Professional Documents
Culture Documents
1 Welcome
Welcome to the Washington State Department of Transportation's Pavement Guide Interactive. This
document (or "Guide" for short) is a multimedia CD-ROM based document whose primary purpose is to
provide a general pavement overview covering all aspects from materials to design to construction to
maintenance. Many sections offer supplemental information beyond this general overview in a series of
in-depth or WSDOT practices pages. It functions as a “Web site” that resides on a CD-ROM and
requires only a PC/Mac and minimal freeware to access the information. All total, it consists of 275
Web pages, 2500 images, 50 animations, 14 videos, and 11,000 hyperlinks. The following subsections
describe the target audience and intended function of the WSDOT Pavement Guide Interactive.
1.2 Function
The WSDOT Pavement Guide Interactive functions both as a learning tool and a ready reference; users
can learn about pavements as well as use it as a reference to look up typical values, methods, practices
and resources. Learning is comparable to that of an instructor-led classroom course and locating
information for reference is relatively straightforward. Specific motivation for the Pavement Guide
Interactive is multifold:
● Provide a multimedia, interactive product that will benefit those who want basic pavement
knowledge. As such, this Guide can be useful for Federal, State and local agencies,
inspectors, design consultants, contractors, operators and students.
● Provide in one interactive product, the major features associated with designing and delivering
pavements.
● Improve the connection between different pavement aspects such as mix design, structural
1. A broad and well-rounded introduction to pavement in the areas of pavement types, materials,
design considerations, mix design, structural design, construction, quality assurance,
specifications, evaluation, maintenance and rehabilitation, and pavement management.
2. The option to explore in-depth information and WSDOT practices where applicable.
3. The option to investigate other sources of information on the Web or in print form through the
use of extensive references.
After covering a pavement concept in the WSDOT Pavement Guide Interactive, you should, in general,
be able to:
The WSDOT Pavement Guide Interactive is also an extensive pavement reference. Users can browse
the Guide or use the search tool to look up typical values, methods, practices and resources. It can also
be used to assemble multimedia pavement presentations or lectures through simple cut-and-paste of text,
pictures, videos and animation.
The rest if this introductory module provides a bit of pavement history as well as a brief overview of
pavements today.
2 History
In its most general sense, a road is an open, generally public Major Topics on this Page
way for the passage of vehicles, people, and animals. The
earliest human road builders predate recorded history by 2.1 Roman Roads
thousands of years. With the advent of modern man, road
building - the purposeful construction of general public ways 2.2 Telford Pavements
- became a common sign of an advancing civilization. 2.3 Macadam Pavements
Covering these roads with a hard smooth surface (pavement)
helped make them durable and able to withstand traffic and 2.4 Early Bituminous Pavements
the environment. Some of the oldest paved roads still in
existence were built by the Roman Empire. 2.5 Early Portland Cement Concrete
Pavements
2.6 Summary
● Nucleus. A kind of base layer composed of gravel and sand with lime cement.
● Rudus. The third layer was composed of rubble masonry and smaller stones also set in lime mortar.
As can be seen, Roman pavements were quite thick - on the order of almost 0.9 m. Roman road construction was not
inexpensive. Updated construction estimates of the Appian Way are about $2,000,000 per km (updated estimates following
Rose, 1935 and Leger, 1875).
Figure 1.3: Typical Telford Road (after Collins and Hart, 1936)
general" of the Bristol roads and was then able to use his design on numerous projects. It proved successful
enough that the term "macadamized" became a term for this type of pavement design and construction. The
term "macadam" is also used to indicate "broken stone" pavement (Baker, 1903). By 1850, about 2,200 km
(1,367 miles) of macadam type pavements were in use in the urban areas of the UK. MacAdam realized that
the layers of broken stone would eventually become "bound" together by fines generated by traffic. With the
introduction of the rock crusher, large mounds of stone dust and screenings were generated (Gillette, 1906).
The increased use of these fines resulted in the more traditional dense graded base materials. The first
macadam pavement in the U.S. was constructed in Maryland in 1823.
Figure 1.5: Typical Macadam Road (after Collins and Hart, 1936)
streets). Coal tar, the binder, had been available in the U.K. from about 1800 as a residue from coal-gas
lighting. Possibly this was one of the earlier efforts to recycle waste materials into a pavement!
Soon after the Nottingham project, tar macadam projects were built in Paris (1854) and Knoxville, Tennessee
(1866) (Hubbard, 1910). In 1871 Washington, D.C. extensively used a "tar concrete" for road construction.
Sulfuric acid was used as a hardening agent and various materials such as sawdust, ashes, etc. were used in the
mixture (Hubbard, 1910). Over a seven-year period, 630,000 square meters (156 acres) were placed. In part,
due to lack of attention in specifying the tar, most of these streets failed within a few years of construction.
This resulted in tar being discredited, thereby boosting the asphalt industry (Hubbard, 1910). However, some
of these tar-bound surface courses in Washington, D.C., survived substantially longer - about 30 years. For
these mixes, the tar binder constituted about 6 percent by weight of the total mix (air voids of about 17
percent). Further, the aggregate was crushed with about 20 percent passing the 2.00 mm (No. 10) sieve. The
wearing course was about 50 mm (2 inches) thick. Hot tar paving products have not been used in the U.S. for
many years.
As a side note, the term "Tarmac" was a proprietary product in the U.K. in the early 1900s (Hubbard, 1910).
Actually it was a plant mixed material, but was applied to the road surface "cold." Tarmac consisted of
crushed blast furnace slag coated with tar, pitch, portland cement and a resin. Today the term "tarmac" is
generic and generally refers to airport pavements (however, inappropriately).
3. Mix the asphalt binder and aggregate together using a tractor-pulled disk or harrow, windrow the
mixed material in the center of the road, turn it, then redistributed across the road and smooth it.
4. Compact the resultant wearing course until no movement is discernible under the roller wheels.
5. After a few weeks to several months, spread a cover coat of fine aggregate over the surface and
apply a seal coat.
These pavements were not true hot mix asphalt pavements because the asphalt was often applied as an
● A wearing course 40 to 50 mm (1.6 to 2 inches) thick composed of asphalt cement and sand.
● A binder course about 40 mm (1.6 inches) thick composed of broken stone and asphalt cement.
● A base layer of hydraulic cement concrete or pavement rubble (old granite blocks, bricks, etc.).
Generally, this layer was 100 mm (4 inches) thick for "light" traffic and 150 mm (6 inches) thick for
"heavy" traffic (Baker, 1903).
Sheet asphalt became popular during the mid-1800s with the first ones being built on the Palais Royal and on
the Rue St. Honore in Paris in 1858 (Abraham, 1929). The first such pavement placed in the U.S. was in
Newark, New Jersey, in 1870. Sheet asphalt pavements are no longer built today.
In 1910 in Topeka, Kansas, a court ruling found that asphalt concrete mixes containing 12.5 mm (0.5 inch)
maximum size aggregate did not infringe on Warren's patent (Steele and Himmelman, 1986). Thus, HMA
mixes thereafter became oriented to the smaller maximum aggregate sizes. A typical "Topeka mix" consisted
of 30 percent graded crushed rock or gravel (all passing 12.5 mm (0.5 inch) sieve, about 58 to 62 percent sand
(material passing 2.00 mm (No. 10) and retained on 0.075 mm (No. 200) sieve) and 8 to 12 percent filler
(material passing 0.075 mm (No. 200) sieve). This mixture required 7.5 to 9.5 percent asphalt cement.
In 1910, Edwin C. Wallace, a retired employee of Warren Brothers, invented Warrenite-Bitulithic. It consisted
of an approximately 25 mm (1 inch) thick layer of "Finely divided mineral matter coated with bitumen rolled
into a lower layer of large stone, small stone, stone dust and bitumen" (Wallace, 1910). This was basically a
sheet asphalt wearing course over hot, uncompacted bitulithic. By adding the thin wearing course, the large
aggregate of the Bitulithic mixes were not exposed directly to heavy, steel rimmed wheels that could cracked
the aggregate and result in mix degradation. By 1920, Warren's original patents had expired in the U.S.
(Oglesby and Hewes, 1962) but the legacy of the Topeka mix lived on as reflected by the U.S. tendency
towards finer mixes.
During the 1920's and 1930's, PCC pavement was usually constructed directly on the
underlying soil. This practice was satisfactory until the 1930s when highway truck traffic
increased to the point where pumping distresses began to appear on roadways carrying heavy
truck traffic. Research into this phenomenon resulted in recommendation of a non-pumping
layer called a subbase (although now this layer is often referred to as the base layer) be placed
under the PCC slabs. Gravel, crushed stone, and slag were commonly used as subbase
material. In the late 1940's, California began using cement-treated granular bases under
concrete pavements. This practice quickly spread to other states.
In 1946, two Iowa highway engineers, James W. Johnson and Bert Myers,
conceptualized the slip form paver. In 1949, the Iowa Highway Department
constructed the first slipformed roadway, a 3 m (9 ft.) wide, 150 mm (6 inch) thick
section of county road. By placing two lanes side-by-side, a typical 6 m (18 ft.) wide
county road could be built. The paver attached to a ready mix concrete truck, which
would discharge its load into the paver, then pull the paver forward. In 1955, Quad
City Construction Company developed an improved, self-propelled, track-mounted
slipform paver capable of placing 8 m (24 ft.) wide slabs up to 250 mm (10 inches)
thick. In just a few years, several equipment manufacturers were marketing slipform
pavers capable of placing concrete up to four lanes wide.
During the same period, central mixing replaced on-site mixing on most paving jobs.
Evaluations by several agencies showed that central-mixed concrete could be hauled
from the mixer to the slipform paver in non-agitating dump trucks with no loss in
workability or quality.
It was also during the late 1940's and early 1950's that paving contractors began
experimenting with sawed concrete joints. Previously, joints were formed in the plastic
concrete with jointing tools. These hand-formed joints often created a rough ride. After
early attempts in Kansas and California, sawing was used on several projects in 1951,
and soon became a standard construction method.
2.6 Summary
Road and pavement building has often been used as a benchmark of a civilizations advancement. The quality
and strength of many of the ancient roads has helped them survive to this very day. The Via Appia in Rome is
now over 2,300 years old and is still used today. As the use of slave labor declined, smaller more economical
roads, such as Telford and Macadam roads, began to arise. Around the beginning of the 19th century, binding
agents began to be used to assist aggregate cohesion and improve the durability of roads. By the end beginning
of the 20th century, the two principal pavement types, flexible and rigid, had taken on many of their modern
qualities and were being built throughout the U.S.
3 Pavement Overview
Major Topics on this Page
2. Smoothness. Pavement material can be placed and maintained much smoother than in situ material. This
helps improve ride comfort and reduce vehicle operating costs.
3. Drainage. Pavement material and geometric design can effect quick and efficient drainage thus
eliminating moisture problems such as mud and ponding (puddles).
● Portland cement concrete (PCC). A combination of aggregate, water and portland cement to form a hard,
strong construction material when set. PCC is known by several names including "cement" and
"concrete". The WSDOT Pavement Guide Interactive makes a conscious effort to consistently refer to
this material as PCC. PCC is distinguished by its design and production methods.
● Concrete. Term often used to describe portland cement concrete. However, in its more generic form
"concrete" refers to any conglomeration or coalescence of materials usually held together by a binding
substance. Thus, asphalt concrete and portland cement concrete are two types of concrete with the
"asphalt" and "portland cement" referring to the binding material.
Much of this country relies on paved roads to move themselves and their products rapidly and reliably throughout the
transportation system. Currently, there are over 3.96 million public centerline road miles (8.28 million lane miles) in
the U.S. and of this, 2.50 million miles (or about 63 percent) are paved (FHWA, 2002). Pavements can be generally
classified into two broad categories:
● Flexible pavements (see Figure 1.6 and 1.8). These are asphalt pavements (sometimes called bituminous
pavements), which may or may not incorporate underlying layers of stabilized or unstabilized granular
materials on a prepared subgrade. These types of pavements are called "flexible" since the total pavement
structure bends (or flexes) to accommodate traffic loads. Flexible pavements comprise about 82.2 percent
of U.S. paved roads.
● Rigid pavements (see Figure 1.7 and 1.8). These are portland cement concrete (PCC) pavements, which
may or may not incorporate underlying layers of stabilized or unstabilized granular materials. Since PCC
has a high modulus of elasticity, rigid pavements do not flex appreciably to accommodate traffic loads.
Rigid pavements comprise 6.5 percent of U.S. paved roads.
Figure 1.6: Flexible Pavement (SR 27) Figure 1.7: Rigid Pavement (I-90)
Figure 1.8: SR 395 South of Ritzville, WA Showing Both Flexible (Southbound - Left)
and Rigid (Northbound - Right) Pavements
The FHWA also identifies a third type of pavement, called a composite pavement. Composite pavements are
combination HMA and PCC pavements. Occasionally, they are initially constructed as composite pavements, but
more frequently they are the result of pavement rehabilitation (e.g., HMA overlay of PCC pavement). Modeling
these pavements depends on the composite action. For instance, an HMA overlay of rubblized PCC is typically
classified as a flexible pavement, while an HMA overlay of a PCC pavement with no fracture preparation typically
responds with rigid pavement characteristics (see Figure 1.9). Officially, the FHWA "composite pavement" category
is defined as a "mixed bituminous or bituminous penetration roadway" of more than 25 mm (1 inch) of compacted
material on a rigid base (FHWA, 2001). Tables 1.1 and 1.2 show a breakdown of U.S. roads.
Note: For the purposes of this Table, composite pavements are roughly defined as more
than 25 mm (1 inch) of compacted HMA on a rigid base. These would most likely be
classified as rigid pavements with a flexible wearing course.
Figure 1.11: Growth of Rural Average Daily Load and Traffic in the U.S. Since 1970 (from FHWA, 2000)
Washington State has 80,256 centerline miles of roadway, of which roughly 70 percent are paved. As of 2000, the
State population is just over 5.8 million (4.4 million or 76.5% of driver age and 4 million drivers licenses) with
over 5.2 million registered vehicles. These vehicles traveled approximately 54 billion miles in 2000 (an increase
of about 90 percent over the last 20 years) at an average of just over 10,600 miles per vehicle.
Note: For the purposes of this Table, only WSDOT roads were
considered. This represents about 7,000 centerline miles and
18,000 lane-miles of pavement.
Data sources:
WSDOT. (January 2001). Key Facts: A Summary of Transportation Information. WSDOT Finance and Administration Service
Center. Olympia, WA. http://www.wsdot.wa.gov/KeyFacts/default.htm.
WSDOT. (May 1999). Washington State Highway Pavements: Trends, Conditions and Strategic Plan. WSDOT Field Operations
Support Service Center, Materials Laboratory. Olympia, WA. http://www.wsdot.wa.gov/biz/mats/pavement/Pavement%20Plan.pdf.
What this means is that design, construction and maintenance practices must be conducted to the highest standard
possible because there is simply no room for error given the amount of money that is allocated and the amount of
traffic that must be accommodated.
One offshoot of this philosophy is an increased attention to pavement education and training. The better educated
and the more well-trained we are, the more likely we are to achieve and surpass even the highest standards; this
Guide was created as a small step.
1 Introduction
Basically, all hard surfaced pavement types can be categorized into two groups, flexible and rigid.
Flexible pavements are those which are surfaced with bituminous (or asphalt) materials. These can be
either in the form of pavement surface treatments (such as a bituminous surface treatment (BST)
generally found on lower volume roads) or, HMA surface courses (generally used on higher volume
roads such as the Interstate highway network). These types of pavements are called "flexible" since the
total pavement structure "bends" or "deflects" due to traffic loads. A flexible pavement structure is
generally composed of several layers of materials which can accommodate this "flexing". On the other
hand, rigid pavements are composed of a PCC surface course. Such pavements are substantially
"stiffer" than flexible pavements due to the high modulus of elasticity of the PCC material. Further,
these pavements can have reinforcing steel, which is generally used to reduce or eliminate joints.
Each of these pavement types distributes load over the subgrade in a different fashion. Rigid pavement,
because of PCC's high elastic modulus (stiffness), tends to distribute the load over a relatively wide area
of subgrade (see Figure 2.1). The concrete slab itself supplies most of a rigid pavement's structural
capacity. Flexible pavement uses more flexible surface course and distributes loads over a smaller area.
It relies on a combination of layers for transmitting load to the subgrade (see Figure 2.1).
Overall, it may be somewhat confusing as to why one pavement is used versus another. Basically, state
highway agencies generally select pavement type either by policy, economics or both. Flexible
pavements generally require some sort of maintenance or rehabilitation every 10 to 15 years. Rigid
pavements, on the other hand, can often serve 20 to 40 years with little or no maintenance or
rehabilitation. Thus, it should come as no surprise that rigid pavements are often used in urban, high
traffic areas. But, naturally, there are trade-offs. For example, when a flexible pavement requires major
rehabilitation, the options are generally less expensive and quicker to perform than for rigid pavements.
This section will discuss flexible and rigid pavements and the basic characteristics and types of each.
http://training.ce.washington.edu/WSDOT/Modules/02_pavement_types/02-1_body.htm4/2/2008 6:26:33 PM
2.2 Pavement Types - Flexible Pavement Basics
In order to take maximum advantage of this property, material layers are usually arranged in order of
descending load bearing capacity with the highest load bearing capacity material (and most expensive)
on the top and the lowest load bearing capacity material (and least expensive) on the bottom. This
section describes the typical flexible pavement structure consisting of:
● Surface course. This is the top layer and the layer that comes in contact with traffic. It may
be composed of one or several different HMA sublayers.
● Base course. This is the layer directly below the HMA layer and generally consists of
aggregate (either stabilized or unstabilized).
● Subbase course. This is the layer (or layers) under the base layer. A subbase is not always
needed.
After describing these basic elements, this section then discusses subsurface drainage and perpetual
pavements.
As seen in Figure 2.4, a flexible pavement structure can vary greatly in thickness. The signs on top of
the pictured cores indicate the State Route (SR) and the Mile Post (MP) where the core was taken. The
scale at the right edge of the photo is in inches.
1. Wearing Course. This is the layer in direct contact with traffic loads. It is meant to take the
brunt of traffic wear and can be removed and replaced as it becomes worn. A properly
designed (and funded) preservation program should be able to identify pavement surface
distress while it is still confined to the wearing course. This way, the wearing course can be
rehabilitated before distress propagates into the underlying intermediate/binder course.
2. Intermediate/Binder Course. This layer provides the bulk of the HMA structure. It's chief
purpose is to distribute load.
1. Aggregate. Base courses are most typically constructed from durable aggregates (see Figure
2.5) that will not be damaged by moisture or frost action. Aggregates can be either stabilized
or unstabilized.
2. HMA. In certain situations where high base stiffness is desired, base courses can be
constructed using a variety of HMA mixes. In relation to surface course HMA mixes, base
course mixes usually contain larger maximum aggregate sizes, are more open graded and are
subject to more lenient specifications.
1. Minimize the intrusion of fines from the subgrade into the pavement structure.
2. Improve drainage.
The subbase generally consists of lower quality materials than the base course but better than the
subgrade soils. A subbase course is not always needed or used. For example, a pavement constructed
over a high quality, stiff subgrade may not need the additional features offered by a subbase course so it
may be omitted from design. However, a pavement constructed over a low quality soil such as a
swelling clay may require the additional load distribution characteristic that a subbase course can offer.
In this scenario the subbase course may consist of high quality fill used to replace poor quality subgrade
(over excavation).
Researchers have used this idea as well as pavement materials research to develop a basic perpetual
pavement structural concept. This concept uses a thick asphalt over a strong foundation design with
three HMA layers, each one tailored to resist specific stresses (TRB, 2001):
1. HMA base layer. This is the bottom layer designed specifically to resist fatigue cracking.
Two approaches can be used to resist fatigue cracking in the base layer. First, the total
pavement thickness can be made great enough such that the tensile strain at the bottom of the
base layer is insignificant. Alternatively, the HMA base layer could be made using an extra-
flexible HMA. This can be most easily accomplished by increasing the asphalt content.
Combinations of the previous two approaches also work.
2. Intermediate layer. This is the middle layer designed specifically to carry most of the traffic
load. Therefore it must be stable (able to resist rutting) as well as durable. Stability can best
be provided by using stone-on-stone contact in the coarse aggregate and using a binder with
the appropriate high-temperature grading.
3. Wearing surface. This is the top layer designed specifically to resist surface-initiated
distresses such as top-down cracking and rutting. Other specific distresses of concern would
depend upon local experience.
In order to work, the above pavement structure must be built on a solid foundation. Nunn (1998) notes
that rutting on roads built on subgrade with a CBR greater than 5 percent originates almost solely in the
HMA layers, which suggests that a subgrade with a CBR greater than 5 percent (resilient modulus
greater than about 7,000 psi (50 MPa)) should be considered adequate. As always, proper construction
techniques are essential to a perpetual pavement's performance. Figure 2.6 shows an example cross-
section of a perpetual pavement design to be used in California on I-710 (the Long Beach Freeway) in
Los Angeles County.
Figure 2.6: Example I-710 Long Beach Freeway Perpetual Pavement Design
(from Monismith and Long, 1999)
Finally, the most important point in this brief perpetual pavement discussion is that it is possible to
design and build HMA pavements with extremely long design lives. In fact, some HMA pavements in
service today are living examples of perpetual pavements. For instance, two sections of Interstate 40 in
downtown Oklahoma City are now more than 33 years old (built in 1967) and are still in excellent
condition. These sections, which support 3 to 3.5 million ESALs per year, have been overlaid but the
base and intermediate courses have lasted since construction without any additional work (APA, no date
given).
Washington State has a significant length of nationally recognized Perpetual Pavement. In 2002,
I-90 in Washington State was awarded one of the inaugural Asphalt Pavement Alliance (APA)
Perpetual Pavement Awards. Criteria for this award included: the nature of the original design;
the overall quality of the pavement; the absence of structural failures; the condition of any long-
lasting overlays; the existence of a history of low overall maintenance; the nature of the efforts
that were made to minimize traffic disruptions during resurfacing; and other factors (APA,
2001b). Since their construction in the early to mid 1970s, all the flexible sections of I-90 are still
performing well and none have required reconstruction (Mahoney, 2001).
Western Washington
Weighted 25.8 years 368 mm 18.5 years 7.4 years 1.0 m/km 5 mm
Average (14.5 in.) (63 in./mi.) (0.20 in.)
Number of 9 9 9 9 9 9
Sections
350 - 472 0.7 - 1.3 m/ 2 - 7 mm
mm km
Range 23 - 29 years (13.8 - 18.6 17 - 22 years 4 - 12 years (44 - 82 in./ (0.08 - 0.28
in.) mi.) in.)
Eastern Washington
Weighted 29.3 years 240 mm 12.4 years 4.7 years 0.8 m/km 5 mm
Average (9.5 in.) (51 in./mi.) (0.20 in.)
Number of 27 27 25 25 25 25
Sections
150 - 350 0.6 - 1.2 m/ 1 - 9 mm
mm km
Range 6 - 35 years (6.0 - 13.9 6 - 21 years 2 - 10 years (38 - 76 in./ (0.04 - 0.35
in.)
in.) mi.)
● WSDOT design practice. Back in the early 1970s, before the Perpetual Pavement
concept was formalized, WSDOT's structural and mix design practices were sound
enough to produce extremely long-lasting flexible pavements.
service.
from each other mainly in maximum aggregate size, aggregate gradation and 3.3 Open-Graded Mixes
asphalt binder content/type. This Guide focuses on dense-graded HMA in
most flexible pavement sections because it is the most common HMA 3.4 Mix Selection Guidance
pavement material in the U.S. This section provides a brief exposure to:
● Dense-graded HMA. Flexible pavement information in this Guide is generally concerned with dense-
graded HMA. Dense-graded HMA is a versatile, all-around mix making it the most common and well-
understood mix type in the U.S.
● Stone matrix asphalt (SMA). SMA, although relatively new in the U.S., has been used in Europe as a
surface course for years to support heavy traffic loads and resist studded tire wear.
● Open-graded HMA. This includes both open-graded friction course (OGFC) and asphalt treated
permeable materials (ATPM). Open-graded mixes are typically used as wearing courses (OGFC) or
underlying drainage layers (ATPM) because of the special advantages offered by their porosity.
This section is taken largely from the NAPA's HMA Pavement Mix Type Selection Guide (2001). In addition to the general
information presented here, the HMA Pavement Mix Type Selection Guide provides specific information on minimum lift
thicknesses, mix selection criteria, mix materials as well as several informative examples.
● Superpave. Dense-graded HMA conforming to Superpave mix design requirements. WSDOT uses
Superpave mixes with the following nominal maximum aggregate sizes: 25 mm (1 inch), 19 mm
(0.75 inch), 12.5 mm (0.5 inch) and 9.5 mm (0.375 inch). 12.5 mm (0.5 inch) Superpave has the
same nominal maximum aggregate size as WSDOT Class A and B mixes and is considered, along
with Class A, as WSDOT's "standard" surface course mix. Minimum lift thicknesses are as follows:
● Class A. Dense-graded HMA with at least 90 percent of the coarse aggregate having at least one
fractured face. Its primary use is as a surface course for locations with high traffic levels or when
the potential for rutting within the HMA layer exists. This is considered, along with 12.5 mm
Superpave, WSDOT's "standard" surface course paving mix. Minimum lift thickness is 36 mm (0.12
ft).
● Class B. Dense-graded HMA with at least 75 percent of the coarse aggregate having at least one
fractured face. Its primary use is as a leveling course or surface course. Minimum lift thickness is
36 mm (0.12 ft).
● Class D. An open-graded HMA. Class D HMA overlays are used primarily to seal and maintain
aged, but otherwise structurally sound flexible pavements. Class D HMA (known previously as
“open-graded asphalt seal coat”) is basically a bituminous surface treatment (BST) aggregate mixed
hot in a plant with a relatively high percentage of asphalt binder. Class D overlays are placed with
an asphalt paver at a compacted depth of 18 mm (0.06 ft.). WSDOT's targeted service life for a
Class D overlay is eight years, although performance of these mixes has varied on the state route
system because Class D HMA is susceptible to studded tire wear. Thus, this mix is seldom used on
high ADT routes or where high studded tire use is expected. Minimum lift thickness is 18 mm (0.06
ft).
● Class E. Dense-graded HMA primarily intended for use as a base course. This is WSDOT's
"standard" base course paving mix for both flexible and rigid pavements. Minimum lift thickness is
60 mm (0.20 ft).
● Class F. Dense-graded HMA similar to Class B but a bit coarser. This is used in lieu of Class B
where aggregate sources cannot meet Class B grading requirements. Thus, it has a higher
performance risk. Minimum lift thickness is 36 mm (0.12 ft).
● Class G. Dense-graded HMA for thin lifts (25 mm (1 inch) or less). The nominal maximum
aggregate size is about 9.5 mm (0.375 inches). This small size allows for its use as a thin, hot
maintenance seal. Minimum lift thickness is 18 mm (0.06 ft).
● Asphalt Treated Base (ATB). Dense-graded HMA with a wide gradation band and lower asphalt
content (2.5 - 4.5 percent by weight of aggregate) intended for use as a base course.
Purpose: Dense-graded mixes are suitable for all pavement layers and for all traffic conditions. They work well for structural, friction,
leveling and patching needs.
Other Particulars about dense-graded HMA are covered by flexible pavement sections in the rest of this Guide.
Info:
Table 2.1: Fine- and Course-Graded Definitions for Dense-Graded HMA (from NAPA, 2001)
37.5 mm (1.5 inches) < 35 % passing the 4.75 mm (No. 4 Sieve) > 35 % passing the 4.75 mm (No. 4 Sieve)
25.0 mm (1.0 inch) < 40 % passing the 4.75 mm (No. 4 Sieve) > 40 % passing the 4.75 mm (No. 4 Sieve)
19.0 mm (0.75 inches) < 35 % passing the 2.36 mm (No. 8 Sieve) > 35 % passing the 2.36 mm (No. 8 Sieve)
12.5 mm (0.5 inches) < 40 % passing the 2.36 mm (No. 8 Sieve) > 40 % passing the 2.36 mm (No. 8 Sieve)
9.5 mm (0.375 inches) < 45 % passing the 2.36 mm (No. 8 Sieve) > 45 % passing the 2.36 mm (No. 8 Sieve)
WSDOT has built several SMA wearing courses both in urban and rural settings. For a brief summary of SMA
in Washington State, see the WSDOT SMA Tech Note. The first SMA project in Washington State was a 1999
resurfacing of SR 524 in Lynwood. Experiences on this project are documented in a WSDOT Research Report
available at:
http://www.wsdot.wa.gov/biz/mats/pavement/pavementresearch.htm
Purpose: Improved rut resistance and durability. Therefore, SMA is almost exclusively used for surface courses on high volume
interstates and U.S. roads.
Materials: Gap-graded aggregate (usually from coarse aggregate, manufactured sands and mineral filler all combined into a final
gradation), asphalt binder (typically with a modifier)
Mix Superpave or Marshall procedures with modifications. Refer to NAPA's Designing and Constructing SMA Mixtures: State-
Design: of-the-Practice, QIP 122 (1999) publication or NCHRP Report 425: Designing Stone Matrix Asphalt Mixtures for Rut-
Resistant Pavements.
Other Because SMA mixes have a high asphalt binder content (on the order of 6 percent), as the mix sits in the HMA storage silos,
Info: transport trucks, and after it is placed, the asphalt binder has a tendency to drain off the aggregate and down to the bottom - a
phenomenon known as "mix draindown". Mix draindown is usually combated by adding cellulose or mineral fibers to keep
the asphalt binder in place. Cellulose fibers are typically shredded newspapers and magazines, while mineral fibers are spun
from molten rock. A laboratory test is run during mix design to ensure the mix is not subject to excessive draindown.
In mix design a test for voids in the coarse aggregate (AASHTO T 19) is used to ensure there is stone-on-stone contact.
Other reported SMA benefits include wet weather friction (due to a coarser surface texture), lower tire noise (due to a coarser
surface texture) and less severe reflective cracking. Mineral fillers and additives are usually added to minimize asphalt
binder drain-down during construction, increase the amount of asphalt binder used in the mix and to improve mix durability.
Figure 2.9: SMA Aggregate Structure. Notice the stone-on- Figure 2.10: Dense-Graded HMA (left) vs. SMA (right). (it is a
stone contact of the larger aggregate particles. bit more shiny from the extra asphalt binder)
Figure 2.11: Dense-Graded HMA (left) vs. SMA (right). Figure 2.12: SMA Pavement Surface
Notice the SMA has a better-defined large aggregate skeleton
(from NAPA, 2001)
1. Open-graded friction course (OGFC). Typically 15 percent air voids, no minimum air voids specified, lower
aggregate standards than PEM.
2. Porous European mixes (PEM). Typically 18 - 22 percent air voids, specified minimum air voids, higher aggregate
standards than OGFC and requires the use of asphalt binder modifiers. See Figure 2.13.
3. Asphalt treated permeable bases (ATPB). Less stringent specifications than OGFC or PEM since it is used only
under dense-graded HMA, SMA or PCC for drainage. See Figure 2.14.
Purpose: OGFC and PEM - Used as for surface courses only. They reduce tire splash/spray in wet weather and typically result in
smoother surfaces than dense-graded HMA. Their high air voids trap road noise and thus reduce tire-road noise by up to 50-
percent (10 dBA) (NAPA, 1995).
Materials: Aggregate (crushed stone or gravel and manufactured sands), asphalt binder (with modifiers)
Mix Less structured than for dense-graded or SMA mixes. Open-graded mix design generally consists of 1) material selection, 2)
Design: gradation, 3) compaction and void determination and 4) asphalt binder drain-down evaluation. NCAT Report 99-3: Design
of New-Generation Open Graded Friction Courses provides a recommended mix design procedure for OGFCs.
Other Both OGFC and PEM are more expensive per ton than dense-graded HMA, but the unit weight of the mix when in-place is
Info: lower, which partially offsets the higher per-ton cost. The open gradation creates pores in the mix, which are essential to the
mix's proper function. Therefore anything that tends to clog these pores, such as low-speed traffic, excessive dirt on the
roadway or deicing sand, should be avoided.
Figure 2.13: Core from a Pavement Using PEM as Figure 2.14: Asphalt Treated Permeable Base
the Wearing Course (from NAPA, 2001)
Table 2.2: General Appropriateness of Mix Types For Each HMA Layer (NAPA, 2001)
Surface
Intermediate
Base
= Moderately Appropriate
1. Determine the total thickness of HMA required. This is accomplished using an appropriate structural design
procedure.
