples with less than 1.5%
(w/w) Elvaloy® AM passed the SHRP viscosity test. AC-2.5 blends, on the other hand,
pass the specification at up to 3% (w/w) Elvaloy® AM due to the softer base asphalt.
Excessive viscosity increases may limit the amount of Elvaloy® AM and/or the hardness

of the base asphalt that can be used to prepare Elvaloy® AM modified binders.

 

 

 

 

 

 

1.00900
0 0%(MN)
+15%(ww) ‘
I
1m ' 949601041)
6‘
9.
E
3 1.00904
>
i
1MB“
1110902 + . . 1 : . 1 4 4 . : , .
20021020230240250200270200200300310320330340300300
Tenperauaem

 

 

 

Figure 4.19 - Melt viscosity of A C -5/Elvaloy® AM modified binders

76

 

 

 

 

 

 

111915
. +044»

3.
3 1.11804
'71
o
O
.2
>
g 1.115113
2

10902 . . 91

an m 20 a) a) an 3:)
Terraauem

 

 

Figure 4.20 - Melt viscosity of A C -2. 5/Elvaloy® AM modified binders

4. 2.5 - Bending Beam Rheometry

Bending beam rheometry was conducted to determine the low temperature
performance grade of Elvaloy® AM modified binders. In all cases, the m-value was the
limiting factor in determining the SHRP PG. Table 4.11 contains the low temperature PG
for Elvaloy® AM modified AC-S and AC-2.5 samples. The AC-S performance grades
are significantly worse than those of the SBR modified AC-5. This is because of the
severe oxidation which occurs during Elvaloy® AM processing. The AC-2.5
performance grades are the same as the SBR modified binders, however this is because

the coarse temperature scale used to determine the PG.

 

77

Table 4.11 - Low temperature SHRP PG of Elvaloy® AM modified binders

 

 

AC-2.5 AC-S
0% (w/w) -28 -16
1% (w/w) -28 -16
2% (w/w) -28 -16
3% (w/w) -28 -l 6

 

Figure 4.21 shows the m-value and stiffness of AC-2.5 binders modified with
SBR and Elvaloy® AM at -l8°C. This figure clearly indicates that the low temperature
properties of SBR binders are better than Elvaloy® AM binders despite the lack of
difference in their performance grades. There was minimal changes in low temperature
properties with increasing Elvaloy® AM content. AC—5 blends showed decreased
stiffness at -12°C with increasing Elvaloy® AM content but showed no differences when
tested at -6°C. AC-2.5 blends showed no changes for contents of 0% (w/w) to 2% (w/w)
with a 20% reduction in stiffness at 3% (w/w) Elvaloy® AM. The low temperature
properties of asphalt, as measured by BBR, are not improved upon addition of Elvaloy®
AM. They are in fact degraded due to the high level of oxidation which occurs during
processing. It is important to note that these tests do not confirm that Elvaloy® AM
modified pavements will have poor low temperature performance. The SHRP
specification ignores many factors, such as tensile properties, plastic deformation, and
toughness which will likely play an important role in determining the low temperature

performance of PMA pavements. This topic is addressed in detail in Chapter Seven.

78

 

120

100

Stiffness (MPa)

40

20

 

140

 

 

 

    

 

 

 

 

 

 

 

 

 

0.7
. 0.65
, 0.55
1+SBR stiffness
t + EAM stiffness 0'5
1 —o— SBR m-value 0.45
-D- EAM m-value
0.4
k G—*— ‘+ I 0.35
J +
"' 0.3
0.25
- . 0.2
0 0.5 1 1.5 2 2.5 3
Polymer Content (% wlw)

 

Figure 4.21 - BBR data of A C -2. 5/SBR and A C-2. 5/EAM binders at -18°C

4. 2.6 - Fourier Transform Infrared Spectroscopy

Fourier—Transform Infrared Spectroscopy (FTIR) was used to analyze the original
binders as well as the unaged Elvaloy® AM modified binders. The characteristic IR
absorption peak for the asphalt is the aryl C-C peak at 1600 cm". The characteristic IR
absorption peak of Elvaloy® AM is 1740 cm", which corresponds to a C=O bond found
in Elvaloy® AM. A typical FTIR curve for EAM/AC-S blends is shown in Figure 4.22.
The absolute ratio of the 1740 cm'1 absorption band to the 1600 cm‘l band of an
Elvaloy® AM modified asphalt can be used to determine the polymer content in the
binder. The calibration curve is valid only for the given asphalt/polymer blend. Table

4.12 details the calibration curves for AC-2.5, AC-5, and AC-10 binders modified with

Elvaloy® AM. In all cases, the R2 regression coefficient was greater than 0.990.

 

79

Table 4.12 - FT IR calibration curves of Elvaloy® AM modified binders

 

Sample Calibration Curve Equation
AC-2.5/EAM Polymer Content = (Ratio - 0.3041) / 0.1680
AC-S/EAM Polymer Content = (Ratio - 0.3551) / 0.1427
AC-lO/EAM Polymer Content = (Ratio - 0.2743) / 0.1657

 

 

 

100 1

80 ~

1030
60‘ 1733 1602

96T

40~
1377

 

201

 

1460

 

 

ti

4000 3000 2000 dfil 1500 1000 400

 

 

Figure 4.22 - Typical AC -5/Elvaloy® AM FTIR spectrum

IR experiments were conducted to investigate the difference between the small size batch
(150g) and the large size batch (300g). In Table 4.13 below, two important ratios of the
characteristic peaks are shown. One is the ratio of C=O peak (1740 cm") of Elvaloy®
AM to the aryl C-C peak (1600 cm") of asphalt and the other is the ratio of the C=O peak

(1700 cm") from oxidation of asphalt to the aryl C-C peak (1600 cm'l).

 

 

80

Table 4.13 - F TIR ratios of diflerent batch size Elvaloy® AM modified binders

AC-5/EAM 1.5% (w/w) (Small Batch)

 

 

Curing Time 1740/ 1600 1700/ 1600
2 hours 0.5740 0.6010
5 hours 0.5576 0.6970
10 hours 0.5858 0.7676

 

AC-S/EAM 1.5 % (w/w) (Large Batch)

 

 

Curing Time 1740/ 1600 1700/ 1600
1 hour 0.5511 0.6079
2 hours 0.5590 0.6521
5 hours 0.5460 0.5570
10 hours 0.4880 0.6547
15 hours 0.5000 0.7012

 

The small batch ratios behave as expected. The 1740/1600 ratio does not change with
additional heating time as expected because the amount of Elvaloy® AM does not
change. The 1700/1600 ratio increases with increased heating time, which is expected
because oxidation occurs for longer time. The large batch ratios do not behave as
expected. The small batch system, which had good mixing efficiency, showed greater
oxidation peak ratios (0.7676 vs. 0.6547 for 10 hours mixing). This was consistent with
the lower storage and loss modulus of the large batch binders. The large batch might
have had poor mixing efficiency which is related to thermal induction and reaction as
well as thermal oxidative degradation occurring simultaneously during the curing. The
differences in both physical properties and chemical properties of the different batch size

mixes should be examined further.

81

4. 2. 7 - Gel Permeation Chromatography

GPC was used to analyze Elvaloy® AM modified binders. Figure 4.23 shows a
typical chromatogram for Elvaloy® AM modified asphalt. The chromatogram has been
divided into two regions. Region A represents the relatively high molecular weight
portion which was formed, possibly by the reaction of the epoxy group in Elvaloy® AM
with a nucleophilic functional group like carboxylic acid in the asphalt. Since straight
AC-10 has more high molecular weight asphaltenes (A region) and more reactive sites
than straight AC-2.5, the ratio of region A to region B is higher for AC-IO binders. It is
important to note that even the unmodified binders subjected to the Elvaloy® AM mixing
procedure show relatively high portion of region A due to the thermal oxidation. Whether
this high MW material in the modified binders was due to oxidation, reaction, or both
could not be determined directly from the GPC results.

In addition, GPC tests for two different batch sizes of AC-5/E1valoy® AM blends
were conducted. The small batch samples showed a much larger Region A than the large

batch samples. This is consistent with both the rheological properties and the F TIR data.

82

 

80

 

70

% Max Signal

 

 

 

 

15 20 25 30 35 40 45 50
Retention Time (min)

 

 

 

Figure 4.23 - Typical AC—5/Elvaloy® AM GPC curve

4. 2.8 - T hermomechanical Analysis

The softening points of straight and Elvaloy® AM modified AC-2.5, AC-5, and
AC-10 binders were measured using thermal mechanical analysis. Table 4.14 is the
complete softening point and melt temperature data for the Elvaloy® AM modified
binders. The softening point and melt temperature showed considerable improvement
upon addition of polymer. There was an average increase of 16°C in the softening point
of the three asphalt grades due to the processing conditions. The addition of 2-3 % (w/w)
resulted in an additional increase of approximately 10-15°C in the softening point. The
melt temperatures measured also showed similar trends as outlined in Table 4.14. TMA

indicates that while much of the high temperature performance improvement is due to

83

processing, the polymer itself does contribute to that increase somewhat. This is contrary
to the dynamic shear results which suggested the polymer's role was minimal in

performance enhancement.

Table 4.14 - Softening and melt temperatures of Elvaloy® AM modified binders

AC-2.5/EAM Blends
Polymer Content (%) Softening Point (°C) Melt Temperature (°C)

 

(by intercept) (by 1 mm depression)
virgin 22 31
0 38 46
l 38 48
2 42 54
3 49 60

 

AC-S/EAM Blends
Polymer Content (%) Softening Point (°C) Melt Temperature (°C)

 

 

virgin 26 33
0 43 51
l 42 49.5
1.5 45 55
2 48 57.5
3 45 56
4 55 56

 

AC-10/EAM Blends
Polymer Content (%) Softening Point (jgmeelt Temperature (°C)

 

 

virgin 35 47
0 41 52
1 55 58

2 56 69

84

4.3 - Recycled Crumb Rubber

There is a great deal of interest in the feasibility of using ground tire rubber to
modify asphalt. Because of the environmental problem of tire disposal any way to reuse
the tire rubber would be greatly beneficial. The amount of crumb rubber (CRM) used for
modification is much higher than the other polymers, with double-digit weight
percentages common. This indicates that recycling CRM pavements may be problematic
and should be considered in any evaluation of CRM modification.

Crumb rubber modifiers which were supplied by Rouse Rubber Industries were
used in this study. The crumb rubber modifier is a mixture of natural and synthetic
rubbers, carbon black, fillers, and oils. Three different modifiers were obtained:
UltrafineTM GF-80A, GF-80AE, and XP-14. These UltrafineTM crumb rubber modifiers
were graded by their size. For example, UltrafineTM GF—80A is an ambiently produced
tire rubber powder that passes the 80 mesh sieve (174 microns) screen with a resulting
200 mesh (74 micron) mean particle size and a surface area of ~2 m2/g. All three were
investigated to determine the best rubber for use with the Amoco asphalt. The
UltrafineTM Wet method was used for blending as recommended by Rouse. This method
can produce all types of rubber modified asphalt binders including membrane interlayers
and surface treatments but requires additional equipment. The ‘Wet’ process goal is to
achieve a finite degree of plasticization or reclaiming without degassification or pyrolysis
Fine rubber reacts and solvates very rapidly for a stable asphalt rubber modified mixture.
The advantage of the use of the fine rubber in ‘Wet’ process is the ease of dispersion and

imparting the properties of the tire rubber readily into the asphalt matrix.

