SP1
SP1
Download as pdf
eeeveeeueunvve”s
eerewweweseeeeeseeseeeueeee
Performance Graded
Asphalt Binder
Specification
and Testing
he
ASPHALT INSTITUTE
‘Superpave Series No. 1 (SP-1)
Third Edition, Revised 2003
ANVddadiis ‘The Asphalt Institute can accept no responsibility for the inappropriate use
of this manual. Engineering judgment and experience must be used to
properly utlize the principles and guidelines contained in this manual,
taking into account available equipment, local materials and conditions,
All reasonable care has been taken in the preparation of this manual;
owever, the Asphalt Institute can accept no responsibilty for the
consequences of any inaccuracies which it may contain,
Copyright © 2003, Asphalt Institute
All Rights Reserved
Printed in the US.A. -ss ewww vvuvuvvvvvuvvuvuuvuuUuUuUuUuUuUuUuUueUei
‘The Strategic Highway Research Program (SHRP) was
established by Congress in 1987 as a five year, $150 million
research program to improve the performance and durability
Of United States roads and to make those roads safer for both
motorists and highway workers. $50 milion of the SHRP
research funds were used for the development of performance
based asphalt specifications to directly relate laboratory
analysis with field performance.
Superpave™ (Superior Performing, Asphalt Pavements) is a
product of the SHRP asphalt research. ‘The Superpave
system incorporates performance based asphalt materials
characterization with the design environmental conditions 10
improve performance by controling rutting, low temperature
‘cracking and fatigue cracking. The three major components
‘of Superpave are the asphalt binder specification, the mixture
design and analysis system, and a computer sofiware system.
‘This manual is devoted to the Supespave asphalt binder
specification, AASHTO M320, and the tests that it uses 10
characterize asphalt binders.
‘Much of the material contained in this manual was originally
produced by the Asphalt Institute under contract with the
Federal Highway Administration (FHWA) as the National
Asphalt ‘Training Center (NATO) for Superpave technology.
‘The NATC developed and conducted week-long training
‘courses in Superpave binder and mixture technology, and.
‘was administered by the FHWA’s Office of Technology
Applications.
A.companion publication, Superpave Mix Design, Superpave
‘Seties No, 2 (SP-2) is also available from the Asphalt Institute,
Superpave™ isa trademark of the Strategic Highway Research
Program, weeweeeeeeeeee eee eee eeeeueewee
Foreword. i
“Table of Contents a
Lis of Figures. vii
Lis of Tables. wii
Chapter 1: Back groundernncrnnneunenemmrememnne 1
Origins of Asphalt 1
Chemical Composition of Asphal. 3
Asohalt Behavior. 6
High Temperature Behavior. 6
Low Temperature Behavior. 9
Intermediate Temperature Behavior. 10
Aging Behavior. a
Asohalt Property Measurements B
Superpave Binder Aging Procedures, 0
Rolling Thin Film Oven (RTFO). 18
Pressure Aging Vessel (PAV). 20
Superpave Binder Tests, 2
Dynamic Shear Rheometer 2
Rotational Viscometer. sane 29
Bending Beam Rheometer. 33
Direct Tension Teser. 38 | Ghoptee’4' Pavensent Porformance In the Siperaave
Binder Specification...
sofey.
Pumping and Handling
Permanent Deformation,
Excessive Aging.
Fatigue Cracking
Low Temperature Cracking,
|
48
Chapter 8: Binder Grade Sel0CtlOAmmmrnnararnn 53 |
Air Temperature Selection.
Pavement Temperanure Selection
Binder Grade Selection Based on Pavement Temperatures.
Adjusting Binder Grade Selection For Tralfic Speed and Loading,
|Member Companies of the Asphalt Insite
56
56
56
58
6 liad ee ee
List of Figures
Figure 11
Figure 1.2
Figure 13
Figure 14
Figure 15
Figure 1.6
Figure 17
Figure 1.8
Figure 1.9
Figure 2.1
Figure 22
Figure 3.1
Figure 3.2
Figure 33
Figure 34
Figue 35
Figure 36
Figure 37
Figure 38
Figure 39
Figure 3.10
Figare 3.11
Figare 3:12
Figure 313
Figure 314
Figure 315
Figare 3.16
Figure 3.17
Figure 3:18
Figure 3:19
Figure 3.20,
Figure 321
Figure 322
Figure 323
Figure 324
‘Annual Asphalt Usage
Component Analysis of Asphalt Cement
‘Types of Asphalt Molecules
Asphalt Cement Flow Behavior
Microscopic View of Liquid Flow Characteristics
Newtonian Fluid Behavior
Pseudoplastic (Shear Thinning) Behavior
Dilatant (Shear Thickening) Behavior
“Spring-Dashpot” Model of Viscoelastic Behavior
Variations of Three Viscesity:Graded Asphalts
‘Superpave Laboratory Tests ~ Relation to Performance
Rolling Thin Film Oven
IFO Bottle
Pressure Aging Vessel
Dynamic Shear Rheometer Operation
Viscoelastic Behavior
DSR Molds and Specimens, Plates and Spindles
Asphalt Sample Configuration in DSR
Dynamic Shear Rheometer
‘Suess Strain Output for a Constant Stress Rheometer
Suess Sain Response of a Viscoelastic Material
[DSR Asphalt Specimen Calculations
Rotational Viscometer
Rotational Viscometer Operation
Rotational Viseometer
Rotational Viscometer Displays
Bending Beam Rheometer
‘Aluminum Mold Asserbly,
ending Beam Test
[BBR Deflection and m-value
‘Typical BBR Deflection and Load Output
Direct Tension Test Specimen,
Determination of Failure Stan inthe Direct Tension Test
Determination of Faure Stes inthe Direct Tension Test
Direct Tension Specimens
16
18
SSRRBERRRERBES
31
jrepeseeece Figure 325. Direct Tension Test Specimen Before and After Trimming... . 43,
Figure 326. Direct Tension Test Loading System ‘4
Figure 327 Thermal Suess Curve for Determining Critical Cracking
Temperature 45
Figure 41 Superpave Binder Specification Format a7
Figure 42. Superpave Specification Rutting Factor Requirement 0
Figure 43. Superpave Specification Fatigue Cracking Factor
Requirements 50
Figure 44 Superpave Specification Low Temperature Requirements... 50
Figure 5.1 Design Temperature Reliability 55
Figure 5.2 Example High and Low Air Temperature Variations 55
Figure 53. Example High and Low Pavement Temperature Variations .. 56
Figure 54 Superpave Binder Grade Selections for Cleveland 57
List of Tables = z
‘Table 1.1 Elemental Analysis of Four Asphalt Cements, 3
‘Teble 21 Superpave Binder Test Equipment 15
Teble 3.1 Superpave Binder Test Aging Condition,
Teble 32 Example DSR Test Information
‘Table 33. Example BBR Test Information
‘Table 34 Example DTT Test Information
RSeR
P> sconce Asphalt is defined by the
American Society for
fesing and Materials (ASTM) as “a dark brown to black:
‘cementitious material in which the predominating constituents
are bitumens which occur in nature or are obtained in
petroleum processing.” As cement, asphalt i especially
valuable to the engineer because its strong, readily adhesive,
highly waterproof, and durable. It provides limited flexibility
to mixtures of mineral aggregates with which itis usually
combined. Although a solid or semi-solid at ordinary
atmospheric temperatures, asphalt may be readily iquefied by
applying heat, by dissolving it in petroleum solvents, or by
emulsifying tin water,
Man has long been aware of the adhesive and
‘waterproofing properties of asphalt. Surface accumulations of
petroleum, forced upward by geological forces, Icave behind,
nnaturally-occurting lakes of asphalt that have hardened after
‘exposure to the elements, Current examples of these deposits
inckide Trinidad Lake Asphalt, on the Island of Trinidad off the
norher coast of Venezuela, and the La Brea “Tar” its near
‘Los Angeles, Natural asphalt is also found impregnated within
porous rock, such as sandstone or limestone, called rock
spilt. Natural asphalts were used by the ancient Babylonians,
Egyptians, Greeks, and Romans as 2 road-building and
‘waterproofing material. Even with the long history of asphalt
usage, the evolution of asphalt as a widespread ingredient in
paving material did not occur until modem petroleum-refining
techniques were developed in the early 1900's.