2. Determine the types of mixtures appropriate for the surface course based on traffic and cost.
❍ From Table 2.2, identify the general traffic category for the pavement in question then select those mix
types that are appropriate for the surface course.
❍ Determine what aggregate size to use for a mix. In general, the higher the traffic loads, the higher the
nominal maximum aggregate size should be.
❍ Consider appearance. Mixes with larger aggregates often have a coarser surface texture and may be more
susceptible to segregation during placement. Therefore, for a city street where appearance is an issue, a
finer mix such as a 9.5 or 12.5-mm (0.375 or 0.5-inch) dense-graded mix may be appropriate. Conversely,
for a heavy industrial area where load resistance is much more important that aesthetic appearance, a 19.0-
mm (0.75-inch) mix may be more appropriate. However, never sacrifice performance for appearance.
❍ Consider traffic flow. Maximum aggregate size can also affect traffic flow during rehabilitation of existing
roadways. In many urban areas off-peak construction is used to minimize traffic impacts. However, for a
road to be released to traffic during peak hours, either the lane drop-off (elevation difference between
adjacent lanes) must be kept below a specified minimum value (typically less than 37.5 mm (1.5 inches)
with proper signage) or all lanes must be brought to the same elevation. Bringing all lanes to the same
elevation at the end of each paving day may require changing traffic control and moving paving equipment,
which can increase construction costs and decrease safety. Therefore it is often better to satisfy the lane
drop-off requirement. However, with larger aggregate mixes the minimum lift thickness may exceed the
maximum lane drop-off allowed. As a result, using a finer gradation may allow paving one lane, then
releasing the road to traffic, then paving the other lane. Again, do not sacrifice performance.
3. Subtract the surface course thickness from the total thickness and determine what mix or mixes are appropriate for the
intermediate and/or base courses using Table 2.2.
4. Continue to subtract intermediate/base course thicknesses from the total thickness until mixes and layer thicknesses
have been selected for the required pavement section.
● As an addition
to regular
HMA.
● As an
aggregate in
cold-mix
asphalt.
● As a granular
base course
when
pulverized.
● As a fill or
embankment
material.
HMA recycling can be divided into two basic categories based on the recycling methods used: hot
recycling and cold recycling. This section presents the basic recycling process as well as typical uses
and considerations for each of these recycling methods.
4.1.1 Recycled
Hot Mix (RHM)
Recycled hot mix (RHM) is the most common way of using RAP. Basically, new HMA is produced at a
batch or drum plant to which a predetermined percentage of RAP is added. RAP addition is typically 10
to 30 percent by weight although additions as high as 80 percent by weight have been done and
additions as high as 90 to 100 percent by weight are feasible (FHWA, 2001c). There is ample evidence
that HMA which incorporates RAP performs as well as HMA without RAP. Figure 2.17 shows two
dense-graded HMA cores, one with RAP and one without.
Mix Design: Superpave, Marshall or Hveem procedures. Blending charts are typically needed
when using high percentages of RAP.
Other Info: When heated, RAP may give off gaseous hydrocarbons. To minimize these
emissions, HMA plants generally heat RAP indirectly (usually it is added after the
aggregate is heated and thus heats up through contact with the already-hot aggregate).
RAP addition may require longer HMA plant heating times. This can sometimes
reduce plant output by as much as half.
In general, state DOTs allow more RAP in base and binder HMA courses than they
do in surface courses.
After milling or crushing, RAP gradation is generally finer than pure virgin
aggregate because of the degradation that occurs during removal and processing.
Table 2.3: Superpave Asphalt Binder Selection Guidelines for RAP Mixtures (from AASHTO, 2001)
The WSDOT 2002 Standard Specifications for Road, Bridge and Municipal Construction
(M 41-10) allows contractors the option of using RAP in the amount up to 20 percent of
total aggregate weight without specifically accounting for it in mix design (e.g., the mix
design is determined using virgin aggregate only even though RAP will be included in the
HMA production). It also states that "Recycled materials shall not be used in asphalt
concrete Class D." There are no relaxations in gradation acceptance requirements for
HMA with RAP.
If contractors desire to add more than 20 percent RAP, WSDOT requires a separate mix
design that specifically accounts for the percentage they want to add. This type of mix
design often involves binder extraction from the RAP so that it can be graded and the
virgin binder grade can be appropriately adjusted.
Figure 2.17: HMA Cores from a RAP Mix and a non-RAP Mix
● Heater scarification (Figure 2.18). This method uses a plant that heats the pavement surface
(typically using propane radiant heaters), scarifies the pavement surface using a bank of
nonrotating teeth, adds a rejuvenating agent to improve the recycled asphalt binder viscosity,
then mixes and levels the recycled mix using a standard auger system. The recycled asphalt
pavement is then compacted using conventional compaction equipment. Heater scarification
is limited in its ability to repair severely rutted pavements, which are more easily rehabilitated
with a conventional HMA overlay.
Figure 2.18: Heater Scarification Train Showing 2 Preheaters, the Heater/Scarifier, the Paver and Rollers.
● Repaving. This method removes (by heating and scarification and/or grinding) the top 25 to
50 mm (1 to 2 inches) of the existing HMA pavement, adds a rejuvenating agent to improve
the recycled asphalt binder viscosity, places the recycled material as a leveling course using a
primary screed, and simultaneously places a thin (usually less than 25 mm (1 inch)) HMA
overlay. Conventional equipment and procedures are used immediately behind the train to
compact both layers of material (Rathburn, 1990 as cited in FHWA, 2001c).
● Remixing. This method is used when additional aggregate is required to improve the strength
or stability. Remixing is similar to repaving but adds new virgin aggregate or new HMA to the
recycled material before it is leveled.
Materials: Asphalt binder rejuvenating agent and possibly new aggregate and HMA.
Mix Design: Not well-defined, but as a minimum cores are usually taken from the existing
pavement to determine the proper amount of rejuvenating agent to add.
Other Info: HIPR is only applicable to specific situations. First, air void content of the existing
asphalt binder must be high enough to accept the necessary amount of asphalt binder
rejuvenator. Second, HIPR can only adequately address shallow surface distress
problems (less than 50 mm (2 inches)). Third, pavements with delaminations
(subsequent layers not binding together) in the top 50 mm (2 inches) should not be
considered for HIPR projects. Finally, pavements that have been rutted, heavily
patched, or chip-sealed are not good candidates for HIPR projects (FHWA, 2001c).
Materials: RAP, asphalt emulsion or foamed asphalt, asphalt rejuvenating agent and possibly
virgin aggregate.
Mix Design: No generally accepted mix design method, but the Asphalt Institute recommends and
most agencies use a variation of the Marshall mix design method (FHWA, 2001b).
Other Info: Since cold in-place recycling has become more commonplace, cold plant mixing has
become less popular.
1. Milling. A milling machine pulverizes a thin surface layer of pavement, usually from 50 to
100 mm (2 to 4 inches) deep.
2. Gradation control. The pulverized material is further crushed and graded to produce the
desired gradation and maximum particle size. On some jobs this step is omitted, however on
others a trailer mounted screening and crushing plant is used to further crush and grade the
pulverized pavement. If needed, virgin aggregate can be added to the recycled material.
3. Additive incorporation. The graded pulverized material is mixed with a binding additive
(usually emulsified asphalt, lime, portland cement or fly ash). On some jobs, this is done by
the milling machine, however on others a trailer mounted pugmill mixer is used.
4. Mixture placement. The pulverized, graded pavement and additive combination is placed
back over the previously milled pavement and graded to the final elevation. Mixture
placement is most often done with a traditional asphalt paver (either through windrow pickup
or by depositing the mixture directly into the paver hopper), however on some very low traffic
applications the mixture can be placed by a motor grader. Because of the larger maximum
aggregate sizes of the graded mixture, the minimum lift thickness for placement is usually
around 50 mm (2 inches).
5. Compaction. The placed mixture is compacted to the desired density. Typical compaction
efforts involve a large pneumatic tire roller and a large vibratory steel wheel roller. If an
emulsion additive is used rolling is typically delayed until the emulsion begins to break. If a
portland cement or fly ash additive is used, rolling should begin immediately after placement.
6. Fog seal. If the newly placed material is to operate as a high quality gravel road then a fog
seal is usually applied over the top to delay surface raveling of the cold recycled mix. A fog
seal is necessary over CIR using a portland cement or fly ash additive not only to delay
surface raveling but also to provide a curing membrane for the additive to properly set.
7. Surface course construction. On higher volume roads, the cold recycled mix is overlaid with
either a BST or a thin HMA overlay. In either case, a tack coat should be used to provide a
good bond between the cold recycled mix and the surface course.
Purpose: Stabilized base course or a low volume road granular surface course.
Materials: Recycled material and a binding additive (usually asphalt emulsion, lime, portland
cement or fly ash).
Mix Design: No generally accepted mix design method, but most methods are based on the
Marshall or Hveem methods and equipment (AASHTO, 1996).
Other Info: CIR is best suited for cracked pavements with structurally sound, well drained bases
and subgrades. CIR is generally not appropriate for repairing pavement failures
caused by:
● Frost action
● Stripping
CIR is generally suitable for lower volume roads that may only require a simple
surface treatment over the resulting stabilized base course, or at most a thin HMA
wearing course (Better Roads, 2001).
WSDOT has been using CIR since 1982 and has place about 167 lane miles on 14
different projects. For those projects that have been subsequently rehabilitated it appears
service life was in the range of 10 years. Some specific WSDOT guidance on CIR in
addition to that listed above is:
● The necessary warm, dry climate for CIR can be found in Eastern Washington
(east of the Cascade mountains) but generally not in Western Washington
(west of the Cascade mountains).
● CIR can drastically reduce aggregate or mixture haul costs because the
recycled material never leaves the project site. Therefore CIR is a more
competitive rehabilitation option where limited supplies of close-by virgin
aggregate would force a contractor to otherwise haul aggregate in from far
away.
● CIR has not been used in urban environments, areas of inconsistent pavement
width and depth or areas of multiple pavement types within the same pavement
structure.
● Production rates can vary but are typically between 2,000 and 5,000 tons/day,
● When used for suitable projects (e.g., no close-in supply of aggregate, low
volume road, warm and dry climate, etc.) CIR costs are comparable to
traditional overlay costs.
2. Moisture conditioning. The road reclaimer or a separate truck adds water to the newly
pulverized RAP to assist in achieving required density.
3. Breakdown roller. A sheepsfoot or pneumatic tire roller is typically used to compact the
recently pulverized RAP to a consistent density.
5. Intermediate roller. A pneumatic tire roller or a steel wheel vibratory roller is used to knead
and seat any loose aggregates left from the shaping process.
6. Finish roller. A 12 to 14-ton static steel wheel roller is used to seat any remaining loose
aggregates and create a smooth surface.
7. Sealant. A fog seal is typically applied to protect the finished reclaimed layer. After the fog
seal sets the reclaimed layer can generally withstand interim traffic loading. Therefore, at this
point the road is often opened to traffic until the contractor is ready to apply the surface
treatment or HMA surface course.
8. Surface treatment or surface course. Finally, a more durable surface treatment or surface
course is applied over the new stabilized base course.
Materials: Recycled material, asphalt emulsion or foamed asphalt, asphalt rejuvenating agent
and possibly virgin aggregate.
Mix Design: No generally accepted mix design method, but the Asphalt Institute recommends and
most agencies use a variation of the Marshall mix design method (FHWA, 2001b).
Other Info: FDR is generally suitable for lower volume roads that may only require a simple
surface treatment over the resulting stabilized base course, or at most a thin HMA
wearing course. However, FDR has been used on major highways including
interstates (Better Roads, 2001).
This section describes the typical rigid pavement structure consisting of:
● Surface course. This is the top layer, which consists of the PCC slab.
● Base course. This is the layer directly below the PCC layer and generally consists of aggregate or stabilized subgrade.
● Subbase course. This is the layer (or layers) under the base layer. A subbase is not always needed and therefore may
often be omitted.
1. Aggregate base. A simple base course of crushed aggregate has been a common option since the early 1900s and is still
appropriate in many situations today.
2. Stabilized aggregate or soil (see Figure 2.23). Stabilizing agents are used to bind otherwise loose particles to one another,
providing strength and cohesion. Cement treated bases (CTBs) can be built to as much as 20 - 25 percent of the surface
course strength (FHWA, 1999). However, cement treated bases (CTBs) used in the 1950s and early 1960s had a tendency
to lose excessive amounts of material leading to panel cracking and settling.
3. Dense-graded HMA. In situations where high base stiffness is desired base courses can be constructed using a dense-
graded HMA layer.
4. Permeable HMA. In certain situations where high base stiffness and excellent drainage is desired, base courses can be
constructed using an open graded HMA. Recent research may indicate some significant problems with ATPB use.
5. Lean concrete (see Figure 2.24). Contains less portland cement paste than a typical PCC and is stronger than a stabilized
aggregate. Lean concrete bases (LCBs) can be built to as much as 25 - 50 percent of the surface course strength (FHWA,
1999). A lean concrete base functions much like a regular PCC surface course and therefore, it requires construction joints
and will crack over time. These joints and cracks can potentially cause reflection cracking in the surface course if they are
not carefully matched.
Figure 2.23: Completed CTB with Curing Seal Figure 2.24: Lean Concrete Base Material
1. Minimize the intrusion of fines from the subgrade into the pavement structure.
2. Improve drainage.
The subbase generally consists of lower quality materials than the base course but better than the subgrade soils. Appropriate
materials are aggregate and high quality structural fill. A subbase course is not always needed or used.
5.2 Joints
Joints are purposefully placed discontinuities in a rigid pavement surface course. The most common types of pavement joints,
defined by their function, are (AASHTO, 1993): contraction, expansion, isolation and construction.
Contraction joints are chiefly defined by their spacing and their method of load transfer. They are generally between 1/4 - 1/3 the
depth of the slab and typically spaced every 3.1 - 15 m (12 - 50 ft.) with thinner slabs having shorter spacing (see Figure 2.25).
Some states use a semi-random joint spacing pattern to minimize their resonant effect on vehicles. These patterns typically use a
repeating sequence of joint spacing (for example: 2.7 m (9 ft.) then 3.0 m (10 ft.) then 4.3 m (14 ft.) then 4.0 m (13 ft.)). Transverse
contraction joints can be cut at right angles to the direction of traffic flow or at an angle (called a "skewed joint", see Figure 2.27).
Skewed joints are cut at obtuse angles to the direction of traffic flow to help with load transfer. If the joint is properly skewed, the
left wheel of each axle will cross onto the leave slab first and only one wheel will cross the joint at a time, which results in lower
load transfer stresses (see Figure 2.28).
Figure 2.25: Rigid Pavement Showing Figure 2.26: Missing Contraction Joint
Contraction Joints (The middle lane contraction joint was not sawed resulting in a
transverse slab crack. The outer lanes have proper contraction joints
and therefore, no cracking)
(The Tining is Perpendicular to the Direction of Travel While the Contraction Joint is Skewed)
The contraction joint spacing used by WSDOT is based on dowel bar use for load transfer. A reasonable joint
spacing when dowels are used is 3.7 m (12 ft.); however, contraction joint spacings up to 4.5 m (15 ft.) can be used
and are specified in the WSDOT Standard Plans.
These contraction joint spacings are, in part, based on prior rigid pavement performance in Washington State and
elsewhere and slab stress calculations. For example:
● Contraction joint spacings of 3.7 - 4.5 m (12 to 15 ft.) result in lower slab stresses due to thermal
gradients.
● A contraction joint spacing of about 3.7 m (12 ft). conforms to the FHWA L/l = 5.0 criterion for
"thinner" slabs of about 228 mm (9 in.) on stiff subbases. A spacing of about 4.5 m (15 ft.) conforms to
the same criterion for "thicker" slabs of about 330 mm (13 in.) on stiff subbases.
● In general, annual joint openings should be limited to no more than 0.6 - 0.9 mm (0.025 - 0.035 in.) to
insure long term joint performance. Using the slab shrinkage/expansion equation and PCC slabs on
stabilized base for annual temperature ranges estimated for eastern and western Washington, the
resulting joint movements are:
This efficiency depends on several factors, including temperature (which affects joint opening), joint spacing, number and magnitude
of load applications, foundation support, aggregate particle angularity, and the presence of mechanical load transfer devices.
Figure 2.32 illustrates the extremes in load transfer efficiency. Most performance problems with concrete
pavement are a result of poorly performing joints (ACPA, 2001). Poor load transfer creates high slab stresses,
which contribute heavily to distresses such as faulting, pumping and corner breaks. Thus, adequate load transfer
is vital to rigid pavement performance. Load transfer across transverse joints/cracks is generally accomplished
using one of three basic methods: aggregate interlock, dowel bars, and reinforcing steel.
Figure 2.32: Load Transfer Efficiency Across a PCC Surface Course Joint
Dowel bars are short steel bars that provide a mechanical connection
between slabs without restricting horizontal joint movement. They increase
load transfer efficiency by allowing the leave slab to assume some of the Figure 2.34: Typical Dowel Bar
load before the load is actually over it. This reduces joint deflection and Location
stress in the approach and leave slabs.
on Transverse Joints
at mid-slab depth and coated with a bond-breaking substance to prevent bonding to the PCC. Thus, the
dowels help transfer load but allow adjacent slabs to expand and contract independent of one another.
Figure 2.36 shows typical dowel bar locations at a transverse construction joint.
Figure 2.35: Stainless Steel-Clad Dowel Bars Figure 2.36: Dowel Bars in Place at a Construction Joint- the Green
(Epoxy Coating on Ends Only) Color is from the Epoxy Coating
WSDOT uses one standard dowel bar for all new construction, reconstruction
and dowel bar retrofits:
Typically, rigid pavement reinforcing steel consists of grade 60 (yield stress of 60 ksi (414 MPa) No. 5 or
No. 6 bars (ERES, 2001). The steel constitutes about 0.6 - 0.7 percent of the pavement cross-sectional area (ACPA,
2001) and is typically placed at slab mid-depth or shallower. At least 63 mm (2.5 inches) of PCC cover should be maintained over
the reinforcing steel to minimize the potential for steel corrosion by chlorides found in deicing agents (Burke, 1983).
Tie bars are typically No. 5 bars, 800 mm (32 in.) long and spaced
900 mm (36 in.) center to center.
Joint Spacing: Typically between 3.7 m (12 ft.) and 6.1 m (20 ft.). Due to the nature of concrete, slabs
longer than about 6.1 m (20 ft.) will usually crack in the middle. Depending upon
environment and materials slabs shorter than this may also crack in the middle.
Reinforcing None.
Steel:
Load Transfer: Aggregate interlock and dowel bars. For low-volume roads aggregate interlock is often
adequate. However, high-volume roads generally require dowel bars in each transverse joint
to prevent excessive faulting.
Other Info: A majority of U.S. State DOTs build JPCP because of its simplicity and proven performance.
(JRCP)
Jointed reinforced concrete pavement (JRCP, see Figure 2.40) uses contraction joints and reinforcing
steel to control cracking. Transverse joint spacing is longer than that for JPCP and typically ranges from
about 7.6 m (25 ft.) to 15.2 m (50 ft.). Temperature and moisture stresses are expected to cause cracking
between joints, hence reinforcing steel or a steel mesh is used to hold these cracks tightly together.
Dowel bars are typically used at transverse joints to assist in load transfer while the reinforcing steel/
wire mesh assists in load transfer across cracks.
Joint Spacing: Longer than JPCP and up to a maximum of about 15 m (50 ft.). Due to the nature of
concrete, the longer slabs associated with JRCP will crack.
Reinforcing A minimal amount is included mid-slab to hold cracks tightly together. This can be in the
Steel: form of deformed reinforcing bars or a thick wire mesh.
Load Transfer: Dowel bars and reinforcing steel. Dowel bars assist in load transfer across transverse joints
while reinforcing steel assists in load transfer across mid-panel cracks.
Other Info: During construction of the interstate system, most agencies in the
Eastern and Midwestern U.S. built JRCP. Today only a handful of
agencies employ this design (ACPA, 2001).
Pavement (CRCP)
Continuously reinforced concrete pavement (CRCP, see Figure 2.41) does not
require any contraction joints. Transverse cracks are allowed to form but are held tightly
together with continuous reinforcing steel. Research has shown that the maximum
allowable design crack width is about 0.5 mm (0.02 inches) to protect against spalling
and water penetration (CRSI, 1996). Cracks typically form at intervals of 1.1 - 2.4 m
(3.5 - 8 ft.). Reinforcing steel usually constitutes about 0.6 - 0.7 percent of the cross-
sectional pavement area and is located near mid-depth in the slab. Typically, No. 5 and
No. 6 deformed reinforcing bars are used.
During the 1970's and early 1980's, CRCP design thickness was typically about 80 percent of the
thickness of JPCP. However, a substantial number of these thinner pavements developed distress sooner
than anticipated and as a consequence, the current trend is to make CRCP the same
thickness as JPCP (FHWA, June 1990). The reinforcing steel is assumed to only
handle nonload-related stresses and any structural contribution to resisting loads
is ignored.
Reinforcing Typically about 0.6 - 0.7 percent by cross-sectional area (ACPA, 2001).
Steel:
Other Info: CRCP generally costs more than JPCP or JRCP initially due to increased
quantities of steel. Further, it is generally less forgiving of construction
errors and provides fewer and more difficult rehabilitation options.
However, CRCP may demonstrate superior long-term performance and
cost-effectiveness. Some agencies choose to use CRCP designs in their
heavy urban traffic corridors (ACPA, 2001).
In 1995, WSDOT began replacing flexible pavement with rigid pavement at selected
intersections. These intersections were severely rutted and distressed due to loads from
slow moving vehicles and warm temperatures. Statewide, ruts of 50 mm (2 inches) or
more occasionally have occurred and resurfacing to restore the intersection to an
acceptable level recurred at intervals of eight years or less. Though WSDOT has
numerous rigid pavement intersections, the unique feature with these particular
intersections was the replacement of existing flexible with rigid pavement only at
intersections.
A major advantage with rigid pavement replacement is that once the rigid pavement is
placed rehabilitation should not be necessary for 40 years. The major disadvantage with
rigid pavement intersections is the initial construction cost; however, these costs appear to
be coming down, particularly as contractors become familiar with this type of
construction. Rigid pavement intersections in the Tri-Cities area completed in 2000
showed that an entire intersection can be paved in a single weekend closure.
The following WSDOT report discusses rigid pavement intersection design and
construction in detail:
The following is a overview of some considerations for rigid pavement intersection design:
● The area of pavement rutting or distress must be well defined. The major
arterial approach legs to intersections may require rigid pavement sections
reaching as long as 60 - 150 m (200 - 500 ft.) back from the crosswalk. The
approach legs on the minor arterial typically require 60 m (200 ft.) or less but
may extend farther. It is desirable to extend the leave legs as far as the
adjacent approach legs but, at a minimum, the leave leg should extend at least
30 m (50 ft.) from the crosswalk.
● Dowel bars should be used for all joints with cross traffic. For slabs without
cross traffic, use dowel bars on the transverse joints and tie bars on the
longitudinal joints. Dowel bars and tie bars should not be placed within 0.61 m
(2 ft.) of new signal detection loops. Tie bars should not be placed within 0.46
m (1.5 ft.) of dowel bars.
Generally, recycling PCC involves breaking it up, removing embedded metal (e.g., steel reinforcing
bars, dowels, etc), then crushing it to a specified size. For small projects, RCM is usually broken up into
large pieces and loaded into dump trucks for removal from the site. This RCM is typically hauled to a
central facility for stockpiling and processing (FHWA, 2001d). The central processing facility crushes,
screens and removes ferrous metal from the RCM. Present crushing systems, with magnetic separators,
are capable of removing reinforcing steel without much difficulty, however welded wire mesh
reinforcement may be difficult or impossible to remove effectively (FHWA, 2001d). For large projects,
RCM is usually processed on site using a mobile plant or processed in place using one or several
machines.
Some general conclusions about RCM material properties from NCHRP Synthesis 154: Recycling of
Portland Cement Concrete Pavements (Yrjanson, 1989) are:
● Aggregate particle characteristics. The aggregate particles produced by crushing have good
particle shape, high absorptions, and low specific gravity compared with conventional mineral
aggregates.
❍ When crushed RCP is used as a fine aggregate, the mixture is less workable and
requires more cement because of its increased water demand. As a result, most
state agencies do not use recycled fines in concrete mixtures, and if they are used
they are limited to a maximum of 30 percent of the fine aggregate portion of the
mixture.
● Durability.
❍ PCC made from RCP aggregate has shown an increase in freeze-thaw resistance
compared with PCC made from normal conventional aggregates.
● Strength. The strength of PCC made with RCP aggregate can be equivalent to conventional
PCC mixtures when recycled fines are omitted or used in small amounts.
The following sections discuss the three major uses of RCM: coarse aggregate, base material and
embankment/fill.
WSDOT does not allow the use of RCM as an aggregate in new PCC or new HMA surface courses. RCM
can be used in asphalt treated base material.
Mix Design: Standard PCC mix design. Most standards consider RCM coarse aggregate to be a
conventional coarse aggregate and treat it as such. If not prewetted, RCM
aggregates will absorb a substantial amount of water.
Standard HMA mix design. RCM is more appropriate for asphalt treated base
materials that generally have fewer restrictions on mix design and aggregate
properties. RCM has functioned well as an asphalt treated base material.
Other Info: PCC incorporating more than about 10 to 20 percent fine aggregate will have low
workability and require more water to maintain reasonable workability. This excess
water will result in an overall strength reduction.
For a given compressive strength (at 28 days), both the static and dynamic moduli of
elasticity for recycled-aggregate concrete are significantly lower (up to 40 percent)
than those for concrete containing virgin aggregate (FHWA, 2001e).
Other Info: Typically, crushing RCM will result in recovery of 55 to 80 percent of the original
pavement volume. In general, the larger the maximum aggregate size, the higher the
recovery rate.
RCM base material has high friction angle, typically in excess of 40° and
consequently demonstrates good stability and little post-compaction settlement
(FHWA, 2001f).
Typical CBRs range from 90 to more than 140 depending on the angularity of the
virgin concrete aggregate and strength of the portland cement matrix (Petrarca and
Galdiero, 1984).
Other Info: RCM is highly alkaline (pH in excess of 11). Therefore, contact with aluminum or
galvanized steel pipes can cause corrosion in the presence of moisture (FHWA,
2001g).
1 Introduction
Pavements are a conglomeration of materials. These materials, their associated properties, and their
interactions determine the properties of the resultant pavement. Thus, a good understanding of these
materials, how they are characterized and how they perform, is fundamental to understanding pavement.
This section will emphasize what each material is, how it is characterized and the typical tests used in
this characterization. This section is meant to provide an overview of these materials and as such,
provides limited in-depth technical analysis. Where needed, the generic term "binder" is used to refer to
either the asphalt binder in HMA or the portland cement paste in PCC.
http://training.ce.washington.edu/WSDOT/Modules/03_materials/03-1_body.htm4/2/2008 6:28:40 PM
3.2 Materials - Aggregate
2 Aggregate
"Aggregate" is a collective term for the mineral materials such as sand, Major Topics on this Page
gravel and crushed stone that are used with a binding medium (such as
water, bitumen, portland cement, lime, etc.) to form compound materials 2.1 Aggregate Sources
(such as asphalt concrete and portland cement concrete). By volume,
aggregate generally accounts for 92 to 96 percent of HMA and about 70 2.2 Aggregate Production
to 80 percent of portland cement concrete. Aggregate is also used for base 2.3 Mineral Properties
and subbase courses for both flexible and rigid pavements.
2.4 Chemical Properties
Aggregates can either be natural or manufactured. Natural aggregates are
2.5 Physical Properties
generally extracted from larger rock formations through an open
excavation (quarry). Extracted rock is typically reduced to usable sizes 2.6 Aggregate as a Base Material
by mechanical crushing. Manufactured aggregate is often the byproduct
of other manufacturing industries. 2.7 Summary
This section will briefly discuss aggregate sources and quarrying operations then describe the basic aggregate mineral,
chemical and physical properties most important to pavements and the typical tests used to determine these properties. The
following source contains more detailed information on aggregate:
● National Stone, Sand & Gravel Association (NSSGA). Aggregate Handbook. National Stone, Sand & Gravel
Association. Arlington, VA. http://www.nssga.org.
This section discusses many different aggregate tests. Rather than list WSDOT specifications for each test in their
respective sections, an overall summary of WSDOT aggregate specifications can be viewed through the below link.
In general, WSDOT uses AASHTO and ASTM testing methods in addition to specific WSDOT testing methods. All
WSDOT testing methods are contained in the WSDOT Materials Manual (M 46-01), which is available for free download
in the online technical manual library through WSDOT Engineering Publications (http://www.wsdot.wa.gov/fasc/
EngineeringPublications).
● Igneous rock. These rocks are primarily crystalline and are formed by the cooling of molten rock material
beneath the earth’s crust (magma).
● Sedimentary rocks. These rocks are formed from deposited insoluble material (e.g., the remains of existing rock
deposited on the bottom of an ocean or lake). This material is transformed to rock by heat and pressure.
Sedimentary rocks are layered in appearance and are further classified based on their predominant mineral as
calcareous (limestone, chalk, etc.), siliceous (chert, sandstone, etc.) or argillaceous (shale, etc.).
● Metamorphic rock. These are igneous or sedimentary rocks that have been subjected to heat and/or pressure
great enough to change their mineral structure so as to be different from the original rock.
Manufactured rock typically consists of industrial byproducts such as slag (byproduct of the metallurgical processing –
typically produced from processing steel, tin and copper) or specialty rock that is produced to have a particular physical
characteristic not found in natural rock (such as the low density of lightweight aggregate).
An aggregate’s mineral composition largely determines its physical characteristics and how it behaves as a pavement
material. Therefore, when selecting an aggregate source, knowledge of the quarry rock’s mineral properties can provide an
excellent clue as to the suitability of the resulting aggregate. Cordon (1979) provides some general guidelines for aggregate
used in HMA (shown in Table 3.1).
Resistance to
Rock Type Hardness, Toughness Surface Texture Crushed Shape
Stripping1,2
Igneous
Granite Fair Fair Fair Fair
Syenite Good Fair Fair Fair
Diorite Good Fair Fair Good
Basalt (trap rock) Good Good Good Good
Diabase (trap rock) Good Good Good Good
Gabbro (trap rock) Good Good Good Good
Sedimentary
Limestone Poor Good Good Fair
Sandstone Fair Good Good Good
Chert Good Fair Poor Good
Shale Poor Poor Fair Fair
Metamorphic
Gneiss Fair Fair Good Good
Schist Fair Fair Good Fair
Slate Good Fair Fair Fair
Quartzite Good Fair Good Good
Marble Poor Good Fair Fair
Serpentine Good Fair Fair Fair
Notes:
1. Aggregates that are hydrophilic (water-loving) tend to strip more readily since water more easily replaces the asphalt film
over each particle.
2. Freshly crushed aggregates with many broken ionic bonds tend to strip more easily.
In general, relationships between mineral and physical properties are quite complex, making it difficult to accurately predict
how a particular aggregate source will behave based on mineral properties alone.
solution. The end result is alkali-silica gel that is formed in place. This gel formation is not expansive itself but
it does destroy the integrity of the aggregate particle.
3. Attraction of water by the gel. The gel attracts considerable amounts of water and expands. If the expansion is
great enough, the resulting stress will crack the now-weakened aggregate and surrounding cement paste.
4. Formation of a gel colloid. After the gel ingests enough water, the water takes over and the substance becomes
an alkali-silica gel disbursed in a water fluid. This fluid then escapes to surrounding cracks and voids and may
partake in secondary reactions.
● Avoiding susceptible aggregates. Local experience may show that certain types of rock contain reactive silica.
Typically rock types that may be susceptible are: siliceous limestone, chert, shale, volcanic glass, synthetic glass,
sandstone, opaline rocks and quartzite. River rock is also typically susceptible.
● Pozzolanic admixture. By reacting with the calcium hydroxide in the cement paste, a pozzolan can lower the pH
of the pore solution. Additionally, the silica contained in a pozzolan may react with the alkali in the cement.
This reaction is not harmful because it essentially skips the expansive water attraction step.
● Low-alkali cement. Less alkali available for reaction will limit gel formation.
● Low water-cement ratio. The lower the water-cement ratio, the less permeable the concrete. Low permeability
will help limit the supply of water to the alkali-silica gel.