85

The major concern is the reactivity of the tire rubber itself that can easily occur
with elevated temperature of 350°F on a relatively short time scale (< 6 hours). Heating
rubber in an asphalt for defined periods of time can result in devulcanization or scission
of long chain macromolecules into shorter chains while preserving the main rubber
polymer, or pyrolysis to complete decomposition of the rubber polymer into its basic
components of oil, gas and carbon black. Reactivity is a direct firnction of surface and
particle size. The smaller the particle size, the greater the reactivity, and thus, the less the
capital and operating costs. The GF-80A crumb rubber modifier, which is ultra-fine,
allows the asphalt producer, distributor, or hot mix plant operator to economically handle,
disperse, react and produce a very uniform, stable and flowable rubber modified asphalt
cements in all types of asphaltic processes and application needs. The physical properties
of UltrafineTM GF-80A also allow the inherent properties of the tire rubber to be impacted
into the asphalt binder itself to achieve asphalt modification. However, reactivity or
solvation can also be hampered by high heating temperatures which can lead to
environmental and polymer degradation problems.

The high temperature properties of CRM modified AC-10 and AC-20 binders
were investigated in this study. The aging characteristics and low temperature properties

will be investigated in the future

4.3.1 - Mixing Procedure
AC-lO or AC-20 are generally recommended for CRM blends but, in colder
climates, lower viscosity AC can be used. High temperatures are not required for

blending. For example, the blending temperature of GF-80A is between 325-350°F.

86

Because it can be a continuous process, time required for incorporation is less than one
hour. For the wet applications, the fine tire rubber is fully dispersed into the asphalt
binder prior to addition to any type of hot mix unit or surface treatment application.
Rubber modified asphalt concrete binders have a shelf life in excess of eleven days and
are stable with minimal settling occurring.

A mixing procedure was developed based upon the improvement of rheological
properties for varying temperatures and mixing times. The method of preparation of
crumb rubber modified asphalt binders followed the suggested guidelines from Rouse
Rubber Industries. Since low viscosity asphalt binders are not recommended, AC-10 was
first mixed with CRM UltrafineTM GF-80A. Mixing temperature was determined based
on the investigation of rheological properties as well as homogeneity of binders. For
determination of the optimum mixing temperature, 10% (w/w) CRM asphalt binders were
mixed for one hour at three different temperatures (250°F, 300°F, 350°F). Figures 4.24
and 4.25 show the dynamic shear modulus and tan 6, respectively, as a function of
temperature for the three different mixing temperatures. Higher temperature showed

slightly higher moduli and slightly lower tan 6 due to possible oxidation of the binders.

87

 

 

 

 

 

g
a .
E I
a 1
T 1.00E+03 3
o

1.006102 j

Temperature (C)

 

 

Figure 4.24 - Dynamic shear modulus of 1 0% (w/w) CRM modified AC-I 0 mixed for
one hour at various temperatures

 

 

 

2° ‘+250F
+300F
+350F
g 1.
5
5.
10
5
0 i ‘
30 40 50 60 70 80 90
Temperature(C)

 

 

 

Figure 4.25 - Tan 5 curves of 1 0% (w/w) CRM modified AC -1 0 mixedfor one hour at
various temperatures

88

Additional samples were mixed using several different mixing times (15, 30, 45,
60, and 120 minute) at 300°F and then tested to examine any rheological property
improvements as well as melt viscosity change due to increased mixing times. Figures
4.26 shows the dynamic shear modulus as a function of temperature of these samples. It
is not clear that longer mixing enhances the rheological properties of the blend. There
was also minimal differences in the melt viscosity of all five samples. Based on the
above studies,‘the best mixing time and temperature were determined to be 30 minutes

and 350°F. This is consistent with the recommendations of the supplier.

 

 

 

 

 

 

 

100900 9 - w- — - — — ~
—o—15 min. 3
1.00805 -o-—30 min.
+45 min.
7" —x—60 min.
$ —o-120 min.
.8.
8
5 1.00804
9
0
1.00803
1.00sz
28 34 40 46 52 58 64 70 76 82 68
Temperature (C)

 

 

 

Figure 4.26 - Dynamic shear modulus of 1 0% (w/w) C RM modified AC -1 0 mixed at
3 00°F for various amounts of time

89

4.3.2 - Dynamic Shear Rheometry

DSR was used to determine optimum polymer contents as well as high
temperature performance grade. Unlike previous systems, CRM showed no optimum
polymer content. That is, there was no content above which the rheological properties
improvement was minimal. CRM modification continued to increase properties with
additional modifier up to 20% (w/w). At this content, the asphalt became untestable
using the SHRP specifications due to excessive viscosity.

The AC-lO/CRM system was studied using CRM contents of 0%, 5%, 10%, 15%,
and 20% (w/w). Contents of 20% (w/w) and greater could not be classified using SHRP
because they did not meet the viscosity specification. Figure 4.27 shows the dynamic
shear modulus of AC-lO/CRM blends after spending an additional hour at elevated
temperature. It clearly shows that there is no optimum content as defined earlier.
Likewise, the tan 6 curves for the same samples displayed continued improvement with

additional CRM as shown in Figure 4.28.

 

1305.03 _ . . .. .... .. .. . . .. . . 53.2.3.2“... .

1.00E005

1.005004

G'Ialn deita (Pa)

 

1.00Eo03

 

1.00E902

 

Tern perature (C)

 

 

 

Figure 4.2 7 - Dynamic shear modulus of A C-I O/CRM binders

90

 

 

 

 

70

40

Tan delta

30

20

10

 

 

 

 

30 40 50 60 70 80 90
Temperature (C)

 

 

 

Figure 4.28 - Tan 5 curves of A C -I 0/C RM binders

The dynamic shear curves were also used to assign high temperature performance
grades to the AC-lO/CRM binders. Table 4.15 details the high temperature PG of each of
the samples. The PG continues to increase with increasing CRM content. The PG for
10% (w/w) CRM can also be achieved using 2% (w/w) SBR or simply heating the asphalt
at 380°F for two hours. Each sample was also subjected to additional time at high
temperature to determine the effects of storage time on the SHRP PG. There was

minimal effect of storage for times up to two hours.

Table 4.15 - SHRP high temperature PG of A C -1 O/CRM binders

 

_- CRM mount---30.911912111361013.____-1-1,.h_r_.§.t.}§9f’£-+ 2.01.43.22fo

 

0 % 64 64 64
5 % 70 70 70
10 % 70 76 76
15 % 82 82 82

20 % 88 88 88

 

91

4. 3.3 - Rotational Viscometry

CRM modified binders were tested for processibility using rotational Viscometry.
The viscosity of the modified binders increased with increases in CRM content. For
example, at 135°C the viscosity of AC-10 increased from 0.32 Pa-s for the 0% (w/w)
CRM to 2.42 Pa-s for the 15% (w/w) CRM. Contents greater than 15% (w/w) were too
viscous to be tested. All binders with 15% (w/w) or less CRM passed the SHRP
specification for viscosity. Figure 4.29 shows the viscosity versus temperature curves for
AC-lO/CRM binders. The dashed lines indicate the SHRP specification temperature and
viscosity value. It is important to note that the plot is linear for all contents on a log

scale.

 

 

 

1.00905 5 ~~ ~~—. -- ----—*-----
l
: i 00% 80A No Oven Time
: I 5% 80A No Oven Time ;
A 9 1 A10% 80A No Oven Time
§ 10080“ ‘ : 115% 80A NoOvenTime
E ' I
'71 _____ I _________________ l. .............
8 I e
e . ‘ .
I I 0
1.00903 1 A
e r
r l A
: e I
I
I
100902
200 220 240 260 280 300 320

Temperature (F)

Figure 4.29 - Viscosity of A C -1 0 binders with various C RM contents (w/w)

92

The viscosity change associated with long term high temperature storage was also
investigated. The samples were stored at 350°F for five hours and 24 hours after being
mixed using the normal procedure. The percentage change in viscosity after this aging is
shown in Figure 4.30. These results are inconclusive. The 24 hour aging data shows that
all of the CRM binders increased more rapidly than the control sample. This indicates
that the increase in viscosity is not simply due to oxidation of the asphalt binder and that
the CRM affects the viscosity increase. The increase is possibly due to solvation of the
rubber, crosslinking of the rubber, or some other phenomenon. This should be studied

further.

 

180% , -.. .... 2- - .. -... .. .... . .. .. . .. .. . . .. .. . .. ... .. .. . - .. .. . - . .. .. . .. s... ..-.-.2._._~__2....__ .._.. .. .. .. . . . ...,-.. -m.mm-w.~.fik. . . .. -.w-..__..._._..._-..._ .._.-..,|

(:15 Hours

160% .24 Hours

140%

120%

100%

80%

Pereentlnereeee

60%

40%

20%

 

 

 

 

 

 

 

 

 

 

 

0%

 

 

5% 10%
Weight Percent CRI

 

 

 

Figure 4.30 - Percent increase in viscosity of A C -1 0/CRM binders after high
temperature storage

93

4. 3.4 - T hermomechanical Analysis

The softening point and melt temperature of AC-lO/CRM binders was
measured using thermal mechanical analysis. Table 4.16 contains the TMA data of AC-
lO/CRM mixtures that have been aged for up to 24 hours at 350°F. Unlike other
polymers, the softening point increase does not level off with increasing polymer content.
This is consistent with the dynamic shear rheometry results. There is a sharp increase in
melting temperature of the 0% (w/w) CRM sample afier 24 hours aging which indicates
thermal oxidation. This effect becomes less pronounced with increasing CRM content.
At a CRM content of 20% (w/w), there is only a four degree increase in the melt
temperature after 24 hours compared with a 10 degree increase in the control. This seems

to indicate that the CRM is either preventing of masking the effects of oxidative aging.

4.4 - Ethylene Vinyl Acetate

Ethylene vinyl acetate (EVA) is a member of the dispersed thermoplastics family.
Polyethylene (PE) is the most common dispersed thermoplastic used for modification.
However, PE has a tendency to phase separate in asphalt which causes undesirable
results. EVA is much more compatible with asphalt.