‘The first documented use of rock asphalt as a sidewalk
surfacing occurred in France in 1802 and later in Philadelphia
in 1838. In 1870, the first asphalt pavement was constructed in
Newark, New Jersey. ‘The first sheet asphalt (fine sand mix)
pavement was built in Washington, D.C. in 1876 with imported
Jake asphalt, Finally, in 1902, asphalt began to be refined from
petroleum, which led to the development of the asphalt
paving industry in the US. This rapid growth in asphalt usage
is shown in Figure 1.1 1980
Figure 1.1 Annual Asphalt Usage
Asphalt is also used in hundreds of products not related to traditional
construction and transportation industries, as described in Introduction to
Asphalt, MS-5, Asphalt Institute. However, about 85 percent of the asphalt
refined from petroleum (the source of nearly all asphalt today) continues to be
wed as a paving material, commonly called asphalt cement, Approximately
50 companies in the US and Canada currently produce asphalt cement refined
from crude oil. Few of these companies produce asphalt cement as their
principal product. In the majority of cases, the asphalt used to pave roads is
‘made from the residuum that remains after the refineries remove the distillates
tomanufacture gasoline, jt fuel, kerosene, lubricating oils and other
commodities.
Before spot market trading of crude oil was common, refiners rarely
changed their crude oil source, or feedstock. This practice led to asphalt
sources that had consistent properties over long periods of time, Out of the
need to react to changes in the world marketplace, modem integrated
refineries can now switch crude feedstock blends in a few days. This,
‘enhanced refining capability has created more variation in asphalt cement
properties, both chemical and physical. New oil exploration in areas such as
the Alaska North Slope and the Maya Peninsula in Mexico has addled to the
variation in asphalt properties. Out of this variation comes a greater need t0
bbe able to evaluate asphalt response to particular conditions and predict
performance in terms of recognized forms of distress
D> vnccnans arcane een
Sarna eenorm
ee eal hes aol acto Ris lide hpccrtons te
Satie Re ene eee acon ee
Fi eee eee Set estas eta Ten
atoms in the asphalt molecular structure. These contribute to many of asphalt's
unique chemical and physical properties by causing much of the interaction.
Sees een in pps ane bea eee ea
Basra eisai Ales eames endl fara eree ne a
inompomte snp oat) Orlane pray por ote ea
Spill robs feces crap Grin) ao Segre seed
wih git Qhowlgecatie ake crete
cl aes aatas es oa ee tent eee te
pera, pay orn on one poem, he maar agubcase tie
Sasi bas toe on ld aoe inleaton oo hegetee te
ak nso ase ne nace racer eae
Record 999, 1984) shows an elemental analysis of four representative
pamicon sph
Table 1.1 Elemental Analysis of Four Asphalt Cements
Asphalt Cement A Bo G oD
Carbon, percent 8377 8578 8290 86.77
Hydrogen, percent 991 1019 1045 10.93
Nitrogen, peroent 0.28 026 078 1.10
Sulfur, pereent 525 341 543 099
Oxygen, percent 077 036 029 020
Vanadium, ppm 100. 7. 1380, 4.
Nicke!, ppm 2 04 108. 6
Asphalt components can be separated and evaluated by using the
solubility ofthe various molecules in different solvents. The two most
frecuently-used fractionation methods are the chromatographic method of
Corbett (ASTM D 4124, Standard Test Method for Separation of Aspalt into be
Four Fractions) and the precipitation method of Rostler (described in
Proceedings, Association of Asphalt Paving Technologists, Volume 31, 1962),
‘The generic fractions determined from these methods are complex mixtures
with varying properties and are not discreet chemical species, which has led to confusing and contradicting predictions of behavior. The two methods
pprocluce different fractions because of the inherent differences in separation
"The Rostler method uses sulfuric acid of varying strength to analytically
separate the groups while the Corbett method uses differential absorption and
desorption. Both methods separate out the insoluble discrete “asphaltenes'
in different quantities using two different solvents (n-heptane or n-pentane)
as shown in Figure 1.2. The remaining groups of resins and oils are defined
differently by each procedure, Corbett refers to this mix as petrolenes, made
Uupof saturates, napthene aromatics, and polar aromatics. Rostler refers to this,
tix as maltenes, composed of paraffins, first and second acidiffins, and
nitogen bases, The figure shows some vague relationships between these two
sets of components; however, overlaps exist and there are no strict dividing,
lines. Some information and relative tendencies have been observed from
these analytical methods, but no definite correlations between component
‘quantities and asphalt performance have been identified
Rostler
n-pentane
saturates. paraffins
napthene aromatics second acidiffins
polar aromatics first scidiffins
nitrogen bases:
Figure 1.2 Component Analysis of Asphalt Cement
‘The molecular structure of asphalt is extremely complex, and varies in
size andl type of chemical bonding with each blend or source. There are three
basic types of molecules: aliphaties, cyclics, and aromatics (Figure 1.3). The
aliphatics, or paraffinic, are linear, three-dimensional, chain-like molecules and
are “oily” or “waxy” in nature. ‘The cydics, or naphthenics, are three-] _ Because of its viscoelastic nature, asphalt cement
Behavior | behavior depends on both temperature and rate
of loading. As shown in Figure 1.4, the amount
an asphalt flows could be the same for one hour at 60°C or 10 hours at 25°C.
In other words, the effects of time and temperature are related; the behavior at
ire over short time periods is equivalent to what occurs at lower
temperatures and long durations. This is often referred to as the time-
temperature shift or superposition concept of asphalt cement,
| Asphatt
Figure 1.4 Asphalt Cement Flow Behavior
>» > High Temperature Behavior
In hot conditions (e.g,, desert climate) or under sustained loads (€.,
slow moving or parked trucks), asphalt cement acts like a viscous liquid.
[Under these circumstances, the aggregate is the part of the hot mix asphalt that
—— eT er ee See eee eee eee ee ee eee
beurs the load} Viscosity isthe physical material characteristic used to describe
the resistance of liquids to flow. If the slow-flow movement of hot asphalt
‘cement could be observed under a powerful microscope, adjacent layers of
molecules would be observed sliding past each other (Figure 1.5). The
resisting force or friction between these layers i related to the relative velocity
at which they slide by each other. The top layer tres to pull the bottom layer
along while the bottom layer tries to hold the top layer back. ‘The relationship
beween the resisting force and relative velocity can be very different for
different liquids,
‘Shear or Resisting Stress, T
between Layers
Layer 4
oe. velocity layer 1
SewtewPe- velocity layer 2
Layer 2
direction of flow
of both layers
Figure 1.5 Microscopic View of Liquid Flow Characteristics
Fortunately, viscosity is one characteristic (as opposed to chemical
‘composition, for example) that can be used to express this difference. This
‘equation describes the situation in Figure 1.5 and shows how the coefficient of
viscosity (i) is used to explain differences in flow characteristics among
diferent liquids:
‘t= Ux rate of shear strain
In this equation, tis the shearing resistance between the layers, and the rate of
shear stain corresponds to the relative speed at which layer 1 slides by layer 2.
scx Shear Viscosity
Stress
in
Rate of Shear Strain Rate of Shear Strain
Figure 1.6 Newtonian Fluid Behavior
Asphalt binders typically display ether Newtonian or non-Newtonian
behavior a shear rate is changed. Newtonian fluids have a linear relationship
between resisting force and relative velocity. In other words, if you apply twice
the force toa Newtonian fluid, it will move twice as fast. Air, water, and hot
asphalt at temperatures greater than 60°C) are common Newtonian fluids. For
these material, viscosity (y) is constant regardless of shear rate as shown in
Figure 16
‘Asphalt binders, particularly modified asphals, may exhibit non-
‘Newtonian behavior. Some asphalt materials may exhibit psuecoplastic or
shear thinning” behavior. This type of behavior, demonstrated in paint, is
Shear Viscosity
Stress
Rate of Shear Strain Rate of Shear Strain
Figure 1.7 Pseudoplastic (Shear Thinning) Flow Behavior
Pao characterized by a decrease in viscosity as shear rate is increased (Figure 1.7).
In other words, the faster you stir the material, the thinner (more fluid) it gets.
‘At moderate temperatures, asphalt exhibits "shear thinning! behaviors since its
viscosity decreases at increasing shear rates.