In sum, alkali-silica reactions are expansive in nature and occur in most PCC. If the reaction is severe enough it can
fracture aggregates and surrounding paste resulting in cracking, popouts and spalling. There are several ways of avoiding
this reaction, the simplest of which is just avoiding susceptible aggregate.
● Specific gravity
These are not the only physical properties of aggregates but rather the most commonly measured. Tests used to quantify
these properties are largely empirical. The physical properties of an aggregate can change over time. For instance, a newly
crushed aggregate may contain more dust and thus be less receptive to binding with an asphalt binder than one that has been
crushed and stored in a stockpile for a year.
Maximum aggregate size can affect HMA, PCC and base/subbase courses in several ways. In HMA, instability may result
from excessively small maximum sizes; and poor workability and/or segregation may result from excessively large
maximum sizes (Roberts et al., 1996). In PCC, large maximum sizes may not fit between reinforcing bar openings, but they
will generally increase PCC strength because the water-cement ratio can be lowered. ASTM C 125 defines the maximum
aggregate size in one of two ways:
● Maximum size. The smallest sieve through which 100 percent of the aggregate sample particles pass. Superpave
defines the maximum aggregate size as "one sieve larger than the nominal maximum size" (Roberts et al.,
1996).
● Nominal maximum size. The largest sieve that retains some of the aggregate particles but generally not more than
10 percent by weight. Superpave defines nominal maximum aggregate size as "one sieve size larger than the
first sieve to retain more than 10 percent of the material" (Roberts et al., 1996).
Thus, it is important to specify whether "maximum size" or "nominal maximum size" is being referenced.
The gradation of a particular aggregate is most often determined by a sieve analysis (see Figure 3.3). In a sieve analysis, a
sample of dry aggregate of known weight is separated through a series of sieves with progressively smaller openings. Once
separated, the weight of particles retained on each sieve is measured and compared to the total sample weight. Particle size
distribution is then expressed as a percent retained by weight on each sieve size. Results are usually expressed in tabular or
graphical format. PCC gradation graphs are traditionally semi-logarithmic, while HMA graphs often employ the standard
0.45 power gradation graph.
Figure 3.4 shows typical gradation graphs. Note that sieve sizes are presented from smallest to largest, left to right. The
number and size of the sieves used in a sieve analysis depend upon specification requirements.
For PCC, aggregate is typically classified as either "coarse" or "fine". Coarse aggregate is generally the fraction retained on
the 4.75 mm (No. 4) sieve while fine aggregate is the fraction passing the 4.75 mm (No. 4) sieve.
● AASHTO T 27 and ASTM C 136: Sieve Analysis of Fine and Coarse Aggregates
● AASHTO T 11 and ASTM C 117: Materials Finer Than 75-µm (No. 200) Sieve in Mineral Aggregate by
Washing
● AASHTO T 30: Mechanical Analysis of Extracted Aggregate (this is used for aggregate extracted from
bituminous mixtures)
Gradation has a profound effect on material performance. But what is the best gradation? This is a complicated question,
the answer to which will vary depending upon the material (HMA or PCC), its desired characteristics, loading,
environmental, material, structural and mix property inputs. Therefore, gradation requirements for specific HMA and PCC
mixes are discussed in their respective pavement type sections. This section presents some basic guidelines applicable to
common dense-graded mixes.
It might be reasonable to believe that the best gradation is one that produces the maximum density. This would involve a
particle arrangement where smaller particles are packed between the larger particles, which reduces the void space between
particles. This creates more particle-to-particle contact, which in HMA would increase stability and reduce water
infiltration. In PCC, this reduced void space reduces the amount of cement paste required. However, some minimum
amount of void space is necessary to:
● Provide adequate volume for the binder (asphalt binder or portland cement) to occupy.
● Promote rapid drainage and resistance to frost action for base and subbase courses.
Therefore, although it may not be the "best" aggregate gradation, a maximum density gradation does provide a common
reference. A widely used equation to describe a maximum density gradation was developed by Fuller and Thompson in
1907. Their basic equation is:
n = parameter which adjusts curve for fineness or coarseness (for maximum particle density n
≈ 0.5 according to Fuller and Thompson)
Figure 3.5: Maximum Density Curves for 0.45 Power Gradation Graph
(each curve is for a different maximum aggregate size)
To illustrate how the maximum density curves in Figure 3.5 are determined, Table 2.2 shows the associated calculations for
a maximum aggregate size of 19.0 mm.
Table 2.2: Calculations for a 0.45 Power Gradation Curve Using 19.0-mm (0.75-inch) Maximum Aggregate Size
19.0
12.5
9.5
2.00
0.300
0.075
Gradation Terminology
Several common terms are used to classify gradation. These are not precise technical terms but rather terms that refer to
gradations that share common characteristics (refer to Figure 3.6):
● Dense or well-graded. Refers to a gradation that is near the FHWA’s 0.45 power curve for maximum density.
The most common HMA and PCC mix designs in the U.S. tend to use dense graded aggregate. Typical
gradations are near the 0.45 power curve but not right on it. Generally, a true maximum density gradation
(exactly on the 0.45 power curve) would result in unacceptably low VMA.
● Gap graded. Refers to a gradation that contains only a small percentage of aggregate particles in the mid-size
range. The curve is flat in the mid-size range. Some PCC mix designs use gap graded aggregate to provide a
more economical mix since less sand can be used for a given workability. HMA gap graded mixes can be prone
to segregation during placement.
● Open graded. Refers to a gradation that contains only a small percentage of aggregate particles in the small
range. This results in more air voids because there are not enough small particles to fill in the voids between the
larger particles. The curve is near vertical in the mid-size range, and flat and near-zero in the small-size range.
● Uniformly graded. Refers to a gradation that contains most of the particles in a very narrow size range. In
essence, all the particles are the same size. The curve is steep and only occupies the narrow size range specified.
● Restricted zone. Note: the restricted zone will be eliminated by late 2002. The restricted zone refers to a
particular area of the FHWA’s 0.45 power gradation graph associated with Superpave mix designs. It was
originally observed that mixes closely following the 0.45 power maximum density line in the finer gradations
sometimes had unacceptably low VMA. Therefore, in an attempt to minimize this problem, Superpave included
a restricted zone through which a typical gradation should not pass as a recommended guideline. However, since
the restricted zone's original inception, NCHRP Report 464: The Restricted Zone in the Superpave Aggregate
Gradation Specification has concluded that "...gradations that violated the restricted zone performed similarly to
or better than the mixes having gradations passing outside the restricted zone; therefore, the restricted zone
requirement is redundant for mixes meeting all Superpave volumetric parameters...It has been recommended to
delete references to the restricted zone as either a requirement or a guideline from the AASHTO specification
(AASHTO MP 2) and practice (AASHTO PP 28) for Superpave volumetric mix design." (Kandhal and Cooley,
2001).
WSDOT experience and analysis has shown that HMA mixes crossing 0.45
power curve in the restricted zone at a severe angle may be susceptible to
rutting.
● Fine gradation. A gradation that, when plotted on the 0.45 power gradation graph, falls mostly above the 0.45
power maximum density line. The term generally applies to dense graded aggregate.
● Coarse gradation. A gradation that, when plotted on the 0.45 power gradation graph, falls mostly below the 0.45
power maximum density line. The term generally applies to dense graded aggregate.
Permeability
Figure 3.7 shows some typical aggregate gradations and their associated permeabilities. This shows that even a small
amount of particles passing the 0.075-mm (#200) sieve results in very low permeability. Therefore, for base and subbase
aggregates where permeability is important for drainage and frost resistance, many agencies will specify a maximum
percent-by-weight passing for this sieve.
WSDOT uses crushed surfacing base course (CSBC) as a frost resistant crushed aggregate because it has a maximum
of only 7.5% passing the 0.075 mm (No. 200) sieve.
Figure 3.7: Typical Aggregate Gradations and Permeabilities (after Ridgeway, 1982)
Table 3.3 and Figure 3.8 show some typical specification bands for aggregate courses taken from the FHWA 1996 Standard
Specifications (FHWA, 1996).
Table 3.3: Some Representative Gradation Specifications for Aggregate Courses from
the 1996 FHWA Standard Specifications for Construction of Roads and Bridges on Federal Highway Projects (FP-96)
Percent Passing
Sieve Size Subbase Course Base Course Surface Course
(Grading A) (Grading B) (Grading F)
63 mm 2.5-inch - 100 -
Notes:
1. Number in parentheses indicates the allowable deviations (± ) from the target value.
2. These are only representative gradations and do not represent a comprehensive list of FHWA specified gradations.
Figure 3.8: Some Representative Gradation Specifications for Aggregate Courses from
the 1996 FHWA Standard Specifications for Construction of Roads and Bridges on Federal Highway Projects (FP-96)
(click on text in the Figure to show plots)
For aggregates used in PCC, another common gradation description for fine aggregate is the fineness modulus. It is
described in ASTM C 125 and is a single number used to describe a gradation curve. It is defined as:
specified = 0.150 mm (No. 100), 0.30 mm (No. 50), 0.60 mm (No. 30), 1.18 mm (No. 16),
sieves 2.36 mm (No. 8), 4.75 mm (No. 4), 9.5 mm (0.375-in.), 19.0 mm (0.75-in.), 37.5
mm (1.5-in.), and larger increasing in the size ratio of 2:1.
The larger the fineness modulus, the more coarse the aggregate. A typical fineness modulus for fine aggregate used in PCC
is between 2.70 and 3.00.
A common test used to characterize toughness and abrasion resistance is the Los Angeles (L.A.) abrasion test. For the L.A.
abrasion test, the portion of an aggregate sample retained on the 1.70 mm (No. 12) sieve is placed in a large rotating drum
that contains a shelf plate attached to the outer wall (the Los Angeles machine – see Figure 3.9). A specified number of
steel spheres are then placed in the machine and the drum is rotated for 500 revolutions at a speed of 30 - 33 revolutions per
minute (RPM). The material is then extracted and separated into material passing the 1.70 mm (No. 12) sieve and material
retained on the 1.70 mm (No. 12) sieve. The retained material (larger particles) is then weighed and compared to the
original sample weight. The difference in weight is reported as a percent of the original weight and called the "percent loss".
Table 3.4 shows some typical test values from the L.A. abrasion test. Unfortunately, the test does not seem to correspond
well with field measurements (especially with slags, cinders and other lightweight aggregates). Some aggregates with high
L.A. abrasion loss, such as soft limestone, provide excellent performance. However, no matter the performance
characteristics, aggregate with high L.A. abrasion loss values will tend to create dust during production and handling, which
may produce environmental and mixture control problems.
Dolomite 18 - 30
Gneiss 33 - 57
Granite 27 - 49
Limestone 19 - 30
Quartzite 20 - 35
● AASHTO T 96 and ASTM C 131: Resistance to Degradation of Small-Size Coarse Aggregate by Abrasion and
Impact in the Los Angeles Machine
● ASTM C 535: Resistance to Degradation of Large-Size Coarse Aggregate by Abrasion and Impact in the Los
Angeles Machine
The most common soundness test involves repeatedly submerging an aggregate sample in a saturated solution of sodium or
magnesium sulfate. This process causes salt crystals to form in the aggregate pores, which simulate ice crystal formation
(see Figure 3.10 and 3.11). The basic procedure is as follows (from Roberts et al., 1996):
● Oven dry the sample and separate it into specific sieve sizes.
● Immerse the sample in a saturated solution of sodium or magnesium sulfate and let it remain at a constant
temperature for 18 hours.
● Remove the sample from the solution and dry to a constant weight at 110 ± 5oC (230 ± 9oF).
● Determine the loss in weight for each specific sieve size and compute a weighted average percent loss for the
entire sample.
The maximum loss values typically range from 10 – 20 percent for every five cycles.
Other soundness tests use relatively the same procedure but substitute actual freezing and thawing in place of the salt
crystallization of the procedure described previously. Cracks in PCC resulting from poor aggregate freeze-thaw resistance
are often called durability cracks or "D cracks".
● AASHTO T 104 and ASTM C 88: Soundness of Aggregates by Use of Sodium Sulfate or Magnesium Sulfate
WSDOT does not use an aggregate soundness test but rather uses its own test to determine a "degradation value".
This degradation value test determines the susceptibility of an aggregate to degrade into plastic fines when abraded in
the presence of water.
Basically, the procedure takes a sample of aggregate retained on the 12.5 mm (0.5 inch) sieve and crushes it so that it
will then pass the 12.5 mm (0.5 inch) sieve. This crushed material is then placed in a container filled with water and
the container is agitated for 20 minutes. The amount of fines generated is measured and the result is reported as a
degradation factor. The more fines generated, the lower the degradation factor. Degradation factor values can range
from 0 - 100 with higher values representing less degradation.
workability. However, the ideal shape for HMA and PCC is different because aggregates serve different purposes in each
material. In HMA, since aggregates are relied upon to provide stiffness and strength by interlocking with one another,
cubic angular-shaped particles with a rough surface texture are best. However, in PCC, where aggregates are used as an
inexpensive high-strength material to occupy volume, workability is the major issue regarding particle shape. Therefore, in
PCC rounded particles are better. Relevant particle shape/texture characteristics are:
● Particle shape. Rounded particles create less particle-to-particle interlock than angular particles and thus provide
better workability and easier compaction. However, in HMA less interlock is generally a disadvantage as
rounded aggregate will continue to compact, shove and rut after construction. Thus angular particles are desirable
for HMA (despite their poorer workability), while rounded particles are desirable for PCC because of their better
workability (although particle smoothness will not appreciably affect strength) (PCA, 1988).
● Flat or elongated particles. These particles tend to impede compaction or break during compaction and thus,
may decrease strength.
● Smooth-surfaced particles. These particles have a lower surface-to-volume ratio than rough-surfaced particles
and thus may be easier to coat with binder. However, in HMA asphalt tends to bond more effectively with rough-
surfaced particles, and in PCC rough-surfaced particles provide more area to which the cement paste can bond.
Thus, rough-surface particles are desirable for both HMA and PCC.
There are several common tests used to identify and quantify aggregate particle shape and surface texture. Among the most
popular are:
● Particle index
Other tests, using automated machines equipped with video cameras and lasers are under development.
Particle Index
The particle index test provides a combined shape-texture characterization. This test requires that an aggregate sample be
divided up into specific size fraction. Each size fraction is placed into a container in three layers. This is done twice; the
first time, each layer is compacted with 10 blows of a tamping rod, and the second time, each layer is compacted with 50
blows of a tamping rod. The particle index is computed from the following equation:
The overall sample particle index is computed as a weighted average of the individual size fraction particles indexes based
on the size fraction weights. Aggregates composed of rounded, smooth particles may have a low particle index of around 6
or 7, while aggregates composed of angular, rough particles may have a high particle index of between 15 and 20 or more.
process. First, the longest dimension is measured on one end of the caliper (see Figure 3.13). Then, based on the position
of the pivot point (numbered holes shown in Figure 3.12), the other end of the caliper (see Figure 3.14) is automatically
sized to the predetermined length-to-width ratio (in Figures 3.13 and 3.14 it is set at 2:1). If the aggregate is able to pass
between the bar and caliper it fails the test.
"…the ratio of the mass of a unit volume of a material at a stated temperature to the mass of the same volume of gas-
free distilled water at a stated temperature."
The commonly used "stated temperature" is 23° C (73.4° F). Given the structure of a typical aggregate particle, there are
several different kinds of specific gravity. This section will first describe the structure of a typical aggregate particle and
then discuss each type of specific gravity and its use.
A typical aggregate particle consists of some amount of solid material along with a certain amount of air voids. These air
voids within the aggregate particle (see Figure 3.15) can become filled with water, binder or both (see Figure 3.16). It takes
a finite amount of time for water/binder to penetrate these pores, so specific gravity test procedures generally contain a 15
to 19-hour (for AASHTO procedures) or a 24-hour (for ASTM procedures) soak period for the purpose of allowing
Depending upon how aggregate voids are dealt with, calculated aggregate specific gravities can vary. If they are excluded
entirely, then the specific gravity is that of the solid portion of the aggregate only, while if they are included entirely then
the specific gravity essentially becomes a weighted average of the specific gravity of the solid aggregate and whatever is in
its voids.
Generally, there are three different aggregate specific gravities used in association with pavements: bulk, apparent and
effective.
The sand equivalent test is a rapid field test to show the relative proportions of fine dust or claylike materials in aggregate
(or soils). A sample of aggregate passing the 4.75-mm (No. 4) sieve and a small amount of flocculating solution are poured
into a graduated cylinder and are agitated to loosen the claylike coatings from the sand particles. The sample is then
irrigated with additional flocculation solution forcing the claylike material into suspension above the sand. After a
prescribed sedimentation period, the height of flocculated clay and height of sand are determined. The sand equivalent is
determined from the below equation:
Cleaner aggregates will have higher sand equivalent values. Agencies often specify a minimum sand equivalent around 25
to 35 (Roberts et al., 1996).
● AASHTO T 176: Plastic Fines in Graded Aggregates and Soils by Use of the Sand Equivalent Test
2.5.6.2 Tests for Deleterious Materials – Clay Lumps and Friable Particles
To test for clay lumps or friable particles, a sample is first washed and dried to remove material passing the 0.075-mm (No.
200) sieve. The remaining sample is separated into different sizes and each size is weighed and soaked in water for 24
hours. Particles that can be broken down into fines with fingers are classified as clay lumps or friable material. The
amount of this material is calculated by percentage of total sample weight. Specifications usually limit clay and friable
particles to a maximum of one percent.
● AASHTO T 112 and ASTM C 142: Clay Lumps and Friable Particles in Aggregate
1. Oven-dry (OD). All moisture is removed by heating the aggregate in an oven at 105°C (221°F) to constant
weight (this usually constitutes heating it overnight). All pores connected to the surface are empty and the
aggregate is fully absorbent.
2. Airdry (AD). All moisture is removed from the surface, but pores connected to the surface are partially filled
3. Saturated surface dry (SSD). All pores connected to the surface are filled with water, but the surface is dry. The
aggregate is neither absorbent nor does it contribute water to the concrete mixture.
4. Wet. All pores connected to the surface are filled with water and there is excess moisture on the surface. The
aggregate contributes water to the concrete mixture.
Note that pores not connected to the surface are not considered.
These conditions are used to calculate various aggregate properties. The moisture content of an aggregate is expressed as:
If the moisture content is positive, the aggregate has surface moisture and will contribute water to the PCC, while if the
moisture content is negative the aggregate is air dry to some degree and will absorb moisture from the PCC.
● AASHTO T 85 and ASTM C 127: Specific Gravity and Absorption of Coarse Aggregate
● AASHTO T 84 and ASTM C 128: Specific Gravity and Absorption of Fine Aggregate
2.7 Summary
Aggregates are a principal material in pavement. Additionally, they are often used in either stabilized or unstabilized base/
subbase courses. They comprise the majority of pavement volume but only account for a minority of total pavement
material costs. Therefore, a knowledge of aggregate properties is crucial to designing a high quality pavement. Aggregates
can be either natural or man-made and are most often characterized by their physical properties, including:
● Specific gravity
● Moisture content
However, aggregate chemical and material properties are also important because:
In sum, accurate aggregate characterization (physical, chemical and material) will not always ensure high quality aggregate,
but it can at least make structural and mix designers aware of a particular aggregate’s characteristics, which may aid in
critical design decisions.
3 Asphalt
Asphalt is one of the two principal constituents of HMA. Asphalt functions as Major Topics on this Page
an inexpensive (typically, $0.05/lb.), waterproof, thermoplastic, viscoelastic
adhesive. In other words, it acts as the glue that holds the road together 3.1 Background
(Anderson, Youtcheff and Zupanick, 2000). But just what is asphalt and how is
it characterized? Like many engineering substances, a vernacular definition of 3.2 Refining
"asphalt" is rather imprecise. For engineering purposes, the definition needs to 3.3 Chemical Properties
be more unequivocal. ASTM D 8 provides the following definitions:
3.4 Physical Properties
asphalt A dark brown to black cementitious material in which the 3.5 Grading Systems
predominating constituents are bitumens, which occur in nature or
3.6 Asphalt Binder Modifiers
are obtained in petroleum processing.
3.7 Other Forms of Asphalt Used in Paving
asphalt cement A fluxed or unfluxed asphalt specially prepared as to quality and
3.8 Summary
consistency for direct use in the manufacture of bituminous
pavements, and having a penetration at 25° C (77° F) of between 5
and 300, under a load of 100 grams applied for 5 seconds.
This section uses the generic term, "asphalt binder", to represent the principal binding agent in HMA. "Asphalt binder" includes
asphalt cement as well as any material added to modify the original asphalt cement properties. The term "asphalt cement" is used
to represent unmodified asphalt cement only.
3.1 Background
The first recorded use of asphalt by humans was by the Sumerians around 3,000 B.C. Statues from that time period used asphalt
as a binding substance for inlaying various shells, precious stones or pearls. Other common ancient asphalt uses were preservation
(for mummies), waterproofing (pitch on ship hulls), and cementing (used to join together bricks in Babylonia). Around 1500 A.D.,
the Incas of Peru were using a composition similar to modern bituminous macadam to pave parts of their highway system. In
more modern times, asphalt paving use first began with foot paths in the 1830s and then progressed to actual asphalt roadways in
the 1850s. The first asphalt roadways in the U.S. appeared in the early 1870s (Abraham, 1929).
In the U.S., Trinidad (near the coast of Venezuela) was the earliest source of asphalt binder. Trinidad supplied about 90 percent of
all asphalt (worldwide) from 1875 to 1900 (Baker, 1903). The asphalt was produced from a "lake" (see Figure 3.18) with a surface
area of 465,000 m2 (46.5 hectares or 115 acres) and a depth of about 24 meters (75 feet). In 1900, Tillson estimated that this
"lake" contained about 8,000,000 tonnes of "asphalt" (compare this against 1990 consumption in Europe and the U.S. of
approximately 40,000,000 tonnes (tons)). This asphalt, once free of water, was too "hard" to use in paving (Krchma and Gagle,
1974). In fact, Trinidad Lake asphalt, when loaded into a ship’s holds for transport, would fuse to the point that removal required
chopping.
Typically, producers added flux, created from petroleum distillation, to Trinidad Lake asphalt to soften it for use in early
pavements. It appears that the earliest use of asphalt binder in the U.S. was about 1874 for a project built in Washington, D.C.
This binder was a combination of Trinidad Lake asphalt and a flux distilled from crude oil. Without question, these early asphalt
binders were quite variable, making pavement mix and structural design somewhat challenging. By the 1880s, asphalt binders
were regularly produced from crude oil in California and by 1902 in Texas as well. In 1907, crude oil-based asphalt production
surpassed "natural" asphalt production (Krchma and Gagle, 1974). Today, asphalt binder for HMA pavements is produced almost
entirely from petroleum refining.
3.2 Refining
In the simplest terms, asphalt binder is simply the residue left over from petroleum refining. Thus, asphalt binders are produced
mainly by petroleum refiners and, to a lesser extent, by formulators who purchase blending stock from refiners. The composition
of base crude oil from which asphalt is refined can vary widely and thus the asphalt yield from different crude oil sources can also
vary widely.
The American Petroleum Institute (API) classifies crude oils by their API gravity. API gravity is an arbitrary expression of a
API gravity can be used as a rough estimate of asphalt yield with lower API gravity crude oils producing more asphalt (see Table
3.5). Figure 3.19 shows the composition of three very different crude oils and their associated API gravities.
Water 10
Asphalts 5 – 10
Gasoline 55
Low API gravity crude oil < 25 (yields high percentages of asphalt)
High API gravity crude oil > 25 (yields low percentages of asphalt)
Crude oil is heated in a large furnace to about 340° C (650° F) and partially vaporized. It is then fed into a distillation tower where
the lighter components vaporize and are drawn off for further processing. The residue from this process (the asphalt) is usually fed
into a vacuum distillation unit where heavier gas oils are drawn off. Asphalt cement grade is controlled by the amount of heavy
gas oil remaining. Other techniques can then extract additional oils from the asphalt. Depending upon the exact process and the
crude oil source, different asphalt cements of different properties can be produced. Additional desirable properties can be obtained
by blending crude oils before distillation or asphalt cements after distillation.
Asphalt binder specifications used to be relatively lenient, and gave refiners a high level of production flexibility. Therefore,
refiners tended to view asphalt as a simple, convenient way to use the residual material from the refinery operation. Partially as a
result of Superpave specifications, asphalt binder specifications are now more stringent and asphalt refiners increasingly perceive
asphalt as a value-added product. Superpave specifications have also caused many refiners to reevaluate their commitment to
asphalt production; some have made a strategic decision to de-emphasize or cease asphalt production, though others have renewed
their efforts to produce high-quality binders (Anderson, Youtcheff and Zupanick, 2000).
The asphalt chemical microstructure model described here is based on SHRP findings on the microstructure of asphalt using
nuclear magnetic resonance (NMR) and chromatography techniques. The SHRP findings describe asphalt microstructure as a
dispersed polar fluid (DPF). The DPF model explains asphalt microstructure as a continuous three-dimensional association of
polar molecules (generally referred to as "asphaltenes") dispersed in a fluid of non-polar or relatively low-polarity molecules
(generally referred to as "maltenes") (Little et al., 1994). All these molecules are capable of forming dipolar intermolecular bonds
of varying strength. Since these intermolecular bonds are weaker than the bonds that hold the basic organic hydrocarbon
constituents of asphalt together, they will break first and control the behavioral characteristics of asphalt. Therefore, asphalt’s
physical characteristics are a direct result of the forming, breaking and reforming of these intermolecular bonds or other properties
associated with molecular superstructures (Little et al., 1994).
The result of the above chemistry is a material that behaves (1) elastically through the effects of the polar molecule networks, and
(2) viscously because the various parts of the polar molecule network can move relative to one another due to their dispersion in
the fluid non-polar molecules.
● Aging. Some aging is reversible, some is not. Irreversible aging is generally associated with oxidation at the molecular
level. This oxidation increases an asphalt’s viscosity with age up until a point when the asphalt is able to quench (or
halt) oxidation through immobilization of the most chemically reactive elements. Reversible aging is generally
associated with the effects of molecular organization. Over time, the molecules within asphalt will slowly reorient
themselves into a better packed, more bound system. This results in a stiffer, more rigid material. This thixotropic
aging can be reversed by heating and agitation.
● Rutting and permanent deformation. If the molecular network is relatively simple and not interconnected, asphalt will
tend to deform inelastically under load (e.g., not all the deformation is recoverable). Additionally, asphalts with higher
percentages of non-polar dispersing molecules are better able to flow and plastically deform because the various polar
molecule network pieces can more easily move relative to one another due to the greater percentage of fluid non-polar
molecules.
● Fatigue cracking. If the molecular network becomes too organized and rigid, asphalt will fracture rather than deform
elastically under stress. Therefore, asphalts with higher percentages of polar, network-forming molecules may be more
susceptible to fatigue cracking.
● Thermal cracking. At lower temperatures even the normally fluid non-polar molecules begin to organize into a
structured form. Combined with the already-structured polar molecules, this makes asphalt more rigid and likely to
fracture rather than deform elastically under stress.
● Stripping. Asphalt adheres to aggregate because the polar molecules within the asphalt are attracted to the polar
molecules on the aggregate surface. Certain polar attractions are known to be disrupted by water (itself a polar
molecule). Additionally, the polar molecules within asphalt will vary in their ability to adhere to any one particular type
of aggregate.
● Moisture damage. Since it is a polar molecule, water is readily accepted by the polar asphalt molecules. Water can
cause stripping and/or can decrease asphalt viscosity. It typically acts like a solvent in asphalt and results in reduced
strength and increased rutting. When taken to the extreme, this same property can be used to produce asphalt
emulsions. Interestingly, from a chemical point-of-view water should have a greater effect on older asphalt. Oxidation
causes aged (or older) asphalts to contain more polar molecules. The more polar molecules an asphalt contains, the
more readily it will accept water. However, the oxidation aging effects probably counteract any moisture-related aging
effects.
In summary, asphalt is a complex chemical substance. Although basic chemical composition is important, it is an asphalt’s
chemical microstructure that is most influential in its physical behavior. Although most basic asphalt binder failure mechanisms
can be described chemically, currently there is not enough asphalt chemical knowledge to adequately predict performance.
Therefore, physical properties and tests are used.
The earliest physical tests were empirically derived tests. Some of these tests (such as the penetration test) have been used for the
better part of the 20th century with good results. Later tests (such as the viscosity tests) were first attempts at using fundamental
engineering parameters to describe asphalt binder physical properties. Ties between tested parameters and field performance were
still quite tenuous. Superpave binder tests, developed in the 1980s and 1990s, were developed with the goal of measuring specific
asphalt binder physical properties that are directly related to field performance by engineering principles. These tests are generally
a bit more complex but seem to accomplish a more thorough characterization of the tested asphalt binder.
This subsection, taken largely from Roberts et al. (1996), describes the more common U.S. asphalt binder physical tests. Asphalt
binder tests specifically developed or adopted by the Superpave research effort are noted by a " – Superpave" in their title.
Sections that discuss Superpave tests also discuss relevant field performance information as well as the engineering principles used
to develop the relationship between test and field performance.
3.4.1 Durability
Durability is a measure of how asphalt binder physical properties change with age (sometimes called age hardening). In general,
as an asphalt binder ages, its viscosity increases and it becomes more stiff and brittle. Age hardening is a result of a number of
factors, the principal ones being (Vallerga, Monismith and Grahthem, 1957 and Finn, 1967 as referenced by Roberts et al., 1996):
● Volatilization. The evaporation of the lighter constituents of asphalt binder. It is primarily a function of temperature
and occurs principally during HMA production.
● Polymerization. The combining of like molecules to form larger molecules. These larger molecules are thought to
cause a progressive hardening.
● Thixotropy. The property of asphalt binder whereby it "sets" when unagitated. Thixotropy is thought to result from
hydrophilic suspended particles that form a lattice structure throughout the asphalt binder. This causes an increase in
viscosity and thus, hardening (Exxon, 1997). Thixotropic effects can be somewhat reversed by heat and agitation.
HMA pavements with little or no traffic are generally associated with thixotropic hardening.
● Syneresis. The separation of less viscous liquids from the more viscous asphalt binder molecular network. The liquid
loss hardens the asphalt and is caused by shrinkage or rearrangement of the asphalt binder structure due to either
physical or chemical changes. Syneresis is a form of bleeding (Exxon, 1997).
● Separation. The removal of the oily constituents, resins or asphaltenes from the asphalt binder by selective absorption
of some porous aggregates.
There is no direct measure for asphalt binder aging. Rather, aging effects are accounted for by subjecting asphalt binder samples
to simulated aging then conducting other standard physical tests (such as viscosity, dynamic shear rheometer (DSR), bending beam
rheometer (BBR) and the direct tension test (DTT)). Simulating the effects of aging is important because an asphalt binder that
possesses a certain set of properties in its as-supplied state, may possess a different set of properties after aging. Asphalt binder
aging is usually split up into two categories:
● Short-term aging. This occurs when asphalt binder is mixed with hot aggregates in an HMA mixing facility.
● Long-term aging. This occurs after HMA pavement construction and is generally due to environmental exposure and
loading.
The thin-film oven (TFO) test simulates short-term aging by heating a film of asphalt binder in an oven for 5 hours at 163° C (325°
F). The effects of heat and air are determined from changes incurred in physical properties measured before and after the oven
treatment by other test procedures.
● AASHTO T 179 and ASTM D 1754: Effects of Heat and Air on Asphalt Materials (Thin-Film Oven Test)
The rolling thin-film oven (RTFO) test (see Figure 3.20) simulates short-term aging by heating a moving film of asphalt binder in
an oven for 85 minutes at 163° C (325° F). The effects of heat and air are determined from changes incurred in physical properties
measured before and after the oven treatment by other test procedures. The moving film is created by placing the asphalt binder
sample in a small jar (see Figure 3.21) then placing the jar in a circular metal carriage that rotates within the oven. The RTFO test
is generally considered superior to the TFO because:
● It achieves the same degree of hardening (aging) in less time (85 minutes vs. 5 hours)
❍ Allows continuous exposure of fresh asphalt binder to heat and air flow
❍ Prevents the formation of a surface skin on the sample, which may inhibit aging
Although it has been in common use by some western states for some time, Superpave adopted the RTFO test to simulate short-
term asphalt binder aging.