Two different EVA polymers (Polybilt 152 and Polybilt 104) were supplied by
Exxon Chemical and were used to modify AC-S asphalt binders. The manufacturer
claims that EVA modification is used to increase binder sofiening point and decrease
penetration, to maintain base asphalt low temperature behavior, and to provide modified
blends with good storage stability and handling characteristics. Polybilt (PB) 152 and

104 are available as free-flowing translucent pellets and can be processed using

94

Table 4.16 - Melt temperature and softening point of A C -1 O/C RM binders

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0% GF80A Softening Point (°C) Melting Point (°C)
No Oven Time" 29 38
One Hour at 350°F 32 39
Two Hours at 350°F 32 39
Five Hours at 350°F 32 40
24 Hours at 350°F 4O 48
5% GF80A Softeninfloint (°C) Meltifl Point (°C)
No Oven Time“ 28 39
One Hour at 350°F 32 41
Two Hours at 350°F 32 41
Five Hours at 350°F 37 41
24 Hours at 350°F 41 50
10% GF80A Softening Point (°C) Melting Point (°C)
No Oven Time“ 36 46
One Hour at 350°F 38 47
Two Hours at 350°F 37 48
Five Hours at 350°F 39 47
24 Hours at 350°F 40 52
15% GF80A Softening Point (°C) Meltilg Point (°C)
No Oven Time“ 37 50
One Hour at 350°F 38 49
Two Hours at 350°F 40 50
Five Hours at 350°F 42 52
24 Hours at 350°F 43 54
20% GF80A Softening Point (°C) MeltiliPoint (°C)
No Oven Time* 41 55
One Hour at 350°F 40 56
Two Hours at 350°F 43 60
Five Hours at 350°F 45 6O
24 Hours at 350°F 46 59

 

 

 

 

95

conventional equipment. Blending time will depend on the shear rate of the equipment
used and temperature. Recommended blending temperature for Polybilt 152 is 330°F to
350°F. The recommended blend temperature for Polybilt 104 is about 10°F lower. Table
4.17 shows typical properties of the EVA modifiers. The EVA dissolves in several
solvents such as tetrahydrofuran, chloroform, carbon tetrachloride, and toluene. THF was
mainly used for the IR work of EVA modified asphalt binders because of its convenience
and availability in the lab. Typical EVA has 30 or 40 % vinyl acetate on a mole/mole
basis. Polybilt 152 is a higher molecular weight polymer than Polybilt 104 as indicated
by the lower melt flow rate. Polybilt 152 also has a higher melt temperature which

requires a higher blending temperature with the asphalt.

Table 4.1 7 - Physical properties of E VA modifiers

 

 

 

Property* Polybilt 152 Polybilt 104 Test Method
Melt Flow Rate, dg/ifii‘fi'fw 2.5 20 ASTM D-1238
Density, g/cm3 0.942 0.941 ASTM D-792
Softening Point, °F 350 260 ASTM E-28
Hardness, Shore A 90 85 ASTM D-2240
1% Scant Modulus, psi 7000 7100 ASTM D-638
Elongation at Break, % 650 675 ASTM D-638

 

* Property information supplied by Exxon Chemical.

4. 4.1 -Mixing Procedure

Polybilt blends have been successfully made using a variety of blend equipment.
Blending requirements are different for the various Polybilt grades with the higher
molecular weight grades (lower melt flow rate) requiring more time. Increasing mixing

temperature and/or shear rate will reduce the blending time required. The selected

96

temperature and time is based on the production of a homogeneous blend of EVA with
asphalt. Polybilt polymers, like most other polymers, have a density of 0.92-0.95 as
compared to 1.0+ for most asphalt. This results in a tendency for the EVA Polybilt to
float and stay at the surface. It is important that Polybilt polymers must melt into the
asphalt bulk at high enough temperature without agglomerating into a molten mass or
skin which becomes very difficult to mix. Laboratory mixes can be made using high or
low shear mixers. To reduce laboratory blending time in the lab, the low shear conditions
used are higher speed than commercial installations. The key consideration in lab
blending is to achieve turn over in the mix so that polymer pellets are pulled below the
surface. The mixing speed for EVA Polybilt 152 and 104 binders was about 1600 rpm.
Low viscosity grade asphalt binder, such as AC-S, with low molecular weight EVA like
Polybilt 104 required milder mixing conditions than that of AC-lO/ EVA Polybilt 152
system. Phase separation is best checked by turning off the mixer and observing for
incorporated polymer floating to surface. Mixing temperature was determined based on
the investigation of rheological properties as well as homogeneity of binders. For
determination of the optimum mixing temperature and time, 5% (w/w) EVA PB152 and
PBIO4 modified asphalt binders were mixed at three different conditions (300°F/ 1hr,
325°F/2hrs, 350°F/2hrs). In Figure 4.31, the dynamic shear modulus of the three samples
is plotted against temperature. It is not clear that longer mixing at higher temperature
enhances the rheological properties of the blend. The mixing conditions of 350°F and
two hours blending was selected to maintain consistency with the SBS/SEBS mixing
procedure. This allows direct comparison to the SBS/SEBS results. The mixing time and

temperature are within the guidelines suggested by Exxon.

97

 

 

 

 

 

1005+“
7;
9:
E
‘L’
C
0
1.00900
+350 F for2 m
+350 F for 1 hr
+325 F for 2 hrs
1.00502 0.-........-_.~,fi____,_____*_____,_,________ 3 _ ‘ -- - _ ., _ . _, g
22 28 34 40 as 52 se 54 70 7s 32 00

Temperature (C)

 

 

 

Figure 4.31 - Dynamic shear modulus of A C -5/E VA binders with different mix conditions

4. 4.2 - Dynamic Shear Rheometry

DSR was used to determine optimum polymer content as well as high temperature
performance grade. The optimum EVA content was determined for AC-5 grade asphalt
binder. The optimum content was determined using the same method as other polymers.
Figure 4.32 shows the tan6 curves for different polymer kinds (P8152 and PB104) and
contents as a function of temperature. In case of AC-S, adding more than 3-5 weight
percent polymer did not significantly enhance the rheological properties in the range of

service temperatures regardless of EVA molecular weight. Figure 4.33 shows the storage

98

modulus values as a function of polymer content. Unlike the tan 8 curves, the storage
modulus of the binders enhanced almost linearly with increasing EVA contents. This
trend is more clear when G' and G" at 60°C are plotted in a function of polymer content
in Figure 4.34. Based on all of these factors, the optimum content is 5% (w/w). This

optimum is not as obvious as that determined for SBS, SBR, or Elvaloy® AM.

 

 

 
 
 
 
 
 
  
 
 
    

 

 

 

 

50
—-x-0% EVA
+15% EVA P3104.
40 1 +15% EVA P8152,
+23% EVA P3104 .
7 +391. EVA PB152
+594. EVA P8104 l
30 —o—5% EVA PB152
5 q+7% EVA 93104 1
2 +791. EVA P8152
i
"' 20
10
0 - .
30 40 50 60 70 80 90
Temperature ('C)

 

 

 

Figure 4.32 - Tan 6 curves of A C -5/E VA binders

99

 

 

 

1.00E+05

1 .00E+04

1.00E+03

6' (Pa)

1.00E+02

1.00E+01

 

     
 

—x-—0% EVA
+15% EVA P8104
+15% EVA P8152
t+3% EVA P8104
1—0—396 EVA P8152
’+5% EVA P8104

t-o—see EVA P3152 ‘

+796 EVA P8104

+791. EVA P3152 _‘

l

l

 

 

1.00E+00
30

 

35 40

45 50 55 60 65 70

Tern peretu re (C)

 

Figure 4.33 - Storage modulus of A C -5/E VA binders

 

1 .00E+05

1 .00E+04

1 .OOE-t-03

1 .00E+02

G' and G" (Pa) at 60C

1 .00E+01

 

 

 

   
 

 

-e—G'(Pa)l
+G"(Pa)

 

 

2 3 4 5 6 7
Polymer Content (%)

 

Figure 4.34 - Storage and loss modulus at 600C of A C -5/E VA (PB! 04) blends

 

 

100

There was very little difference between the two molecular weight polymers. This is
evident in the dynamic shear modulus plot used to assign a performance grade. Figure
4.35 shows the dynamic shear modulus as a function of temperature. At all polymer
contents, the SHRP PG of the blends was the same for P8104 and P8152. P8152 had
slightly better high temperature properties in each case. However, the ability to mix
P8104 at a lower temperature, although not done in this study, could help preserve the
low temperature properties of the base asphalt by reducing oxidation. Table 4.18 details

the SHRP high temperature PG of the AC-S/EVA samples.

Table 4.18 - High temperature SHRP PG of A C ~5/E VA blends

 

EVAEsntsnttw/W) SHRP PG (PBIQ4_)_...__.SHB£BQJB_B.152)

 

0% 64 64
1.5% 64 64
3% 70 70
5% 76 76

7% 82 82

 

101

 

 

 

 

1.00E+05 - -~~---— fl
-x—O% EVA - 4
+15% EVA P3104 ~

. 1+1.5% EVA P3152

A 1.005...“ ‘ ° ‘ \ +3%EVAPB104 ‘
g \ \. +395 EVA P3152 ‘
I; . ' \ +590 EVA P3104 _
3
3 - . . , -o—5% EVA P3152 1
3 . , 1+790EVAP3104 1
E 1.00E+04 . . +791, EVA P3152
8 0
S
U)
.2
E
a
C
E 1.00E+03

1.00E+02

 

 

 

 

 

 

28 34 40 46 52 58 64 7O 76 82 88
Temperature (C)

 

 

 

Figure 4.35 - Dynamic shear modulus of A C-5/E VA blends

4. 4.3 - Rotational Viscometry

The viscosity of AC-S/EVA P8152 and AC-5/EVA P8104 blends was measured
at various polymer contents and temperatures using a Brookfield viscometer. Figure
4.36 is the collected data for some AC-5/EVA polymer modified binder samples. All
EVA modified AC-5 samples met the SHRP specification (maximum 3 Pas) at 135°C
and exhibited a linear fit on a semi-log plot. The viscosity of the binders was a strong
function of polymer molecular weight. This is expected because polymer viscosity
increases with molecular weight. This effect was most dramatic with larger polymer

contents. Overall, the viscosity of the modified binders was relatively low compared to

102

the other polymers studied. Even a high content of 7% (w/w) easily met the SHRP

 

   

 

 

specification.
1m ,E--- .--. .-.--- .. .___.,._._ __~____-___ -2... __... .-. ____.__.._.._ .--, _ -----. .....E---” ---..- ...___,,,--_. ..
i
i
i
a
1L 1000
0
+311 EVA 33104 ,
+311 EVA P3152
+515 EVA P3104 5
+616 EVA P8152 E
+711. BIA 33104 ‘
+711 EVA P3152 ’
100 . - .
200 220 240 200 200 300 320
Turpentine (F)

 

 

 

Figure 4.36 - Viscosity of A C -5/E VA binders

4. 4.4 - Fourier Transform Infrared Spectroscopy

Fourier-Transform Infrared Spectroscopy (FTIR) has been used to examine AC-
S/EVA. It has been determined that there are several defining wavenumbers of interest
including 965, 1245, 1375, 1600, and 1740 cm". These wavenumbers correspond to the
characteristic functional groups in the blends. Ratios of the absorbence intensities of
these functional groups can yield useful information when fingerprinting different asphalt
grades and binders. The IR characteristic absorption peak for aromatic C-C in asphalt is

at wavelength of 1600 cm". The IR characteristic absorption peaks for EVA are at

103

wavelengths of 1740 and 1240 cm", which are due to the C=O and C-O bond,
respectively. The absolute ratio of the 1740 or 1240 cm" absorption band to the 1600
CMI band of an EVA modified asphalt can be used as fingerprinting characteristic tools
to determine the polymer content in the binder. Figure 4.37 shows the calibration curves
for AC-S/EVA P8104. The calibration curve is valid only for the given asphalt/polymer
blend. The system shows a good linear fit for both the 1740/1600 and 1240/1600 ratios.
P8152 showed similar results with R2 values for the 1740/1600 and the 1240/1600 curves

of 0.998 and 1.00 respectively. FTIR is a useful tool for determining the polymer content