‘Another type of non-Newtonian behavior, although less common, is
llstant or "shear thickening” behavior. This type of behavior, demonstrated in
y slurries, is characterized by an increase in viscosity as shear rate is
increased (Figure 1.8). In other words, the faster you stir the material, the
thicker (less fluid) it gets,
|
Rate of Shear Strain Rate of Shear Strain
Figure 1.8 Dilatant (Shear Thickening) Flow Behavior
> > Low Temperature Behavior
In cold climates (e.g, winter days) or under rapidly-applied loads (e.,,
fast moving trucks), asphalt cement behaves like an elastic solid. Elastic solids
are like rubber bands; when loaded they deform, and when unloaded, they
retum to their original shape.
If sessed beyond material capacity or strength, clastic solids may
break. Even though asphalt cement is an elastic solid at low temperatures, it
‘may become too britle and crack when excessively loaded, For this reason,
low temperature cracking sometimes occurs in asphalt pavement during cold
‘weather, In these cases, loads are applied by internal stresses that accumulate
in the asphalt pavement when it tries to shrink while being restrained by the
lower pavement layers (eg,, when temperatures fall during and after a sudden i
cold front),
BACKGROUND q >> Intermediate Temperature Behavior
‘Most environmental conditions lie between the extreme hot and cold
situations. In these climates, asphalt binders exhibit the characteristics of both
Tiscous liquids and elastic solids. Because of this range of behavior, asphalt is
fan excellent adhesive material t0 use in paving, but an extremely complicated
rmiaeral to understand and explain, When heated, asphalt acts asa lubricant,
allowing the aggregate to be mixed, coated, and tightly-compacted to form a
‘smooth, dense surface, After cooling, the asphalt acts as the glue to hold the
aggregate together in a solid matrix. In this finished state, the behavior of the
asphalt is termed viscoelastic; it has both elastic and viscous characteristics,
depending on the temperature and rate of loading. Conceptually, this kind of
response to load can be shown using the “spring- > Aging Behavior
Because asphalt cements are composed of organic molecules, they
react with oxygen from the environment. ‘This reaction is called oxidation and
it changes the structure anc! composition of asphalt molecules. Oxidation
uses the asphalt cement to become more brite, generating the term
oxidative hardening or age hardening, Oxidative hardening happens at a
relatively slow rate in a pavement, although it occurs faster in warmer climates
and during warmer seasons. Because of this hardening, old asphalt pavements
ate more susceptible o cracking. Improperly compacted asphalt pavements
ray exhibit premature oxidative hardening. In this cise, inadequate
compaction leaves higher percentage of interconnected air voids, which
alows more air to penetrate into the asphalt mixture, leading to more
‘oxidative hardening,
In practice, a considerable amount of oxidative hardening occurs before
the asphalt is placed, At the hot mixing facility, asphalt cement is added to the
hot aggregate and the mixture is maintained at elevated temperatures for a
period of time. Because the asphalt cement exists in thin films covering the
aggregate, the oxidation reaction occurs at a much faster rate.
(Other forms of hardening include volatilization ancl physical
hardening, Volatilzation occurs during hot mixing and construction, when
vlatile components tend to evaporate from the asphalt. Physical hardening
‘occurs when asphalt cements have been exposed to low temperatures for long
periods. When the temperature stabilizes at a constant low value, the asphalt
cement continues to shrink and harden. Physical hardening is more
pronounced at temperatures less than 0°C and must be considered when
testing asphalt cements at very low temperatures. <<< wwe wee www HHH HHT aaa
eRe EES
Because of its chemical complexities, asphalt
specifications have been developed around physical property
tess such as penetration, viscosity, and ductility, These
physical propery tess are performed at standard test
temperatures, and the test results are used to determine if the
muterial meets the specification enteia, However, there are
limitations in what the results ofthese test procedures provide
Many of these tests are empirical, meaning that pavement
Performance experience is required before the test results yield
‘meaningful information, Penetration is an example of this.
‘Tre penetration test indicates the stiffness of the asphalt, but
any relationship between asphalt penetration and performance
has to be gained by experience. An additional drawback of
empiricism is thatthe relationship between the test result anc
performance may not be very good
Another limitation to these tests and specifications is
thatthe tests do not give information for the entire range of
‘ypical pavement temperatures. although viscosity is a
fundamental measure of flow, it only provides information
about higher temperature viscous behavior—the standard test
temperatures are 60°C and 135°C. As well, penetration
«describes only the consistency at a medium temperature
250). Lower temperature elastic behavior cannot be
realistically determined from this data to predict low
temperature performance.
Penetration and viscosity specifications can classify
different asphals with the same grading, when infact these
asphalt may have very diferent temperature and performance
characteristics. As an example, Figure 2.1 shows three asphalts
that have the same viscosity grade because they are within the
specified viscosity limits at 60°C, have the minimum penetration
325°C, and reach the minimum viscosity at 135°C, While
Asphalts A and B display the same temperature dependency, they have much different consistency at all temperatures, Asphalts A and C
have the same consistency at low temperatures, but remarkably different high
temperature consistency. Asphalt B has the same consistency at 60°C, but
‘shares no other similarities with Asphalt C, Because these asphalts meet the
‘same grade specifications, one might erroneously expect the same
Chanacteristics during construction and the same performance during hot
and cokl weather conditions,
Consistency
(pen or vis)
hard
Recognizing the
deficiencies in the
cexisting system, state
highway agencies decided to fund a research program to develop a new
stem for specifying paving asphalt. In 1987, the Strategic Highway Research
Program (SHRP) began developing new test for measuring the physical
properties of asphalt. A major result ofthis $50 million research effor was the
Superpave binder specification. It is called a “binder” specification because itis
intended for modified 2s well as unmoxtified asphalts. A unique feature of the
‘Superpave specification is that the specified criteria remain constant, but the
‘temperature at which the criteria must be achieved changes for the various
grades, As an example, consider two construction projecs—one at the
equator and one at the Arctic Circle, Good asphalt performance is expected in
PD> cruursameanoss beth locations, but the temperature conditions under which specified binder
properties must be achieved are vastly different,
‘The Superpave binder specification and the test methods used to,
characterize asphalt are being evaluated by both AASHTO andl ASTM,
AASHTO M320, Standard Specification for Performance-Graded Asphalt
‘Binder, standardizes the specification requirements, AASHTO's provisional
standard, MP-1a, incorporates an alternate procedure for determining the
criical low cracking temperature of an asphalt binder using a combination of
‘wo test procedures. Some researchers believe that the alternate approach in
MP-la is more rigorous and inclusive of most modified asphalts. These
specifications may be revised as additional experience is gained
‘The Superpave tests measure physical properties that can be related
directly to field performance by engineering principles. ‘The Superpave binder
teas are also conducted at temperatures that are encountered by in-service
pavements. Table 2,1 lists the binder test equipment and a brief description
of why each testis used in the Superpave specification, Figure 2.2 desenbes
how each test provides some indication of binder performance, however, the
pavement structure and mixture properties will have some additional bearing
on this performance,
Rolling Thin Film Oven (RTEO) Simulate binder aging (hardening)
Pressure Aging Vessel (PAV) ——_characterstios
Dynamic Shear Rheometer (DSR) Measure binder sifness and
elasticity properties at high
‘and intermediate temperatures
Ga)
Rotational Viscometer (RV) ‘Measure binder viscosity at high
temperatures
Bending Beam Rheometer (BBR) Measure low temperature stfiness
and failure properties
Direct Tension Tester (OTT)
ssmnersmcmesons Performance Property
Test Equipment
Handling
Rotational Viscometor eae
Permanent
Deformation|
Dynamic Shear Rheometer
mY
‘Structural
Cracking
Fatigue
Cracking
Bending Beam Rheometer ———f They Low Temp
Direct Tension Tester. —————|_Cracking Cracking
Figure 2.2 Superpave Laboratory Tests -
Relation to Performance
‘The central theme of the Superpave binder specification is its reliance
on testing asphalt binders in conditions that simulate the three critical stages
luring the binders life. Tests performed on the original asphalt represent the
first stage of transpor, storage, and handling. ‘The second stage represents the
asphalt during mix production and construction and is simulated for the
specification by aging the binder in a rolling thin film oven, ‘This procedure
exposes thin binder films to heat and air and approximates the aging of the
asphalt during mixing and construction. The third stage occurs as the binder
ages over a long period as part of the hot mix asphalt pavement layer. This
stage is simulated for the specification by the pressure aging vessel. This
procedure exposes binder samples to heat and pressure in order to simulate
years of in-service aging in a pavement.