● AASHTO T 240 and ASTM D 2872: Effects of Heat and Air on a Moving Film of Asphalt ( Rolling Thin-Film Oven
Test)
The pressure aging vessel (PAV) (see Figure 3.22) was adopted by Superpave to simulate the effects of long-term asphalt binder
aging that occurs as a result of 5 to 10 years HMA pavement service (Bahia and Anderson, 1994). Prior to Superpave, the general
concept of the pressure aging vessel had been used for many years in rubber product aging. The PAV is an oven-pressure vessel
combination that takes RTFO aged samples (see Figure 3.23) and exposes them to high air pressure (2070 kPa (300 psi)) and
temperature (90° C (195° F), 100° C (212° F)° or 110° C (230° F) depending upon expected climatic conditions) for 20 hours.
● The oxidation process can be accelerated without resorting to extremely high temperatures
● AASHTO PP1: Practice for Accelerated Aging of Asphalt Binder Using a Pressurized Aging Vessel
3.4.2 Rheology
Rheology is the study of deformation and flow of matter. Deformation and flow of the asphalt binder in HMA is important in
determining HMA pavement performance. HMA pavements that deform and flow too much may be susceptible to rutting and
bleeding, while those that are too stiff may be susceptible to fatigue or thermal cracking. HMA pavement deformation is closely
related to asphalt binder rheology. Since the rheological properties of asphalt binder vary with temperature, rheological
characterization involves two key considerations:
● To compare different asphalt binders, their rheological properties must be measured at some common reference
temperature.
● To fully characterize an asphalt binder, its rheological properties must be examined over the range of temperatures that
it may encounter during its life.
3.4.2.1 Chewing
Originally, the degree of asphalt binder softening was determined by chewing (Halstead and Welborn, 1974). A sample of asphalt
binder was literally chewed to subjectively determine its softness. This method is no longer in use today for obvious reasons.
Aside from chewing, the penetration test is the oldest asphalt test. In 1888, H.C. Bowen of the Barber Asphalt Paving Company
invented the forerunner to the penetration test, the Bowen Penetration Machine (Halstead and Welborn, 1974). It’s basic principle,
and the basic principle of the penetration test, was to determine the depth to which a truncated No. 2 sewing needle penetrated an
asphalt sample under specified conditions of load, time and temperature. In 1915, ASTM even went as far as specifying the brand
of needle (R.J. Roberts Parabola Sharps No. 2) (Halstead and Welborn, 1974).
The current penetration test (see Figure 3.24), first published in 1959, describes the following basic procedure:
● Melt and cool the asphalt binder sample under controlled conditions.
● Measure the penetration of a standard needle into the asphalt binder sample under the following conditions:
❍ Time = 5 seconds
The depth of penetration is measured in units of 0.1 mm and reported in penetration units (e.g., if the needle penetrates 8 mm, the
asphalt penetration number is 80). Penetration grading is based on the penetration test.
The softening point is defined as the temperature at which a bitumen sample can no longer support the weight of a 3.5-g steel ball.
Although it is commonly used in Europe, it is mostly used for roofing asphalts in the U.
S. Basically, two horizontal disks of bitumen, cast in shouldered brass rings (see Figure
3.25), are heated at a controlled rate in a liquid bath while each supports a steel ball. The
softening point is reported as the mean of the temperatures at which the two disks soften
enough to allow each ball, enveloped in bitumen, to fall a distance of 25 mm (1.0 inch)
(AASHTO, 2000).
τ = shear stress
γ = shear rate
Asphalt binder viscosity is typically measured at 60° C (140° F) because it approximates the maximum HMA pavement surface
temperature during summer in the U.S.
The basic absolute viscosity test measures the time it takes for a fixed volume of asphalt binder to be drawn up through a capillary
tube by means of vacuum, under closely controlled conditions of vacuum and temperature (ASTM, 2001). Although absolute
viscosity is an improvement over the penetration test, it still only measures viscosity at one temperature and thus does not fully
characterize an asphalt binder’s consistency over the expected range of construction and service conditions.
● AASHTO T 202 and ASTM D 2171: Viscosity of Asphalts by Vacuum Capillary Viscometer
The kinematic viscosity of a liquid is the absolute (or dynamic) viscosity divided by the density of the liquid at the temperature of
measurement. The 135° C (275° F) measurement temperature was chosen to simulate the mixing and laydown temperatures
typically encountered in HMA pavement construction.
The basic kinematic viscosity test measures the time it takes for a fixed volume of asphalt binder to flow through a capillary
viscometer under closely controlled conditions of head and temperature (ASTM, 2001).
The ductility test (see Figure 3.26) measures asphalt binder ductility by stretching a standard-sized briquette of asphalt binder (see
Figure 3.27) to its breaking point. The stretched distance in centimeters at breaking is then reported as ductility. Like the
penetration test, this test has limited use since it is empirical and conducted at only one temperature (25° C (77° F)).
The rotational viscometer (RV) (see Figures 3.28 and 3.29) is used in the Superpave system to test high temperature viscosities
(the test is conducted at 135° C (275° F)). The basic RV test measures the torque required to maintain a constant rotational speed
(20 RPM) of a cylindrical spindle while submerged in an asphalt binder at a constant temperature (see Figure 3.30). This torque is
then converted to a viscosity and displayed automatically by the RV.
The RV high-temperature viscosity measurements are meant to simulate binder workability at mixing and laydown temperatures.
Since the goal is to ensure the asphalt binder is sufficiently fluid for pumping and mixing, Superpave specifies a maximum RV
viscosity. The RV is more suitable than the capillary viscometer (used for kinematic viscosity) for testing modified asphalt
binders because some modified asphalt binders (such as those containing crumb rubber particles) can clog the capillary viscometer
and cause faulty readings.
● AASHTO TP 48 and ASTM D 4402: Viscosity Determination of Asphalt Binder Using Rotational Viscometer
The dynamic shear rheometer (DSR) (see Figure 3.31) is used in the Superpave system for testing medium to high temperature
viscosities (the test is conducted between 46° C (115° F) and 82° C (180° F)). The actual temperatures anticipated in the area
where the asphalt binder will be placed determine the test temperatures used.
Figure 3.31: Dynamic Shear Rheometer Figure 3.32: Dynamic Shear Rheometer Samples
The basic DSR test uses a thin asphalt binder sample (see Figure 3.32) sandwiched between two plates. The lower plate is fixed
while the upper plate oscillates back and forth across the sample at 1.59 Hz to create a shearing action (see Figure 3.33). These
oscillations at 1.59 Hz (10 radians/sec) are meant to simulate the shearing action corresponding to a traffic speed of about 90 km/
hr (55 mph) (Roberts et al., 1996). The following equations are then used to determine a complex shearing modulus, G* and a
phase angle, δ:
δ = phase angle. This is the time lag (expressed in radians) between the maximum applied
shear stress and the maximum resulting shear strain. For a typical neat asphalt (no
modifiers) the phase angle is about 88 - 89°, while some modified binders can
have phase angles as low as 60°.
Asphalt binders in the medium to high temperature range behave partly like an elastic solid (deformation due to loading is
recoverable – it is able to return to its original shape after a load is removed) and a viscous liquid (deformation due to loading is
non-recoverable – it cannot return to its original shape after a load is removed). By measuring G* and δ , the DSR is able to
determine the total complex shear modulus as well as its elastic and viscous components (see Figure 3.34).
1. Rutting. In order to resist rutting, an asphalt binder should be stiff (not deform too much) and it should be elastic (it
should be able to return to its original shape after load deformation). Therefore, the complex shear modulus elastic
portion, G*cosδ (see Figure 3.34), should be large. Therefore, when rutting is of greatest concern (during an HMA
pavement’s early and mid life), Superpave specifies a minimum value for the elastic component of the complex shear
modulus. Intuitively, the higher the G* value, the stiffer the asphalt binder is (able to resist deformation), and the lower
the δ value, the greater the elastic portion of G* is (able to recover its original shape after being deformed by a load).
2. Fatigue. In order to resist fatigue cracking, an asphalt binder should be elastic (able to dissipate energy by rebounding
and not cracking) but not too stiff (excessively stiff substances will crack rather than deform-then-rebound). Therefore,
the complex shear modulus viscous portion, G*sinδ (see Figure 3.34), should be small. Therefore, when fatigue
cracking is of greatest concern (late in an HMA pavement’s life), Superpave specifies a maximum value for the viscous
component of the complex shear modulus. This relationship between G*sinδ and fatigue cracking is more tenuous than
the rutting relationship discussed in #1.
Note that although they appear similar, specifying a large G*cosδ and a small G*sinδ are not the same. They both involve small
phase angles (δ ) but the key is getting an asphalt binder whose complex shear modulus (G*) is neither too large nor too small.
● AASHTO TP 5: Determining the Rheological Properties of Asphalt Binder Using a Dynamic Shear Rheometer (DSR)
The bending beam rheometer (BBR, see Figure 3.35) is used in the Superpave system to test asphalt binders at low temperatures
where the chief failure mechanism is thermal cracking. The BBR basically subjects a simple asphalt beam to a small (100-g) load
over 240 seconds (see Figure 3.36). Then, using basic beam theory, the BBR calculates beam stiffness (S(t)) and the rate of change
of that stiffness (m-value) as the load was applied.
The m-value is simply the rate of change of the stiffness at time, t = 60 seconds and is used to describe how the asphalt binder
relaxes under load.
The BBR test is meant to simulate asphalt binder stiffness after two hours of loading at the minimum HMA pavement design
temperature. Creep stiffness (S(t)) is related to thermal stresses in an HMA pavement due to shrinking while the m-value is related
to the ability of an HMA pavement to relieve these stresses. Thus, Superpave binder specifications require a maximum limit on
creep stiffness (thermal stress not too great) and a minimum limit on m-value (must have some minimum ability to relieve thermal
stresses without cracking).
● AASHTO TP1: Method for Determining the Flexural Creep Stiffness of Asphalt Binder Using the Bending Beam
Rheometer
The DTT basically loads a small sample of asphalt binder in tension until it breaks (see Figure 3.38). The failure strain is then
calculated from the following equation:
Le = effective length
If a particular asphalt binder has a high BBR creep stiffness (indicating high thermal stress), it must have a minimum failure strain
(indicating it will stretch rather than crack) to meet Superpave binder specifications.
● AASHTO TP 3: Method for Determining the Fracture Properties of Asphalt Binder in Direct Tension (DT)
A typical flash point test involves heating a small sample of asphalt binder in a test cup. The temperature of the sample is
increased and at specified intervals a test flame is passed across the cup. The flash point is the lowest liquid temperature at which
application of the test flame causes the vapors of the sample to ignite. The test can be continued up to the fire point – the point at
which the test flame causes the sample to ignite and remain burning for at least 5 seconds.
● AASHTO T 48 and ASTM D 92: Flash and Fire Points by Cleveland Open Cup (more common for asphalt cement used
in HMA)
3.4.4 Purity
Asphalt cement, as used for HMA paving, should consist of almost pure bitumen. Impurities are not active cementing constituents
and may be detrimental to asphalt cement performance. Mineral impurities can be quantified by dissolving a sample of asphalt
cement in trichloroethylene or 1,1,1 trichloroethane through a filter mat. Anything remaining on the mat is considered an
impurity. Water impurities are quantified through distillation.
● AASHTO T 55 and ASTM D 95: Water in Petroleum Products and Bituminous Materials by Distillation
● AASHTO T 110 and ASTM D 1461: Moisture or Volatile Distillates in Bituminous Paving Mixtures
● AASHTO T 228 and ASTM D 70: Specific Gravity and Density of Semi-Solid Bituminous Materials
Basically, the spot test is a form of paper chromatography (a method for analyzing complex mixtures by separating them into the
chemicals from which they are made). A small drop of prepared asphalt cement is dropped onto a filter paper. If the spot formed
is uniformly brown then the test is negative. If the spot formed is brown with a black center then the test is positive. Today, the
spot test is rarely used.
WSDOT uses the Superpave asphalt binder performance grading system and specifications. Therefore, asphalt
binder must meet the requirements of AASHTO MP 1. WSDOT uses three baseline asphalt binder performance
grades based on geography. These baseline grades are typically used and then adjusted as necessary.
Previously, WSDOT had used the aged residue (AR) viscosity grading. The commonly used grade in this old
system was AR-4000W.
● Solubility in trichloroethylene
❍ Retained penetration
Penetration grading’s basic assumption is that the less viscous the asphalt, the deeper the needle will penetrate. This penetration
depth is empirically (albeit only roughly) correlated with asphalt binder performance. Therefore, asphalt binders with high
penetration numbers (called "soft") are used for cold climates while asphalt binders with low penetration numbers (called "hard")
are used for warm climates. Penetration grading key advantages and disadvantages are listed in Table 3.6.
Advantages Disadvantages
May also provide a better correlation with low-temperature Shear rate is variable and high during the test. Since
asphalt binder properties than the viscosity test, which is asphalt binders typically behave as a non-Newtonian fluid
performed at 60° C (140° F). at 25° C (77° F), this will affect test results.
The test is quick and inexpensive. Therefore, it can easily The test does not provide information with which to
be used in the field. establish mixing and compaction temperatures.
Penetration grades are listed as a range of penetration units (one penetration unit = 0.1 mm) such as 120 – 150. Penetration grades
specified in AASHTO M 20 and ASTM D 946 are listed in Table 3.7.
40 – 50 Hardest grade.
60 - 70
Typical grades used in the U.S.
85 - 100
120 – 150
A few states still have provisions for the penetration grading system. These will most likely disappear as the Superpave PG
system becomes more prevalent.
In the early 1960s an improved asphalt grading system was developed that incorporated a rational scientific viscosity test. This
scientific test replaced the empirical penetration test as the key asphalt binder characterization. Viscosity grading quantifies the
following asphalt binder characteristics:
● Solubility in trichloroethylene
● Thin film oven test (accounts for the effects of short-term aging)
:
Viscosity grading can be done on original (as-supplied) asphalt binder samples (called AC grading) or aged residue samples
(called AR grading). The AR viscosity test is based on the viscosity of aged residue from the rolling thin film oven test. With AC
grading, the asphalt binder is characterized by the properties it possesses before it undergoes the HMA manufacturing process.
The AR grading system is an attempt to simulate asphalt binder properties after it undergoes a typical HMA manufacturing process
and thus, it should be more representative of how asphalt binder behaves in HMA pavements. Table 3.8 lists key advantages and
disadvantages of the viscosity grading system.
Advantages Disadvantages
Unlike penetration depth, viscosity is a fundamental The principal grading (done at 25° C (77° F)) may not
engineering parameter. accurately reflect low-temperature asphalt binder rheology.
Viscosity is measured in poise (cm-g-s = dyne-second/cm2, named after Jean Louis Marie Poiseuille). The lower the number of
poises, the lower the viscosity and thus the more easily a substance flows. Thus, AC-5 (viscosity is 500 ± 100 poise at 60° C (140°
F)) is less viscous than AC-40 (viscosity is 4000 ± 800 poise at 60° C (140° F)). Table 3.9 shows standard viscosity grades for the
AC and AR grading systems from AASHTO M 226 and ASTM D 3381. Typical grades used for HMA paving in the U.S. are AC-
10, AC-20, AC-30, AR-4000 and AR 8000.
Standard Grading based on Original Asphalt (AC) Grading based on Aged Residue (AR)
AASHTO
AC-2.5 AC-5 AC-10 AC-20 AC-30 AC-40 AR-10 AR-20 AR-40 AR-80 AR-160
M 226
ASTM
AC-2.5 AC-5 AC-10 AC-20 AC-30 AC-40 AR-1000 AR-2000 AR-4000 AR-8000 AR-16000
D 3381
Superpave performance grading (PG) is based on the idea that an HMA asphalt binder’s properties should be related to the
conditions under which it is used. For asphalt binders, this involves expected climatic conditions as well as aging considerations.
Therefore, the PG system uses a common battery of tests (as the older penetration and viscosity grading systems do) but specifies
that a particular asphalt binder must pass these tests at specific temperatures that are dependant upon the specific climatic
conditions in the area of use. Therefore, a binder used in the Sonoran Desert of California/Arizona/Mexico would have different
properties than one used in the Alaskan tundra. This concept is not new – selection of penetration or viscosity graded asphalt
binders follows the same logic – but the relationships between asphalt binder properties and conditions of use are more complete
and more precise with the Superpave PG system. Information on how to select a PG asphalt binder for a specific condition is
contained in Module 5, Section 5, Superpave Method. Table 3.10 shows how the Superpave PG system addresses specific
penetration, AC and AR grading system general limitations.
Table 3.10: Prior Limitations vs. Superpave Testing and Specification Features
(after Roberts et al., 1996)
Limitations of Penetration, AC and AR Grading Superpave Binder Testing and Specification Features
Systems that Address Prior Limitations
Penetration and ductility tests are empirical and not The physical properties measured are directly related to
directly related to HMA pavement performance. field performance by engineering principles.
Asphalt binders can have significantly different Grading is more precise and there is less overlap
characteristics within the same grading category. between grades.
Superpave performance grading is reported using two numbers – the first being the average seven-day maximum pavement
temperature (°C) and the second being the minimum pavement design temperature likely to be experienced (°C). Thus, a PG 58-22
is intended for use where the average seven-day maximum pavement temperature is 58°C and the expected minimum pavement
temperature is -22°C. Notice that these numbers are pavement temperatures and not air temperatures (these pavement
temperatures are estimated from air temperatures using an algorithm contained in the LTPP Bind program). As a general rule-of-
thumb, PG binders that differ in the high and low temperature specification by 90°C or more generally require some
sort of modification (see Table 3.11).
WSDOT uses the Superpave asphalt binder performance grading system and specifications. Therefore, asphalt
binder must meet the requirements of AASHTO MP 1. WSDOT uses three baseline asphalt binder performance
grades based on geography. These baseline grades are typically used and then adjusted as necessary.
Previously, WSDOT had used the aged residue (AR) viscosity grading. The commonly used grade in this old
system was AR-4000W.
● AASHTO PP6: Practice for Grading or Verifying the Performance Grade of an Asphalt Binder
● Increased demand on HMA pavements. Traffic volume, loads and tire pressures have increased substantially in recent
years, which can cause increased rutting and cracking. Many modifiers can improve the asphalt binder's stiffness at
normal service temperatures to increase rut resistance, while decreasing its stiffness at low temperatures to improve its
resistance to thermal cracking.
● Superpave asphalt binder specifications. Superpave asphalt binder specifications developed in the 1990s require asphalt
binders to meet stiffness requirements at both high and low temperatures. In regions with extreme climatic conditions
this is not possible without asphalt binder modification.
● Environmental and economic issues. It is both environmentally and economically sound to recycle waste and industrial
byproducts (such as tires, roofing shingles, glass and ash) in order to gain added benefit. Thus, when they can benefit
the final product without creating an environmental liability they are often used as additives in HMA.
● Public agency willingness to fund higher-cost asphalt additives. Modified asphalt cement is usually higher in initial cost
than unmodified asphalt cement, but it should provide a longer service life with less maintenance.
There are numerous binder additives available on the market today. The benefits of modified asphalt cement can only be realized
by a judicious selection of the modifier(s); not all modifiers are appropriate for all applications. In general, asphalt cement should
be modified to achieve the following types of improvements (Roberts et al., 1996):
● Lower stiffness (or viscosity) at the high temperatures associated with construction. This facilitates pumping of the
liquid asphalt binder as well as mixing and compaction of HMA.
● Higher stiffness at high service temperatures. This will reduce rutting and shoving.
● Lower stiffness and faster relaxation properties at low service temperatures. This will reduce thermal cracking.
● Increased adhesion between the asphalt binder and the aggregate in the presence of moisture. This will reduce the
likelihood of stripping. Figure 3.39 shows two aggregate samples from the same source after they have been coated
with asphalt binder. The asphalt binder used with the sample on the left contain no anti-stripping modifier, which
resulted in almost no aggregate-asphalt binder adhesion. The asphalt binder used with the sample on the right contains
0.5% (by weight of asphalt binder) of an anti-stripping modifier, which results in good aggregate-asphalt binder
adhesion.
Generally, emulsions appear as a thick brown liquid when initially applied (see Figure 3.40). When the asphalt cement starts to
adhere to the surrounding material (aggregate, existing surface, subgrade, etc.) the color changes from brown to black (see Figure
3.41) and the emulsion is said to have "broken" (see Figure 3.42). As water begins to evaporate, the emulsion begins to behave
more and more like pure asphalt cement. Once all the water has evaporated, the emulsion is said to have "set". The time required
to break and set depends upon the type of emulsion, the application rate, the temperature of the surface onto which it is applied and
environmental conditions (TRB, 2000). Under most circumstances, an emulsion will set in about 1 to 2 hours (TRB, 2000).
ASTM D 3628 contains guidance on selection and use of emulsified asphalt.
● Environmental regulations. Cutback asphalts contain volatile chemicals that evaporate into the atmosphere. Emulsified
asphalts evaporate water into the atmosphere.
● Loss of high energy products. The petroleum solvents used require higher amounts of energy to manufacture and are
expensive compared to the water and emulsifying agents used in emulsified asphalts.
In many places, cutback asphalt use is restricted to patching materials for use in cold weather.
WSDOT does not specify cutback asphalts because of their potential effect
on the environment.
3.8 Summary
Humans have used asphalt for thousands of years. In the HMA paving industry, asphalt functions as an adhesive that holds
aggregate together. Currently, HMA use exceeds 500,000,000 tonnes (tons) per year at a cost of almost $3 billion per year
(Anderson, Youtcheff and Zupanick, 2000). Although natural sources still exist, today’s asphalt is almost entirely produced from
petroleum refining. Asphalt cement can also be modified using certain chemical and organic products to alter its behavior.
Modern asphalt binder produced using the PG system is often modified.
Asphalt binders can be characterized by chemical and physical properties. Chemically, asphalt is a mixture of polar and non-polar
complex organic molecules. The microstructure of these molecules tends to govern asphalt’s physical behavior. Since chemical
knowledge and testing is limited, asphalt is most commonly described by its physical attributes. Over the years many tests have
been developed to fully characterize asphalt’s physical attributes. To date, these tests have reached an apogee with the Superpave
binder tests. Superpave tests measure specific asphalt binder physical properties that are directly related to field performance by
engineering principles. Thus, theoretically they offer the best and most complete asphalt binder characterization. They are also
the most complex and the most expensive.
Using the tests discussed in this section, asphalt binders are classified for use (graded) based on their physical properties as
measured through testing. The most common asphalt binder classifications are: penetration grade, viscosity grade and performance
grade (from Superpave). These asphalt grades are what is generally specified in HMA mix design.
Although this section has concentrated on asphalt binder characterizations and tests associated with HMA, asphalt binder is also
used in other road-related products: emulsions, cutbacks and foamed asphalt. These products are often used in an HMA
pavement's supporting layers as well as by themselves for low-volume roads.
Of all the HMA pavement constituents, we have the most control over the asphalt binder. Generally, roads will be built where
they can or need to be regardless of the subgrade, and aggregate is usually taken from the closest source as long as it meets
minimum standards. However, we generally specify asphalt binder characteristics for each and every HMA pavement. This is
reflected in the substantial level of effort put forth to accurately characterize asphalt binder.
4 Portland Cement
Portland cement is the chief ingredient in cement paste - the binding agent in Major Topics on this Page
portland cement concrete (PCC). It is a hydraulic cement that, when
combined with water, hardens into a solid mass. Interspersed in an 4.1 Background
aggregate matrix it forms PCC. As a material, portland cement has been
used for well over 175 years and, from an empirical perspective, its behavior 4.2 Manufacturing
is well-understood. Chemically, however, portland cement is a complex 4.3 Chemical Properties
substance whose mechanisms and interactions have yet to be fully defined.
ASTM C 125 and the Portland Cement Association (PCA) provide the 4.4 Types of Portland Cement
following precise definitions:
4.5 Physical Properties
hydraulic cement An inorganic material or a mixture of inorganic materials that 4.6 Summary
sets and develops strength by chemical reaction with water by
formation of hydrates and is capable of doing so under water.
4.1 Background
Although the use of cements (both hydraulic and non-hydraulic) goes back many
thousands of years (to ancient Egyptian times at least), the first occurrence of
"portland cement" came about in the 19th century. In 1824, Joseph Aspdin, a Leeds
mason took out a patent on a hydraulic cement that he coined "Portland" cement
(Mindess and Young, 1981). He named the cement because it produced a concrete
that resembled the color of the natural limestone quarried on the Isle of Portland, a
peninsula in the English Channel (see Figure 3.43 and 3.44). Since then, the name
"portland cement" has stuck and is written in all lower case because it is now
recognized as a trade name for a type of material and not a specific reference to
Portland, England.
Figure 3.43: Portland, England
Today, portland cement is the most widely used building material in the world with
about 1.56 billion tonnes (1.72 billion tons) produced each year. Annual global
production of portland cement concrete hovers around 3.8 million cubic meters (5
billion cubic yards) per year (Cement Association of Canada, 2001). In the U.S.,
rigid pavements are the largest single use of portland cement and portland cement
concrete (ACPA, 2002).
4.2 Manufacturing
Although there are several variations of commercially manufactured portland cement, they
each share many of the same basic raw materials and chemical components. The chief
chemical components of portland cement are calcium, silica, alumina and iron. Calcium is
derived from limestone, marl or chalk, while silica, alumina and iron come from the sands,
clays and iron ore sources. Other raw materials may include shale, shells and industrial
byproducts such as mill scale (Ash Grove Cement Company, 2000).
The basic manufacturing process heats these materials in a kiln to about 1400 to 1600°
C (2600 - 3000°F) - the temperature range in which the two materials interact
chemically to form calcium silicates (Mindess and Young, 1981). This heated
substance, called "clinker" is usually in the form of small gray-black pellets about
12.5 mm (0.5 inches) in diameter. Clinker is then cooled and pulverized into a fine
powder that almost completely passes through a 0.075 mm (No. 200) sieve and Figure 3:44: Limestone at the
fortified with a small amount of gypsum. The result is portland cement. The Portland Portland Bill near Weymouth
Cement Association (PCA) has an excellent interactive illustration of this process on
their website.
Table 3.12: Main Constituents in a Typical Portland Cement (Mindess and Young, 1981)
4.3.2 Hydration
When portland cement is mixed with water its chemical compound constituents undergo a series of chemical reactions that
cause it to harden (or set). These chemical reactions all involve the addition of water to the basic chemical compounds listed
in Table 3.12. This chemical reaction with water is called "hydration". Each one of these reactions occurs at a different time
and rate. Together, the results of these reactions determine how portland cement hardens and gains strength.
● Tricalcium silicate (C3S). Hydrates and hardens rapidly and is largely responsible for initial set and early strength.
Portland cements with higher percentages of C3S will exhibit higher early strength.
● Dicalcium silicate (C2S). Hydrates and hardens slowly and is largely responsible for strength increases beyond one
week.
● Tricalcium aluminate (C3A). Hydrates and hardens the quickest. Liberates a large amount of heat almost
immediately and contributes somewhat to early strength. Gypsum is added to portland cement to retard C3A
hydration. Without gypsum, C3A hydration would cause portland cement to set almost immediately after adding
water.
● Tetracalcium aluminoferrite (C4AF). Hydrates rapidly but contributes very little to strength. Its use allows lower
kiln temperatures in portland cement manufacturing. Most portland cement color effects are due to C4AF.
Figure 3.46 shows rates of heat evolution, which give an approximate idea of hydration times and when a typical portland
cement initially sets.
Figure 3.46: Rate of Heat Evolution During Hydration of a Typical Portland Cement
The result of the two silicate hydrations is the formation of a calcium silicate hydrate (often written C-S-H because of is
variable stoichiometry). C-S-H makes up about 1/2 - 2/3 the volume of the hydrated paste (water + cement) and therefore
dominates its behavior (Mindess and Young, 1981).
WSDOT specifies that portland cement shall conform to the requirements for Types I, II or III cement as listed
in AASHTO M 85. Type II cement shall additionally meet the requirements for setting time by the Vicat
method.
Used as a precaution against moderate sulfate attack. It will usually generate less heat
II Moderate Sulfate Resistance
at a slower rate than Type I cement.
Used when high early strength is needed. It is has more C3S than Type I cement and
III High Early Strength has been ground finer to provide a higher surface-to-volume ratio, both of which speed
hydration. Strength gain is double that of Type I cement in the first 24 hours.
Used when hydration heat must be minimized in large volume applications such as
IV Low Heat of Hydration gravity dams. Contains about half the C3S and C3A and double the C2S of Type I
cement.
Keep in mind that these tests are, in general, performed on "neat" cement pastes - that is, they only include portland cement
and water. Neat cement pastes are typically difficult to handle and test and thus they introduce more variability into the
results. Cements may also perform differently when used in a "mortar" (cement + water + sand). Over time, mortar tests have
been found to provide a better indication of cement quality and thus, tests on neat cement pastes are typically used only for
research purposes (Mindess and Young, 1981). However, if the sand is not carefully specified in a mortar test, the results may
not be transferable.
4.5.1 Fineness
Fineness, or particle size of portland cement affects hydration rate and thus the rate of strength gain. The smaller the particle
size, the greater the surface area-to-volume ratio, and thus, the more area available for water-cement interaction per unit
volume. The effects of greater fineness on strength are generally seen during the first seven days (PCA, 1988).
● AASHTO T 128 and ASTM C 184: Fineness of Hydraulic Cement by the 150-µm (No. 100) and 75-µm (No. 200)
Sieves
● AASHTO T 153 and ASTM C 204: Fineness of Hydraulic Cement by Air Permeability Apparatus
● AASHTO T 192 and ASTM C 430: Fineness of Hydraulic Cement by the 45-µm (No. 325) Sieve
4.5.2 Soundness
When referring to portland cement, "soundness" refers to the ability of a hardened cement paste to retain its volume after setting without
delayed destructive expansion (PCA, 1988). This destructive expansion is caused by excessive amounts of free lime (CaO) or magnesia
(MgO). Most portland cement specifications limit magnesia content and expansion. The typical expansion test places a small sample of
cement paste into an autoclave (a high pressure steam vessel). The autoclave is slowly brought to 2.03 MPa (295 psi) then kept at that
pressure for 3 hours. The autoclave is then slowly brought back to room temperature and atmospheric pressure. The change in specimen
length due to its time in the autoclave is measured and reported as a percentage. ASTM C 150, Standard Specification for Portland
Cement specifies a maximum autoclave expansion of 0.80 percent for all portland cement types.
2. Final set. Occurs when the cement has hardened to the point at which it can sustain some load.
These particular times are just arbitrary points used to characterize cement, they do not have any fundamental chemical
significance. Both common setting time tests, the Vicat needle and the Gillmore needle, define initial set and final set based
on the time at which a needle of particular size and weight either penetrates a cement paste sample to a given depth or fails to
penetrate a cement paste sample. The Vicat needle test is more common and tends to give shorter times than the Gillmore
needle test. Table 3.14 shows ASTM C 150 specified set times.
● AASHTO T 131 and ASTM C 191: Time of Setting of Hydraulic Cement by Vicat Needle
4.5.4 Strength
Cement paste strength is typically defined in three ways: compressive, tensile and flexural. These strengths can be affected by
a number of items including: water-cement ratio, cement-fine aggregate ratio, type and grading of fine aggregate, manner of
mixing and molding specimens, curing conditions, size and shape of specimen, moisture content at time of test, loading
conditions and age (Mindess and Young, 1981). Since cement gains strength over time, the time at which a strength test is to
be conducted must be specified. Typically times are 1 day (for high early strength cement), 3 days, 7 days, 28 days and 90
days (for low heat of hydration cement). When considering cement paste strength tests, there are two items to consider:
● Cement mortar strength is not directly related to concrete strength. Cement paste strength is typically used as a
quality control measure.
● Strength tests are done on cement mortars (cement + water + sand) and not on cement pastes.
The most common strength test, compressive strength, is carried out on a 50 mm (2-inch) cement mortar test specimen. The
test specimen is subjected to a compressive load (usually from a hydraulic machine) until failure. This loading sequence must
take no less than 20 seconds and no more than 80 seconds. Table 3.15 shows ASTM C 150 compressive strength
specifications.
Table 3.15: ASTM C 150 Portland Cement Mortar Compressive Strength Specifications in MPa (psi)
● AASHTO T 106 and ASTM C 109: Compressive Strength of Hydraulic Cement Mortars (Using 50-mm or 2-in.
Cube Specimens)
● ASTM C 349: Compressive Strength of Hydraulic Cement Mortars (Using Portions of Prisms Broken in Flexure)
Although still specified by ASTM, the direct tension test does not provide any useful insight into the concrete-making
properties of cements. It persists as a specified test because in the early years of cement manufacture, it used to be the most
common test since it was difficult to find machines that could compress a cement sample to failure.