 

 

  
  
 
  

 

 

 

of EVA blends.
2.5 ,_.___-_.-_ -m.__-...._ m..- .. .. ..-._ -.- --..EE .-. .. .. ..-- .- ... .-.... -.. _ _ -..--- .. .......__..___-..-_ .................,.___ ..----
2 y = 02405:: + 0.9459
32 = 0.9532
2 1.5 y=0.2811x+0_224
g 1::2 = 09861
E
'6
g 1
—0— 1740/1600
0.5 +1240/1600
0
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
Polymer Content (%)

 

 

 

Figure 4.38 - FT IR calibration curves for AC—5/E VA (P3104) blends

104

4. 4.5 - Thermal Mechanical Analysis

The softening points of unmodified and EVA modified AC-S binders were
measured using Thermal Mechanical Analysis. Table 4.19 details the melt temperature
and softening point of the AC-S/EVA blends. There is minimal increase with addition of
1.5% (w/w) polymer. At contents of 1.5% (w/w) to 7% (w/w), the increase in the
softening point and melting point is relatively linear with an approximate increase of 5°C
per percent polymer. This indicates that at contents of 1.5% (w/w) and less, the polymer
is too dispersed to affect the softening point. At contents greater than 1.5% (w/w), the
polymer acts as a dispersed thermoplastic with the particles becoming more concentrated
with increasing molecular weight. TMA data indicates that there is not an internal
network forming. If that were the case, a sharp increase in softening point would be seen
at the critical content for network formation. When 7% (w/w) EVA was mixed with AC-
5, there was an increase of more than 25°C in softening point for both EVA polymers.
The increase of the softening point is the largest among the modifiers which were used in
this project. This indicates good rutting resistance. There was minimal effect of polymer

molecular weight on either sofiening point or melting point.

Table 4.19 - Melt temperature and softening point of A C -5/E VA blends.

 

 

 

 

 

EVA Softening Softening Melting Melting
Content Point (°C) Point (°C) Point (°C) Point (°C)
PBIO4 ______ PB152 “P8104 P8152
0% (w/w) 28 28 36 36
1.5% (w/w) 27 3O 37 38
3% (w/w) 32 34 44 46
5% (w/w) 45 44 56 57

7% (w/w) 52 56 63 68

 

Chapter Five

Modeling

The task of modeling the interaction between polymer and asphalt to predict and
explain the properties of the blend is daunting. The difficulty stems from two problems
encountered in the use of asphalt. The first is that the asphalt is black, making visual
inspections of the polymer and asphalt phases difficult. This is partially overcome by the
use of fluorescence and scanning electron microscopy. Additional insights can be gained
by examining the effect of polymer content on rheological properties. These techniques
are used to develop phenomenological models of how the polymer and binder interact.
These models are presented in Section 5.]. The second problem is the complex chemical
makeup of the asphalt. It is not known how the different asphalt chemical properties
affect physical properties in neat asphalt. Without completing this first step, any
predictions of how these chemical properties affect the blend physical properties would
be at best speculation. An attempt at correlation modeling with neat asphalt is presented

in Section 5.2.

5. I - Phenomenological Modeling

Phenomenological models are presented for SBR and crumb rubber that explain
the experimental results and are consistent with the visual evidence gathered by advanced
microscopy. These models are not intended for use as quantitative predictive models,

rather they are qualitative models which present a theory of how the polymer improves

105

106

the blend performance. At best, they are a hypothesis. They should be considered a

starting point for further research, rather than a conclusion of this research.

5.1. I - Styrene-Butadiene Rubber

SBR has been shown to improve the high temperature rheological properties of
asphalt binders. The purpose of this section is to investigate and explain the rheological
behavior of SBR modified binders. The behavior of SBR modified binders appears to be
the result of two independent phenomena: percolation theory and network formation.
This is suggested by the rheological properties and supported visually using fluorescence
microscopy.

The storage and loss moduli of the SBR modified binders was were measured
using DSR. Figure 5.1 shows the important rheological properties of the binders at a
number of temperatures for various SBR contents. A large dynamic shear modulus is
desired for rutting resistance. This can be accomplished by either increasing the total
complex modulus (G*) or increasing the elastic nature of the binder (lowering 8). The
addition of SBR reduces the phase angle. This indicates that the binder is becoming more
elastic. SBR is much more elastic than asphalt at 60°C. The linear nature of the phase
angle reduction suggests that the phase angle can be modeled by a simple parallel model
of the form Sblcnd=vng*653R + (l-vng)*Baspha.L The relationship between complex
modulus and SBR content is much more complicated. There is virtually no increase at a
content of 1% (w/w). At this concentration the SBR acts as a dispersed polymer. The
polymer concentration is small enough that it does not significantly affect the rheological

properties. At contents greater than 3% (w/w), a continuous network forms throughout

107

the binder. The result is a relatively linear increase in the complex modulus with
increasing SBR content. The network acts as a support structure for the asphalt, resisting
deformation. It serves a similar purpose to the reinforcing fibers in a fiber-matrix
composite.

There is a sharp increase in all of the modulus values at 2% (w/w). This is likely
due to a phenomenon called percolation theory. In simple terms, percolation theory states
that there exists a critical content at which a sharp peak is seen in the properties. At this
critical content, the dispersed phase, in this case SBR, changes from dispersed behavior to
network behavior without actually forming a network. The SBR is still in a dispersed
state, however it is present in sufficient quantity that it behaves similar to a network. At
contents higher than the critical content, this balance is disrupted and the system acts like
a dispersed system with the corresponding drop in modulus values.

This behavior has not been reported in the literature for polymer modified binders.
However, similar behavior is seen in other systems. One system of note is a water, oil,
surfactant microemulsion. The microemulsion consists of spherical droplets of water
surrounded by a monomolecular layer of surfactant dispersed in the continuous oil phase.
The microemulsions have a macroscopic conductivity because of the ionic surfactants.
At a critical content, the conductivity increases sharply. The conductivity rise is due to
the fact that charge carriers are able to move along connected paths in the microemulsion.
In the SBR modified binders, the modulus values rise because the SBR has similarly
connected paths that allow for support of the binder. This effect is minimized as
temperature is increased because the asphalt is practically a liquid and there is nothing for

the SBR to support.

108

Laser scanning microscopy (LSM) was used to investigate the distribution of
polymer in the asphalt binder. A Zeiss 10 high performance light microscope operating
in fluorescence mode was used in this study. The microscope uses an argon laser
emitting 488 nm (blue) light. The microscope operates in scanning mode, rastering the
laser back and forth across the surface of the sample. The polymer absorbs the laser light
and re-emits it as orange-red length with a wavelength of >520 nm. A filter is used to
remove the reflected 488 nm light leaving only the fluorescence light. Features as small
as 0.1 pm may be resolved using the confocal LSM.

The progression from dispersed polymer to global network can be visually seen by
using a Laser Scanning Confocal Microscope (LSM). In the LSM, the sample is exposed
to a blue laser. The polymer fluoresces allowing it to be seen against the dark asphalt
background. Figure 5.2 shows the LSM micrographs of 1% (w/w), 3% (w/w), and 5%
(w/w) SBR binders. The micrographs have been contrast reversed so that the polymer
appears dark while the asphalt appears light.

SBR forms a polymer network phase inside the asphalt. At very low polymer
concentrations, the SBR acts as a dispersed polymer and does not significantly affect
properties. At 3% (w/w) SBR, a global network has not formed. This supports the
percolation theory for the sharp increase at 2% (w/w). Also, the SBR has changed shape
from spherical to fibril, which may account for the reduction in modulus values. At
higher concentrations, the global network begins to form. The density of this global
network increases with increasing polymer content. This causes a gradual increase in the
rheological properties consistent with normal network formation phenomena. Figure 5.3

shows a pictorial representation of this phenomenological model.

109

‘—D—-G' (Exp) +G' (Exp) A G'(Exp) +d (exp)

 

 

 

 

 

 

 

 

 

 

 

120000
100000
80000
‘=
E.
3 60000
a
3
2
40000
20000
11
Ah 1‘
o 1
00 10 20 30 40 50
SBR Content
12000
.9
10000
0
.7
8000
§- .0
3 6000 .5
§
5 .4
4000
.3
2000 '2
1
0 .
00 r0 20 30 40 50

SBR Content

 

Pheee Angle

Hodulue (Pa)

 

 

A

 

0.0 1.0 2,0 3 0 4.0 5.0
SBR Content

Figure 5.1 - Rheological properties of A C -5/SBR blends

Pheee Angle

Pheee Angle

110

 

 

 

ZOOHIGB Tl?! Cl447 il43? Fl L089

(a) 1% (w/w)

 

ZOOM-68 TIES CIQZZ Il432 Fl L488

(b) 3% (w/w)

       

a
. __. .k\,_ :
ZOOM-60 Tl?!

 

. a;
C'422

     

'I

3'433 Fl L‘OB

(c) 5% (w/w)

 

Figure 5.2 - Laser Scanning Microscope micrographs of SBR modified binders

 

111

 

- - - - Percolation

------- Network

——-Corrbination
200

 

 

 

 

 

150
3
g
a
:I
1'3
E 100
0 . \
' n
I. ‘
l ‘\
50 ,’ ‘\
I |
l \
I \
I t
0
0 1 2 3 4 5
SBR Content

 

 

 

Figure 5.3 - Pictorial representation of SBR behavior model

5.1.2 - Recycled Crumb Rubber

The rheological properties of recycled crumb rubber at various temperatures are
presented in Figure 5.4. The modulus values are basically linear when plotted against
rubber volume fraction. A rubber density of 60 lb./ft. and an asphalt density of 81 lb./fi.
were used to determine volume fraction. Volume fraction was chosen because it is the

generally accepted method for two-phase mixing rules. Recycled crumb rubber acts as a

112

simple dispersed, two-phase system. The rheological properties follow a simple mixing
rule of the form:

GM 2 GRVR + GA(1'VR)

GM = mixture property

GR = rubber property

G A = asphalt property

VR = volume fraction rubber

It is likely that GA is a function of the rubber volume fraction. This is because the rubber
tends to absorb some of the oils in the asphalt, increasing the stiffness. This accounts for

the deviations from linearity seen in the modulus values, especially at higher

temperatures.

Scanning electron micrographs validate this theory. It is easy to see in Figure 5.5 that the
rubber particles are simply acting as a dispersed particle in the asphalt phase. This lends

credibility to the use of a dispersed two-phase mixing rule.

5.2 - Correlation Based Modeling

A great deal of work has been done in an attempt to relate the fundamental
chemical properties of asphalt to its physical and rheological properties. The majority of
the work has concentrated on correlating molecular size distributions with properties such
as penetration, viscosity temperature susceptibility, and penetration viscosity number.
While there have been some minor statistical successes, there has been no consistent

results.