> ASPHALT SPECIFICATIONS ‘This chapter introduces the concepts andl test
procedures for the asphalt binder tests performed to determine
compliance with the applicable specification, "This information
intended to provide only an overview of the test procedures.
Specific details are given in the appropriate AASHTO or ASTM
test methods.
As discussed
in Chapter
1, asphalt
binders age primarily due to two different mechanisms:
volatilization of light oils present in the asphalt and oxidation
by reacting with the oxygen in the environment. The blending
and agitation inthe hot mixing facility and during placement
ages the binder by both mechanisms because of the high
temperature and air flow involved in the process. ‘The
‘Superpave specification uses the rolling thin film oven (RTFO)
procedure to simulate this form of aging. After the asphalt
pavement is constructed, aging continues, but the oxidation
mechanism dominates because of the relatively moderate
temperatures of the environment. The Superpave specication
uses the pressure aging vessel (PAV) procedure to simulate tis
in-service aging, It should be noted that binder samples aged
in the PAV have already been aged in the RTFO. Consequently,
PAV residue represents binder that has been exposed to all of
the environmental conditions to which binders are subjected
during production and service >» Rolling Thin Film Oven Procedure
“The RTFO procedure serves two purposes. The first is to provide an
‘aged asphalt binder that can be used for further testing of physical properties
‘The second is to determine the mass quantity of volatiles lost from the asphalt
during the process, Volatile mass loss is an indication of the aging that may
‘occur in the asphalt during mixing and construction operations. Some asphalts
‘even gain weight during the RTFO procedure because of the oxidative products
formed. ‘The RTPO continually exposes fresh films of binder to heat and air
flow. In some cases, its rolling action keeps modifiers (eg., some polymers)
dispersed in the asphalt
‘The RTFO procedure requires an electrically heated convection oven
(Figure 3.D. The oven contains a circular, vertical carriage that holds the
sample bottles and rotates about its center, An air jt blows air into each
sample boitle at its lowest position as it circulates in the carriage. The RTFO
‘oven must be preheated to the aging temperature, 163°C, for a minimum of 16
hours prior to use. ‘The thermostat should be set so that the oven will return to
ths temperature within 10 minutes after the sample bottles are placed and the
door closed.
bottle carriage
air jet
Figure 3.1 Rolling Thin Film Oven oll
Before Filling Alter Filling Coated Bottle
After Testing
Figure 3.2 RIFO Bottle
‘To prepare for RTFO aging, a binder sample is heated until fluid, not
exceeding 163°C, RIFO botles are loaded with 35 grams of binder (Figure
3.2), Immediately ater pouring the sample into the RTFO bottle, the bottle is
tumed on is side to a horizonal position and placed in a cooling rack. The
bottes are allowed to cool 60 to 180 minutes. Eight sample bottles are
required for Superpave binder testing, Two botles are required to make the
‘mass loss determination, and six bottles are used for Further testing. These two
imass-loss sample bottles are weighed to the nearest 0.001 gram.
Sample bottles are placed in the carriage and rotated at arate of 15
revolutions per minute, The air low is set ata rate of 4000 ml/min, and the
samples are subjected to these conditions for 85 minutes
After aging, the two botles containing the mass loss samples are
cooled, weighed to the nearest 0.001 gram, and the samples discarded. The
IFO residue from the other bottle is poured into a single container and.
sired to achieve homogeneity. Scraping the coated bottles is necessary to
‘ensure that enough material can be removed and a homogeneous sample is
‘obtained, ‘The RTFO-aged binder is then used for DSR testing, transferred into
PAV pans for additional aging, or equally proportioned into small containers
and stored for future use.
‘Mass change is reported as the average percent loss of the two samples
after RTFO aging. Iti calculated using this equation:
‘Aged mass - Original mass
‘Mass Change, % =
ass * ‘Original Mass
x10
smear asia osoen ress > Pressure Aging Vessel
[As described previously, the effects of long term in-service aging of
asphalt were not incorporated in specifications for asphalt binders prior to the
Superpave specifications. The PAV, used for many years in asphalt research,
twas modified by SHRP and a new procedure was developed to simulate in-
service aging The PAV exposes the binder to high pressure and temperature for
2ohhours to simulate the effects of long term aging. Since asphalt exposed t0
Jong term aging has also been through the mixing and construction process, the
PAV procedure uses binder aged in the RTFO.
The pressure aging apparatus consists of the pressure aging vessel
and a forced draft oven. Air pressure is provided by a cylinder of clean, dry
‘compressed air with a pressure regulator, release valve, and a slow-release
bleed valve. The pressure vessel is designed to operate under the pressure and
temperature conditions of the test (2070 kPa and either 90°, 100°, or 110°C),
The vessel must accommodate at least 10 sample pans and uses a sample rack
that fits conveniently into the vessel. A continuous temperature monitor is
required during the aging period. The oven should be able to control the
internal temperature of the PAV to within + 0.5°C during the aging period.
{A digital proportional control and readout of oven temperature is required.
Figure 3.3
Pressure
Aging Vessel -~-e ew ewww www weeweeweweweweeeseeeeseeweweweEe
To prepare for the PAV, RIFO-aged binder is heated until fluid and
stimed to ensure homogeneity. ‘Three PAV sample pans of 50 grams each are
prepared and the PAV pans are placed in the sample rack.
‘The unpressurized PAV is preheated to the desired tes temperature
Afier preheating, the sample rack with the samples is placed in the hot vessel
and the lid is quickly secured to avoid excessive heat loss. The aging process is
‘conducted at diferent temperatures depending on the design climate. When
the vessel temperature enters the range from 20°C to 2°C of the required
temperature, the pressure is applied and the timing for the aging perio! begins
After 20 hours, the pressure is gradually released — 8 to 10 minutes are
usually required. If the pressure is released too quickly, foaming may oc
The sample rack is then removed from the PAV and the sample pans are
removed from the rack and placed in an oven at 163°C for 15 minutes. ‘The
sample i then transferred to a storage container and degassed in a 170°C
vacuum oven for 30 minutes at a pressure of 15 kPa absolute.
‘A report for the PAV includes sample identification; aging tet
temperature to the nearest 0.5°C; maximum and minimum aging temperature
recorded fo the nearest 0.1°C; total time during aging that temperature was
ouside the specified range to the nearest 0.1 minute; and total aging time in
hhours and minutes
Rare Figi] As discussed previously, the Superpave
ee ae: binder tests are used to quantify the
asphalt’s performance at three stages of its
life: in is original state, after mixing and construction, and after in-service
aging. The RTFO testis used to simulate the binder aging that occurs during
‘mixing and construction, and the PAV procedure is used t0 simulate the in-
service aging, ‘Table 3.1 shows the binder’s aging condition used in the
Superpave binder test,
Table 3.1 Superpave Binder Test Aging Condition
‘Superpave Binder Test ‘Binder Condition
Dynamic Shear Rheometer (DSR) Original binder
RIFO-aged binder
PAV-aged binder
Rotational Viscometer (RV) Original binder
Bending Beam Rheometer (BBR) PAV-aged binder
Direct Tension Tester (OTT) PAV-aged binder
SUPERPAVE ASPHALT BINDER TESTS q “The relationships among the Superpave binder test result, the binder
‘aging condition, and asphalt performance are further discussed in Chapter 4
>> Dynamic Shear Rheometer
‘Since asphalt behavior depends on both loading time and temperature,
the ideal test for asphalt binders should include both factors, Fortunately,
testing equipment with this capability already exists and is generically known
asdynamic rheometers, dynamic shear rheometers, or oscillatory shear
theometers. Rheometers are adaptable for use with asphalt, so both time and
temperature effects can be evaluated. When used to test asphalt binders, the
dynamic shear rheometer (DSR) measures the theological properties (complex
shear modulus and phase angle) at intermediate to high temperatures.