Flexural strength (actually a measure of tensile strength in bending) is carried out on a 40 x 40 x 160 mm (1.57-inch x 1.57-
inch x 6.30-inch) cement mortar beam. The beam is then loaded at its center point until failure.
Loss on ignition is calculated by heating up a cement sample to 900 - 1000°C (1650 - 1830°F) until a constant weight is
obtained. The weight loss of the sample due to heating is then determined. A high loss on ignition can indicate prehydration
and carbonation, which may be caused by improper and prolonged storage or adulteration during transport or transfer (PCA,
1988).
4.6 Summary
Portland cement, the chief ingredient in cement paste, is the most widely used building material in the world. In the presence
of water, the chemical compounds within portland cement hydrate causing hardening and strength gain. Portland cement can
be specified based on its chemical composition and other various physical characteristics that affect its behavior. ASTM
specifies eight basic types of portland cement concrete. Tests to characterize portland cement, such as fineness, soundness,
setting time and strength are useful in quality control and specifications but should not be substituted for tests on PCC.
1 Introduction
When designing pavements (both mix design and structural design), there are three fundamental external
design parameters to consider: the characteristics of the subgrade upon which the pavement is placed,
the applied loads and the environment. First, the subgrade upon which the pavement is placed will have
a large impact on structural design. Subgrade stiffness and drainage characteristics help determine
pavement layer thickness, the number of layers, seasonal load restrictions and any possible
improvements to subgrade stiffness and drainage itself. Second, the expected traffic loading is a primary
design input (both in mix design and structural design). Traffic loads are used to determine pavement
composition, layer type and thickness, all of which affect pavement life. Third, the environment has a
large impact on pavement material performance. Environmental factors such as temperature, moisture
and ice formation can affect pavement durability, binder rheology, structural support and ultimately
pavement life and failure.
This section provides an overview of subgrade characteristics, pavement loading concepts and
environmental factors.
http://training.ce.washington.edu/WSDOT/Modules/04_design_parameters/04-1_body.htm4/2/2008 6:29:42 PM
4.2 Materials - Subgrade
2 Subgrade
Although a pavement's wearing course is most prominent, the success or Major Topics on this Page
failure of a pavement is more often than not dependent upon the underlying
subgrade (see Figures 4.1 and 4.2) - the material upon which the pavement 2.1 Subgrade Performance
structure is built. Subgrades be composed of a wide range of materials
although some are much better than others. This subsection discusses a few 2.2 Stiffness/Strength Tests
of the aspects of subgrade materials that make them either desirable or 2.3 Modulus of Subgrade Reaction
undesirable and the typical tests used to characterize subgrades.
2.4 Summary
1. Load bearing capacity. The subgrade must be able to support loads transmitted from the pavement structure. This
load bearing capacity is often affected by degree of compaction, moisture content, and soil type. A subgrade that
can support a high amount of loading without excessive deformation is considered good.
2. Moisture content. Moisture tends to affect a number of subgrade properties including load bearing capacity,
shrinkage and swelling. Moisture content can be influenced by a number of things such as drainage, groundwater
table elevation, infiltration, or pavement porosity (which can be assisted by cracks in the pavement). Generally,
excessively wet subgrades will deform excessively under load.
3. Shrinkage and/or swelling. Some soils shrink or swell depending upon their moisture content. Additionally, soils
with excessive fines content may be susceptible to frost heave in northern climates. Shrinkage, swelling and frost
heave will tend to deform and crack any pavement type constructed over them.
Poor subgrade should be avoided if possible, but when it is necessary to build over weak soils there are several methods
available to improve subgrade performance:
● Removal and replacement (over-excavation). Poor subgrade soil can simply be removed and replaced with high
quality fill. Although this is simple in concept, it can be expensive. Table 4.1 shows typical over-excavation
depths recommended by the Colorado Asphalt Pavement Association (CAPA).
● Stabilization with a cementitious or asphaltic binder. The addition of an appropriate binder (such as lime, portland
cement or emulsified asphalt) can increase subgrade stiffness and/or reduce swelling tendencies. Table 4.2
summarizes the Colorado Asphalt Pavement Association recommendations.
● Additional base layers. Marginally poor subgrade soils may be compensated for by using additional base layers.
These layers (usually of crushed stone – either stabilized or unstabilized) serve to spread pavement loads over a
larger subgrade area. This option is rather perilous; when designing pavements for poor subgrades the temptation
may be to just design a thicker section with more base material because the thicker section will satisfy most design
equations. However, these equations are at least in part empirical and were usually not intended to be used in
extreme cases. In short, a thick pavement structure over a poor subgrade will not necessarily make a good
pavement.
In sum, subgrade characteristics and performance are influential in pavement structural design. Characteristics such as load
bearing capacity, moisture content and expansiveness will influence not only structural design but also long-term
performance and cost.
WSDOT uses a modified version of AASHTO T 292 (Resilient Modulus of Subgrade Soils and Untreated
Base/Subbase Materials) to characterize subgrade soil and untreated base/subbase material stiffness.
Therefore, WSDOT uses the resilient modulus rather than CBR or R-value for design purposes. WSDOT
uses R-value to characterize aggregate pit sources for material approval.
The basic CBR test involves applying load to a small penetration piston at a rate of 1.3 mm (0.05") per minute and recording
the total load at penetrations ranging from 0.64 mm (0.025 in.) up to 7.62 mm (0.300 in.). Figure 4.3 is a sketch of a typical
CBR sample.
Values obtained are inserted into the following equation to obtain a CBR value:
The R-value test was developed by F.N. Hveem and R.M. Carmany of the California Division of Highways and first reported
in the late 1940's. During this time rutting (or shoving) in the wheel tracks was a primary concern and the R-value test was
developed as an improvement on the CBR test. Presently, the R-value is used mostly by State Highway Agencies (SHAs) on
the west coast of the U.S.
The test procedure to determine R-value requires that the laboratory prepared samples are fabricated to a moisture and
density condition representative of the worst possible in situ condition of a compacted subgrade. The R-value is calculated
from the ratio of the applied vertical pressure to the developed lateral pressure and is essentially a measure of the material's
resistance to plastic flow. The testing apparatus used in the R-value test is called a stabilometer (identical to the one used in
Hveem HMA mix design) and is represented schematically in Figure 4.4.
Values obtained from the stabilometer are inserted into the following equation to obtain an R-value:
● MH silts: 15-30
● AASHTO T 190 and ASTM D 2844: Resistance R-Value and Expansion Pressure of Compacted Soils
WSDOT uses R-value to characterize aggregate pit sources for material approval. WSDOT Test Method 611 is very
similar to AASHTO T 190. However, WSDOT uses a 300 psi exudation pressure while AASHTO T 190 uses a 400
psi exudation pressure. WSDOT and AASHTO T 190 R-values may differ due to this exudation pressure difference.
● Typical values
. Although they measure the same stress-strain relationship, the load application rates are different, thus resilient modulus is
considered an estimate of elastic modulus.
Elastic modulus is sometimes called Young's modulus after Thomas Young who published the concept back in 1807. An
elastic modulus (E) can be determined for any solid material and represents a constant ratio of stress and strain (a stiffness):
A material is elastic if it is able to return to its original shape or size immediately after being stretched or squeezed. Almost
all materials are elastic to some degree as long as the applied load does not cause it to deform permanently. Thus, the
"flexibility" of any object or structure depends on its elastic modulus and geometric shape.
The modulus of elasticity for a material is basically the slope of its stress-strain plot within the elastic range (as shown in
Figure 4.5). Figure 4.6 shows a stress versus strain curve for steel. The initial straight-line portion of the curve is the elastic
range for the steel. If the material is loaded to any value of stress in this part of the curve, it will return to its original shape.
Thus, the modulus of elasticity is the slope of this part of the curve and is equal to about 207,000 MPa (30,000,000 psi) for
steel. It is important to remember that a measure of a material's modulus of elasticity is not a measure of strength.
Strength is the stress needed to break or rupture a material (as illustrated in Figure 4.5), whereas elasticity is a measure of
how well a material returns to its original shape and size.
The nomenclature and symbols from the 1993 AASHTO Guide is generally used in referring to pavement moduli. For
example:
Changes in stress can have a large impact on resilient modulus. "Typical" relationships are shown in Figures 4.7 and 4.8.
Figure 4.7: Resilient Modulus vs. Bulk Stress for Unstabilized Coarse Grained Materials
Figure 4.8: Resilient Modulus vs. Deviator Stress for Unstabilized Fine Grained Materials
Tables 4.4 shows typical values of modulus of elasticity for various materials.
Elastic Modulus
Material
MPa psi
Rubber 7 1,000
WSDOT uses resilient modulus to characterize base and subbase materials as well as the subgrade (CBR was
used up until 1951 after which R-Values were used).
A series of resilient modulus triaxial tests were conducted at the WSDOT Materials Laboratory in July 1988,
April 1989 and May 1989 on disturbed (i.e., not in situ) samples from 14 sites:
Base Material
Subgrade
Keep in mind that this was not a comprehensive study of all Washington State granular materials but it does
give an idea of the range and typical values of base and subgrade stiffness in Washington State.
There are two fundamental approaches to estimating elastic moduli – laboratory tests and field deflection data/
backcalculation. This section discusses laboratory tests. Of the laboratory tests, two are noted:
● Diametral resilient modulus. This test is typically used on HMA and is covered in Module 5, Section 6, HMA
Performance Tests.
● Triaxial resilient modulus. This test is typically used on unbound materials such as soil and aggregate and is
covered here.
In a triaxial resilient modulus test a repeated axial cyclic stress of fixed magnitude, load duration and cyclic duration is
applied to a cylindrical test specimen. While the specimen is subjected to this dynamic cyclic stress, it is also subjected to a
static confining stress provided by a triaxial pressure chamber. The total resilient (recoverable) axial deformation response of
the specimen is measured (see Figure 4.9) and used to calculate the resilient modulus using the following equation:
where: MR = resilient modulus (or elastic modulus since resilient modulus is just an estimate of
(or ER) elastic modulus)
● AASHTO T 292: Resilient Modulus of Subgrade Soils and Untreated Base/Subbase Materials
This equation is restricted to fine grained materials with soaked CBR values of 10 or less. Like all such correlations, it
The proposed new AASHTO Design Guide will likely use the following relationship:
MR = 2555 x CBR0.64
The 1993 AASHTO Guide offers the following correlation equation between R-value and elastic modulus for fine-grained
soils with R-values less than or equal to 20.
A WSDOT developed relationship between the R-value and resilient modulus is shown below. This graph
was developed using WSDOT samples which ranged from silty materials (A-7) to coarse aggregate (A-1).
The samples were tested according to Washington Test Method 611 (Determination of the Resistance (R-
Value) of Untreated Bases, Subbases, and Basement Soils by the Stabilometer) and AASHTO T 274. Note
that WSDOT Test Method 611 “design R-Values” are determined at an exudation pressure of 400 psi.
AASHTO T 190 allows the use of a 300 psi exudation pressure. Thus, R-Values may differ due to the
exudation pressure.
with other tests. There is no direct laboratory procedure for determining k-value.
The modulus of subgrade reaction came about because work done by Westergaard during the 1920s developed the k-value as
a spring constant to model the support beneath the slab (see Figure 4.10).
The reactive pressure to resist a load is thus proportional to the spring deflection (which is a representation of slab deflection)
and k (see Figure 4.11):
∆ = slab deflection
Figure 4.11: Relation of Load, Deflection and Modulus of Subgrade Reaction (k)
The value of k is in terms of MPa/m (pounds per square inch per inch of deflection, or pounds per cubic inch - pci) and
ranges from about 13.5 MPa/m (50 pci) for weak support, to over 270 MPa/m (1000 pci) for strong support. Typically, the
modulus of subgrade reaction is estimated from other strength/stiffness tests, however, in situ values can be measured using
P = applied pressure (load divided by the area of the 762 mm (30 inch) diameter plate)
The 1993 AASHTO Guide offers the following relationship between k-values from a plate bearing test and resilient modulus
(MR):
● AASHTO T 222 and ASTM D 1196: Nonrepetitive Static Plate Load for Soils and Flexible Pavement
Components, for Use in Evaluation and Design of Airport and Highway Pavements
2.4 Summary
Subgrade properties are essential pavement design parameters. Materials typically encountered in subgrades are
characterized by their strength and their resistance to deformation under load (stiffness). In the U.S. the CBR, R-value and
resilient modulus are commonly used to characterize subgrade materials. Although each method is useful, the resilient
modulus is most consistent with other disciplines and is gaining widespread use in pavement design. The modulus of
subgrade reaction (k) is the subgrade characterization used in rigid pavement design. It can be estimated from CBR, R-value
or elastic modulus, or calculated from field tests like the plate bearing test.
3 Loads
One of the primary functions of a pavement is load distribution. Therefore, Major Topics on this Page
in order to adequately design a pavement something must be known about
the expected loads it will encounter. Loads, the vehicle forces exerted on 3.1 Tire Loads
the pavement (e.g., by trucks, heavy machinery, airplanes), can be
characterized by the following parameters: 3.2 Axle and Tire Configurations
Loads, along with the environment, damage pavement over time. The simplest pavement structural model asserts that each
individual load inflicts a certain amount of unrecoverable damage. This damage is cumulative over the life of the pavement
and when it reaches some maximum value the pavement is considered to have reached the end of its useful service life.
Therefore, pavement structural design requires a quantification of all expected loads a pavement will encounter over its
design life. This quantification is usually done in one of two ways:
1. Equivalent single axle loads (ESALs). This approach converts wheel loads of various magnitudes and repetitions
("mixed traffic") to an equivalent number of "standard" or "equivalent" loads.
2. Load spectra. This approach characterizes loads directly by number of axles, configuration and weight. It does
not involve conversion to equivalent values. Structural design calculations using load spectra are generally more
complex than those using ESALs.
Both approaches use the same type and quality of data but the load spectra approach has the potential to be more accurate in
its load characterization.
States generally limit the allowable load per inch width of tire. Based on a slightly dated survey (Sharma, Hallin and
Mahoney, 1983), this tire load limitation varies from a high of 140 N/mm (800 lbs/inch) to a low of 79 N/mm (450 lbs/inch).
Figure 4.14: FHWA Class 9 Five-Axle Tractor – Semi trailer (18 Tires Total) A typical tire load is
18.9 kN (4,250 lbs) with an inflation pressure of
689 kPa (100 psi.)
3.2.1 Descriptions
Tire-axle combinations are typically described as (see Figure 4.15):
Single Axle with Single Tires Single Axle with Dual Tires
Tandem Axles with Single Tires Tandem Axles with Dual Tires
Figure 4.16:
Tandem
Drive Axle on
a Tractor
Frame During
Manufacturing
Federal and State laws establish maximum axle and gross vehicle weights to limit pavement damage. The range of weight
limits in the U.S. vary a bit based on various Federal and State laws. Figure 4.17 shows the range of maximum limits for
single axle, tandem axle and gross vehicle weight (GVW) established by the states and the FHWA.
Item Limit
105 N/mm (600 lb/inch)
Tire Load
of tire width
89 kN
Single Axle
(20,000 lbs)
151 kN
Tandem Axle
(34,000 lbs)
469 kN
Gross Vehicle Weight
(105,500 lbs)
Figure 4.17: Range of Allowable Axle and Truck Weights in the U.S.
(based on data from USDOT, 2000)
Although each state and the FHWA have established maximum axle-tire load combinations, there are other restrictions as
well. One of the most common is the FHWA bridge formula (sometimes called the Federal Bridge Formula B).
1. Equivalent single axle load (ESAL). Based on AASHO Road Test results, the most common approach is to
convert wheel loads of various magnitudes and repetitions ("mixed traffic") to an equivalent number of "standard"
or "equivalent" loads. The most commonly used equivalent load in the U.S. is the 80 kN (18,000 lbs) equivalent
single axle load (normally designated ESAL).
2. Load spectra. The 2002 Guide for the Design of New and Rehabilitated Pavement Structures (NCHRP 1-37A)
essentially does away with the ESAL and determines loading directly from axle configurations and weights. This
is a more precise characterization of traffic but relies on the same input data used to calculate ESALs. A typical
load spectrum input would be in the form of a table that shows the relative axle weight frequencies for each
common axle combination (e.g. single axle, tandem axle, tridem axle, quad axle) over a given time period (see
Figure 4.18). Often, load spectra data can be obtained from weight-in-motion stations.
Figure 4.18: Example Load Spectra Input Screen from NCHRP 1-37A
Typically, designers must not only calculate ESALs or load spectra for various vehicles but also must forecast the expected
number of ESALs or load spectra a pavement will encounter over its entire design life. This information then helps
determine the structural design. Highway design in most states is based on the ESAL traffic input anticipated over a future 10
to 50 year period.
D = a directional distribution factor, expressed as a ratio, that accounts for the distribution of loads by direction (e.g.,
D
east-west, north-south). For instance, one direction may carry a majority of the heavy truck loads and thus it
would either be designed differently or, at a minimum, it would control the structural design. Generally taken as
0.5 (50%) for most roadways unless more detailed information is known.
a lane distribution factor, expressed as a ratio, that accounts for the distribution of loads when two or more lanes
D = are available in one direction. For instance, on most interstate routes, the outside lane carries a majority of the
L
heavy truck traffic.
1 100
2 80 – 100
3 60 – 80
4 50 – 75
^
= the cumulative two-directional loads predicted for a specific section of highway during the design period.
w18
The equation outputs are load equivalency factors (LEFs) or ESAL factors. This factor relates various axle load
combinations to the standard 80 kN (18,000 lbs) single axle load. It should be noted that ESALs as calculated by the ESAL
equations are dependent upon the pavement type (flexible or rigid) and the pavement structure (structural number for flexible
and slab depth for rigid). As a rule-of-thumb, the 1993 AASHTO Design Guide, Part III, Chapter 5, Paragraph 5.2.3
recommends the use of a multiplier of 1.5 to convert flexible ESALs to rigid ESALs (or a multiplier of 0.67 to convert rigid
ESALs to flexible ESALs). Using load spectra (as proposed in the 2002 Guide for the Design of New and Rehabilitated
Pavement Structures) will eliminate the need for flexible-rigid ESAL conversions. Table 4.5 shows some typical LEFs for
various axle-load combinations.
Assumptions:
● pt = 2.5
1. The relationship between axle weight and inflicted pavement damage is not linear but exponential. For instance, a
44.4 kN (10,000 lbs) single axle needs to be applied to a pavement structure more than 12 times to inflict the
same damage caused by one repetition of an 80 kN (18,000 lbs) single axle. Similarly, a 97.8 kN (22,000 lbs)
single axle needs to be repeated less than half the number of times of an 80 kN (18,000 lbs) single axle to have an
equivalent effect.
❍ An 80 kN (18,000 lbs) single axle does over 3,000 times more damage to a pavement than an 8.9 kN
(2,000 lbs) single axle (1.000/0.0003 ≈ 3,333).
❍ A 133.3 kN (30,000 lbs) single axle does about 67 times more damage than a 44.4 kN (10,000 lbs)
single axle (7.9/0.118 ≈ 67).
❍ A 133.3 kN (30,000 lb) single axle does about 11 times more damage than a 133.3 kN (30,000 lb)
tandem axle (7.9/0.703 ≈ 11).
❍ Heavy trucks and buses are responsible for a majority of pavement damage. Considering that a typical
automobile weighs between 2,000 and 7,000 lbs (curb weight), even a fully loaded large passenger van
will only generate about 0.003 ESALs while a fully loaded tractor-semi trailer can generate up to about
3 ESALs (depending upon pavement type, structure and terminal serviceability).
2. Determining the LEF for each axle load combination on a particular roadway is possible through the use of weigh-
in-motion equipment. However, typically this type of detailed information is not available for design. Therefore,
many agencies average their LEFs over the whole state or over different regions within the state. They then use a
standard "truck factor" for design which is simply the average number of ESALs per truck. Thus, an ESAL
determination would involved counting the number of trucks and multiplying by the truck factor.
❍ This method allows for ESAL estimations without detailed traffic measurements, which is often
appropriate for low volume roads and frequently must be used for lack of a better alternative for high
volume roads.
❍ When using this method, there is no guarantee that the assumed truck factor is an accurate representation
of the trucks encountered on the particular roadway in question.
forecasting the traffic the pavement will be subjected to over its design life then converting the traffic to a specific number of
ESALs based on its makeup. A typical ESAL estimate consists of:
1. Traffic count. A traffic count is used as a starting point for ESAL estimation. Most urban areas have some
amount of historical traffic count records. If not, simple traffic tube counts are relatively inexpensive and quick. In
some cases, designers may have to use extremely approximate estimates if no count data can be obtained.
2. A count or estimate of the number of heavy vehicles. This usually requires some sort of vehicle classification
within the traffic count. The simplest classifications divide vehicles into two categories: (1) heavy trucks and (2)
others. Other, more elaborate schemes can also be used such as the FHWA's vehicle classification.
First, the WSDOT Pavement Management System (PMS) uses a simplified version of the FHWA vehicle classification system. Like
many other states WSDOT uses three categories and assumes the following ESAL values:
The WSDOT PMS equation for annual ESALs on any given roadway is:
This equation implies that passenger automobile contributions to total ESAL counts are negligible.
Second, data collected between 1960 and 1983 provides a rough estimate of ESALs divided up into single units, combination units, buses
and an overall truck factor.
ESAL Factors
● All panel trucks and pickups were excluded from the calculations if they had two axles with four tires (i.e., two axle, six tire trucks or larger were
used).
● The ESAL calculations are for flexible pavements (LEFs from Appendix D, 1993 AASHTO Guide, SN = 5, pt = 2.5) only.
● Results are based on weight data from a limited number of weigh stations (typically 5 to 15) which operated for a maximum period of 24 hours for
no more than five days per year. Thus, the samples and hence the summary may be biased (either high or low).
● The above ESAL factors may appear to be "low"; however, about one-half of the trucks weighed at weigh stations were empty. Thus, an ESAL/
axle factor = 0.25 corresponds to a single axle load of about 12,700 lb (56.5 kN) (assumes SN = 5, pt = 2.5).
Third, initial WSDOT weigh-in-motion (WIM) analysis reveals the following ESALs per vehicle:
Trains 1.22
Note that these assumptions agree rather well with WSDOT PMS assumptions for all vehicles except "trains". For the 10 initial WSDOT
WIM sites analyzed, the ESAL per vehicle for trains ranged from a low of 0.43 to a high of 1.79.
3. An estimated traffic (and heavy vehicle) growth rate over the design life of the pavement. A growth rate estimate
is required to convert a single year traffic count into the total traffic experienced over the pavement design life.
Typically, multiplying the original traffic count by the pavement design life (in years) will grossly underestimate
total ESALs. For example, Interstate 5 at mile post 176.35 (near Shoreline, Washington) has experienced a growth
from about 200,000 ESALs per year in 1965 (original construction) to about 1,000,000 ESALs per year in 1994.
Thus, over a 30 year period, the ESALs per year have increased by a factor of five or an annual growth rate of
about six percent.
The WSDOT Pavement Management System (PMS) calculates ESAL growth rate using the following equation:
Where: G = traffic growth rate. If the truck growth rate is greater than zero then G is assumed to be
equal to the truck growth rate. Otherwise, G is assumed to be equal to the ADT growth
rate. A minimum G of 2 percent is assumed.
0.016 = an additional growth rate assumed to account for the increase in per-tire load. The ESAL
accounts for the loading on each axle, while this additional factor is an attempt to account
for how that load is actually transmitted to the pavement through the tires. For instance,
an 18,000 lb. axle load supported by four tires at 80 psi each is less damaging than the
same 18,000 lb. load supported by two tires at 125 psi each.
Whereas traffic growth rate is important for capacity issues, ESAL growth rate is the critical growth factor in pavement
structural design.
The Total number of ESALs over a number of years is calculated by using the Annual ESAL estimate (at the time of the
traffic count) and compounding it annually over the total number of years using the "total ESAL growth rate" determined
from the equation above.
4. Select appropriate LEFs to convert truck traffic to ESALs. Different regions may experience different types of
loads. For instance, a particular area may experience a high number of trucks but they may be mostly empty thus
lowering their LEF. For instance, the statewide LEF for Washington State is about 1.028 ESALs/truck. However,
this may be drastically different from local LEFs.
5. An ESAL estimate. An ESAL estimate can be made based on the preceding steps. Depending upon circumstances
these estimates may vary widely. Figure 4.19 shows an example of a pavement that was built for an estimated
ESAL loading but is experiencing a much higher loading due to a marked increase in bus traffic.
The ESAL calculator presented below uses standard 2002 Washington State Pavement Management System
(WSPMS) assumptions about load equivalencies and growth rates. These standards may not apply in all situations.
1. It is compatible with the FHWA's Traffic Monitoring Guide (TMG) and thus many agencies are already
collecting the appropriate data.
2. It offers a hierarchical approach to traffic data input depending upon the users needs and resources. There are
three levels of potential input:
• Level 1 Inputs – Use of volume/classification and axle load spectra data directly related to the project.
• Level 2 Inputs – Use of regional axle load spectra data and project-related volume/classification data.
• Level 3 Inputs – Use of regional or default classification and axle load spectra data.
3. It already includes information on traffic distribution including directional, lane and temporal distribution (if
needed) as well as traffic growth rates.
3.8 Summary
Loading is a fundamental pavement design parameter. In order to fully characterize a load, the following parameters should
be known:
● Tire loads
● Repetition of loads
● Vehicle speed
Pavement damage caused by a particular load is roughly related to the load by about a power of four (for reasonably strong
surfaces). This means that, generally speaking, a vehicle weighing twice as much as another (and having the same axle/tire
arrangement) will cause 16 times as much damage to the pavement.
Given the number and types of vehicles in the world today, there are many different types of loads and load configurations.
The most common load characterization approach is to convert all loads into an equivalent number of 80 kN (18,000 lbs)
axle loads (ESALs). ESALs can then be used in pavement structural design. The 2002 Guide for the Design of New and
Rehabilitated Pavement Structures dispenses with ESAL calculations and deals directly with traffic load spectra, however
the general load vs. damage concepts are the same. Loads work in conjunction with materials, subgrade and the environment
to determine pavement design inputs.
4 Environment
A pavement must be able to function within the Major Topics on this Page
environment in which it is built. The environment can
vary greatly across the globe at any one time and it can 4.1 Temperature Variations
also vary greatly across time at any one place.
Environmental variations can have a significant impact 4.2 Frost Action
on pavement materials and the underlying subgrade, 4.3 Moisture
which in turn can drastically affect pavement
performance. Certainly every environmental constituent 4.4 Summary
(e.g., solar flux, heat, wind, humidity, etc.) can have an
incremental effect on pavement. However, there are several constituents that exert an overriding
influence. These variables (considered in this section) are:
● Temperature Variations
● Frost action
● Moisture
In 1935, measurements reported by Teller and Southerland of the Bureau of Public Roads showed that the
maximum temperature differential (hence, maximum warping) is much larger during the day than during
the night. Further, during the day, the upper surface of the slab is at a higher temperature than the bottom
resulting in tensile stresses at the bottom of the slab.
1. Frost heave. An upward movement of the subgrade resulting from the expansion of accumulated
soil moisture as it freezes.
2. Thaw weakening. A weakened subgrade condition resulting from soil saturation as ice within
A series of questions were posed to Washington State, City, County, and WSDOT Regions
concerning pavement frost design issues. These results were compiled by the Washington
State Policy Plan Subcommittee on Weight Restrictions and Road Closures. One of the
questions asked was: "What do you feel are the most important factors in causing road
deterioration in your jurisdiction?" Of a wide range of possible responses, three of the top-
ranked factors (frost heaving, road use during freeze/thaw, excess subgrade moisture (which
may not be frost related)) constituted the following percentages of all responses:
● Counties: 46%
● Cities: 48%
● State: 30%
Source: WSDOT. (April 15, 1994). "Questionnaire Results with Comments," Washington
State Policy Plan Subcommittee on Weight Restrictions and Road Closures. Washington
State Department of Transportation. Olympia, WA.
The three elements necessary for ice lenses and thus frost heave are:
2. Subfreezing temperatures (freezing temperatures must penetrate the soil and, in general, the
thickness of an ice lens will be thicker with slower rates of freezing).
3. Water (must be available from the groundwater table, infiltration, an aquifer, or held within the
voids of fine-grained soil).
Remove any of the three conditions above and frost effects will be eliminated or at least minimized. If the
three conditions occur uniformly, heaving will be uniform; otherwise, differential heaving will occur
resulting in pavement cracking and roughness. Differential heave is more likely to occur at locations such
as:
● Where subgrades change from clean not frost susceptible (NFS) sands to silty frost susceptible
materials.
● Abrupt transitions from cut to fill with groundwater close to the surface.
● Drains, culverts, etc., frequently result in abrupt differential heaving due to different backfill
material or compaction and the fact that open buried pipes change the thermal conditions (i.e.,
remove heat resulting in more frozen soil).
Additional factors which will affect the degree of frost susceptibility (or ability of a soil to heave):
● Temperature gradient
In 1932, Dr. Arthur Casagrande proposed the following widely known rule-of-thumb criterion for
identifying potentially frost susceptible soils:
"Under natural freezing conditions and with sufficient water supply one should expect considerable
ice segregation in non-uniform soils containing more than 3% of grains smaller than 0.02 mm, and in
very uniform soils containing more than 10 percent smaller than 0.02 mm. No ice segregation was
observed in soils containing less than 1 percent of grains smaller than 0.02 mm, even if the
groundwater level is as high as the frost line."
Application of the Casagrande criterion requires a hydrometer test of a soil suspension (in water) to
determine the distribution of particles passing the 0.075 mm sieve and to compute the percentage of
particles finer than 0.02 mm.
WSDOT uses crushed surfacing base course (CSBC) as a frost resistant crushed aggregate because it has a maximum
of only 7.5% passing the 0.075 mm (No. 200) sieve.
Figure 4.24: Typical Pavement Deflections Illustrating Seasonal Pavement Strength Changes
(on a portion of State Route 172 in Washington State)
Thawing can proceed from the top downward, or from the bottom upward, or both. How this occurs
depends mainly on the pavement surface temperature. During a sudden spring thaw, melting will proceed
almost entirely from the surface downward. This type of thawing leads to extremely poor drainage
conditions. The frozen soil beneath the thawed layer can trap the water released by the melting ice lenses
so that lateral and surface drainage are the only paths the water can take.
"The effects of refreezing after a thaw are also accentuated by the fact that the first freeze leaves the
soil in a more or less loosened or expanded condition."
This observation shows that (1) the reduced density of base or subgrade materials helps to explain the long
recovery period for material stiffness or strength following thawing, and (2) refreezing following an initial
thaw can create the potential for greater weakening when the "final" thaw does occur.
1. Surface water. Enters the pavement primarily by infiltration through surface cracks and joints,
and through adjacent unpaved surfaces, during periods of rain and melting snow and ice. Many
crack-free pavements are not entirely impermeable to moisture.
• Moisture held in soil voids or drawn upward from a water table by capillary forces.
• Moisture that moves laterally beneath a pavement from an external source (e.g., pervious
water bearing strata, etc.).
● Limit the depth of frost into the subgrade soils. This is typically accomplished by specifying the
depth of pavement to be some minimum percentage of the frost depth. By extending the
pavement section well into the frost depth, the depth of frost-susceptible subgrade under the
pavement (between the bottom of the pavement structure and frost depth) is reduced. The
assumption is that a reduced depth of soil under frost action will cause correspondingly less
damage.
● Removing and replacing frost-susceptible subgrade. Ideally the subgrade will be removed at
least down to the typical frost depth. Removing frost-susceptible soils removes frost action.
● Design the pavement structure based on reduced subgrade support. This method simply
increases the pavement thickness to account for the damage and loss of support caused by frost
action.
● Providing a capillary break. By breaking the capillary flow path, frost action will be less severe
because as Tabor (1930) noted, frost heaving requires substantially more water than is naturally
available in the soil pores.
● Where to apply load restrictions. If pavement surface deflections are available to an agency,
spring thaw deflections greater than 45 to 50 percent of summer deflections suggest a need for
load restriction. Further, considerations such as depth of freezing (generally areas with air
Freezing Indices of 400 °F-days or more), pavement surface thickness, moisture condition, type
of subgrade, and local experience should be considered. Subgrades with Unified Soil
Classifications of ML, MH, CL, and CH will result in the largest pavement weakening.