.—D—G‘ (Exp) +G' (Exp)

113

lb G'(Exp)-‘F-d

(exp)

 

500000
450000
400000
350000
300000
250000
200000

Modulm (9.)

150000

100000

50000 ‘

 

30°C

 

 

 

 

OOOOO

 

O
0

100
CRM Content % (VIV)

200

 

60000

50000

50°C

 

40000

30000

Modulus (Pe)

 

50

100
CRM Contentfit (VIV)

 

 

200

 

6000

5000

 

4000

3000

Modulus (Fe)

2000

 

1000

A

 

 

 

Figure 5.4 - Rheological properties of A C -1 O/C RM blends

100
CRM Content it (VIV)

L a. a

9

been

h in o: m
PheseAngle

K) 5 do b

N is a. b

Phase Angle

Phase Angle

RUBBER

.

45 rm

10E GHEE
l-liuzhig '9 ' ' _ '-.':"51t.y H‘SEIILHJIET

 

Figure 5.5 - Scanning electron micrograph of 20% (w/w) CRM modified AC -1 0

115

The most common method of empirical modeling involves relating physical
properties to the molecular size distribution determined using I-IPGPC 3848. The
molecular size distributions (MSD) are generally divided into arbitrary slices and the
percentage of molecules in certain slices is correlated to different physical properties.
This is a very simplistic view of asphalt chemistry. Some of the papers reported success
in correlating a few properties, but the results tend to be random and are likely a
coincidence.

Price and Burati chose to partition the GPC distribution into equal time tenths. 38
They correlated the GPC distribution of the asphalt binder to properties of the asphalt
pavement. This is one of the few studies that used modified binders. They showed good
correlation for specific gravity, indirect tensile strength, and resilient modulus. They
showed no correlation for penetration, Marshall stability, Marshall flow, tensile strength
at extreme temperatures, or resilient modulus at extreme temperatures. There are a
number of problems with this study. The correlations were only good because they
incorporated a type factor based on the modifier used which limits it's usefulness. The
results are based on a modifications of a single AC-20. As GPC distributions as well as
polymer/asphalt interactions are highly dependent on asphalt source this was a poor
choice. Also, the use of THF as the GPC solvent may not be a good choice for SBR and
other rubbers.

Bishara, et al had a slightly better study with slightly worse results. 39 They
studied 20 different virgin asphalts from various sources. The GPC distributions were
divided into fifths. They found that MSD correlates well to viscosity for asphalts from

the same source, but that it is not applicable to asphalts from different sources. In a later

116

paper, they also looked at Corbett fractions in addition to MSD. 49 They found that by
including the Corbett fractions the correlations were improved. However, they offered no
theoretical basis for including the Corbett fractions.

Garrick looked at the problem backwards. 4' He attempted to predict the MSD
from measured physical properties. He created his model from 23 different asphalts from
six suppliers. He tested his model using 11 different asphalts. He concluded that
although he had limited success, the relationship between MSD and physical properties is
dependent on asphalt rheological type.

Jennings, et al. attempted to correlate MSD with pavement rutting and cracking
potentials. They found some average correlation to rutting however the study was not
without flaws. The major flaw is that binder properties simply do not have as great of an
effect on pavement performance. as the authors think. Another minor flaw is that the low
molecular size cutoff was determined arbitrarily, which could make the results look better
than they are.

The large majority of other papers on this subject had similar flaws. The major
flaw with all of these papers is the arbitrary division of the MSD into a number of pieces.
In Gaussian distributions, there are only two independent variables: average and standard
deviation. A division of a Gaussian distribution into ten partitions still has only two
independent variables. Typical asphalt MSD are a little more complicated than a
Gaussian distribution, however it is unlikely that there are ten independent variables in an
asphalt MSD. An orthoganality analysis on a typical asphalt MSD shows that only three
of the sections are of large importance, one is of medium importance, and the rest are

almost completely described by the first four 38. An attempt was made to correlate MSD

117

to physical properties without using more independent variables than exist in an MSD.
Weight average molecular weight (Mw) and number average molecular weight (Mn) were
correlated to a number of properties. In addition, asphaltene content was correlated to
those properties as well. Some simple combinations of the above independent variables

were also correlated.

5. 2. I - Materials Tested

This study used 19 different asphalts obtained from the Michigan Department of
Transportation (MDOT). The samples are from eight different sources and range in
viscosity from A025 to AC-20. The samples had asphaltene contents ranging between
11.8% and 20.4%. The molecular weight range was 586 to 805 for Mn and 1518 to 2284

for Mw.

5. 2.2 - Experiments

The majority of the tests were conducted by the MDOT Materials and Technology
laboratory. The following properties were measured by MDOT:

o Penetration (4°C)

0 Penetration (25°C)

0 Specific Gravity

0 Flash Point

0 Ductility

- Absolute Viscosity (140°F)

o Kinematic Viscosity (140°F, 225°F, 250°F, 275°F)

118

e TFOT and PAV aging

o Aged Penetration (4°C)

0 Aged Penetration (25°C)

0 Aged Absolute Viscosity (140°F)

All chemical tests and some physical tests were conducted at Michigan State
University. These tests include:

o Fourier Transform Infrared Spectroscopy

0 Gel Permeation Chromatography

o Asphaltene/Resin Compatibility

0 Asphaltene Extraction

0 Softening Point/Melting Point by TMA

o Rheological Properties

A number of standard asphalt indexes used to predict temperature susceptibility
were determined. These include:

o Penetration Viscosity Number (PVN) - lower PVN indicates greater
temperature susceptibility

o Viscosity Temperature Susceptibility (VTS) - larger VTS indicates
greater temperature susceptibility

o Penetration Ratio (PR) - lower PR indicates greater temperature
susceptibility

Two other indexes that are used to predict aging effects were investigated. These are:

0 Aging Index - measure of the change in viscosity and penetration due
to aging. Higher numbers indicate less aging effects.

0 Viscosity Ratio - measure of the change in viscosity due to aging.
Higher numbers indicate greater aging.

119

5.2.3 - One Variable Correlations

The first attempt at correlation was to plot all of the important properties against
Mw, Mn, and asphaltene fraction. Asphaltene/Resin compatibility was not used because
the variation in the compatibility index values is very small, likely due to the subjective
nature of the test. The following pages include the plots of viscosity, penetration, PVN,
VTS, aging index, viscosity ratio, penetration ratio, sofiening point, and melting point
plotted against the three variables. It is very clear that there are no direct correlations

between any of the pairs of variables. The best correlation has an R2 value of only 0.66

5. 2.4 - Multivariable Correlation

It may be possible to correlate physical properties to a combination of chemical
properties. Molecular weight, asphaltene content, polarity, and asphaltene/resin
compatibility all seem to play in important role in determining the physical properties of
asphalt. A simple visual examination of the raw data generated in this section shows
clearly that asphaltene content alone cannot explain the variation in properties with
respect to molecular weight. The compatibility index used in this study is problematic.
Because the tests requires determination of the initiation of flocculation in a dark solution
it is dependent on the experimenter's judgment of when flocculation begins. A new
procedure, utilizing light scattering or some other technique for determining the presence
of solids in the solution, would make the compatibility measurement much more accurate.
An increase in accuracy is required before compatibility can be a useful measurement

tool. A possible measure of polarity is to look at ratio's of polar Corbett fractions to non-

120

polar Corbett fractions. This is a time consuming process which makes rapid
classification of asphalt impossible. While a multivariable correlation may give good
statistical results, the reality is that asphalt is so complex that a few simple measure will
probably not be able to fully describe the wide range of physical and thermodynamic
properties. Performance based binder specifications, however flawed, are infinitely more

practical than chemical based specifications.

._q_. " ."

121

 

 

 

 

 

 

 

 

45 I i
U TMA SO“ Point . . i
I I TMA Melt Temp . '- 3' I I
40 I
I
E . El D I
35 u I a D 0 c1
2 n n
i D D
a I I Cl C)
E 3° .9
.—
i
25 D
u a
20
500 550 600 650 700 750 800 850
In (glmol)
Figure 5.6 - Softening and melt temperature versus Mn.
I I
I I
40 .
' I
A o a
I
8, 35 c: u I :1: D
2 o
i '3 D a
a I I
g 30 a u
...
a TMA Soft Point
25 IITMA Melt Temp 0
n u
20 - 5
1500 1600 1700 1800 1900 2000 2100 2200 2300
Mw (glmol)

Figure 5. 7 - Softening and melt temperature versus Mw.

45 .....SN.WWW_MU..,..... . .. .....M .. . . . ___.. _. .. .. _.___..._

 

..- ~mm—‘q

 

 

90

‘80

‘70

~60

_- 50

‘40

-30

-20

'10

D TMA Soft Point I ‘ I
I I
I TMA Melt Temp . I
40 . I
. I
6 9 I Cl
9., 35 c: I a D D a
g n n
'5 ° 0
a u I D I
§ 30 a n
.—
25 D
a n
20
10.0% 12.0% 14.0% 16.0% 18.0% 20.0% 22.0%
Asphaltene Fraction
Figure 5.8 - Softening and melt temperature versus asphaltene fraction.
300 I
I Pen 25°C a
250 - i :3 Pen 4°C
I E
g; 200 ' ’3
D
9‘.
c I
O 150 ‘ a
. I a
‘5 a
8 100 - '3 2
a I - D a D
' '-
. I
50 ~ 5 '3 n c: E
500 550 600 650 700 750 800 850
Mn (glmol)

Figure 5.9 — Penetration versus Mn.

Penetration (4°C)

Penetratlon (25°C)

Penetration (25°C)

 

123

 

 

300 I 90
I Pen 25°C a . 80
250 , El Pen 4°C
70
g H
200 * ‘ 60 g
l 3.
1 50
I
150 . n l 5
a do:
I
. n n
100 : 2 D t 30 a
a a . D . a 1
I . 20
I I I
50 - a E u f a
' 10
1500 1600 1700 1800 1900 2000 2100 2200 2300
Mw (glmol)
Figure 5. 10 — Penetration versus Mw.
300 Mamifl... ........-..-.. ._._.__...._.._.......-V...a.....,.R.._--__..,...-..._- .-.....--. --..-. . _. ---...N. ... -zv. , ._ 90
- Pen 25°C a 80
250 _ [3 Pen 4°C
70
B
200 0 60
150 . i 50
n
' n 40
a 3 n
100 D 0 3°
5 I ' - a I D
. D
I I l 20
50 '3 B a D E
10
0 0
10.0% 12.0% 14.0% 16.070 18.0% 20.0% 22.0%
Asphaltene Fraction

Figure 5.11 — Penetration versus asphaltene fraction.

POMI'ltlon (4‘6)

 

124

 

 

 

 

 

 

 

 

250000 3000
i I Kinematic Viscosity (140°C) I
- D Viscosity (140°C) . 2500
3 200000
.5 U
-
8 l a
E I 2000
_ 150000 a on
g. D
1 500
1 00000 a D
2 I
15 I . I a 1000
g n a u
c I
“' 50000 I I
x a o ‘3 500
a 5 a
0 v f 0
500 550 600 650 700 750 800 850
In (glmol)
Figure 5. 12 — Viscosity versus Mn.
250000 3000
I Kinematic Viscosity (140°C) '
A D Viscosity (140°C) I 2500
3 200000 a
g I I I
g I D 2000
g. D
8 1 500
.3
I
1 000
§ I I . u
S n n a -
_ . I
5 a a
0 0
1500 1600 1700 1800 1900 2000 21 00 2200 2300
m (gmoi)

Figure 5.13 — Viscosity versus Mw.