Spindle
Position
8 c ‘One cycle
A
Figure 3.4 Dynamic Shear Rheometer Operation
“The DSR operation is basi; asphalt is “sandwiched” between two
parallel plates, one that is fixed and one that oscillates (Figure 3.4. As the
plate oscillates, the centerline ofthe plate at point A (indicated by the dark
Yertcal line) moves to point B. From point B, the plate centerline moves back
and passes point A to point C. From point C the plate centerline moves back.
to point A. This oscillation is one cycle and is continuously repeated during,
the DSR operation. ‘The speed of oscillation is frequency. All Superpave DSR
binder tests are performed at a frequency of 10 radians per second, which is
equal to approximately 1.59 Hz (cycles per second), During each cycle, both
saress and strain are measured
>
)PERPAVE ASPHALT BINDER TESTS <<< e eww www wwe eweewnweewnwewwee eee)
‘The DSR is used to characterize both viscous and elastic behavior by
measuring the complex shear moxulus (G*) and phase angle (8) of asphalt
binders. G* is a measure of the total resistance of a material to deformation
when exposed to repeated pulses of shear stress. IC consists of two
components: elastic (recoverable) and viscous (non-recoverable). 5 is an
indicator of the relative amounts of recoverable and non-recoverable
«deformation. ‘The value of G* (°G star") and 8 Celia") for asphalt are highly
dependent on the temperature and frequency of loading. At high
temperatures, asphalts behave like viscous fluids with no capacity for
recovering or rebounding. In this case, the asphalt could be represented by
the vertical axis (viscous component only) in Figure 35; there would be no
lasic component of G*, since 8 = 90°. At very low temperatures, asphalts
behave like elastic solids which rebound from deformation completely. This
condition is represented by the horizontal axis (elastic component only) in
Figure 35. In this case, there is no viscous component of G*, since 8 = 0°
Under normal pavement temperatures and traffic loadings, asphalt
binders act with the characteristics of both viscous liquids and elastic solids.
BBy measuring G* and 8, the DSR provides a more complete picture of the
behavior of asphalt at pavernent service temperatures. The vector arrows in
Figure 35, G*, and Gt, represent the complex moduli of Asphalts 1 and 2.
‘When these asphalts are loaded, part oftheir deformation is elastic (E) and part
is viscous (V); therefore, asphalt isa viscoelastic material, Even though both
asphals in Figure 35 are viscoelastic and have the same G*, Asphalt 2is more
clastic than Asphalt 1, because of its smaller 8. Because Asphalt 2 has a larger
elastic component, it will recover much more deformation from an applied
load. ‘This example clearly shows that G* alone cannot describe asphalt
behavior. The value of 8 is also needed
Viscous Behavior
Figure 3.5
Si ttwbonmer’ Viscoelastic
Behavior
7
Elastic Behavior
SUPERPAVE ASPHALT BINDER TESTS
somone sm men ros ed After the asphalt isin place, the specimen is trimmed! flush with the parallel plates, and the extra 50 microns is “dialed out so that the gap is ‘exactly at the desired value, The specimen will bulge slightly as shown in Figure 3,7. Oscillating Plate \ Proper Amount a a of Asphalt Fixed Plate Figure 3.7 Asphalt Sample Configuration in DSR Since asphalt binder properties are temperature dependent, theometers ‘must have a precise means of controlling the sample temperature. “This is ‘accomplished using either a circulating fluid bath or a forced air oven. Fluid baths typically use water to surround the sample. The water is circulated through a temperature controller that precisely adjusts and maintains the desired sample temperature, Air ovens surround the sample with air during testing. In either case, the temperature must be controlled so that the sample {emperature is uniform and vaties by no more than 0.1°C across the gap. Test Procedure. Aji the asphalt specimen is in place and the test temperature stabilizes, a minimum of ten additional minutes is allowed for the specimen temperature to equilibrate. The actual temperature equilibration time is equipment depenclent and should be checked using a calibration specimen ‘with very accurate temperature sensing capabilites. A thermistor, wrapped with very thin scone rubber sheeting material, s placed between the parallel plates to verify temperature ‘A computer controls the DSR test parameters and records test results, ‘Testing consists of setting the DSR to apply a constant oscillating stress and recording the resulting strain and time lag, 8. The Superpave specifications ‘ requite that the oscillation speed is 10 radians/second, The operator sets the approximate value of shear strain Cstrain amplitude"), Shear stain values vary {rom about one to 12 percent and depend on the aged state ofthe binder being sommes asus noes tested. Original Conaged) binders and RTFO aged binders are tested at strain values of approximately ten to twelve percent. PAV-aged binders are tested at seain values of about one percent. In all cases, strain values must be small ‘enough that the response of the binder (G*) remains in the linear viscoelastic range. In this range, G" is virually unaffected by changes in strain level “To begin the test, the sample is first conditioned by loading the specimen for 10 cycles. During this conditioning period, the theometer measures the stress required to achieve the set shear strain and then precisely rraintains this stress during the test, The shear strain can vary small amounts from the set value to achieve this Constant sires. Variation in shear strain is controlled by the rheometer software. After the 10 conditioning cycles, ten. additional cycles are applied to obtain test data. The sheometer software automatically computes and reports G* and 8, using the relationship between the applied stress and the resulting stain. Figure 3.8 Dynamic Shear Rheometer | Data Analysis. Figure 3.9 shows the two extreme types of behavior, totaly elastic and totally viscous. ‘The complex shear modulus, G*, i the rao of total shear stress guy ~ ug) 1 total shea strain CY = Youg)- The time lag between the applied stress and the resulting strain (for constant stress rheometers as shown in Figure 3:7) ‘or the applied strain and resulting stress (constant strain rheometer) is related to the phase angle, 6. For a perfectly elastic material, an applied load causes an immediate response; thus, the ime lag or phase angle is zero. A viscous material (uch as hot asphalt at mixing temperatures) has a relatively lage time lag between load and response; inthis case, the angle approaches 90 degrees. Asphalt binders are viscoelastic at normal pavement temperatures, they behave somewhere between the two extremes and the DSR displays a response resembling that shown in Figure 3.10 ‘The formulas used by the rheometer software to calculate Tag, and Yue ae shown in Figure 3.11 Although the DSR is capable of providing much more information for analysis, only G* and 8 are required for Superpave specifications, ‘Two forms of G* and 8 are used in the binder specification, Permanent deformation is ‘governed by limiting Gr/sin 5 at the test temperatures to values greater than Elastic: 8 = 0 deg Viscous: = 90 deg Trae Toe time ag Yon “Yon Figure 3.9 Stress-Strain Output for a Constant Stress Rheometer SUPERPAVE ASPHALT BINDER TESTS q VisccelesiccoSpecimen Calculations radius () > ovens mcrr enn ‘DSR Project- Sunemmave Binder Test Strain Controlled Parameters. Measurement Type Intermediate Temperature Range Target Temperature 220°C Strain Amplitude 4.00 percent Pate Diameter 80mm Plate Gap 2.000 mm Equilbration Time 10.0 minutes Aacilry Info Operator 10 Par. Sample 1D RPS Sample Type Pressure Aging Vessel Residue Test Number: 001 ‘Measurement Results: Completed : 41403 3:18 PM Modulus (G") 3787 kPa Phase Angle (delta) 44.9 degrees sin (deta): 2671 kPa Strain Ampitude : 1.00 percent Final Temperature 219°C Ose. Frequency: 10.08 radians/second Test Status PASSED 1.00 kPa for original binder and 2.20 kPa after RTFO aging. Fatigue cracking is governed by limiting Gtsin 8 of pressure aged material in the PAV) to values tess than 5000 kPa atthe test temperature. More information describing how G* and Bare used inthe Superpave binder specification is provided in Chapter 4 Table 3.2 shows the format of test information for one DSR system, > > Rotational Viscometer A rotational viscosity testis used to determine the flow characteristics (of the asphalt binder to provide some assurance that it can be pumped and handled atthe hot mixing facility. rotational coaxial cylinder viscometer as shown in Figure 3.12 and described in AASHTO T316, Viscosity Determination of Asphalt Binder Using Rotational Viscometer, is necessary to evaluate the various types of asphalt binders. Unlike capillary tube viscometers, the sermon asmassonom ners Brookfield viscometer readout control keys, spindle extension temperature controller thermo- container Figure 3.12 Rotational Viscometer rotational viscometers have larger clearances between the components and therefore, ae applicable to modified and unmodified asphalt aa sional viscometer automatically calculates the viscosity at the ‘ex temperature, The rotational viscosity i determine by measuring He Me eee vo maintain a constant rotational speed ofa cylindrical sae or erged in an asphak binder sample ata constant temperature Cre See STs true is direc related tothe binder viscosty, which fs calouted torque Figure 3.13 Rotational Viscometer Operation sample. sample spindle chamber PD sceve