● Amount of load reduction. The minimum load reduction level should be 20 percent. Load
reductions greater than 60 percent generally are not warranted based on potential pavement
damage. A load reduction range of 40 to 50 percent should accommodate a wide range of
pavement conditions.
The duration can also be estimated by use of TI and the following rough relationship:
TI = 0.3 (FI)
1. Frost heave. Results from accumulation of moisture in the soil during the freezing period.
These accumulations (ice lenses) expand perpendicular to the direction of heat flow and push the
pavement up, often causing severe cracking.
2. Thaw weakening. Once a subgrade is frozen it can be severely weakened when it thaws (usually
in the spring time). Therefore, loading that would not normally damage a given pavement may be
quite detrimental during thaw periods.
Frost action can be further characterized by the typical depth to which the subgrade freezes in a particular
area. This depth can be estimated by several equations including the Stefan formula and the modified
Berggren formula. Once this depth is known, it can be used as a pavement structural design input to
mitigate the detrimental effects of frost action. Mitigation techniques can be classified into four broad
categories:
If frost action cannot be adequately mitigated, severe pavement damage (in the case of frost heave) or a
loss of bearing capacity (in the case of thaw weakening) can result. Maintenance options to correct these
problems are limited to pavement repair or replacement (in the case of frost heave) or limiting pavement
loading during spring thawing (in the case of thaw weakening).
4.3 Moisture
Moisture (in the form of accumulated water or rainfall) affects pavements in a number of ways. This
section just briefly lists some of these ways:
● Design. Certain types of soils can be highly expansive when wet. Structural design must
account for this expansiveness.
● Construction.
● Driving Conditions. Rainfall reduces skid resistance and can cause hydroplaning in severely
rutted areas.
4.4 Summary
The environment has a large influence on pavement performance and thus, pavement design. Temperature
extremes cause pavements to expand and contract, frost action may cause them to crack and fail and
moisture is a prime consideration in drainage, construction and driving safety.
5 Drainage
Proper drainage is important to ensure a high quality Major Topics on this Page
long lived pavement; moisture accumulation in any
pavement structural layer can cause problems. 5.1 Surface Drainage
Moisture in the subgrade and aggregate base layer can
weaken these materials by increasing pore pressure 5.2 Subsurface Drainage
and reducing the materials' resistance to shear.
Additionally, some soils expand when moist, causing differential heaving. Moisture in the HMA layers
can cause stripping because it, instead of the asphalt binder, will adhere to aggregate particles.
Moisture sources are typically rainwater, runoff and high groundwater. These sources are prevented
from entering the pavement structure or accumulating in the subgrade through surface drainage and
subsurface drainage. Usually, it is more cost effective and less risky to prevent moisture entry and
accumulation using surface drainage than to effect moisture removal using subsurface drainage.
● Impermeable pavement surface. An impermeable surface will protect the underlying subgrade
from water sources above. Permeability concerns are different for flexible and rigid
pavements.
❍ Flexible pavements. When HMA air voids are greater than about 8 - 9 percent they
are likely to be interconnected with one another, making the HMA water permeable
(Kandhal and Koehler, 1984). Proper compaction practices should be followed to
ensure an impermeable pavement. Also, minor cracks in the HMA should be
promptly sealed.
● Slope. The pavement section should be sloped to allow rainwater to sheet flow (see Figure
4.27) quickly to the edge where it is typically collected in a curb and gutter system or a
roadside ditch. A generally accepted standard is a 2 percent cross slope.
● Grade. The curb and gutter or roadside ditch must be properly graded to allow flow to central
collection points such as catch basins or detention ponds. A generally accepted standard is a
grade of 0.5 percent or more although lesser grades have been used effectively.
● Gravity flow. Water from surrounding areas can be absorbed by the soil then flow by gravity
to areas underneath the pavement structure. In pavement with high air voids (above 8 - 9
percent), water can percolate down through the pavement structure itself.
● Capillary rise. Capillary rise is the rise in a liquid above the level of zero pressure due to a
net upward force produced by the attraction of the water molecules to a solid surface (e.g.,
soil). Capillary rise can be substantial, up to 6 m (20 ft.) or more. In general, the smaller the
soil grain size, the greater the potential for capillary rise. Often, capillary rise is a problem in
areas of high groundwater tables.
Most pavements have performed adequately without considering these effects. However, HMA
pavements can fail because of subgrade support deterioration as a result of excessive moisture or other
water-related problems. While the best solution is usually to prevent water infiltration with surface
drainage measures, subsurface drainage can be useful, however it needs to be done judiciously, because
it may be somewhat akin to treating the symptom rather than the problem. Subsurface drainage consists
of three basic elements (see Figures 4.28, 4.29 and 4.30):
1. A permeable base to provide for rapid removal of water which enters the pavement structure.
Based on recent research from California, asphalt treated permeable base layers may strip and
become clogged with fines thus weakening the overall pavement structure.
1. A method of conveying the removed water away from the pavement structure. At the least,
this may consist of a base sloped towards a drainage ditch. At the most, this may consist of a
pipe collector system.
2. A filter layer (such as a geotextile, graded aggregate layer or HMA) to prevent the migration
of fines into the permeable base from the subgrade, subbase or shoulder base material. Excess
fines in the permeable base will clog its drainage routes and render it ineffective. Depending
upon the subgrade and pavement structure a filter layer may not be used.
Figure 4.29: Rigid Pavement Subsurface Drainage with PCC Tied Shoulder
1 Introduction
The two key components of pavement design are mix design and structural design. This section deals
with HMA and PCC mix design. The goal of mix design is to determine the optimum mixture of
component materials for a given application. This includes detailed evaluations of aggregate, asphalt
and portland cement as well as a determination of their optimum blending ratios. This section covers the
following for HMA and PCC mix design:
● Mix design fundamentals. These are the fundamental philosophies and parameters of mix
design such as (1) why it is done, (2) what basic assumptions are made and (3) the specific
goals of mix design.
● Mix design methods. These sections cover the various mix design procedures used. For
HMA, the Hveem, Marshall and Superpave methods are covered. For PCC, the ACI method
is covered.
Currently, WSDOT uses both the Hveem and Superpave mix design
methods. However, Superpave is slated to eventually phase out the older
Hveem method.
● Performance Tests. These are the tests performed on laboratory designed mixes (or field
samples) to characterize their performance. They can consist of basic physical property
measurements (such as stiffness or strength) or laboratory simulation of field conditions (such
as rutting potential or chloride penetration).
This section is only meant to provide a brief overview of mix design methods as well as their
assumptions, inputs and outputs. Resources that provide a detailed description and analysis of each mix
design method are listed in the beginning of each section.
2 HMA - Fundamentals
HMA consists of two basic ingredients: aggregate and Major Topics on this Page
asphalt binder. HMA mix design is the process of
determining what aggregate to use, what asphalt binder 2.1 Concepts
to use and what the optimum combination of these two
2.2 Variables
ingredients ought to be. HMA mix design has evolved
as a laboratory procedure that uses several critical tests 3.3 Objectives
to make key characterizations of each trial HMA
blend. Although these characterizations are not 3.4 Basic Procedure
comprehensive, they can give the mix designer a good
3.5 Summary
understanding of how a particular mix will perform in
the field during construction and under subsequent traffic loading.
This section covers mix design fundamentals common to all mix design methods. First, two basic
concepts (mix design as a simulation and weight-volume terms and relationships) are discussed to set a
framework for subsequent discussion. Second, the variables that mix design may manipulate are
presented. Third, the fundamental objectives of mix design are presented. Finally, a generic mix design
procedure (which Hveem, Marshall and Superpave methods all use) is presented.
2.1 Concepts
Before discussing any mix design specifics, it is important to understand a couple of basic mix design
concepts:
Being a simulation, mix design has its limitations. Specifically, there are substantial differences
between laboratory and field conditions. Certainly, a small laboratory setup consisting of several 100 -
150 mm (4 - 6 inch) samples, a compaction machine and a couple of testing devices cannot fully recreate
actual manufacturing, construction and performance conditions. For instance, mix design compaction
should create the same general density (void content) to which the traffic will finally compact a mix in
the field under service conditions (Roberts et al., 1996). However, it is difficult to calibrate a number of
tamper blows (laboratory compaction) to a specific construction compaction and subsequent traffic
loading (field compaction). Currently used correlations between these densities are empirical in nature
and extremely rough (e.g., high, medium and low traffic categories). However, despite limitations such
as the preceding, mix design procedures can provide a cost effective and reasonably accurate simulation
that is useful in making mix design decisions.
2.2 Variables
HMA is a rather complex material upon which many different, and sometimes conflicting, performance
demands are placed. It must resist deformation and cracking, be durable over time, resist water damage,
provide a good tractive surface, and yet be inexpensive, readily made and easily placed. In order to meet
these demands, the mix designer can manipulate all of three variables:
1. Aggregate. Items such as type (source), gradation and size, toughness and abrasion resistance,
durability and soundness, shape and texture as well as cleanliness can be measured, judged
and altered to some degree.
2. Asphalt binder. Items such as type, durability, rheology, purity as well as additional
modifying agents can be measured, judged and altered to some degree.
3. The ratio of asphalt binder to aggregate. Usually expressed in terms of percent asphalt binder
by total weight of HMA, this ratio has a profound effect on HMA pavement performance.
Because of the wide differences in aggregate specific gravity, the proportion of asphalt binder
expressed as a percentage of total weight can vary widely even though the volume of
asphalt binder as a percentage of total volume remains quite constant.
2.3 Objectives
Before embarking on a mix design procedure it is important to understand what its objectives. This
section presents the typical qualities of a well-made HMA mix. By manipulating the variables of
aggregate, asphalt binder and the ratio between the two, mix design seeks to achieve the following
qualities in the final HMA product (Roberts et al., 1996):
1. Deformation resistance (stability). HMA should not distort (rut) or deform (shove) under
traffic loading. HMA deformation is related to one or more of the following:
❍ Asphalt binder content. Excess asphalt binder content tends to lubricate and push
aggregate particles apart making their rearrangement under load easier. The
optimum asphalt binder content as determined by mix design should prevent this.
❍ Asphalt binder viscosity at high temperatures. In the hot summer months, asphalt
binder viscosity is at its lowest and the pavement will deform more easily under
load. Specifying an asphalt binder with a minimum high temperature viscosity (as
can be done in the Superpave asphalt binder selection process) ensures adequate
high temperature viscosity.
2. Fatigue resistance. HMA should not crack when subjected to repeated loads over time.
HMA fatigue cracking is related to asphalt binder content and stiffness. Higher asphalt binder
contents will result in a mix that has a greater tendency to deform elastically (or at least
deform) rather than fracture under repeated load. The optimum asphalt binder content as
determined by mix design should be high enough to prevent excessive fatigue cracking. The
use of an asphalt binder with a lower stiffness will increase a mixture's fatigue life by
providing greater flexibility. However, the potential for rutting must also be considered in the
selection of an asphalt binder. Note that fatigue resistance is also highly dependent upon the
relationship between structural layer thickness and loading. However, this section only
addresses mix design issues.
3. Low temperature cracking resistance. HMA should not crack when subjected to low ambient
temperatures. Low temperature cracking is primarily a function of the asphalt binder low
temperature stiffness. Specifying asphalt binder with adequate low temperature properties (as
can be done in the Superpave asphalt binder selection process) should prevent, or at least
limit, low temperature cracking.
4. Durability. HMA should not suffer excessive aging during production and service life. HMA
durability is related to one or more of the following:
❍ The asphalt binder film thickness around each aggregate particle. If the film
thickness surrounding the aggregate particles is insufficient, it is possible that the
aggregate may become accessible to water through holes in the film. If the
aggregate is hydrophilic, water will displace the asphalt film and asphalt-aggregate
cohesion will be lost. This process is typically referred to as stripping. The
optimum asphalt binder content as determined by mix design should provide
adequate film thickness.
❍ Air voids. Excessive air voids (on the order of 8 percent or more) increase HMA
permeability and allow oxygen easier access to more asphalt binder thus
accelerating oxidation and volatilization. To address this, HMA mix design seeks
to adjust items such as asphalt content and aggregate gradation to produce design
air voids of about 4 percent. Excessive air voids can be either a mix design or a
construction problem and this section only addresses the mix design problem.
5. Moisture damage resistance. HMA should not degrade substantially from moisture
penetration into the mix. Moisture damage resistance is related to one or more of the
following:
❍ Air voids. When HMA air voids exceed about 8 percent by volume, they may
become interconnected and allow water to easily penetrate the HMA and cause
moisture damage through pore pressure or ice expansion. To address this, HMA
mix design adjusts asphalt binder content and aggregate gradation to produce design
air voids of about 4 percent. Excessive air voids can be either a mix design or a
construction problem and this section only addresses the mix design problem.
6. Skid resistance. HMA placed as a surface course should provide sufficient friction when in
contact with a vehicle's tire. Low skid resistance is generally related to one or more of the
following:
❍ Asphalt binder content. Excessive asphalt binder can cause HMA bleeding. Using
the optimum asphalt binder content as determined by mix design should prevent
bleeding.
7. Workability. HMA must be capable of being placed and compacted with reasonable effort.
Workability is generally related to one or both of the following:
❍ Aggregate texture, shape and size. Flat, elongated or angular particles tend to
interlock rather than slip by one another making placement and compaction more
difficult (notice that this is almost in direct contrast with the desirable aggregate
properties for deformation resistance). Although no specific mix design tests are
available to quantify workability, tests for particle shape and texture can identify
possible workability problems.
❍ Aggregate gradation. Gradations with excess fines (especially in the 0.60 to 0.30
mm (No. 30 to 50) size range when using natural, rounded sand) can cause a tender
mix. A gradation resulting in low VMA or excess asphalt binder content can have
the same effect. Gradation specifications are used to ensure acceptable aggregate
gradation.
❍ Asphalt binder content. At laydown temperatures (above about 120 °C (250 °F))
asphalt binder works as a lubricant between aggregate particles as they are
compacted. Therefore, low asphalt binder content reduces this lubrication resulting
in a less workable mix. Note that a higher asphalt binder content is generally good
for workability but generally bad for deformation resistance.
Knowing these objectives, the challenge in mix design is then to develop a relatively simple procedure
with a minimal amount of tests and samples that will produce a mix with all the above HMA qualities.
1. Aggregate selection. No matter the specific method, the overall mix design procedure begins
with evaluation and selection of aggregate and asphalt binder sources. Different authorities
specify different methods of aggregate acceptance. Typically, a battery of aggregate physical
tests is run periodically on each particular aggregate source. Then, for each mix design,
gradation and size requirements are checked. Normally, aggregate from more than one source
is required to meet gradation requirements.
2. Asphalt binder selection. Although different authorities can and do specify different methods
of asphalt binder evaluation, the Superpave asphalt binder specification has been or will be
adopted by most State DOTs as the standard (NHI, 2000).
WSDOT uses the Superpave asphalt binder performance grading system and
specifications. Therefore, asphalt binder must meet the requirements of AASHTO MP 1.
WSDOT uses three baseline asphalt binder performance grades based on geography.
These baseline grades are typically used and then adjusted as necessary.
Previously, WSDOT had used the aged residue (AR) viscosity grading. The commonly
used grade in this old system was AR-4000W.
3. Optimum asphalt binder content determination. Mix design methods are generally
distinguished by the method with which they determine the optimum asphalt binder content.
This process can be subdivided as follows:
❍ Compact these trial mixes in the laboratory. It is important to understand that this
step is at best a rough simulation of field conditions.
❍ Run several laboratory tests to determine key sample characteristics. These tests
represent a starting point for defining the mixture properties but they are not
comprehensive nor are they exact reproductions of actual field conditions.
❍ Pick the asphalt binder content that best satisfies the mix design objectives.
Table 5.1: Example Job Mix Formula (JMF) with Specification and Tolerance Bands
Sieve Size (U.S. units) 3/4 inch 1/2 inch 3/8 inch No. 8 No. 200
Figure 5.2: Job Mix Formula (JMF) with Specification Bands Example
2.5 Summary
HMA mix design is a laboratory process used to determine the appropriate aggregate, asphalt binder and
their proportions for use in HMA. Mix design is a process to manipulate three variables: (1) aggregate,
(2) asphalt binder content and (3) the ratio of aggregate to asphalt binder with the objective of obtaining
an HMA that is deformation resistant, fatigue resistant, low temperature crack resistant, durable,
moisture damage resistant, skid resistant and workable. Although mix design has many limitations it has
proven to be a cost-effective method to provide crucial information that can be used to formulate a high-
performance HMA.
Currently, WSDOT uses both the Hveem and Superpave mix design 3.1 History
methods. However, Superpave is slated to eventually phase out the older
3.2 Procedure
Hveem method.
3.3 Summary
The basic concepts of the Hveem mix design method were originally developed by Francis Hveem when he was a
Resident Engineer for the California Division of Highways in the late 1920s and 1930s. Currently, the Hveem
method is used by several western states. The basic philosophy surrounding the Hveem method can be summarized
in the following three points (Vallerga and Lovering, 1985):
1. HMA requires enough asphalt binder to coat each aggregate particle to an optimum film thickness
(allowing for its absorption into the aggregate).
2. HMA requires sufficient stability to resist traffic loading. This stability is generated by internal friction
between aggregate particles and cohesion (or tensile strength) created by the binder.
Based on this philosophy, the design asphalt content is selected as that asphalt content resulting in the highest
durability without dropping below a minimum allowable stability. In other words, as much asphalt binder as possible
should be used while still meeting minimum stability requirements.
This section consists of a brief history of the Hveem mix design method followed by a general outline of the actual
method. This outline emphasizes general concepts and rationale for the specific procedures. Detailed procedures can
vary from state-to-state but typical procedures are available in the following documents:
● Roberts, F.L.; Kandhal, P.S.; Brown, E.R.; Lee, D.Y. and Kennedy, T.W. (1996). Hot Mix Asphalt
Materials, Mixture Design, and Construction. National Asphalt Pavement Association Education
Foundation. Lanham, MD.
● Asphalt Institute. (1997). Mix Design Methods for Asphalt, 6th ed., MS-02. Asphalt Institute. Lexington,
KY.
performance HMA used on major urban streets and highways and the cheaper low performance penetrative method
(asphalt oil sprayed on a roadway surface of unbound particles) used on low-volume rural highways. An oil mix
consisted of a combination of aggregate and asphaltic oil that was mixed either in a plant or on the road itself (called
a "road mix"), spread by blade, then compacted by traffic. Unfortunately, there was no method available for
designing these oil mixes. Based on his research, and that of others, Francis N. Hveem developed a method for
determining the correct amount of oil based on aggregate surface area, which could be determined from gradation. It
also became evident that even given the right oil content, roads containing aggregates with "hard, glassy surface
texture" tended to deform excessively under load while roads containing aggregates with a "rough, irregular surface
texture" were more stable. Therefore, Hveem worked to develop a device that would measure stability, which
eventually became the Hveem Stabilometer. One more problem existed: specimens compacted in the laboratory for
the Stabilometer did not produce the same readings as those taken from field cores. Therefore, a new compaction
machine, which eventually became the California Kneading Compactor, was developed to more closely simulate the
compaction produced by rollers in the field.
3.2 Procedure
The Hveem mix design method consists of 6 basic steps:
1. Aggregate selection.
● AASHTO T 246: Resistance to Deformation and Cohesion of Bituminous Mixtures by Means of Hveem
Apparatus
● AASHTO T 247: Preparation of Test Specimens of Bituminous Mixtures by Means of the California
Kneading Compactor
mix design methods includes three basic steps (Roberts et al., 1996):
1. Determine aggregate physical properties. This consists of running various tests to determine properties
such as:
2. Determine other aggregate descriptive physical properties. If the aggregate is acceptable according to step
#1, additional tests are run to fully characterize the aggregate. These tests determine:
3. Perform blending calculations to achieve the mix design aggregate gradation. Often, aggregates from
more than one source or stockpile are used to obtain the final aggregate gradation used in a mix design.
Trial blends of these different gradations are usually calculated until an acceptable final mix design
gradation is achieved. Typical considerations for a trial blend include:
❍ All gradation specifications must be met. Typical specifications will require the percent retained
by weight on particular sieve sizes to be within a certain band.
❍ The gradation should not be too close to the FHWA's 0.45 power maximum density curve. If it
is, then the VMA is likely to be too low. Gradation should deviate from the FHWA's 0.45 power
maximum density curve, especially on the 2.36 mm (No. 8) sieve.
WSDOT uses the Superpave asphalt binder performance grading system and specifications. Therefore,
asphalt binder must meet the requirements of AASHTO MP 1. WSDOT uses three baseline asphalt
binder performance grades based on geography. These baseline grades are typically used and then
adjusted as necessary.
Previously, WSDOT had used the aged residue (AR) viscosity grading. The commonly used grade in
this old system was AR-4000W.
1. Select the asphalt binder content for each trial blend from a predetermined list. Many agencies have
predetermined lists that specify the asphalt content for each trial blend. It is assumed that the optimum
asphalt binder content will lie within the range of specified trail blend values.
2. Estimate the optimum asphalt binder content then select trail blends with asphalt binder contents at, above
and below the estimated optimum content. One common estimation method is the centrifuge kerosene
equivalent test (CKE), although this procedure has been discontinued by AASHTO (AASHTO, 2000a).
WSDOT does not use the CKE test. WSDOT uses a predetermined set of six asphalt binder
contents (at 0.5% intervals) for each class of mix. Through experience it is generally known
that these six trial blends will bracket the optimum asphalt binder content. Trial blends can be
adjusted depending upon the aggregate used and its specific gravity.
The centrifuge kerosene equivalent (CKE) test, used to estimate optimum asphalt content, involves three basic steps
(ASTM, 2000; AASHTO, 2000; Roberts et al., 1996):
1. Determine the centrifuge kerosene equivalent (CKE). A small fine aggregate sample (passing the 4.75 mm
(No. 4) sieve) is first weighed then submerged in kerosene. Once the sample is saturated with kerosene it is
placed in a centrifuge for 2 minutes to remove excess kerosene, then reweighed. The difference in these
weights gives an estimate of the fine aggregate's ability to absorb asphalt binder.
2. Determine the coarse aggregate surface capacity. A small coarse aggregate sample (passing the 9.5 mm
(0.375 inch) sieve but retained on the 4.75 mm (No. 4) sieve) is first weighed then submerged in SAE 10
oil for 5 minutes. The sample is then drained and placed in an oven for 15 minutes after which it is
reweighed. The difference in these weights gives an estimate of the coarse aggregate's ability to absorb
asphalt binder.
3. Estimate the optimum asphalt content. Results from the first two steps are corrected for aggregate specific
gravity then entered on a chart to determine the percent oil recommended for an asphalt cutback (specific
cutback types referenced are RC-250, MC-250 and SC-250). This percent oil is then corrected for the
increased viscosity of the HMA asphalt binder used.
● AASHTO T 270: Centrifuge Kerosene Equivalent and Approximate Bitumen Ratio (Discontinued)
Based on the results of the CKE test, samples are typically prepared with the following asphalt binder contents
(Roberts et al., 1996):
● 0.5, 1.0, 1.5 and 2.0 percent above the CKE value (at least one set of specimens should have enough
asphalt binder to flush after compaction)
Each sample is then heated to the anticipated compaction temperature and compacted with the California kneading
compactor (see Figure 5.3), a device that applies pressure to a sample through a hydraulically operated tamper foot.
Key parameters of the California kneading compactor are:
● Sample size = 102 mm (4-inch) diameter cylinder approximately 64 mm (2.5 inches) in height (corrections
can be made for different sample heights)
● Compaction pressure = Ranges from 2.4 to 3.4 MPa (350 to 500 psi)
● Number of blows = 150 (plus any preparatory blows at 1.7 MPa (250 psi) )
● Simulation method = The tamper foot strikes the sample on the top near the edge. The base rotates 1/6 of a
revolution after each blow. This helps achieve a sample particle orientation that is somewhat like that
achieved in the field after roller compaction.
● AASHTO T 247 and ASTM D 1561: Preparation of Test Specimens of Bituminous Mixtures by Means of
the California Kneading Compactor
Note: a correction to the stabilometer value is made if the sample height is not 64 mm
(2.5 inches)
With this equation, the stabilometer value can range from 0 to 90. Zero would represent a condition where lateral
pressure is equal to vertical pressure (e.g., a liquid). Ninety would represent a condition where there is no lateral
pressure no matter what the vertical pressure is (e.g., an incompressible solid). Table 5.2 shows typical stabilometer
criteria.
Table 5.2: Typical Hveem Design Criteria (from Asphalt Institute, 1979)
The cohesiometer (see Figures 5.5 and 5.6) attempts to measure cohesive strength across the diameter of a sample on
which the stability test had already been conducted. This is intended to provide some prediction about the ability of
the HMA sample to resist raveling under traffic loading. Basically the sample is bent as a cantilevered beam until it
fails. Although it was useful for oil mixes, HMAs tend to have large cohesion values as measured by the
cohesiometer and rarely, if ever, fail. As a result, the cohesiometer has fallen out of favor and is rarely used (Roberts
et al., 1996).
Basic WSDOT Hveem mix design specifications are shown in the table below. This table does not
list all specifications. These specifications are taken from the WSDOT 2002 Standard Specifications
for Road, Bridge and Municipal Construction (M 41-10).
Mix Class
Item
A B D E F G
Stabilometer 37 35 - 35 35 35
These densities are then used to calculate the volumetric parameters of the HMA. Measured void expressions are
usually:
● Air voids (Va), sometimes expressed as voids in the total mix (VTM)
● Density will generally increase with increasing asphalt content. The curve may or may not reach a
maximum.
Recall that the Hveem mix design method strives to select the asphalt content resulting in the highest durability
without falling below a minimum allowable stability. The "pyramid" method is a common method of selecting the
optimum asphalt binder content (see Figure 5.7).
3.3 Summary
The Hveem mix design method was developed to address specific mix design issues confronting Francis Hveem and
the California Division of Highways in the late 1920s and 1930s. Since then, it has been modified and supplemented
to address new concerns but the basic testing apparatus and selection criteria are still the same. The Hveem method is
based on three basic assumptions:
1. Optimum asphalt binder content is dependent upon aggregate surface area and absorption.
The two biggest differentiating aspects of the Hveem method when compared to other mix design methods are the
kneading compactor and the Hveem stabilometer. The kneading compactor uses a special rotating base to simulate
actual field compaction while the stabilometer measures HMA deformation under load. The design asphalt content is
selected as that asphalt content resulting in the highest durability without going below a minimum allowable
stability.
4.1 history
The basic concepts of the Marshall mix design method
4.2 procedure
were originally developed by Bruce Marshall of the
Mississippi Highway Department around 1939 and 4.3 summary
then refined by the U.S. Army. Currently, the
Marshall method is used in some capacity by about 38 states. The Marshall method seeks to select the
asphalt binder content at a desired density that satisfies minimum stability and range of flow values
(White, 1985).
This section consists of a brief history of the Marshall mix design method followed by a general outline
of the actual method. This outline emphasizes general concepts and rationale over specific procedures.
Detailed procedures vary from state-to-state but typical procedures are available in the following
documents:
● Roberts, F.L.; Kandhal, P.S.; Brown, E.R.; Lee, D.Y. and Kennedy, T.W. (1996). Hot Mix
Asphalt Materials, Mixture Design, and Construction. National Asphalt Pavement
Association Education Foundation. Lanham, MD.
● Asphalt Institute. (1997). Mix Design Methods for Asphalt, 6th ed., MS-02. Asphalt
Institute. Lexington, KY.
"...a simple apparatus suitable for use with the present California Bearing Ratio (CBR) equipment
The most promising method eventually proved to be the Marshall Stability Method developed by Bruce
G. Marshall at the Mississippi Highway Department in 1939. WES took the original Marshall Stability
Test and added a deformation measurement (using a flow meter) that was reasoned to assist in detecting
excessively high asphalt contents. This appended test was eventually recommended for adoption by the
U.S. Army because:
1. It was designed to stress the entire sample rather than just a portion of it.
WES continued to refine the Marshall method through the 1950s with various tests on materials, traffic
loading and weather variables. Today the Marshall method, despite its shortcomings, is probably the
most widely used mix design method in the world. It has probably become so widely used because (1) it
was adopted and used by the U.S. military all over the world during and after WWII and (2) it is simple,
compact and inexpensive.
4.2 Procedure
The Marshall mix design method consists of 6 basic steps:
1. Aggregate selection.
1. Determine aggregate physical properties. This consists of running various tests to determine
properties such as:
3. Perform blending calculations to achieve the mix design aggregate gradation. Often,
aggregates from more than one source or stockpile are used to obtain the final aggregate
gradation used in a mix design. Trial blends of these different gradations are usually
calculated until an acceptable final mix design gradation is achieved. Typical considerations
for a trial blend include:
❍ All gradation specifications must be met. Typical specifications will require the
percent retained by weight on particular sieve sizes to be within a certain band.
❍ The gradation should not be too close to the FHWA's 0.45 power maximum density
curve. If it is, then the VMA is likely to be too low. Gradation should deviate from
the FHWA's 0.45 power maximum density curve, especially on the 2.36 mm (No.
8) sieve.
The Marshall mix design method can use any suitable method for estimating optimum asphalt content
and usually relies on local procedures or experience.
Based on the results of the optimum asphalt binder content estimate, samples are typically prepared at
0.5 percent by weight of mix increments, with at least two samples above the estimated asphalt binder
content and two below.
Each sample is then heated to the anticipated compaction temperature and compacted with a Marshall
hammer, a device that applies pressure to a sample through a tamper foot (see Figure 5.8). Some
hammers are automatic and some are hand operated. Key parameters of the compactor are:
● Sample size = 102 mm (4-inch) diameter cylinder 64 mm (2.5 inches) in height (corrections
can be made for different sample heights)
● Tamper foot = Flat and circular with a diameter of 98.4 mm (3.875 inches) corresponding to
an area of 76 cm2 (11.8 in2).
● Compaction pressure = Specified as a 457.2 mm (18 inches) free fall drop distance of a
hammer assembly with a 4536 g (10 lb.) sliding weight.
● Number of blows = Typically 35, 50 or 75 on each side depending upon anticipated traffic
loading.
● Simulation method = The tamper foot strikes the sample on the top and covers almost the
entire sample top area. After a specified number of blows, the sample is turned over and the
procedure repeated.
● AASHTO T 245: Resistance to Plastic Flow of Bituminous Mixtures Using the Marshall
Apparatus
During the loading, an attached dial gauge measures the specimen's plastic flow as a result of the loading
(see Figure 5.9). The flow value is recorded in 0.25 mm (0.01 inch) increments at the same time the
maximum load is recorded.
Typical Marshall design stability and flow criteria are shown in Table 5.3.
Table 5.3: Typical Marshall Design Criteria (from Asphalt Institute, 1979)
● AASHTO T 245: Resistance to Plastic Flow of Bituminous Mixtures Using Marshall Apparatus
These densities are then used to calculate the volumetric parameters of the HMA. Measured void
expressions are usually:
● Air voids (Va), sometimes expressed as voids in the total mix (VTM)
Table 5.4: Typical Marshall Minimum VMA (from Asphalt Institute, 1979)
Nominal Maximum
Minimum VMA
Particle Size
(percent)
(mm) (U.S.)
63 2.5 inch 11
50 2.0 inch 11.5
37.5 1.5 inch 12
25.0 1.0 inch 13
19.0 0.75 inch 14
12.5 0.5 inch 15
9.5 0.375 inch 16
4.75 No. 4 sieve 18
2.36 No. 8 sieve 21
1.18 No. 16 sieve 23.5
❍ Asphalt binder content vs. density. Density will generally increase with increasing
asphalt content, reach a maximum, then decrease. Peak density usually occurs at a
higher asphalt binder content than peak stability.
❍ Asphalt binder content vs. Marshall stability. This should follow one of two trends:
■ Stability decreases with increasing asphalt binder content and does not
show a peak. This curve is common for some recycled HMA mixtures.
❍ Asphalt binder content vs. air voids. Percent air voids should decrease with
increasing asphalt binder content.
❍ Asphalt binder content vs. VMA. Percent VMA should decrease with increasing
asphalt binder content, reach a minimum, then increase.
❍ Asphalt binder content vs. VFA. Percent VFA increases with increasing asphalt
binder content.
2. Determine the asphalt binder content that corresponds to the specifications median air void
content (typically this is 4 percent). This is the optimum asphalt binder content.