W (90“.)

W (PO...)

125

250000 rm“---.-.---_............---.-----“..- . 77-777.77.77..- 7 . -7 7 ..--..--_._.7 ...7 . 77--. 77-7.7777“ .-.”.--w- 72.... ._7.. _ 3000

 

 

 

 

 

 

 

I Kinematic Viscosity (140°C) I
,._ a Viscosi 140°C 2500
3 200000 ty ( ) '
D
g . . .
.. . D 2000 ..
E 150000 a a a i
D
g 1500
I
E g o . 1000
D
- 50000 I' I
x a D a 500
a B
o 1 0
10.0% 12.0% 14.0% 16.0% 180% 20.0% 22.0%
Asphaltene Fractlon
Figure 5. l4 — Viscosity versus asphaltene fraction.
500 550 600 650 700 750 800 850
0 i
i
0.2 g
o PVN (60°C)
PVN 135°C
04 ' ( )
U
I
-0.6 n
g I
I
o. n I g G
-0.8 D I
E l . I
2' a D
D
-1 2 I P
15’ a
I
-14 - --.7-_-..... .. 77 - 7 7. .. _ 7 ..--” ..- __ 7. -7.-......7_W-7 7_.-.-7.7.. .. . . 77 .7-7. 77- .1 7 ..77._-7_....... . 777. - .7 _ _ ,, 7 s-“ m ..-. _
Mn (glrnol)

Figure 5.15 —- PVN versus Mn.

126

 

 

 

 

 

 

 

 

 

 

 

 

1400 1500 1600 1700 1800 1900 2000 2100 2200 2300
o
'02 1 n PVN (60°C)
I PVN (135°C)
04 “—Poly. (PVN (135°C))
0
-0.6
2
>
O.
.03
-1
-1 .2
-1 4 _. .7 .... .. ..- ..............__ . .-- ..7--...7_....7 .7 ..-.._ 7. 7 --H _- .-..-. .7 .-... .. ...-.-.-..-_____........_..... .-- .. ...-....-. - _7_ 7. ..-- _.
MW (glmol)
Figure 5.16 — P VN versus MW.
100% 12.0% 14.0% 16.0% 180% 20.0% 22.0%
0
-02
g n PVN (60°C)
PVN 135°C
04 ' ( )
C]
I
-o.e D
2 I
g I 3 g B
-o.a I .
I l i
D D D E
_ El
1 2 I I
D D D
I
-14 - -... -7 ..... J
Asphaltono Fraction

Figure 5.1 7 — PVN versus asphaltene fraction.

 

127

 

 

 

 

 

 

 

 

 

290 WWW—”F“ M“. I“... .-__..._._-.~.__.-....,.._.. .- v 0.40
I
2.80 n . 0.38
‘6 I I
a
g 2.70 0'36
g 0.34
x 2.60
E n a 0.32
o 2.50 5
§ 0 ' 0.30
V 2.40 n I
O
I '3 ' 0.28
230 i D ' . a I
g ’ I U f, 0.26
D
I
3 2.20 ‘ B I D a 0.24
> a ViS Rat10(60°C) D
2.10 i 7' Pen Ratio (4°C/25C) '3 0.22
2.00 0.20
500 550 600 650 700 750 800 850
In (glmol)
Figure 5.18 — Viscosity and penetration ratios versus Mn.
2.90 ----- ~ ~ -.. w— - - 0.40
DVis Ratio (60°C) ' o 33
23° 0 I Pen Ratio (4°C/25C) ' '
a I I
g 2.70 0.36
I
0.34
“'" 2.60 n
3 n D 0.32
2.50
2 a I a 0.30
" 2.40 D . u '
0.20
g 2 30 . i a ' . ‘3
g ' I ' a D 0.26
D
I
g 2'20 E u o ' 0.24
D
2.10 0 0,22
2.00 0720
1500 1600 1700 1800 1900 2000 2100 2200 2300
Mw (glmol)

Figure 5.19 — Viscosity and penetration ratios versus Mw.

PM Me

Ratio

Viscosity RItlo (IbIoluulltlnomauc)

VTS

128

 

 

 

 

 

 

 

 

2.90 0.40
I
2.80 a . 0.38
civis Ratio (60°C) . .
270 . I Pen Ratio (4°C/25C) 036
0.34
2.60 7
n a 0.32
2.50
0' a 0.30
2.40 g D D
0.28
l ‘ - a D I
2.30 I n
' n 0.26
I
2.20 T D 0 I 0.24
D
2.10 D 022
2.00 . 0.20
10.0% 12.0% 14.0% 16.0% 18.0% 20.0% 22.0%
Asphaltene Fraction
Figure 5.20 - Viscosity and penetration ratios versus asphaltene fraction.
3.6400 5- 0.300
0
3.6200
I 0.250
3.6000 nVTS .
' I I
3.5800 ‘lAging ndex ‘ I I - .- . I
- . . 0.200
3.5600 0 . .
3.5400 a 0.150
3 5200 D
' Do
a 0.100
3.5000 0 D a a
3.4600 g D
07050
3.4600 a
3.4400 9 07000
500 550 600 650 700 750 800 850
Mn (glmol)

Figure 5.21 — VTS and aging index versus Mn.

Rflo

P

Aging Inde

 

 

VTS

VTS

3.6400 0 -. 7------7- .-....W--77..7.7-----77..--77_..77---...-_-7-.---__77 0.300
U
3.6200
I 0.250
3.6000
- I
3.5600 5 I 3 I ‘l I I I.
. l . 0.200
3.5600 In I g
3.5400 0 0.150 E,
D '5
3.5200 n n <
o 0.100
I in Index. :1 D
3.4600 ' Ag 9 a
0.050
3.4600 a
3.4400 . c— 0.000
1500 1600 1700 1800 1900 2000 2100 2200 2300
Mw(glmol)
Figure 5. 22 - VTS and aging index versus Mw.
3.6400 ——— E1 .. — - 0.300
Cl
3.6200
I 0.250
3.6000
I
I I .
3.5600 5. . = a I I I I
. . I 0.200
3.5600 I a I g
iaVTS ..
35400 1IAginglndex D 0150 g
7 a __
3.5200 n D 2'
0 0.100
3.5000 0 0 1:1 ,3
U
3.4600 g
0.050
3.4600 a
3.4400 5% 0.000
10.0% 12.0% 14.0% 16.0% 18.0% 20.0% 22.0%
Asphaltene Content

 

 

129

 

 

 

 

 

 

Figure 5.23 - VTS and aging index versus asphaltene content.

Chapter Six

Applicability of SHRP Performance Grading to Modified Binders

The SHRP performance grading system was created using unmodified asphalt
binders. The binder properties were correlated to pavement performance to determine the
specification values. This section addresses the applicability of those specification values

to polymer modified asphalts and recommends changes in the SHRP protocol where

necessary.

6. I - Flash Point Test
The flash point specification calls for a minimum flash temperature of 230°C.
Because this specification is a safety measure, there is no reason it should not apply to

modified binders.

6.2 - Viscosity Test

The viscosity test calls for a maximum rotational viscosity of 3 Pas at 135°C.
This specification ensures that the asphalt can be pumped and handled at the hot mix
facility. Polymer modified asphalt is processed using conventional paving equipment.
For this reason, the viscosity specification should not be changed for polymer modified

binders.

130

 

 

131

6.3 - Rutting Resistance

Rutting resistance is measured by dynamic shear rheometry on the unaged and
RTFO aged asphalt binder. While aggregate properties and aggregate-binder interfacial
properties are important in preventing rutting, the binder itself plays a significant role.
This is seen by the use of harder asphalts in hot climates to prevent rutting. An AC-O.5
will likely rut in Arizona regardless of the aggregate and mixture properties. Rutting is a
result of repeated loading cycles. Some of the deformation is recovered while the rest is
dissipated as permanent deformation and heat. To prevent rutting, this dissipated work
should be minimized. This work is inversely proportional to the dynamic shear modulus
(G*/sin8). Therefore a larger dynamic shear modulus is desired to prevent rutting.
Modification of binders should increase the rutting resistance of most asphalts. However,
the specification should not be altered because the same specification limits are

applicable to both unmodified an modified binders.

6.4 - Excessive Aging

Excessive aging is measured by determining the mass loss of an asphalt after
aging. The specification calls for a maximum mass loss of one percent. This
specification should be eliminated. It does not adequately represent the aging chemistry.
Mass is lost through volatilization and is gained through oxidation. Unless the ability to
measure these two processes independently was developed, overall mass loss is a

misleading measurement.

6.5 - Fatigue Cracking

132

Fatigue cracking resistance is a function of complex modulus and phase angle.
Unlike rutting resistance, which calls for a stiff binder, fatigue cracking calls for a soft,
elastic binder. The specification requires a maximum G*sinS of 5000 kPa. This value
should not be changed for modified asphalts because the assumptions built in to the

specification are equally valid for modified and unmodified binders.

6.6 - Low Temperature Cracking

Low temperature cracking is measured using a bending beam rheometer. The
specification calls for a minimum m-value of 0.300 and a maximum stiffness of 300
MPa. There is some question as to whether these values are sufficient for specifying low
temperature cracking properties for modified binders.

When the temperature drops, the asphalt concrete shrinks. Tensile stresses
accumulate in the pavement because fiiction against the lower pavement layers inhibits
movement. When these stresses exceed the tensile strength of the material, low
temperature cracks form. The two properties chosen by SHRP to characterize low
temperature performance are m-value and stiffness. Stiffness is a measurement of the
amount of stress that occurs for a given strain. The m-value is a measurement of the
ability of the asphalt to relax tensile stresses through flow mechanisms. To prevent
cracking, the accumulation of stresses needs to be prevented. This can be done two ways.
A low stiffness indicates the amount of stress will be reduced for a given strain. A large
m-value means indicates that the stresses will be dissipated faster. The ideal binder will

have a low stiffness to minimize stress and a high m-value to dissipate the stress quickly.

133

The SHRP stiffness specification of 300 MPa was selected based on previous
studies that correlated thermal cracking with stiffness. There are some problems inherent
in this process, namely the fact that stiffness is not the only factor that controls thermal
cracking. However, this study has shown that m-value, not stiffness, is the limiting
factor. Therefore the stiffness specification value was not scrutinized as closely.

The SHRP specification for the m-value of a minimum of 0.300 was selected
based on the experience of the Expert Task Group and on various asphalts tested during
the SHRP development. There is much debate about what the specification value should
be. The correlation to performance data is not as good for m-value as it is for stiffness.
There is very little correlation between m-value and stiffness, however, which suggests
that both properties need to be controlled.