snnuramom ree -—--ssssee ee www ewww eweweseeeweeweweeee awromatically by the viscometer. Since this binder viscosity is used to ensure thet the asphalt is fluid enough at normal operating temperatures to pump and mix with aggregate, it is measured on original or “tank” asphalt. The viscometer can also be used to develop temperature-viscosity charts for ceximating mixing and compaction temperatures for use in mixture design, ‘Specimen Preparation. Approximately 30 grams of binder are heated in an oven until sufficiently fluid to pour. The sample should be stired ‘occasionally during heating to remove entrapped air. Asphalt is weighed into the preheated sample chamber. The amount of asphalt used is typically 8 to 11 grams and varies with the size of spindle. The sample chamber containing the binder sample is placed in the preheated thermo-container, the preheated spindle is lowered into the sample, and the binder is ready to test when the temperature stabilizes, usually within 30 minutes. Test Procedure. The apparatus used to measure rotational viscosity consists of two parts: rotational viscometer and the temperature control system, ‘The rotational viscometer consists of a motor, spindle, control keys, and digital Rotational Viscometer readout. ‘The motor powers the spindle rotation through a torsional spring, “The spring is wound as the torque increases. “The torque in the spring is measured by a rotary transducer, For specification testing, the motor is set to feat 20 pm. ‘The spindle resembles a plumb bob and spindle rotation is resisted by the viscous binder, Many spindles are available for the rotational viscometer, the proper spindle is selected based on the viscosity of the binder being tested. The temperature control system consists of the sample chamber, thermo-container, and temperature controller. The sample chamber is a stainless steel or aluminum cup; the thermo-container holds the sample chamber and consists of electric heating elements that are used to maintain ot cchange the test temperature. The controller allows the test temperature to be set at the required 135°C. ‘To function properly, the viscometer and thermo-container must be leveled using bubble levels and leveling screws. Control keys are used to input test parameters such as spindle number, set rotational speed, and turn. the motor on and off. The spindle is lowered into the chamber containing the hhot sample and the spindle is coupled with the viscometer ‘A waiting period of about 15 minutes may be needed to reach a uniform sample temperature of 135°C. During this period, the viscometer motor is tumed on and the viscosity reading and the percent torque (should be berween 2 and 98 percent) can be observed on the digital display. If the percent torque is outside of this range, a different size spindle will be required. {As the temperature equalizes, the viscosity reading will stabilize and test results axe recorded, ‘The digital display is set to show the information that is needed for the repon: viscosity, test temperature, spindle number, and speed. Three viscosity readings are recorded at minute intervals, Figure 3.15 shows the four possible displays; only the upper lft item in the display changes, ~~ oe P3765 SP21) [%6.0 SP21 2oRPM__135.0C} | 20RPM 135.0 SREB SP21) [SS25.5 SP21 2ORPM__136.0C J fl 20RPM 135.0. \ \ hear ate sea/srogs mots ‘erect rs) (Gimecon vent Figure 3.15 Rotational Viscometer Displays > SUPERPAVE ASPHALT BINDER TESTS It may be desirable to determine the binder viscosity at tempenstures other than 135°C. For example, most agencies use equiviscous temperatures for mixing and compaction during mix design. Regardless of the grade used, the binder temperatures are adjusted to obtain the same specified range of binder viscosity when mixing with aggregate and compacting specimens in the laboratory. Therefore, a viscosity-temperature relationship for the binder is needed, To accomplish this, the temperature controler is reset to a higher desired temperature (e.g, 165°C), and the testis performed as before. Data Analysis. ‘The viscosity at 135°C is reported as the average of three readings. ‘The digital output of some rotational viscometers isin units oF centipoise (cP) while the Superpave binder specification uses Pascal-seconds, Pas, The conversion used is 1000 eP = 1 Pars. Therefore, to obtain the viscosity in Pas, the rotational viscosity in cP is multiplied by 0.001. As mentioned previously, the test temperature, spindle number, and speed are also required report items. The Superpave binder specification requirement of ‘4 maximum of 3 Pas is applied at the discretion of the specifying agency and may be waived if the binder supplier guarantees that the binder can be handled and pumped at the necessary temperatures. >> Bending Beam Rheometer Asphalt binders at low temperatures are too stiff to reliably measure properties using the parallel plate geometry of the DSR. ‘Therefore, SHRP researchers developed the bending beam rheometer (BBR) to accurately sermeve surnames evaluate binder properties at low pavement temperatures. Used together, the ‘dynamic shear and the bending beam tests provide siffness behavior of ‘asphalt binders over a wide range of temperatures. Although stiffness can also he used to estimate failure or strength properties, for some asphalt binders (especially moxified ayphalts,) the relationship between stiffness and strength properties is nox well known. This is why some researchers believe that the alternate procedure for determining a critical low cracking temperture specified in AASHTO MP-la represents a more rigorous approach than simply using. [BBR data, In the MP-la alternate procedure, an additional tes, the direct tension text, is conducted to measure strength and ability to stretch before breaking (strain at failure.) The direct tension testis covered in a later section, “The BBR is used to measure how much a binder deflects or creeps under a constant load at a constant temperature. The BBR test temperatures are related to 2 pavement’s lowest service temperature, when the asphalt binder acts more like an clastic solid. Furthermore, the testis performed on ‘binders that have been aged in both a rolling thin film oven and the pressure aging vessel. Therefore, the test measures the performance characteristics of binders as if they had been exposed to hot mixing in a mixing facility and some in-service aging, ‘The BBR gets its name from the test specimen geometry anc! loading, method used during testing. The key elements of the BBR (Figure 3.16) are a loading frame, temperature-controlled fluid bath, and the computer control and ‘data acquisition system. A blunt-nosed shaft applies load to the midpoint of the simply-supported asphalt beam. A load cell is mounted on the loading. aluminum mold ‘Figure 3.17 Rubber O-rings [ Aluminum Mold Assembly 12.7 mm. acetate I strips asphalt beam 6.35 mm. -—-—s-seseeeeeeeseueuweu” f{< SUPERPAVE ASPHALT BINDER TESTS screen for observation. After 240 seconds, the test load is automatically removed and the sheometer software calculates creep stifiness and creep rate. Data Analysis, Beam theory is used to obtain creep stifiness of the asphalt in this test. The equation for calculating creep stiffness, $C, is; where, ‘St)= cep stnass (MPa) at time, P= applied constant load, N, | stance between beam supports, 102 mm, beboam wich, 125mm, hh=beam thickness, 6.25 mm, and {it = detection (mm) atte, ‘The BBR software makes this calculation using deflection versus time. Figure 3.19 shows a sample graph and the procedure used to obtain data (0, ‘The desired value of creep stiffness is when the asphalt has been loaded for two hours at the minimum pavement design temperature However, using the concept of time-temperature superposition, SHRP sesearchers confirmed that by raising the test temperature 10°C, an equal creep stiiness can be obtained after only a 60 second loading. The Obvious benefit §s that a test result can be measured in a much shorter testing time. Using the > Direct Tension Tester "Numerous past studies have shown that there is a strong relationship between the stiffness of conventional (unmodified) asphalt binders and the amount of stretching they undergo before breaking. Asphalts that undergo considerable stretching before failure are called “ductile; those that break without much stretching are called “brite”, Its important that an asphalt binder be capable of a minimal amount of elongation. ‘Typically, siffer asphaks are more brittle and softer asphalts are more ductile Greep stifiness as measured by the BBR is not adequate to completely characterize the capacity of asphalt to stretch before breaking. For example, some binders exhibit high creep stifness, but can also stretch farther before breaking. Consequently, SHRP researchers developed a specification system to accommodate these stiff but-ductile binders. In the Superpave binder SUPERPAVE ASPHALT BINDER TESTS <<< ss e-eweweww eee eeeeeweeesesewewewuews Proj: SPEXt Target Temp: -18.0¢ Conf Test: 2.102408, Oper: PAT. Actual Temp: -18.0 C Date: 04/05/03 Spee: IN-10 ‘Soak time: 60.0 min Load Const: 0.243112 Time: 15:53:47 ‘Beam Width: 12.70 mm Defl Const: 002417 Date: 04/05/03 Thickness: 6.35 mm Date: 04/05/03 File: 0405037, DAT RESULTS: Time, Force, Defl, Measured Estimated Diff, % —mvalue se’ ON mm —Stifness—_Stifness MPa MPa 8 965818774148 4143-1208 261 15 96782031 3498349700374 277 30.9684 27302860888 2003295 6) 9656 336423152823 3678313 120 96654152 187.7 = 18599626331, 20 966253271482 1469 An? 348 Regression Coefficients