3. Determine properties at this optimum asphalt binder content by referring to the plots.
Compare each of these values against specification values and if all are within specification,
then the preceding optimum asphalt binder content is satisfactory. Otherwise, if any of these
properties is outside the specification range the mixture should be redesigned.
4.3 Summary
The Marshall mix design method was developed to address specific mix design issues confronting the
USCOE during World War II. Therefore, it was developed to be simple, light, quick, and reasonably
accurate for the wheel loading of the time. Since then it has been modified and supplemented to
address new concerns but the basic testing apparatus and selection criteria remain the same.
The biggest differentiating aspects of the Marshall method are the Marshall hammer and the Marshall
stability and flow apparatus. Both are probably overly simplistic for high-end or high-load pavements
but they are simple, light, portable and inexpensive.
Currently, WSDOT uses both the Hveem and Superpave mix design 5.1 History
methods. However, Superpave is slated to eventually phase out the older
5.2 Procedure
Hveem method.
5.3 Summary
One of the principal results from the Strategic Highway Research Program (SHRP) was the Superpave mix design method. The
Superpave mix design method was designed to replace the Hveem and Marshall methods. The volumetric analysis common to the
Hveem and Marshall methods provides the basis for the Superpave mix design method. The Superpave system ties asphalt binder
and aggregate selection into the mix design process, and considers traffic and climate as well. The compaction devices from the
Hveem and Marshall procedures have been replaced by a gyratory compactor and the compaction effort in mix design is tied to
expected traffic.
This section consists of a brief history of the Superpave mix design method followed by a general outline of the actual method.
This outline emphasizes general concepts and rationale over specific procedures. Typical procedures are available in the following
documents:
● Roberts, F.L.; Kandhal, P.S.; Brown, E.R.; Lee, D.Y. and Kennedy, T.W. (1996). Hot Mix Asphalt Materials, Mixture
Design, and Construction. National Asphalt Pavement Association Education Foundation. Lanham, MD.
● Asphalt Institute. (2001). Superpave Mix Design. Superpave Series No. 2 (SP-02). Asphalt Institute. Lexington, KY.
● American Association of State Highway and Transportation Officials (AASHTO). (2000 and 2001). AASHTO
Provisional Standards. American Association of State Highway and Transportation Officials. Washington, D.C.
5.1 History
Under the Strategic Highway Research Program (SHRP), an initiative was undertaken to improve materials selection and mixture
design by developing:
1. A new mix design method that accounts for traffic loading and environmental conditions.
When SHRP was completed in 1993 it introduced these three developments and called them the Superior Performing Asphalt
Pavement System (Superpave). Although the new methods of mixture performance testing have not yet been established, the mix
design method is well-established.
5.2 Procedure
The Superpave mix design method consists of 7 basic steps:
1. Aggregate selection.
4. Performance Tests.
As of 2002, once aggregate source properties are tested and prove satisfactory,
aggregate sources are approved for 5 years.
Property Value
Degradation Factor
Aggregate gradation influences such key HMA parameters as stiffness, stability, durability, permeability, workability, fatigue
resistance, frictional resistance and resistance to moisture damage (Roberts et al., 1996). Additionally, the maximum aggregate
Gradation Specifications
Superpave mix design specifies aggregate gradation control points, through which aggregate gradations must pass. These control
points are very general and are a starting point for a job mix formula.
WSDOT uses 9.5 mm (0.375 inch), 12.5 mm (0.5 inch), 19.0 mm (0.75 inch) and 25.0 mm (1 inch) Superpave mixes. WSDOT
gradation requirements are the same as the AASHTO requirements except that the upper and lower control points on the 0.075
mm (No. 200) sieve for the 9.5 mm (0.375 inch), 12.5 mm (0.5 inch) and 19.0 mm (0.75 inch) Superpave mixes are 2.0 and 7.0
percent respectively. The WSDOT upper and lower control points on the 0.075 mm (No. 200) sieve for the 25.0 mm (1 inch)
mix are 1.0 and 7.0 respectively.
Aggregate Blending
It is rare to obtain a desired aggregate gradation from a single aggregate stockpile. Therefore, Superpave mix designs usually draw
upon several different aggregate stockpiles and blend them together in a ratio that will produce an acceptable final blended
gradation. It is quite common to find a Superpave mix design that uses 3 or 4 different aggregate stockpiles (see Figure 5.11).
Figure 5.11: Screen Shot from HMA View Showing a Typical Aggregate Blend from 4 Stockpiles
Typically, several aggregate blends are evaluated prior to performing a complete mix design. Evaluations are done by preparing
an HMA sample of each blend at the estimated optimum asphalt binder content then compacting it. Results from this evaluation
can show whether or not a particular blend will meet minimum VMA requirements and Ninitial or Nmax requirements.
where: P0.075 = mass of particles passing the 0.075 mm (No. 200) sieve
Pbe = effective binder content = the total asphalt binder content of a paving
mixture less the portion of asphalt binder that is lost by absorption into the
aggregate particles.
Dust-to-binder ratio specifications are normally 0.6 - 1.2, but a ratio of up to 1.6 may be used at an agency's discretion (AASHTO,
2001).
The WSDOT Superpave dust-to-binder ratio must fall between 0.6 and
1.6.
"Consensus requirements" came about because SHRP did not specifically address aggregate properties and it was thought that
there needed to be some guidance associated with the Superpave mix design method. Therefore, an expert group was convened
and they arrived at a consensus on several aggregate property requirements - the "consensus requirements". This group
recommended minimum angularity, flat or elongated particle and clay content requirements based on:
● The anticipated traffic loading. Desired aggregate properties are different depending upon the amount of traffic
loading. Traffic loading numbers are based on the anticipated traffic level on the design lane over a 20-year
period regardless of actual roadway design life (AASHTO, 2000b).
● Depth below the surface. Desired aggregate properties vary depending upon their intended use as it relates to depth
below the pavement surface.
These requirements are imposed on the final aggregate blend and not the individual aggregate sources.
WSDOT uses a 15-year traffic loading instead of the 20-year period listed in the
consensus requirement tables because WSDOT typically designs overlays for a
15-year design life.
Property Value
Coarse aggregate angularity is important to mix design because smooth, rounded aggregate particles do not interlock with one
another nearly as well as angular particles. This lack of interlock makes the resultant HMA more susceptible to rutting. Coarse
aggregate angularity can be determined by any number of test procedures that are designed to determine the percentage of
fractured faces. Table 5.5 lists Superpave requirements.
Fine aggregate angularity is important to mix design for the same reasons as coarse aggregate angularity - rut prevention. Fine
aggregate angularity is quantified by an indirect method often called the National Aggregate Association (NAA) flow test. This
test consists of pouring the fine aggregate into the top end of a cylinder and determining the amount of voids. The more voids, the
more angular the aggregate. Voids are determined by the following equation:
An excessive amount of flat or elongated aggregate particles can be detrimental to HMA. Flat/elongated particles tend to
breakdown during compaction (giving a different gradation than determined in mix design), decrease workability, and lie flat after
compaction (resulting in a mixture with low VMA) (Roberts et al., 1996). Flat or elongated particles are typically identified using
ASTM D 4791, Flat or Elongated Particles in Coarse Aggregate. Table 5.7 shows the Superpave recommended flat or elongated
particle requirements.
Maximum Percentage of
20-yr Traffic Loading
Particles with Length/Thickness
(in millions of ESALs)
>5
< 0.3 -
0.3 to < 3
3 to < 10
10
10 to < 30
≥ 30
Clay Content
The sand equivalent test measures the amount of clay content in an aggregate sample. If clay content is too high, clay could
preferentially adhere to the aggregate over the asphalt binder. This leads to a poor aggregate-asphalt binder bonding and possible
stripping. To prevent excessive clay content, Superpave uses the sand equivalent test requirements of Table 5.8.
< 0.3
40
0.3 to < 3
3 to < 10
45
10 to < 30
≥ 30 50
1. Pavement temperature. The designer inputs the design pavement temperatures directly.
2. Air temperature. The designer inputs the local air temperatures, then the software converts them to pavement
temperatures.
3. Geographic area. The designer simply inputs the project location (i.e. state, county and city). From this, the software
retrieves climate conditions from a weather database and then converts air temperatures into pavement temperatures.
Once the design pavement temperatures are determined they can be matched to an appropriate PG asphalt binder.
WSDOT uses the Superpave asphalt binder performance grading system and specifications. Therefore, asphalt
binder must meet the requirements of AASHTO MP 1. WSDOT uses three baseline asphalt binder performance
grades based on geography. These baseline grades are typically used and then adjusted as necessary.
Previously, WSDOT had used the aged residue (AR) viscosity grading. The commonly used grade in this old
system was AR-4000W.
The Superpave mix design method determines both a high and a low design pavement temperature. These temperatures are
determined as follows:
● High pavement temperature - based on the 7-day average high air temperature of the surrounding area.
● Low pavement temperature - based on the 1-day low air temperature of the surrounding area.
Using these temperatures as a starting point, Superpave then applies a reliability concept to determine the appropriate PG asphalt
binder. PG asphalt binders are specified in 6°C increments.
Design pavement temperature calculations are based on HMA pavements subjected to fast moving traffic (Roberts et al., 1996).
Specifically, the Dynamic Shear Rheometer (DSR) test is conducted at a rate of 10 radians per second, which corresponds to a
traffic speed of about 90 km/hr (55 mph) (Roberts et al., 1996). Pavements subject to significantly slower (or stopped) traffic such
as intersections, toll booth lines and bus stops should contain a stiffer asphalt binder than that which would be used for fast-
moving traffic. Superpave allows the high temperature grade to be increased by one grade for slow transient loads and by two
grades for stationary loads. Additionally, the high temperature grade should be increased by one grade for anticipated 20-year
loading in excess of 30 million ESALs. For pavements with multiple conditions that require grade increases only the largest grade
increase should be used. Therefore, for a pavement intended to experience slow loads (a potential one grade increase) and greater
than 30 million ESALs (a potential one grade increase), the asphalt binder high temperature grade should be increased by only one
grade. Table 5.9 shows two examples of design high temperature adjustments - often called "binder bumping".
WSDOT uses the following guidance when considering adjustments to the design high temperature of a PG asphalt binder
(sometimes referred to as "binder bumping"):
The Superpave gyratory compactor (Figure 5.12) was developed to improve mix design's ability to simulate actual field
compaction particle orientation with laboratory equipment (Roberts, 1996).
Each sample is heated to the anticipated mixing temperature, aged for a short time (up to 4 hours) and compacted with the gyratory
compactor, a device that applies pressure to a sample through a hydraulically or mechanically operated load. Mixing and
compaction temperatures are chosen according to asphalt binder properties so that compaction occurs at the same viscosity level
for different mixes. Key parameters of the gyratory compactor are:
● Sample size = 150 mm (6-inch) diameter cylinder approximately 115 mm (4.5 inches) in height (corrections can be
made for different sample heights). Nnote that this sample size is larger than those used for the Hveem and Marshall
methods (see Figure 5.13).
● Load = Flat and circular with a diameter of 149.5 mm (5.89 inches) corresponding to an area of 175.5 cm2 (27.24 in2)
● Simulation method = The load is applied to the sample top and covers almost the entire sample top area. The sample is
inclined at 1.25° and rotates at 30 revolutions per minute as the load is continuously applied. This helps achieve a
sample particle orientation that is somewhat like that achieved in the field after roller compaction.
1. Ninitial. The number of gyrations used as a measure of mixture compactability during construction. Mixes that compact
too quickly (air voids at Ninitial are too low) may be tender during construction and unstable when subjected to traffic.
Often, this is a good indication of aggregate quality - HMA with excess natural sand will frequently fail the Ninitial
requirement. A mixture designed for greater than or equal to 3 million ESALs with 4 percent air voids at Ndesign should
have at least 11 percent air voids at Ninitial.
2. Ndesign. This is the design number of gyrations required to produce a sample with the same density as that expected in
the field after the indicated amount of traffic. A mix with 4 percent air voids at Ndesign is desired in mix design.
3. Nmax. The number of gyrations required to produce a laboratory density that should never be exceeded in the field. If
the air voids at Nmax are too low, then the field mixture may compact too much under traffic resulting in excessively low
air voids and potential rutting. The air void content at Nmax should never be below 2 percent air voids.
Typically, samples are compacted to Ndesign to establish the optimum asphalt binder content and then additional samples are
compacted to Nmax as a check. Previously, samples were compacted to Nmax and then Ninitial and Ndesign were back calculated.
Table 5.10 lists the specified number of gyrations for Ninitial, Ndesign and Nmax while Table 5.11 shows the required densities as a
percentage of theoretical maximum density (TMD) for Ninitial, Ndesign and Nmax. Note that traffic loading numbers are based
on the anticipated traffic level on the design lane over a 20-year period regardless of actual roadway design life (AASHTO,
2001).
Table 5.10: Number of Gyrations for Ninitial, Ndesign and Nmax (from AASHTO, 2001)
≥ 30 9 125 205
* When the estimated 20-year design traffic loading is between 3 and < 10
million ESALs, the agency may, at its discretion, specify
Ninitial = 7, Ndesign = 75 and Nmax = 115.
WSDOT gyration requirements are the same as those shown in Table 5.10.
WSDOT does not use the discretionary values between < 3 and 10 million ESALs.
Table 5.11: Required Densities for Ninitial, Ndesign and Nmax (from AASHTO, 2001)
● AASHTO TP4: Preparing and Determining the Density of Hot-Mix Asphalt (HMA) Specimens by Means of the
Superpave Gyratory Compactor
These densities are then used to calculate the volumetric parameters of the HMA. Measured void expressions are usually:
● Air voids (Va), sometimes called voids in the total mix (VTM)
VMA and VFA must meet the values specified in Table 5.12. Note that traffic loading numbers are based on the anticipated
traffic level on the design lane over a 20-year period regardless of actual roadway design life (AASHTO, 2000b).
Table 5.12: Minimum VMA Requirements and VFA Range Requirements (from AASHTO, 2001)
0.3 to < 3 65 - 78
3 to < 10 15.0 14.0 13.0 12.0 11.0
10 to < 30 65 - 75
≥ 30
9.5 mm (0.375 inch) 12.5 mm (0.5 inch) 19 mm (0.75 inch) 25 mm (1.0 inch)
Superpave Superpave Superpave Superpave
Item
Min. Max. Min. Max. Min. Max. Min. Max.
< 0.3 70 80 70 80 70 80 67 80
0.3 to < 3 65 78 65 78 65 78 65 78
≥3 73 76 65 75 65 75 65 75
1. Air voids at Ninitial > 11 percent (for design ESALs ≥ 3 million). See Table 5.11 for specifics.
2. Air voids at Nmax > 2 percent. See Table 5.11 for specifics.
If requirements 1,2 or 3 are not met the mixture needs to be redesigned. If requirement 4 is not met but close, then asphalt binder
content can be slightly adjusted such that the air void content remains near 4 percent but VFA is within limits. This is because
VFA is a somewhat redundant term since it is a function of air voids and VMA (Roberts et al., 1996). The process is illustrated in
Figure 5.14 (numbers are chosen based on 20-year traffic loading of ≥ 3 million ESALs).
Moisture susceptibility testing is the only performance testing incorporated in the Superpave mix design procedure as of early
2002. The modified Lottman test is used for this purpose.
5.3 Summary
The Superpave mix design method was developed to address specific mix design issues with the Hveem and Marshall methods.
Superpave mix design is a rational method that accounts for traffic loading and environmental conditions. Although not yet fully
complete (the performance tests have not been implemented), Superpave mix design produces quality HMA mixtures. As of 2000,
39 states have adopted, or are planning to adopt, Superpave as their mix design system (NHI, 2000).
2. The use of the PG asphalt binder grading system and its associated asphalt binder selection system.
Even given its many differences when compared to the Hveem or Marshall methods, Superpave still uses the same basic mix
design steps and still strives for an optimum asphalt binder content that results in 4 percent design air voids. Thus, the method is
quite different but the ultimate goals remain fairly consistent.
6 HMA - Testing
When aggregate and asphalt binder are combined to produce a Major Topics on this Page
homogenous substance, that substance, HMA, takes on new
physical properties that are related to but not identical to the 6.1 Mixture Characterization Tests
physical properties of its components. Mechanical laboratory tests
can be used to characterize the basic mixture or predict mixture 6.2 Performance Tests
properties. 6.3 Summary
● Asphalt content/gradation
There are several different ways to measure bulk specific gravity, all of which use slightly different ways to
determine specimen volume:
1. Water displacement methods. These methods, based on Archimedes Principle, calculate specimen volume
by weighing the specimen (1) in a water bath and (2) out of the water bath. The difference in weights can
then be used to calculate the weight of water displaced, which can be converted to a volume using the
specific gravity of water.
❍ Saturated Surface Dry (SSD). The most common method, calculates the specimen volume by
subtracting the mass of the specimen in water from the mass of a saturated surface dry (SSD)
specimen. SSD is defined as the specimen condition when the internal air voids are filled with
water and the surface (including air voids connected to the surface) is dry. This SSD condition
allows for internal air voids to be counted as part of the specimen volume and is achieved by
soaking the specimen in a water bath for 4 minutes then removing it and quickly blotting it dry
with a damp towel. One critical problem with this method is that if a specimen's air voids
are high, and thus potentially interconnected (for dense-graded HMA this occurs at about 8
to 10 percent air voids), water quickly drains out of them as the specimen is removed from
its water bath, which results in an erroneously low volume measurement and thus an
erroneously high bulk specific gravity.
❍ Paraffin. This method determines volume similarly to the water displacement method but uses a
melted paraffin wax instead of water to fill a specimen's internal air voids (see Figure 5.15).
Therefore, after the wax sets there is no possibility of it draining out and, theoretically, a more
accurate volume can be calculated. In practice, the paraffin is difficult to correctly apply and test
results are somewhat inconsistent.
❍ Parafilm. This method wraps the specimen in a thin paraffin film (see Figure 5.16) and then
weighs the specimen in and out of water. Since the specimen is completely wrapped when it is
submerged, no water can get into it and a more accurate volume measurement is theoretically
possible. However, in practice the paraffin film application is quite difficult and test results are
inconsistent.
❍ CoreLok. This method calculates specimen volume like the parafilm method but uses a vacuum
chamber (see Figure 5.17) to shrink-wrap the specimen in a high-quality plastic bag (see Figure
5.18) rather than cover it in a paraffin film. This method has shown some promise in both
accuracy and precision.
2. Dimensional. This method, the simplest, calculates the volume based on height and diameter/width
measurements. Although it avoids problems associated with the SSD condition, it is often inaccurate
because it assumes a perfectly smooth surface thereby ignoring surface irregularities (i.e., the rough surface
texture of a typical specimen).
3. Gamma ray. The gamma ray method is based on the scattering and absorption properties of gamma rays
TMD is determined by taking a sample of oven-dry HMA in loose condition (versus compacted condition), weighing
it and then completely submerging it in a 25°C water bath. A vacuum is then applied for 15 minutes
(see Figure 5.20) to remove any entrapped air. The sample volume is then calculated by
subtracting its mass in water from its dry mass. The formula for calculating TMD is:
Figure 5.20: Containers Used to Agitate and Draw a Vacuum on Submerged TMD Samples
● AASHTO T 209 and ASTM D 2041: Theoretical Maximum Specific Gravity and Density of Bituminous
Paving Mixtures
Solvent extraction, the oldest of the three test methods, uses a chemical solvent (trichloroethylene, 1,1,1-
trichloroethane or methylene chloride) to remove the asphalt binder from the aggregate. Typically, a loose HMA
sample is weighed and then a solvent is added to disintegrate the sample. The asphalt binder/solvent and aggregate
are then separated using a centrifuge (see Figures 5.21 and 5.22) and the aggregate is weighed. The initial and final
weights are compared and the difference is assumed to be the asphalt binder weight. Using this weight and the
weight of the original sample a percent asphalt binder by weight can be calculated. A gradation test can then be run
on the aggregate to determine gradation.
Today, the solvent extraction method is only sparingly used due to the hazardous nature of the specified solvents.
Figure 5.21: Open Centrifuge Used in Figure 5.22: Secondary Centrifuge Used in Solvent
Solvent Extraction Extraction
● AASHTO T 164 and ASTM D 2172: Quantitative Extraction of Bitumen from Bituminous Paving Mixtures
A nuclear asphalt content gauge (see Figure 5.23) measures asphalt content by estimating the actual number of
hydrogen atoms contained within a sample. Similar in theory to a nuclear moisture content gauge used in
construction, the nuclear asphalt content gauge uses a neutron source (such as a 100 μCi specimen of Californium-
252) to emit high energy, “fast” neutrons, which then collide with various nuclei in the sample. Due to momentum
conservation, those neutrons that collide with hydrogen nuclei slow down much quicker than those that collide with
other, larger nuclei. The gauge detector counts only thermal (low energy) or “slow” neutrons thereby making the
detector count proportional to the number of hydrogen atoms in the sample. Since asphalt is a hydrocarbon, the more
hydrogen atoms, the more asphalt. A calibration factor is used to relate thermal neutron count to actual asphalt
content.
The nuclear asphalt content gauge offers a relatively quick (4 to 16 minutes depending upon desired accuracy)
method for measuring asphalt content. Since the gauge actually measures hydrogen nuclei and then correlates their
number with asphalt content, anything affecting the number of hydrogen nuclei in the sample can be a potential
source of error. Because water contains a significant amount of hydrogen (H2O), anything that adds moisture to the
sample (e.g., moisture in the aggregate pores) is a potential error source (Black, 1994).
The ignition furnace test, developed by NCAT to replace the solvent extraction method, determines asphalt binder
content by burning off the asphalt binder of a loose HMA sample. Basically, an HMA sample is weighed and then
placed in a 538°C (1072°F) furnace (see Figure 5.24) and ignited. Once all the asphalt binder has burned off
(determined by a change in mass of less than 0.01 percent over 3 consecutive minutes), the remaining aggregate is
weighed. The initial and final weights are compared and the difference is assumed to be the asphalt binder weight.
Using this weight and the weight of the original sample, a percent asphalt binder by weight can be calculated. A
gradation test can then be run on the aggregate to determine gradation.
A correction factor must be used with the ignition furnace because a certain amount of aggregate fines may be burned
off during the ignition process. The correction factor is determined by placing a sample of known asphalt binder
content in the furnace and comparing the test result with the known asphalt binder content.
Based on a limited National Center for Asphalt Technology (NCAT) study (Prowell, 2002), both traditional and
infrared ignition furnaces, if properly calibrated, should produce statistically similar asphalt contents and recovered
aggregate gradations.
● AASHTO T 308: Determining the Asphalt Binder Content of Hot Mix Asphalt (HMA) by the Ignition
Method
As with asphalt binder characterization, the challenge in HMA performance testing is to develop physical tests that
can satisfactorily characterize key HMA performance parameters and how these parameters change throughout the
life of a pavement. These key parameters are:
● Deformation resistance (rutting). A key performance parameter that can depend largely on HMA mix
design. Therefore, most performance test efforts are concentrated on deformation resistance prediction.
● Fatigue life. A key performance parameter that depends more on structural design and subgrade support
than mix design. Those HMA properties that can influence cracking are largely tested for in Superpave
asphalt binder physical tests. Therefore, there is generally less attention paid to developing fatigue life
performance tests.
● Tensile strength. Tensile strength can be related to HMA cracking - especially at low temperatures. Those
HMA properties that can influence low temperature cracking are largely tested for in Superpave asphalt
binder physical tests. Therefore, there is generally less attention paid to developing tensile strength
performance tests.
● Moisture susceptibility. Certain combinations of aggregate and asphalt binder can be susceptible to
moisture damage. Several deformation resistance and tensile strength tests can be used to evaluate the
moisture susceptibility of a HMA mixture.
● Static creep tests. Apply a static load to a sample and measure how it recovers when the load is removed.
Although these tests measure a specimen's permanent deformation, test results generally do not correlate
will with actual in-service pavement rutting measurements.
● Repeated load tests. Apply a repeated load at a constant frequency to a test specimen for many repetitions
(often in excess of 1,000) and measure the specimen's recoverable strain and permanent deformation. Test
results correlate with in-service pavement rutting measurements better than static creep test results.
● Dynamic modulus tests. Apply a repeated load at varying frequencies to a test specimen over a relatively
short period of time and measure the specimen's recoverable strain and permanent deformation. Some
dynamic modulus tests are also able to measure the lag between the peak applied stress and the peak
resultant strain, which provides insight into a material's viscous properties. Test results correlate
reasonably well with in-service pavement rutting measurements but the test is somewhat involved and
difficult to run.
● Empirical tests. Traditional Hveem and Marshall mix design tests. Test results can correlate well with in-
service pavement rutting measurements but these tests do not measure any fundamental material parameter.
● Simulative tests. Laboratory wheel-tracking devices. Test results can correlate well with in-service
pavement rutting measurements but these tests do not measure any fundamental material parameter.
Each test has been used to successfully predict HMA permanent deformation characteristics however each test has
limitations related to equipment complexity, expense, time, variability and relation to fundamental material
parameters.
A static creep test (see Figure 5.25) is conducted by applying a static load to an HMA specimen and then measuring
the specimen's permanent deformation after unloading (see Figure 5.26). This observed permanent deformation is
then correlated with rutting potential. A large amount of permanent deformation would correlate to higher rutting
potential.
Creep tests have been widely used in the past because of their relative simplicity and availability of equipment.
However, static creep test results do not correlate well with actual in-service pavement rutting (Brown et al., 2001).
Figure 5.25: Unconfined Static Creep Test Figure 5.26: Static Creep Test Plot
● AASHTO TP 9: Determining the Creep Compliance and Strength of Hot Mix Asphalt (HMA) Using the
Indirect Tensile Test Device
A repeated load test applies a repeated load of fixed magnitude and cycle duration to a cylindrical test specimen (see
Figure 5.27). The specimen's resilient modulus can be calculated using the its horizontal deformation and an assumed
Poisson's ratio. Cumulative permanent deformation as a function of the number of load cycles is recorded and can be
correlated to rutting potential. Tests can be run at different temperatures and varying loads. The load varies is
applied in a short pulse followed by a rest period. Repeated load tests are similar in concept to the triaxial resilient
modulus test for unconfined soils and aggregates.
Repeated load tests correlate better with actual in-service pavement rutting than static creep tests (Brown et al., 2001).
Most often, results from repeated load tests are reported using a cumulative axial strain curve like the one shown in
Figure 5.28. The flow number (FN) is the load cycles number at which tertiary flow begins. Tertiary flow can be
differentiated from secondary flow by a marked departure from the linear relationship between cumulative strain and
number of cycles in the secondary zone. It is assumed that in tertiary flow, the specimen's volume remains constant.
The flow number (FN) can be correlated with rutting potential.
diametral plane. Diametral testing has two critical shortcomings that hinder its ability to determine permanent
deformation characteristics (Brown et al., 2001):
1. The state of stress is non-uniform and strongly dependent on the shape of the specimen. At high
temperature or load, permanent deformation produces changes in the specimen shape that significantly
affect both the state of stress and the test measurements.
2. During the test, the only relatively uniform state of stress is tension along the vertical diameter of the
specimen. All other states of stress are distinctly nonuniform.
● AASHTO TP 7: Determining the Permanent Deformation and Fatigue Cracking Characteristics of Hot Mix
Asphalt (HMA) Using the Superpave Shear Tester (SST) - Procedure F
● AASHTO TP 31: Determining the Resilient Modulus of Bituminous Mixtures by Indirect Tension
● ASTM D 4123: Indirect Tension Test for Resilient Modulus of Bituminous Mixtures
Dynamic modulus tests apply a repeated axial cyclic load of fixed magnitude and cycle duration to a test specimen
(see Figure 5.25). Test specimens can be tested at different temperatures and three different loading frequencies
(commonly 1, 4 and 16 Hz). The applied load varies and is usually applied in a haversine wave (inverted cosine
offset by half its amplitude - a continuous haversine wave would look like a sine wave whose negative peak is at
zero). Figure 5.29 is a schematic of a typical dynamic modulus test.
Dynamic modulus tests differ from the repeated load tests in their loading cycles and frequencies. While repeated
load tests apply the same load several thousand times at the same frequency, dynamic modulus tests apply a load over
a range of frequencies (usually 1, 4 and 16 Hz) for 30 to 45 seconds (Brown et al., 2001). The dynamic modulus test
is more difficult to perform than the repeated load test since a much more accurate deformation measuring system is
necessary.
The dynamic modulus test measures a specimen's stress-strain relationship under a continuous sinusoidal loading.
For linear (stress-strain ratio is independent of the loading stress applied) viscoelastic materials this relationship is
defined by a complex number called the “complex modulus” (E*) (Witczak et al., 2002) as seen in the equation
below:
i = imaginary number
The absolute value of the complex modulus, |E*|, is defined as the dynamic modulus and is calculated as follows
(Witczak et al., 2002):
The dynamic modulus test can be advantageous because it can measure also measure a specimen's phase angle (φ),
which is the lag between peak stress and peak recoverable strain. The complex modulus, E*, is actually the
summation of two components: (1) the storage or elastic modulus component and (2) the loss or viscous modulus. It
is an indicator of the viscous properties of the material being evaluated.
1. Superpave shear tester (SST). The SST FSCH test is a is a constant strain test (as opposed to a constant
stress test). Test specimens are 150 mm (6 inches) in diameter and 50 mm (2 inches) tall (see Figure 5.32).
To conduct the test the HMA sample is essentially glued to two plates (see Figures 5.33 through 5.35) and
then inserted into the SST. Horizontal strain is applied at a range of frequencies (from 10 to 0.1 Hz) using
a haversine loading pattern, while the specimen height is maintained constant by compressing or pulling it
vertically as required. The SST produces a constant strain of about 100 microstrain (Witczak et al., 2002).
The SST is quite expensive and requires a highly trained operator to run thus making it impractical for field
use and necessitating further development.
2. Field shear tester (FST). The FST FSCH test is a is a constant stress test (as opposed to a constant strain
test). The FST is a derivation of the SST and is meant to be less expensive and easier to use. For instance,
rather than compressing or pulling the sample to maintain a constant height like the SST, the FST maintains
constant specimen height using rigid spacers attached to the specimen ends. Further, the FST shears the
specimen in the diametral plane.
Figure 5.32: Superpave Shear Tester (SST) Figure 5.33: Loading Chamber
Figure 5.34: Prepared Sample Figure 5.35: Prepared Sample (left) and Sample After Test
(middle and right).
● Shear dynamic modulus. AASHTO TP 7: Determining the Permanent Deformation and Fatigue Cracking
Characteristics of Hot Mix Asphalt (HMA) Using the Simple Shear Test (SST) Device, Procedure E -
Frequency Sweep Test at Constant Height.
The Hveem stabilometer and cohesiometer and Marshall stability and flow tests are empirical tests used to quantify
an HMA's potential for permanent deformation. They are discussed in their mix design sections.
Laboratory wheel-tracking devices (see Video 5.1) measure rutting by rolling a small loaded wheel device repeatedly
across a prepared HMA specimen. Rutting in the test specimen is then correlated to actual in-service pavement
rutting. Laboratory wheel-tracking devices can also be used to make moisture susceptibility and stripping predictions
by comparing dry and wet test results Some of these devices are relatively new and some have been used for upwards
of 15 years like the Laboratoire Central des Ponts et Chausées (LCPC) wheel tracker - also known as the French
Rutting Tester (FRT). Cooley et al. (2000) reviewed U.S. loaded wheel testers and found:
● Results obtained from the wheel tracking devices correlate reasonably well to actual field performance
when the in-service loading and environmental conditions of that location are considered.
● Wheel tracking devices can reasonably differentiate between binder performance grades.
● Wheel tracking devices, when properly correlated to a specific site’s traffic and environmental conditions,
have the potential to allow the user agency the option of a pass/fail or “go/no go” criteria. The ability of the
wheel tracking devices to adequately predict the magnitude of the rutting for a particular pavement has not
been determined at this time.
● A device with the capability of conducting wheel-tracking tests in both air and in a submerged state, will
offer the user agency the most options of evaluating their materials.
In other words, wheel tracking devices have potential for rut and other measurements but the individual user must be
careful to establish laboratory conditions (e.g., load, number of wheel passes, temperature) that produce consistent
and accurate correlations with field performance.
One of the typical ways of estimating in-place HMA fatigue properties is the flexural test (see Figures 5.36 and
5.37). The flexural test determines the fatigue life of a small HMA beam specimen (380 mm long x 50 mm thick x 63
mm wide) by subjecting it to repeated flexural bending until failure (see Figure 5.38). The beam specimen is sawed
from either laboratory or field compacted HMA. Results are usually plotted to show cycles to failure vs. applied
stress or strain.