There are some fundamental assumptions that are made when using these two
values to predict thermal cracking. The first is that the cracks are caused by tensile stress
exceeding tensile strength. This seems reasonable. Accepting this assumption leads to
the conclusion that there are four variables which are involved in thermal cracking. Three
variables are used to characterize the accumulation of tensile stress due to thermal
contraction. Coefficient of thermal expansion (CTE) determines the strain occurring for a
given temperature change. Stiffness determines the stress that occurs as a result of that
strain. M-value determines the ability of the material to relax that stress. The three
variables together determine the tensile stress that occurs. These three variables, with
tensile strength, determine thermal cracking potential. The SHRP specification assumes

that CTE and tensile strength of all asphalts will be approximately equal. This makes m-

 

 

134

value and stiffness the only two parameters that need to be characterized. These
assumptions do not hold for polymer modified binders.

Polymer modification may alter the CTE and tensile strength of the binders. It is
likely that the CTE will remain relatively unchanged while the tensile strength will
increase. SBR has a cubical linear expansion coefficient of 37 X 10'5 ft/°F while a
generic asphalt as a coefficient of 35 X 10’5 fi/"F.SO For a mixture of these two materials,
the CTE would be expected to remain constant. Other modifiers have a CTE of the same
order of magnitude, although they are not as close to that of asphalt as SBR. There may
be some small changes in the CTE of the blend compared to the virgin asphalt, however
the effect should be minor. An increase in tensile strength will allow for greater stress
accumulation before cracking. This means that modified pavements with a larger tensile
strength and a stiffness or m-value slightly off specification will likely perform better
than a normal asphalt that barely meets the specifications. The specification, as written,
only considers direct tension if the stiffness is 300 - 600 MPa and the m-value is greater
than 0.300. The specification should be changed to incorporate tensile strength
measurements. A likely solution would be to first measure the tensile strength. The
tensile strength would then determine the cut-off values for stiffness and m-value. A
separate study should be done to determine the effect of tensile strength on low
temperature pavement performance in order to determine the cut-off values. The
specification, as written, is likely over conservative. That is, a pavement which meets the
specification should perform as specified, and likely will perform better than specified.
More work needs to be done to determine the applicability of the SHRP low temperature

tests for modified binders.

 

 

Chapter Seven

Conclusions

This study show that polymer modified asphalt can be used to extend the
temperature range over which asphalt pavements show good performance. This is
accomplished by using a sofi asphalt to give good low temperature performance and
modifying it with polymer to extend the high temperature performance. Each of the
modifiers studied to date showed some different characteristics which are addressed

below.

7. I - Styrene Butadiene Rubber

Styrene butadiene rubber is one of the most common polymers used for
modification. An SBR content of 3% (w/w) improves the high temperature performance
one grade, while a content of 5% (w/w) improves it by two grades. There is minimal
improvement with contents of less than 3% (w/w). At low contents, SBR acts as a
dispersed polymer while at higher contents a network forms inside the asphalt. This
network lends support to the asphalt at high temperatures improving rutting resistance. In
each case, the low temperature grade remained the same as the base asphalt used for
modification. There was some slight improvement in the low temperature creep stiffness
of the SBR modified binders, however the SHRP performance grading scale is too coarse
to register this improvement. The viscosity of SBR modified binders was sufficiently low
at contents less than 6% (w/w) to be processed using standard equipment. FTIR was an

excellent tool for determining polymer content in SBR modified samples. The sharp peak

135

 

136

at 965 cm‘1 in the SBR is a good indicator of polymer in modified binders. GPC was not
as useful. Due to the low solubility of SBR in THF, GPC results consistently
underestimated polymer content. The molecular weight of the SBR modifier plays an
important role in the properties of the modified binder. There is little change in initial
softening point with molecular weight, however there is significant variation in final melt
temperature. Dynamic shear modulus shows a range of about one performance grade
from the low molecular weight SBR to the high molecular weight SBR. No such effects
were seen in low temperature properties. This is because the low temperature properties
are mainly governed by the base asphalt and not the polymer. All in all, SBR appears to

be a promising modifier.

7.2 - Elvaloy® AM

Elvaloy® AM is a reacting polymer which bonds to the asphalt. Rather than
forming a one-phase Elvaloy® AM/Asphalt network, it is believed that the Elvaloy® AM
simply attaches to some of the functional groups in the asphaltenes creating larger
asphaltene micelles. This is seen clearly in the GPC curves which show a large increase
in the high molecular weight region of the asphalt.

The high temperature performance grade was the highest of any modifier used in
this study. An improvement of one grade over the control sample was seen with addition
of 2% (w/w) Elvaloy® AM. One of the reasons for the large increase in performance
grade compared to the virgin asphalt is the high temperature and long time required for
blending the asphalt and the modifier. These extreme mixing conditions cause severe

oxidation which leads to stiffer binders and better high temperature properties. Because

137

this is a reactive process, the mixing procedure is very sensitive to thermal gradients and
three dimensional flow patterns. Direct scale up to batch sizes of 300 g from 150 g
resulted in significant differences in properties. Because of the difficulty in producing
consistent samples, any Elvaloy® AM pavement work needs to be monitored closely.

Elvaloy® AM modified samples had very high viscosity. Samples with 2% (w/w)
and greater Elvaloy® AM had viscosity that did not meet the SHRP specification for
handling. Because 2% (w/w) is the optimum content, this means that to get the most
benefit from of Elvaloy® AM, different processing equipment may need to be developed.

The low temperature properties of Elvaloy® AM were poor. This degradation is
likely caused by the excessive oxidation of the binders during processing. The low
temperature SHRP performance grade was one to two grades worse than that of the base
asphalt. There was no improvement shown with increasing Elvaloy® AM content.

FTIR was a somewhat useful tool for Elvaloy® AM detection. The characteristic
peak for Elvaloy® AM is a C=O peak at 1740 cm". This peak overlaps slightly with
asphalt which forms C=O bonds during oxidation. The calibration curves showed good
linear results with Elvaloy® AM content, however the presence of a peak at 1740 cm‘l is
not a definitive indicator of polymer.

Thermomechanical analysis shows that Elvaloy® AM improves the sofiening
point of the asphalt significantly. Unlike the rheological properties, which showed most
of the improvement came from oxidation, the softening properties show significant
improvement both from oxidation and the polymer. For example, the optimum 2% (w/w)

Elvaloy® AM modified AC-S shows an improvement in melt temperature of 18°C due to

 

138

oxidation and an additional 7°C due to polymer. Similar results are seen at all contents
and with all three asphalts tested.

Elvaloy® AM appears to show some benefits, however there are some significant
drawbacks. The largest drawback is the degradation of low temperature binder properties
due to the severe mixing procedure. Another drawback is the difficulty encountered in
scale up of the mixing procedure. Without a good scale-up procedure, the lab results are
not very meaningful to a real world pavement application. Other polymers such as SBS

and SBR are better candidates for asphalt modification.

7.3 - Recycled Crumb Rubber

There is a great deal of interest in using recycled crumb rubber in asphalt. This is
driven by the environmental problem of disposing of used tires. Because the goal of
CRM modification is not necessarily improvement of the pavement, but rather as a
dumping ground for old tires, the basis for success is simply maintaining the performance
of the base asphalt upon CRM modification.

Only the high temperature properties of CRM modified pavements have been
studied. The mixing procedure is a low temperature, low time procedure which should
result in minimal oxidation of the asphalt. The use of ultrafine CRM aids in dispersion in
the asphalt, however, CRM particles do not dissolve completely and are still visible in the
final mixture. Dynamic shear rheometry shows that addition of CRM results in an
improvement of the high temperature performance grade by one grade for every 5%
(w/w) CRM. Viscosity tests show that up to 15% (w/w) CRM can be added to the

binders while maintaining a processable viscosity. Studies on the long term (24 hr.) high

 

139

temperature storage of the CRM binders indicate that the presence of CRM either masks
or retards the aging process. The increase in softening point after storage is less for the
CRM modified binders than it is for the control.

CRM looks to be a promising modifier, resulting in significant increases in the
high temperature binder performance. This result is very preliminary. Before any
recommendations can be made, the low temperature performance needs to be
investigated. There are also issues such as how the macroscopic CRM particles will
effect the binder-aggregate adhesion and how CRM will affect the long term aging of the
asphalt. There is also the issue of recyclability of the CRM pavement that will probably

become impossible due to the large CRM contents required for modification.

7.4 - Ethylene Vinyl Acetate

EVA appears to be a promising modifier which doesn't suffer from some of the
problems of the other modifiers. It dissolves well in asphalt at the same conditions as
SBS blending. It improves the high temperature performance one grade with 3% (w/w)
polymer and two grades with 5% (w/w). The viscosity remains low enough to process
7% (w/w) EVA modified binders. It has a characteristic IR peak at 1240 cm'1 that makes
polymer detection easy. It shows softening and melt properties that are better than those
of SBR. It does not suffer from the phase separation problems that plague polyethylene.

Two different molecular weight EVA polymers were studied. There was minimal
differences in the high temperature properties of the two polymers. For this study, both
were blended with asphalt using the same procedure. However, in reality, the low MW

polymer can be blended at lower temperatures for shorter times. This can improve the

140

low temperature properties of the modified binder. While these results are still

preliminary, it appears that EVA is an excellent choice for asphalt modification.

7.5 - Recommendations for Further Study

This study attempts to present a comprehensive investigation of polymer modified
asphalt. As good as it was, there are still many areas than need to be investigated.

One of the main areas that should be investigated is the effect of asphalt source on
the properties of the modified asphalt. This study used four different asphalt grades from
the same supplier. In order to generalize these results, the modifiers need to be tested
with asphalts from different sources. It will be more beneficial to use the same asphalt
grade from different suppliers than using different asphalt grades from the same supplier.
The majority of the differences in properties between modified AC-2.5 and modified AC-
20 are due simply to the differences in the base asphalt properties.

Another area to investigate is to develop a better model of how the asphalt and
modifier interact to yield the properties seen in the blend. A first step is to determine
what variables are important in determining neat asphalt properties. This study
investigated molecular weight and asphaltene content. While it is still believed that these
play an important role, the inability to correlate properties using only these factors
indicates that other variables are also important. Two possible factors are asphalt polarity
and asphaltene/resin compatibility. Asphalt polarity can likely be determined by looking
at different ratios of Corbett fractions. Asphaltene/resin compatibility is currently

measured by looking for the onset of flocculation. Because asphalt forms dark solutions,

141

a more quantitative method than visual inspection is required to determine the actual
onset.

Another view of modeling was presented in Chapter 6, namely phenomenological
modeling. The complex chemical nature of asphalt make it unlikely that good predictive
models can be developed for the physical properties based on chemistry alone. This is
complicated further by the addition of modifiers. One way to deal with this is to
concentrate more on what is happening rather than why it is happening. For example,
with SBR modification, the two main phenomena are network formation and percolation
effects. The rheological properties seen in SBR blends is a result of the interaction of
these two phenomena. Similar models should be developed for the other modifiers.

If the pavement performance studies of Elvaloy® AM modified binders show
improvement, further work needs to be done to better understand the mixing procedure.
This could include kinetics studies, studies of thermal gradients in the mixing vessel, etc.