b=-2080 c=.02949 _R*2=.999821 Deflection, mm, 0.35 ° 250 Figure 3.20 Typical BBR Deflection and Load Output o 250 SUPERPAVE ASPHALT BINDER TESTS SUPERDAVE ASPHALT BINDER TESTS Figure 3.22 Determination of Failure Strain in the Direct ange in length (A : tao strain 6) = ces gauge lens, Tension Test the specimen breaks. Failure stress (6) is the failure load divided by the Cniginal cross section of the specimen (36 mim). Figure 3.23 illustrates the stress-strain curve from a test conducted using the DTT. Although stifiness has been used to estimate failure or surength properties, for some asphalt binders (especialy modified asphalts) the relationship between stifiness and strength properties is not well known. ‘This is why some researchers believe that the alternate procedure for determining a critical low cracking temperature specified in AASHTO MP-1a represents a more rigorous approach than AASHTO M320. In the MP-1a altemate procedure, the BBR «dats is used at ll loading times, and with at least one temperature, to generate Load (P) of eross section (A) = 36mm? Constant Strain Rate Figure 3.23 Determination of Failure Stress in the Direct Tension Test scree asm asoes ress 1a predicted thermal stress curve as a function of temperature, The DTT is used to determine the failure stress in an asphalt binder specimen when tensile fracture occurs. ‘The failure tensile stress is plotted to determine whether, in the case of a verification test, a binder meets low temperature requirements oF, in the case of a classification, the temperature at which the thermal stress curve from the BBR data exceeds the failure sues from the DTT, This intersection is termed the “critical cracking temperature” in the AASHTO MP-la specification. asphalt poured here Figure 3.24 Direct Tension Specimens ‘Specimen Preparation. Direct tension specimens are formed in aluminum ‘molds as shown in Figure 3.24. Six individual specimens are formed One test result is produced from an average value of four specimens ater discarding the two lowest stresyat failure values. End insert, made from phenolic G-10 and containing a 10-mm diameter hole inset with a sainless See! ring, are placed in the ‘mold, the assembly and heat sink are preheated, and hot asphalt binder is then poured between the inserts in the space shown in Figure 3.24. Test specimens are poured and allowed to remain in the heat sink fora short period of ime G minutes), “Test specimens have a mass of approximately 3 grams and are 100 mm Jong, including the end inserts. The dimensions ofthe plastic inserts and the formed binder test specimen are shown in Figure 321. ‘The nominal binder cross section in the mictength portion is 6 mm by 6 mm. A 12mm radius is used to sradually widen the specimen to 20 mm at the end insert. ‘After the specimens are poured, they are allowed to cool at ambient temperature for 30-60 minutes before trimming (Figure 3.25). After trimming, the specimen is again allowed to cool for 10-15 minutes before demokding and Figure 3.25 Direct Tension Test Specimen Before and After ‘Trimming ‘otesting, ‘Ths testing tolerance requires the equipment and specimen preparation to be carefully coordinated > > > > > > > > > 2 > » % conditioning inthe fui bath a the specified test temperature for 60 minutes prior > > ® Test Procedure, Although the direct tension concept is simple, the > Catipment used requires complex features to accurately measure the very small stains involved in the test. The DTT consists of three components: > (Q) an electromechanical test device to apply load, > (2) a transducer system to measure and control grip separation, and ) an environmental control system, usualy in the form of a » controlled temperature liquid bath > ‘The loading device must be capable of applying atleast 500 N with a jy ™aximum crosshead travel of 20 millimeters at arate of 1.0 mm/minute ‘Minimum and maximum deformation rates are required t0 be 0,001 9 mm/minute and 600 mm/minute, respectively, ‘The measurement system for » determining the elongation must have a resolution of 0.001 millimeter. A key feature of the testing machine is the gripping system used to ) atch specimens to the alignment rods that apply the tensile load (Figure 3.26). » The grips have a ball joint connection that ensures no bending within the binder test specimen. q ‘The environmental control system includes an environmental chamber 9 orliguid bath, and mechanical refrigeration unit capable of producing and » Precisely maintaining temperatures as low as-36°C. Most available liquid i cooling units have been found to control adequately 10 only -36— with the 9 assistance of auniliary stiring in the test chamber. In the case of an »_eAvironmental chamber, air is used as the cooling medium, If a liquid bath is Used, Solution of potassium acetate and water is used as the cooling medium. » SUPERPAVE ASPHALT BINDER TESTS Side View Figure 3.26 Direct Tension Test Loading System [At the end of the conditioning period (60 + 5 minutes), each of the six specimens is tested individually. Before the test begins, the equipment is inkialized and a small seating load is applied after mounting the specimen on the end grips. A tensile load is applied by pulling one end ata strain rate of 3% per minute (1.0 mmv/minute) until the specimen fails. A test typically requires less than a minute from load application until failure, A test result is usually considered legitimate when fracture occurs within the 18 mm effective ‘gauge section, which is the midsection of the specimen with the constant cross-section, A specimen that breaks near the end insert should be reported asan “end break”, Data Analysis. Afr dropping the two lowest stress-at-failure values, a sirgle test result is generated by averaging the results from the four remaining test specimens. This practice is based on the concept that itis much more ‘common in direct tension tests to have an early fracture because of specimen a Figure 3.27 Thermal Stress Curve for Cessupecomtenr Determining Critical Cracking BR 6 ‘Temperature ‘Temperature, C ‘Thegmal Suess Curve (from BBR) 5 i i = ‘ procedural imperfections than an exceptionally "strong test specimen that does not fracture early, Other analysis procedures may dictate different ‘methods for evaluating the quality ofthe test data and its suitability for inclusion in the general average. A single test result consists of the average strain to failure of multiple DIT specimens , reported to the nearest 0.01 percent and the average stress at failure reponed 10 the nearest 0.01MPa. ‘Table 34 shows typical DTT test output. In Table 3.4 after dropping the two low values, the average stain at failure is 0.979% and the peak stress is 4.61 MPa. If the direct tension testis being performed for AASHTO M320, only the average strain at failure is needed to determine specification compliance. In AASHTO MP-la, the peak ‘ress is used as an input to calculate the critical cricking temperature of the asphalt binder, To determine critical cracking temperature, BBR test data is used to generate a thermal stress curve asa function of temperature, ‘The DTT peak stress is then used to determine when tensile fracture occurs in an asphalt binder specimen. ‘The peak tensile stress is ploted to determine at what temperature the thermal stress curve from the BBR data exceeds the tensile stress from the DTT. This intersection is termed the “critical cracking temperature” in the AASHTO MP-1a specification (Figure 3.27), and is used as an alternate method of determining the low temperature grace of an asphalt binder. Although not used to determine specification compliance, these values are also required reporting items: ‘Test temperature to the nearest 0.1C, * Average rate of elongation to the nearest 0.01 mm/min, + Peak load to the nearest N, and * Type of break observed (brite, britle- ductile, oF no break). suranve secrnmon res The Superpave asphalt binder specification is intended immprove performance by limiting te potential for the asphal bine to contnbute toward permanent deformation, low =o temperature cracking and fatigue cracking in asphalt pavements = ‘The specification provides fortis improvement by designating ‘various physical properties with the equipment described in Chapter 3. This section briefly explains each ofthe new test parameters as they relate to pavement performance ‘One important distinction between typical asphalt = specifications and the Superpave speciation i the over foomat ofthe requirements. ‘The required physical properties Pr remain constant for all of the performance grades (PG). However, the temperatures at which these properties must be reached vary depending on the climate in which the binder is expected to be used. For example, the partial view of the Superpave specification format shown in Figure 4.1 shows that PG 52-40 grade binder is designed to sustain the condlions of an environment where the average sevenwoos momen ans nom ner Se Figure 4.2 Superpave fasgrnrerg Specification Rutting Factor Requirements "Ch mem 40 8s ‘Spec Requirements to are Address Rutting his, sim ‘hn Minne, 220 Pa Teen @ 1radee,C ‘on a Moving Film of Aspbait (Rolling Thin Film Oven Test). The mass loss for any grade must not exceed 1.00 percent, Like permanent deformation, G* and & are also Used! in the Superpave binder specification to help ‘control the fatigue of asphalt pavements, Since fatigue generally occurs at low to moderate pavement temperatures after the ‘pavement has been in service for a period of time, the specification addresses these properties using