The final area of work is to synthesize the data collected by the author with data
collected by the microstructure group and the engineering properties group to get a better
understanding of how, on a fimdamental level, polymer modification improves pavement
performance. This is especially important when considering low temperature cracking
behavior. The binder properties have only shown slight improvement in low temperature
behavior as measured by SHRP. One thing that SHRP ignores is the crack propagation
behavior and how it is affected by modification. Another is the failure mode of the
asphalt at low temperatures. These factors need to be considered when evaluating the use

of modified binders to prevent low temperature failure.

LIST OF REFERENCES

142

LIST OF REFERENCES

. Baladi, G. Y. Fatigue Life and Permanent Deformation Characteristics of Asphalt

Concrete Mixes. Transportation Research Record 1227, TRB, National Research
Council, Washington, DC, 1989.

. Baladi, G. Y., R. S. Harichandran, and R. W. Lyles. Asphalt Mix Design - An
Innovative Approach, Transportation Research Record 11 71, TRB, National
Research Council, Washington, DC, 1989.

. Baladi, G.Y. Integrated Material and Structural Design Method for Flexible
Pavements. Final Report RD-88-109 and 110. F HWA, US. Department of
Transportation, 1987.

. Newcombe, D.E., M. Stroup-Gardiner, and J. A. Epps, Laboratory and Field Studies
of Polyolefin and Latex Modifiers for Asphalt Mixtures, in Polymer Modified Asphalt
Binders, AST M STP 1108, K. R. Wardlaw and S. Shuler, Editor. 1992, American
Society for Testing Materials: Philadelphia.

. Rogge, D. R., R. L. Terrel, and A. J. George, Polymer Modified Hot Mix Asphalt -
Oregon Experience, in Polymer Modified Asphalt Binders, A S TM S TP 1108, K. R.
Wardlaw and S. Shuler, Editor. 1992, American Society for Testing Materials:
Philadelphia.

. Kholsa, P. N., Effect of the Use of Modifiers on Performance of Asphaltic Pavements.
Transportation Research Record 1317, TRB, National Research Council, Washington,
DC, 1991.

. Shull, J. C., An Investigation of the Fundamental Chemical, Physical, and
Thermodynamic Properties of Polymer Modified Asphalt Cement, MS Thesis,
Michigan State University, 1995

. Material Reference Library Asphalt Data, in Strategic Highway Research Program,
National Reserach Council, Washington, DC. 1992

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

143

Highway Material Engineering - Asphalt Materials and Paving Mixtures. Vol
FHWA-l-90-008, Washington, DC, 1990, U.S.D.O.T., Federal Highway

Administration, National Highway Institute
Barth, E. J ., Asphalt - Science and Technology, 1962, New York, Gordon & Breach.

Goodrich, J .L., Asphaltic Binder Rheology, Asphalt Concrete Rheology and Asphalt
Concrete Mix Properties, in Asphalt Paving Technology. 1990.

Polymers in Bituminous Materials - Final Report, Michigan State University, contract
#92-1595DAB, 1993

Goodrich, J. L., What is the Purpose of Modifi/ing Asphalt with Polymers?, Chevron
Research and Technology Company, 1990.

Newcomb, D.E., M. Stroup-Gardiner, and J .A. Epps, Laboratory and Field Studies of
Polyolefin and Latex Modifiers for Asphalt Mixtures, in Polymer Modified Asphalt
Binders, ASTM STP 1108, K.R. Wardlaw and S. Shuler, Editor. 1992, American
Society For Testing Materials: Philadelphia

Jain, P.K., S.B. Sangita, and LR. Arya, Characterization of Polymer Modified Asphalt
Binders for Roads and Airfields, in Polymer Modified Asphalt Binders, ASTM STP
1108, K.R. Wardlaw and S. Shuler, Editor. 1992, American Society For Testing
Materials: Philadelphia.

Hesp, 8., Personal Communication. 1992,

Little, D., Performance Assessment of Binder-Rich Polyethylene-Modified Asphalt
Concrete Mixtures (No Vophalt). TRR - Transportation Research Record, 1991. 1317.

Daly, W.H., Z. Qiu, and I. Negulescu, Preparation and Characterization of
Polyethylene Modified Asphalt. Transportation Research Board 72nd Annual Meeting
January 10-14, 1993 Washington, DC, 1993.

Anderson, D.A., et al. , Rheological Properties 0f Polymer Modified Emulsion
Residue, in Polymer Modified Asphalt Binders, AS T M S TP [108, K.R. Wardlaw and
S. Shuler, Editor. 1992, American Society For Testing Materials: Philadelphia.

Bouldin, MG. and J .H. Collins, Influence of Binder Rheology on Rut Resistance of
Polymer Modified and Unmodified Hot Mix Asphalt, in Polymer Modified Asphalt
Binders, ASTM STP 1108, K.R. Wardlaw and S. Shuler, Editor. 1992, American
Society For Testing and Materials: Philadelphia. p. 50-60.

Linde, S. and U. Johansson, Thermo©0xidative Degradation of Polymer Modified
Bitumen, in Polymer Modified Asphalt Binders, ASTM STP 1108, K.R. Wardlaw and
S. Shuler, Editor. 1992, American Society For Testing Materials: Philadelphia.

22.

23

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

144

Schwager, B.P., Viscosity Measurements of Polymer Modified Asphalt, in Polymer
Modified Asphalt Binders, ASTM STP 1108, K.R. Wardlaw and S. Shuler, Editor.

1992, American Society For Testing Materials: Philadelphia.

. F arrar, M. and R. Harvey, Performance of Binder-Modifiers in Recycled Asphalt

Pavement. Transportation Research Board 72nd Annual Meeting January 1993, 1993.

Mayarna, M., M. Yshino, and K. Hasegawa, An Evaluation Of Heavy Duty Binders, in
Polymer Modified Asphalt Binders, ASTM STP 1108, K.R. Wardlaw and S. Shuler,
Editor. 1992, American Society For Testing Materials: Philadelphia.

Chollar, B.H., et al., Characteristics of the F urfural Modified Asphalt Reaction.
Transportation Research Board 72nd Annual Meeting January 10-14, 1993
Washington, DC.

Stuart, K.D., Asphalt Mixtures Containing Chemically Modified Binders.
Transportation Research Board 72nd Annual Meeting January 10-14, 1993
Washington, DC, 1993.

Huang, SC, et al. Evaluation of the eflects of a few Modifiers on the Temperature
Susceptibility» and Aging Characteristics of Typical Asphalt Cements. in 72nd TRB
Annual Meeting. 1993. Washington, DC.

Stroup-Gardiner, M., D.E. Newcomb, and B. Tanquist, Asphalt-Rubber Interactions.
Transportation Research Board 72nd Annual Meeting January 10-14, 1993
Washington, DC.

Jones, D.R., T.W. Kennedy, and HF. Torshizi, Field Performance of
Polymer-Modified Asphalts. Transportation Research Board 72nd Annual Meeting
January 10-14, 1993 Washington, DC, 1993.

Khosla, N.P. Laboratory Characterization of Crumb Rubber Modified Mixtures. in
72nd T RB Annual Meeting. 1993. Washington, DC.

Aurilio, V., Crumb Rubber Modified Asphalt Materials, Recycling Crumb Rubber
Modified Mixtures. 1993, Washington, DC.

Polymers in Bituminous Materials Phase I] -Year One Executive Summary, Michigan
State University, 1995

Performace Graded Asphalt Binder Specification and Testing, Superpave Series No.
1 (SP-1), The Asphalt Institute, Lexington, KY

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

145

Heithaus, J. J ., Measurement and Significance of Asphaltene Peptization, Symposium
on Fundamental Nature of Asphalts, Division of Petroleum Chemistry Preprints,
ACS, Vol. 5, No. 4, 1960, pp. A23-A37.

Glover, C.J., R.R. Davision, et al. Chemical Characterization of Asphalt Cement and
Performance-Related Properties, Transportation Research Record 1171, TRB, 1988,
pp. 71-81

ASTM, Standard Test Methods for Separation of Asphalt into Four Fractions, ASTM
D 4124-86.

D. A. Anderson and T. W. Kennedy, Asphalt Paving Technol. 61, 67 (1992).

Price, R. P., and J. L. Burati, Jr. Predicting Laboratory Results for Modified Asphalts
using HP-GPC, Asphalt Paving Technology, Vol. 58, 1989, pp. 1-32.

Bishara, S. W., R. L. McReynolds, and E. R. Lewis, Interrelationship Between
Performance Related Properties of Asphalt Cement and Their Correlation with
Molecular Size Distribution, Transportation Research Record 1323, TRB, National
Research Council, Washington, DC, 1991, pp. 1-9.

Kim, K. W., J. L. Burati, Jr., and S. N. Amirkhanian, Relation of HP-GPC Profile
with Mechanical Properties of AC Mixtures, Journal of Materials in Civil
Engineering, Vol. 5, No. 4, November, 1993, pp. 447-459.

Garrick, N. W., Use of Gel-Permeation Chromatography in Predicting Properties of
Asphalt, Journal of Materials in Civil Engineering, Vol. 6, No. 3, August 1994, pp.
376-389.

Price, R. P. and J. L. Burati, Jr., A Quantitative Method Using HP-GPC to Predict
Laboratory Results of Asphalt Cement Tests, Asphalt Paving Technology, Vol. 58,
1989, pp.182-219.

Robertson, R. E., et. al., Chemical Properties of Asphalts and Their Relationships to
Pavement Performance, Asphalt Paving Technology, Vol. 60, 1991 , pp. 413-436.

Jennings, W., Workshop - Prediction of Asphalt Performance by HP-GPC, Asphalt
Paving Technology, Vol. 54, 1985, pp. 635-646.

Jennings, W., et. al., Predicting the Performance of Montana Test Sections by
Physical and Chemical Testing, Asphalt Paving Technology, Vol. 57, 1988, pp. 412-
423.

Thenoux, G., C. A. Bell, and J. E. Wilson, Evaluation of Physical and Fractional
Properties of Asphalt and Their Interrelationship, Transportation Research Record
1171, TRB, National Research Council, Washington, DC, 1988, pp. 82—97.

146

47. Stock, A. F ., and J. Button, Relationship Among Parameters Derived fi'om High
Performance Liquid Chromatography, Physical Properties, and Pavement

Performance in Texas, Transportation Research Record 1115, TRB, National
Research Council, Washington, DC, 1987, pp. 33-45.

48. Garrick, N. W., and L. E. Wood, Relationship Between High-Pressure Gel
Permeation Chromatography Data and the Rheological Properties of Asphalts,
Transportation Research Record 1096, TRB, National Research Council, Washington,
DC, 1986, pp. 35-41.

49. Bishara, S. W., and R. L. McReynolds, Correlation Between Performance-Related
Characteristics of Asphalt Cement and Its Physiochemical Parameters Using
Corbett ’s Fractions and HPGPC, Transportation Research Record 1342, TRB,
National Research Council, Washington, DC, 1992, pp. 35-49.

50. Perry, R. And D. Green, Perry ’s Chemical Engineering Handbook, Sixth Edition,
McGraw-Hill, New York, 1984, pp. 3.71-3.72.