binder aged in both the RTFO and PAV. ‘The DSR is again used to generate G* and 8. However, instead of dividing the two parameters, the two are multiplied to produce a factor related to fatigue. ‘The fatigue cracking factor is G*sin 8, which is stated °G star sine deta” Itis the product of the complex shear modulus, G*, and the sine of the phase angle, 3. The Superpave binder specification has a maximum value of 5000 kPa for G'sin 8 CFigure 43). Low values of G* and 6 are considered desirable ateibutes from the standpoint of resistance to fatigue cracking. Thus, the Superpave specification promotes the use of compliant, elastic binders (PAV aged) to address fatigue cracking When the pavement temperature decreases, asphalt concrete shrinks. As the temperature drops, the asphalt binder contracts to a much greater degree than the aggregate in an asphalt concrete pavement, thus causing thermal stresses to develop in the ‘pavement. When these stresses exceed the tensile strength of the asphalt ‘mixture then a low temperature crack develops. Em” AV Ating Temp. C Pe 5 een secoe Specification requirement |_Tetenpa onde. | to address fatigue cracking Pins arden creep sites ‘Msi, 200 MPa Figure 4.3 Superpave vale, Mini, 0.300, tate asdee.c Specification Fatigue Cracking Diret Tension Tittore rome’ Factor Requirements pv tung Teng. Pra ein 5004s ‘Specification requirements Fan tenpa@ ordi. to address low temperature Posie cracking asim 30 Pa sae Sime, 300 Tertennawse Figure 4.4 Superpave Specification Low Temperature Requirements In AASHTO M320, the main way of examining the propensity of an asphalt binder to develop thermal stresses at a specified temperature is to use lava generated from the bending beam theometer (BBR). ‘The BBR is used to apply a small creep load, at a specified temperature, to an asphalt binder sample molded into a beam shape and then measure the deformation under load as a function of time. By knowing the load and the beam dimensions, the creep stiffness (binder’s resistance to load) can be calculated at any time. Ifthe ‘siifiness is too high, the asphalt binder sample will behave in a britle manners, indicating that cracking is more likely to occur at that temperature. To prevent thermal cracking, creep stiffness has 2 maximum limit of 300 MPa (Figure 4.4. Since low temperature cracking usually occurs after the pavement has been in ‘service for some time, this part of the specification addresses these low temperature properties using asphalt binder that has been aged in both the CFO and PAV. PAVEMENT PERFORMANCE IN THE SUPERPAYE BINDER SPECIFICATION -—--ssesse eee ee eweeee ‘The rate at which binder stifiness changes with time at low temperatures is regulated through the m-value. In the AASHTO M320 specification, a high m-value is desirable, because as the temperature decreases and contraction occurs, the asphalt binder will respond as a material that is less Stil, ‘This decrease in sifiness (ratio of stress over strain) leads to smaller tensile stresses in the asphalt binder and less chance for low temperature cracking. minimum m-value of 0300 aftr 60 seconds of loading is required by the Superpave asphalt binder specification (Figure 4.4). Past studies have also indicated that ia binder can stretch at least one Percent of its original length during this thermal contraction period. cracks are Jess likely t0 occur. As a result, the dlect tension test (DTD is included in AASHTO MB20 as an alterative requirement to simply using ereep stiffness ard m-value. ‘The DTT is used to pull an asphalt binder specimen in tension at 2 very slow rate, simulating the pavement condition as shrinkage occurs. ‘The amount of strain that occurs before the specimen breaks isthe failure, or peal, stain. To meet the requirements of AASHTO M320, this failure strain must be 4 minimum of 1.0 percent (Figure 4.4). In this specification, the DTT requirement only applies to asphalt binders that have a creep stiffness greater than 300 MPa and less than 600 MPa, with an m-value of 0.300 or greater, IF the creep stifness at the specified temperature is 300 MPa or less, then the dltect tension testis not require. Although sifiness can also be used to estimate failure or strength Properties, for some asphalt binders (especially mocified asphalt) the Teationship between stifness and strength properties is not well known, ‘This is why some researchers belive thatthe alternate procedure for determining a PAVEMENT PERFORMANCE IN TE SUPERPAVE BINDER SPECIFICATION -—---sseew eevee neueueue Contrary to the previous grading systems, the Superpave binder specification is theoretically based directly on performance ‘ther than on empirical relationships between basic physical properties and observed performance, Performance graded binders are selected based on the climate in which the pavement will serve. Unlike all other systems, the physical property requirements are constant among all performance grades, The distinction among the various binder grades is the specified ‘minimum and maximum temperatures at which the requirements must be met. For example, a binder classified as a PG 58.34 ‘means that the binder will meet the high temperature physical Property requirements up to a temperature of 58°C and the low temperature physical property requirements down to -34°C, AASHTO M320 contains a listing of the more common PG grdes, However, the PG grades are not limited to those given classifications. In actuality, the specification temperatures are Unlimited, extending unbounded in both directions. The high and low temperatures extend as far as necessary in the standard sixdegree increments. For example, even though a PG 58-10 is not shown, it exists as a legitimate grade in the system, Even with binder grades classified according to high and Jow temperature categories, more information is needed to select a binder for a particular location. The LTPPbind Superpave sofiware, developed by the Federal Highway Administration, ‘assists users in Selecting binder grades. The Superpave sofiware cortains three methods by which the user can select an asphalt binder grade: Geographic Area: An agency would develop a map showing binder grade to be used based on weather and/or policy decisions Pavement Temperature: The designer determines design pavement temperature. Air Temperature; ‘The designer determines design air temperatures, which are converted to design pavement temperatures “The Superpave soliware assists in the thd method by providing a database (of weather information for 6092 reporting weather stations in the US and Canada that allows users to select binder grades forthe climate atthe project location. For ‘each year that these weather staions have been in operation, the hottest seven- BINDER GRADE SELECTION 98% reliability 32 36 7-Day Maximum Air Temporature, Cleveland, OH 32 Air Temperatures, Cleveland, OH 40 -30 -20 -10 0 10 20 30 40 50 60 Figure 5.2 Example High and Low Air Temperature Variations probability in a single year thatthe actual temperature (one-day low or seven- day average high) will nt exceed the design temperatures, A higher reliability ‘means lower risk, For example, consider summer air temperatures in Cleveland, Ohio, which has a mean seven-day maximum of 32°C and a standard deviation Of 2. In an average year, there is a 50 percent chance that the seven-day, ‘maximum air temperature will exceed 32°C. However, assuming a normal Statistical frequency distribution, there is only a two percent chance tat the seven-day maximum will exceed 36° (mean plus two standard deviations); therefore, as shown in Figure 5.1, a design air temperature of 36°C will provide 98 percent reliability. BINDER GRADE SELECTION [BINDER GRADE SELECTION temperature grade happens to match the design temperature, PG 52. Using the same reasoning, the low temperature grade is a PG -16 to atain 50 percent relabilty. Coincidentally, the low temperiture grade again happens to match the design temperature, -16. As shown in Figure 5.4, to obtain at least 98. percent reliability, it is necessary to select a high temperature grade of PG 58 10 protect above 56°C andl a low temperature grade of PG -28 to protect below 23°C. in both the high and low temperature cases of the PG 58-28 binder ‘gre, the actual reliability exceeds 99 percent because of the “rounding up" caused by te six degree difference between standard grades, This “rounding up" introduces conservatism into the binder selection process. Another possible source of conservatism occurs when considering the sane steps encountered during asphalt binder tet classification, Although a specific binder may pass all ofthe criteria when tested at lower or higher temperatures, it will nevertheless be classified by “rounding down’ to the next "sixdegree” step of the grading system. ‘The net result is that a significant facior of safety is included in the binder selection scheme. For example, iti possible that the PG 52-16 binder, selected previously for a minimum of 50 percent reliability for Cleveland may actually have been graded 2s a PG 56-20, hhad such a grade existed. Users of this temperature-based stepped grading system for binder selection should recognize that considerable safeguards are already included in the process. Because of these factors, it may not be necessary or cos effective to require indiscriminately high values of reliability cr abnormally conservative high of low temperature grades PG 58-28 40 -30 -20 -10 0 10 20 30 40 50 60 70 Figure 5.4 Superpave Binder Grade Selections for Cleveland soem ose sescnonYou might also like