● AASHTO TP 8: Determining the Fatigue Life of Compacted Hot-Mix Asphalt (HMA) Subjected to
Repeated Flexural Bending
The indirect tensile test uses the same testing device as the diametral repeated load test and applies a constant rate of
vertical deformation until failure. It is quite similar to the splitting tension test used for PCC.
● AASHTO TP 9: Determining the Creep Compliance and Strength of Hot Mix Asphalt (HMA) Using the
Indirect Tensile Test Device
The thermal cracking test determines the tensile strength and temperature at fracture of an HMA sample by measuring
the tensile load in a specimen which is cooled at a constant rate while being restrained from contraction. The test is
terminated when the sample fails by cracking.
● AASHTO TP 10: Method for Thermal Stress Restrained Specimen Tensile Strength
Of particular note, temperature has a profound effect on HMA stiffness. Table 5.13 shows some typical HMA
resilient modulus values at various temperatures. Figure 5.39 shows that HMA resilient modulus changes by a factor
of about 100 for a 56 °C (100 °F) temperature change for "typical" dense-graded HMA mixtures. This can affect
HMA performance parameters such as rutting and shoving. This is one reason why the Superpave PG binder grading
system accounts for expected service temperatures when specifying an asphalt binder.
Table 5.13: Typical Resilient Modulus Values for HMA Pavement Materials
will be described.
The modified Lottman test basically compares the indirect tensile strength test results of a dry sample and a sample
exposed to water/freezing/thawing. The water sample is subjected to vacuum saturation, an optional freeze cycle,
followed by a freeze and a warm-water cycle before being tested for indirect tensile strength (AASHTO, 2000a). Test
results are reported as a tensile strength ratio:
Generally a minimum TSR of 0.70 is recommended for this method, which should be applied to field-produced
rather than laboratory-produced samples (Roberts et al., 1996). For laboratory samples produced in accordance with
AASHTO TP 4 (Method for Preparing and Determining the Density of Hot-Mix Asphalt (HMA) Specimens by
Means of the Superpave Gyratory Compactor), AASHTO MP 2 (Specification for Superpave Volumetric Mix
Design) specifies a minimum TSR of 0.80.
In addition to the modified Lottman test, some state agencies use the Hamburg Wheel Tracking Device (HWTD) to
test for moisture susceptibility since the test can be carried out in a warm water bath.
6.3 Summary
All pavements can be described by their fundamental characteristics and performance. Thus, HMA tests are an
integral part of mix design because they provide (1) basic HMA characteristics and (2) the means to relate mix design
to intended performance. Without performance tests, mix design has no proven relationship with performance
(Roberts et al., 1996). The Hveem and Marshall mix design methods use two basic performance tests (Hveem
stabilometer and the Marshall stability and flow), but these tests are empirical and limited in their predictive ability.
New and better performance tests are still being developed and evaluated. In fact, Superpave has yet to implement
performance testing because of this. The performance tests presented in this section are those that are most
commonly used in the industry today, although it is quite likely that these tests will change in the future as better
methods and equipment are developed.
7 PCC - Fundamentals
PCC consists of three basic ingredients: aggregate, Major Topics on this Page
water and portland cement. According to the Portland
Cement Association (PCA, 1988): 7.1 Concepts
7.2 Variables
"The objective in designing concrete mixtures is
to determine the most economical and practical 7.3 Objectives
combination of readily available materials to
produce a concrete that will satisfy the 7.4 Basic Procedure
performance requirements under particular 7.5 Summary
conditions of use."
PCC mix design has evolved chiefly through experience and well-documented empirical relationships.
Normally, the mix design procedure involves two basic steps:
1. Mix proportioning. This step uses the desired PCC properties as inputs then determines the
required materials and proportions based on a combination of empirical relationships and local
experience. There are many different PCC proportioning methods of varying complexity that
work reasonably well.
2. Mix testing. Trial mixes are then evaluated and characterized by subjecting them to several
laboratory tests. Although these characterizations are not comprehensive, they can give the
mix designer a good understanding of how a particular mix will perform in the field during
construction and under subsequent traffic loading.
This section covers mix design fundamentals common to all PCC mix design methods. First, two basic
concepts (mix design as a simulation and weight-volume terms and relationships) are discussed to set a
framework for subsequent discussion. Second, the variables that mix design may manipulate are
presented. Third, the fundamental objectives of mix design are presented. Finally, a generic mix design
procedure is presented.
7.1 Concepts
Before discussing any mix design specifics, it is important to understand a couple of basic mix design
concepts:
Being a simulation, mix design has its limitations. Specifically, there are substantial differences
between laboratory and field conditions. For instance, mix testing is generally done on small samples
that are cured in carefully controlled conditions. These values are then used to draw conclusions about
how a mix will behave under field conditions. Despite such limitations mix design procedures can
provide a cost effective and reasonably accurate simulation that is useful in making mix design
decisions.
7.2 Variables
PCC is a complex material formed from some very basic ingredients. When used in pavement, this
material has several desired performance characteristics - some of which are in direct conflict with one
another. PCC pavements must resist deformation, crack in a controlled manner, be durable over time,
resist water damage, provide a good tractive surface, and yet be inexpensive, readily made and easily
placed. In order to meet these demands, mix design can manipulate the following variables:
1. Aggregate. Items such as type (source), amount, gradation and size, toughness and abrasion
resistance, durability and soundness, shape and texture as well as cleanliness can be measured,
judged and altered to some degree.
2. Portland cement. Items such as type, amount, fineness, soundness, hydration rate and
additives can be measured, judged and altered to some degree.
3. Water. Typically the volume and cleanliness of water are of concern. Specifically, the
volume of water in relation to the volume of portland cement, called the water-cement ratio, is
of primary concern. Usually expressed as a decimal (e.g., 0.35), the water-cement ratio has a
major effect on PCC strength and durability.
4. Admixtures. Items added to PCC other than portland cement, water and aggregate.
Admixtures can be added before, during or after mixing and are used to alter basic PCC
properties such as air content, water-cement ratio, workability, set time, bonding ability,
coloring and strength.
7.3 Objectives
By manipulating the mixture variables of aggregate, portland cement, water and admixtures, mix design
seeks to achieve the following qualities in the final PCC product (Mindess and Young, 1981):
1. Strength. PCC should be strong enough to support expected traffic loading. In pavement
applications, flexural strength is typically more important than compressive strength (although
both are important) since the controlling PCC slab stresses are caused by bending and not
compression. In its most basic sense, strength is related to the degree to which the portland
cement has hydrated. This degree of hydration is, in turn, related to one or more of the
following:
❍ Water-cement ratio. The strength of PCC is most directly related to its capillary
porosity. The capillary porosity of a properly compacted PCC is determined by its
water-cement ratio (Mindess and Young, 1981). Thus, the water-cement ratio is an
easily measurable PCC property that gives a good estimate of capillary porosity and
thus, strength. The lower the water-cement ratio, the fewer capillary pores and thus,
the higher the strength. Specifications typically include a maximum water-cement
ratio as a strength control measure.
entrained air (by volume of the total mixture) increases, the voids-cement ratio
(voids = air + water) decreases. This generally results in a strength reduction.
However, air-entrained PCC can have a lower water-cement ratio than non-air-
entrained PCC and still provide adequate workability. Thus, the strength reduction
associated with a higher air content can be offset by using a lower water-cement
ratio. For moderate-strength concrete (as is used in rigid pavements) each
percentile of entrained air can reduce the compressive strength by about 2 - 6
percent (PCA, 1988).
❍ Cement properties. Properties of the portland cement such as fineness and chemical
composition can affect strength and the rate of strength gain. Typically, the type of
portland cement is specified in order to control its properties.
3. Durability. PCC should not suffer excessive damage due to chemical or physical attacks
during its service life. As opposed to HMA durability, which is mainly concerned with aging
effects, PCC durability is mainly concerned with specific chemical and environmental
conditions that can potentially degrade PCC performance. Durability is related to:
through osmosis (PCA, 1988). Entrained air voids act as expansion chambers for
freezing and migrating water and thus, specifying a minimum entrained air content
can minimize freeze-thaw damage.
4. Skid resistance. PCC placed as a surface course should provide sufficient friction when in
contact with a vehicle's tire. In mix design, low skid resistance is generally related to
aggregate characteristics such as texture, shape, size and resistance to polish. Smooth,
rounded or polish-susceptible aggregates are less skid resistant. Tests for particle shape and
texture can identify problem aggregate sources. These sources can be avoided, or at a
minimum, aggregate with good surface and abrasion characteristics can be blended in to
provide better overall characteristics.
5. Workability. PCC must be capable of being placed, compacted and finished with reasonable
effort. The slump test, a relative measurement of concrete consistency, is the most
common method used to quantify workability. Workability is generally related to one
or more of the following:
❍ Water content. Water works as a lubricant between the particles within PCC.
Therefore, low water content reduces this lubrication and makes for a less workable
mix. Note that a higher water content is generally good for workability but
generally bad for strength and durability, and may cause segregation and bleeding.
Where necessary, workability should be improved by redesigning the mix to
increase the paste content (water + portland cement) rather than by simply adding
more water or fine material (Mindess and Young, 1981).
❍ Aggregate texture, shape and size. Flat, elongated or angular particles tend to
interlock rather than slip by one another making placement and compaction more
difficult. Tests for particle shape and texture can identify possible workability
problems.
❍ Aggregate gradation. Gradations deficient in fines make for less workable mixes.
In general, fine aggregates act as lubricating "ball bearings" in the mix. Gradation
specifications are used to ensure acceptable aggregate gradation.
❍ Aggregate porosity. Highly porous aggregate will absorb a high amount of water
leaving less available for lubrication. Thus, mix design usually corrects for the
anticipated amount of absorbed water by the aggregate.
❍ Air content. Air also works as a lubricant between aggregate particles. Therefore,
low air content reduces this lubrication and makes for a less workable mix. A
volume of air-entrained PCC requires less water than an equal volume of non-air-
entrained PCC of the same slump and maximum aggregate size (PCA, 1988).
❍ Cement properties. Portland cements with higher amounts of C3S and C3A will
hydrate quicker and lose workability faster.
Knowing these objectives, the challenge in mix design is then to develop a relatively simple procedure
with a minimal amount of tests and samples that will produce a mix with all the qualities discussed
above.
1. Aggregate selection. No matter the specific method, the overall mix design procedure begins
with evaluation and selection of aggregate and asphalt binder sources. Different authorities
specify different methods of aggregate acceptance. Typically, a battery of aggregate physical
tests is run periodically on each particular aggregate source. Then, for each mix design,
gradation and size requirements are checked. Normally, aggregate from more than one source
is required to meet gradation requirements.
2. Portland cement selection. Typically, a type and amount of portland cement is selected based
on past experience and empirical relationships with such factors as compressive strength (at a
given age), water-cement ratio and chemical susceptibility.
4. Testing. Run laboratory tests on properly prepared samples to determine key mixture
characteristics. It is important to understand that these tests are not comprehensive nor are
they exact reproductions of actual field conditions.
The selected PCC mixture should be the one that, based on test results, best satisfies the mix design
objectives.
7.5 Summary
PCC mix design is a laboratory process used to determine appropriate proportions and types of
aggregate, portland cement, water and admixtures that will produce desired PCC properties. Typical
desired properties in PCC for pavement are adequate strength, controlled shrinkage, durability, skid
resistance and workability. Although mix design has many limitations it had proven to be a cost-
effective simulation that is able to provide crucial information that can be used to formulate a high-
performance PCC.
8.7 any
Contractors provide their own PCC mix designs for WSDOT jobs. Therefore, Fine Aggregate Content
mix design method can be used as longs as specifications are met. 8.8 Adjustments for Aggregate Moisture
8.9 Summary
This section is a general outline of the ACI proportioning
method with specific emphasis on PCC for pavements. It emphasizes general concepts and rationale over
specific procedures. Typical procedures are available in the following documents:
● The American Concrete Institute's (ACI) Standard Practice for Selecting Proportions for Normal,
Heavyweight, and Mass Concrete (ACI 211.1-91) as found in their ACI Manual of Concrete Practice
2000, Part 1: Materials and General Properties of Concrete.
● The Portland Cement Association's (PCA) Design and Control of Concrete Mixtures, 14th edition
(2002) or any earlier edition.
The standard ACI mix design procedure can be divided up into 8 basic steps:
1. Choice of slump
4. Water-cement ratio
5. Cement content
WSDOT specifies ACI 211.1 as a guide to determine mix proportions. Additionally, some of the WSDOT 2002 Standard
Specifications for Road, Bridge and Municipal Construction (M 41-10) specified properties are shown in the table below.
None explicitly set. In slipform paving, slab edges that slump down
Slump
below 0.25 inches of their plan height shall be corrected.
Not less than 1000 lbs. less than that established in the mix design as
the arithmetic mean of the five sets of 28 day compressive strength
Compressive Strength
cylinders (cast at the same time as the flexural strength beams used to
pre-qualify the mix design), or 3000 psi, whichever is higher.
1Many of these specification limits are for use in a statistical analysis of multiple random samples. Therefore, the limits
listed above are not absolute (i.e., all samples must be above or below the specified limit); rather, they should be used in
conjunction with WSDOT statistical acceptance specifications.
8.1 Slump
The choice of slump is actually a choice of mix workability. Workability can be described as a combination of
several different, but related, PCC properties related to its rheology:
● Ease of mixing
● Ease of placing
● Ease of compaction
● Ease of finishing
Generally, mixes of the stiffest consistency that can still be placed adequately should be used (ACI, 2000).
Typically slump is specified, but Table 5.14 shows general slump ranges for specific applications. Slump
specifications are different for fixed form paving and slip form paving. Table 5.15 shows typical and extreme
state DOT slump ranges.
Table 5.14: Slump Ranges for Specific Applications (after ACI, 2000)
Slump
Type of Construction
(mm) (inches)
Table 5.15: Typical State DOT Slump Specifications (data taken from ACPA, 2001)
Note that the use of water-reducing and/or set-controlling admixtures can substantially reduce the amount of
mixing water required to achieve a given slump.
Mixing Water Quantity in kg/m3 (lb/yd3) for the listed Nominal Maximum Aggregate Size
Non-Air-Entrained PCC
Air-Entrained PCC
● The calculated amount based on the selected mixing water content and water-cement ratio.
● The specified minimum cement content, if applicable. Most state DOTs specify minimum cement
contents in the range of 300 - 360 kg/m3 (500 - 600 lbs/yd3).
An older practice used to be to specify the cement content in terms of the number of 94 lb. sacks of portland
cement per cubic yard of PCC. This resulted in specifications such as a "6 sack mix" or a "5 sack mix". While
these specifications are quite logical to a small contractor or individual who buys portland cement in 94 lb.
sacks, they do not have much meaning to the typical pavement contractor or batching plant who buys portland
cement in bulk. As such, specifying cement content by the number of sacks should be avoided.
Notes:
1. Aggregate weights. Aggregate volumes are calculated based on oven dry unit weights, but aggregate
is typically batched based on actual weight. Therefore, any moisture in the aggregate will increase
its weight and stockpiled aggregates almost always contain some moisture. Without correcting for
2. Amount of mixing water. If the batched aggregate is anything but saturated surface dry it will absorb
water (if oven dry or air dry) or give up water (if wet) to the cement paste. This causes a net change
in the amount of water available in the mix and must be compensated for by adjusting the amount of
mixing water added.
8.9 Summary
The ACI mix design method is one of many available methods. It has been presented here to give a general
idea of the types of calculations and decisions that are typical in PCC mix design.
9 PCC - Testing
When aggregate, water and portland cement paste are combined to produce Major Topics on this Page
a homogenous substance, that substance takes on new physical properties
that are related to but not identical to the physical properties of its 9.1 Workability
components. Thus, several common mechanical laboratory tests are used to
characterize the basic mixture and predict mixture properties. Unlike 9.2 Strength
HMA, it is difficult to draw a clean distinction between characterization 9.3 Durability
tests and performance tests. Typically, PCC is characterized by slump, air
content and strength. However, these characteristics can also be used as 9.4 Early Age Behavior
performance predictors for workability, durability and strength
respectively. Therefore, this section does not distinguish between mixture 9.5 Summary
characterization tests and performance tests.
Whereas HMA tests are often scale simulations of actual field conditions (such as rut tests), PCC tests are directed more at
the basic physical properties of PCC as a material.
The challenge in PCC testing is to develop physical tests that can satisfactorily characterize key PCC performance parameters
and the nature of their change throughout the life of a pavement. These key parameters are:
● Workability. This parameter, typically measured by slump, is indicative of fresh concrete rheology.
● Strength. This parameter is related to a rigid pavement's ability to support loads. Flexural strength is commonly
used in design and then correlated to compressive strength for use in field tests.
● Durability. Several tests can be conducted to determine susceptibility to freeze-thaw or chemical attack damage.
● Early age behavior. HIPERPAV, a software program, can be used to predict early-age PCC behavior.
Although there are many different PCC tests, only those typically used on pavement PCC are discussed in this Guide.
9.1 Workability
Slump Test
The slump test (see Figure 5.40) is the most common test for
workability. The slump test involves hand placing an amount of fresh
concrete into a metal cone and then measuring the distance the fresh PCC
falls (or "slumps") when the cone is removed.
"The bottom line is that the slump test is useful only as a comparative tool. If changes in slump are
greater than 2 inches on a given job, one can conclude that there was likely a change in the mix.
Variation in slump less than 2 inches is more than likely from a combination of the testing and typical
concrete variability. No conclusion can be drawn from slump tests to the quality of the material.
Strength measurements must be used to indicate quality."
9.2 Strength
Strength is probably the most well-known PCC performance parameter. Compressive and tensile strength are fundamental to
any building material in order to properly proportion and design structural items made from that material. Although PCC is
most often known for its compressive strength, it is typically its tensile strength (or more exactly, its flexural strength) that
governs its use in rigid pavements. However, given the popularity and relative ease of the compressive test, both tests are
typically used in pavement applications. Strength concepts covered are:
● Compressive strength
● Tensile strength (including splitting tension tests and flexural strength tests)
measured. Most often, 28-day strength is specified although other strengths such as 1-day, 7-day and 90-day strength can be
used as well. For pavement applications, strength at a particular age is quite important because typically, rigid pavements
cannot be opened to traffic until the PCC reaches a certain strength. Curing methods can play a major role in PCC strength
gain. Often, PCC maturity is used to estimate strength at a particular time.
Compressive strength is most often measured by forming 150 mm diameter, 300 mm long (6 inch diameter, 12 inches long)
test cylinders and then breaking them at a specified age (typically 28 days) although it can also be performed on specimens of
different sizes and origins (such as field cores or the remnants of a flexural test).
Some state agencies use compressive strength as a field quality assurance measurement of a flexural strength specification.
Flexural strength is first correlated to compressive strength based on mix design test results. Then, using this correlation,
quality assurance field tests can use the easier and more widely known compressive strength test, which can be converted
back to flexural strength through the previously determined correlations.
WSDOT correlates PCC compressive strength to flexural strength and then uses
compressive strength in acceptance testing.
Most pavement PCC has a compressive strength between 20.68 and 34.47 MPa (3000 and 5000 psi) (ACPA,
2001). High-strength PCC (usually defined as PCC with a compressive strength of at least 41.37 MPa (6000
psi)) has been designed for compressive strengths of over 137.90 MPa (20,000 psi) for use in building
applications.
● AASHTO T 140 and ASTM C 116: Compressive Strength of Concrete Using Portions of Beams Broken in Flexure
tensile stresses at the bottom of the PCC surface course as it is subjected to loading. These stresses are typically the
controlling structural design stresses. Tensile strength is difficult to directly measure because of secondary stresses induced
by gripping a specimen so that it may be pulled apart. Therefore, tensile stresses are typically measured indirectly by one of
two means: a splitting tension test or a flexural strength test.
A splitting tension test uses a standard 150 mm diameter, 300 mm long (6-inch diameter, 12" long) test cylinder laid on its
side. A diametral compressive load is then applied along the length of the cylinder until it fails (see Figure 5.41). Because
PCC is much weaker in tension than compression, the cylinder will typically fail due to horizontal tension and not vertical
compression.
● AASHTO T 198 and ASTM C 496: Splitting Tensile Strength of Cylindrical Concrete Specimens
Flexural strength (sometimes called the modulus of rupture) is typically used in PCC mix design for pavements because it
best simulates slab flexural stresses as they are subjected to loading. Because the flexural test involves bending a beam
specimen, there will be some compression involved, and thus flexural strength will generally be slightly higher than tensile
strength measured using a split tension test. Usually, mix designs are typically tested for both flexural and compressive
strength; they must meet a minimum flexural strength, which is then correlated to measured compressive strengths so that
compressive strength (an easier test) can be used in field acceptance tests.
There are two basic flexural tests: the third-point loading (Figure 5.42) and the center-point loading (Figure 5.43). For
maximum aggregate sizes less than 50 mm (2 inches), each test is conducted on a 152 x 152 x 508 mm (6 x 6 x 20 inch) PCC
beam (see Figures 5.44 and 5.45). The beam is supported on each end and loaded at its third points (for the third-point
loading test) or at the middle (for the center-point loading test) until failure. The modulus of rupture is then calculated and
reported as the flexural strength. The third-point loading test is preferred because, ideally, in the middle third of the span the
sample is subjected to pure moment with zero shear (Mindess and Young, 1981). In the center-point test, the area of eventual
failure contains not only moment induced stresses but also shear stress and unknown areas of stress concentration. In general,
the center-point loading test gives results about 15 percent higher (ACPA, 2001).
● AASHTO T 97 and ASTM C 78: Flexural Strength of Concrete (Using Simple Beam with Third-Point
Loading)
● AASHTO T 177 and ASTM C 293: Flexural Strength of Concrete (Using Simple Beam with Center-
Point Loading)
9.3 Durability
Durability is a measure of how PCC performs over time. Durability is one factor in PCC pavement performance. Typically,
the two major factors that affect PCC pavement durability are freeze-thaw cycles and chemical attack. Fortunately, steps can
be taken to mitigate these factors and tests are available to determine PCC vulnerability to them.
9.3.1 Freeze-Thaw
Freeze-thaw resistance is important in order to avoid excessive cracking, scaling and crumbling. As water freezes it increases
in volume by about 9 percent. Thus, as the water in PCC freezes and expands it exerts osmotic and hydraulic pressures on
capillaries and pores within the cement paste. If these pressures exceed the tensile strength of the cement paste, the paste will
dilate and rupture (PCA, 1988). As this process repeats itself over a number of freeze-thaw cycles, the result can be cracking,
In the late 1930s it was discovered that purposefully increasing PCC air content (called "air entrainment") mitigates the
effects of freeze-thaw damage. This occurs because the greater air content provides extra void space within the PCC into
which the freezing water can expand. Thus, hydraulic and osmotic pressures on the cement paste are minimized, which
effectively prevents dilation and rupture. The total air content of the mortar (cement paste + fine aggregate) required to give
optimum freeze-thaw protection is about 9 percent, which results in an air content by volume of PCC of between 4 and 8
percent (Mindess and Young, 1981). In addition to the total volume, the distribution of air within the cement paste is also
important for freeze-thaw resistance. A properly air-entrained PCC contains a uniform dispersion of tiny bubbles throughout
the cement paste. As these bubbles get larger and farther apart, it becomes more difficult for the freezing water to migrate
through the cement paste into them. In general, the smaller the bubbles and more uniform their distribution, the better.
Actions such as excessive vibration or pumping can adversely affect both total air volume and air distribution. Today, most
PCC for exterior use (this includes pavements) is entrained with air to mitigate freeze-thaw effects.
Laboratory testing of PCC freeze-thaw resistance involves subjecting a specimen to a series of rapid freeze-thaw cycles, then
reporting a durability factor. First, specimens are created such that they are between 75 - 125 mm (3 - 5 inches) in width and
depth or diameter and between 280 - 400 mm (11 - 16 inches) long (see Figure 5.46). Specimens are then subjected to a
number of freeze-thaw cycles in the following manner (AASHTO, 2000a):
❍ The specimen dynamic modulus of elasticity reaches 60 percent of its initial value.
Typically, a DF < 40 indicates a PCC that may have poor freeze-thaw resistance, while a DF > 60 indicates a PCC that has
good freeze-thaw resistance (Mindess and Young, 1981). However, there are several limitations to this test. First, it uses 2 -
4 hour freeze-thaw cycles, which are much more rapid than will be experienced in the field. ASTM C 671 solves this issue
by using only one freeze-thaw cycle every 2 weeks. Second, even though these cycles are rapid when compared to field
conditions, the test can take between 600 and 1200 hours to complete (if the full 300 cycles are tested).
● AASHTO T 161 and ASTM C 666: Resistance of Concrete to Rapid Freezing and Thawing
● AASHTO T 121: Mass Per Cubic Meter (Cubic Foot), Yield, and Air Content (Gravimetric) of Concrete
Although it is actually the air content within the mortar (cement paste + fine
aggregate) that is of concern, cement paste air content is usually what is
measured. This air content can be measured in several ways, the most
common of which is the pressure method. Using the pressure method, a
sample of fresh PCC is placed in a pressure vessel (see Figure 5.47). The
remaining volume of the vessel is filled with water and then the vessel is
pressurized. The water level is read once, then the vessel is depressurized and
the water level is read again. Finally, using Boyle's law (The principle that at a
constant temperature the volume of a confined ideal gas varies inversely with
its pressure) the difference in water levels (which corresponds to a volume) is
converted into a volume of air.
These chloride ion concentrations are compared to the average chloride ion concentration of the control slab to determine the
amount and extent of chloride ion penetration. Critical chloride ion concentrations for reinforcing steel corrosion are on the
order of 0.6 - 1.2 kg Cl-/m3 (1.0 - 2.0 lb Cl-/yd3) of PCC.
Although sulfate attack is a PCC concern, it is generally not an issue in PCC pavement.
● AASHTO T 277 and ASTM C 1202: Electrical Indication of Concrete's Ability to Resist Chloride Ion Penetration
● AASHTO T 303 and ASTM C 227: Accelerated Detection of Potentially Deleterious Expansion of Mortar Bars
Due to Alkali-Silica Reaction
The FHWA and the Transtec Group, Inc. have produced a software package, termed HIgh PERformance PAVing
(HIPERPAV), that is capable of assessing the influence of mix design, structural design, construction methods and
environmental conditions on the early-age behavior of rigid pavements. HIPERPAV was originally produced for an FHWA
study of fast-track rigid pavements. The goal of this project was to develop high early strength rigid pavements that could be
rapidly opened to traffic upon construction completion. What this project discovered was that rapid-setting high early
strength PCC created a new set of concerns including: uncontrolled slab cracking, spalling and excessive plastic shrinkage.
HIPERPAV addresses these issues and others by modeling early-age PCC pavement performance (see Figure 5.50).
Figure 5.48: PCC Early Age Crack in Palmdale, CA Figure 5.49: Close-Up of Early Age Crack
Figure 5.50: One Output of HIPERPAV Showing Early Age Tensile Strength vs. Time
(screen shot courtesy of Transtec Group, Inc.)
9.5 Summary
All pavements can be described by their fundamental characteristics and performance. Thus, PCC tests are an integral part of
mix design because they can describe PCC characteristics and provide the means to relate mix design to intended
performance. Typically, PCC performance tests concentrate on basic physical properties such as strength and durability.
Early age behavior modeling can also be beneficial in predicting early strength gain, excessive plastic shrinkage, cracking and
spalling. PCC performance modeling provides the crucial link between laboratory mix proportioning and field performance.
1 Introduction
The goal of structural design is to determine the number, material composition and thickness of the
different layers within a pavement structure required to accommodate a given loading regime. This
includes the surface course as well as any underlying base or subbase layers. This module is focused on
the structural design of new pavement. Structural design for rehabilitation is covered in Module 10,
Section 3 (flexible pavements) and Section 5 (rigid pavements).
Design Life
Pavements are typically designed for a specified "design life". Design life (or "design period") is the
time from original construction to a terminal condition for a pavement structure. Structural design is
carried out so that the pavement structure is sufficient to withstand the traffic loading encountered over
the pavement's design life. It is recognized that intermittent maintenance and rehabilitation efforts may
be needed to preserve a pavement's surface quality and ensure that the structure lasts through the design
life.
For flexible pavements, structural design is mainly concerned with determining appropriate layer
thickness and composition. Calculations are chiefly concerned with traffic loading stresses; other
environmentally related stresses (such as temperature) are accounted for in mix design asphalt binder
selection. The two principal methods of flexible pavement structural design in use today, empirical and
mechanistic-empirical, are covered.
For rigid pavements, structural design is mainly concerned with determining the appropriate slab
thickness based on traffic loads and underlying material properties, and joint design. This is done by
considering a variety of stresses which affect rigid pavement performance: curling (temperature
stresses), warping (moisture stresses), wheel load and shrinkage/expansion. The two principal methods
of rigid pavement structural design in use today, empirical and mechanistic-empirical, are covered.
These tables provide an overview of typical flexible pavement layer thicknesses used by WSDOT for
design.
Specific WSDOT structural design policy is contained in the WSDOT Pavement Guide,
Volume 1. In general, WSDOT uses the following structural design procedures:
● Rehabilitation.
Overall, this section is only meant to provide a brief overview of the different structural design
techniques as well as their assumptions, inputs and outputs. Detailed analysis of the design methods
presented here can be found in:
● Empirical Method: The 1993 AASHTO Guide for Design of Pavement Structures.
● Mechanistic Method: The 2002 AASHTO Guide for Design of Pavement Structures, the
WSDOT Pavement Guide (1998), and other state design procedures.
There are a variety of ways to calculate or at least account for these stresses in design. The empirical
approach uses the AASHO Road Test results to correlate measurable parameters (such as subgrade
resilient modulus) and derived indices (such as the structural number and pavement serviceability index)
to pavement performance. The mechanistic-empirical approach relates calculated pavement stresses to
empirically derived failure conditions.
2.1 Stress
The stresses that occur in a flexible pavement under load are quite complex. Although rigid pavement
stresses have been routinely calculated since the 1920s, routine calculation of flexible pavement stresses
is a more recent development. First, two-dimensional layered elastic programs offered desktop
computers the ability to calculate these stresses. More recently, three-dimensional finite element
programs have allowed more exact and detailed calculations.
Figure 6.1: Typical Two-Layer Flexible Pavement Stresses as Calculated by a Two-Dimensional Linear Elastic
Model. Click the yellow boxes to view different stresses. Note that "E" refers to a layer's stiffness.
Figure 6.2: Typical Two-Layer Flexible Pavement Stresses as Calculated by EverFlex (Wu, 2001), a Three-
Dimensional Finite Elements Program. Click the yellow boxes to view different stresses. Note that "E" refers
to a layer's stiffness.
2.2 Deflection
HMA pavements are often described as "flexible" because they deflect under load. Figure 6.3 shows
schematically how pavements deflect under load. FWDs can be used to accurately determine deflection
characteristics of in-service pavements.
Figure 6.3: Schematic Showing Deflections for Different Pavement Thicknesses. The same HMA material
characteristics are assumed for each graph - only the thickness varies.
Many pavement design procedures use an empirical approach. This means that the relationship between
design inputs (e.g., loads, materials, layer configurations and environment) and pavement failure were
arrived at through experience, experimentation or a combination of both. Empirical design methods can
range from extremely simple to quite complex. The simplest approaches specify pavement structural designs
based on what has worked in the past. For example, local governments often specify city streets to be
designed using a given cross section (e.g., 100 mm (4 inches) of HMA over 150 mm (6 inches) of crushed
stone) because they have found that this cross section has produced adequate pavements in the past. More
complex approaches are usually based on empirical equations derived from experimentation. Some of this
experimentation can be quite elaborate. For example, the empirical equations used in the 1993 AASHTO
Guide are largely a result of the original AASHO Road Test.
● The empirical equation – using the 1993 AASHTO Guide flexible pavement equation as an
example
● An empirical computer program - using the 1993 AASHTO Guide equation for flexible pavements
∆PSI = difference between the initial design serviceability index, po, and the design
terminal serviceability index, pt
This equation is not the only empirical equation available but it does give a good sense of what an empirical
equation looks like, what factors it considers and how empirical observations are incorporated into an
empirical equation. The rest of this section will discuss the specific assumptions, inputs and outputs
associated with the 1993 AASHTO Guide flexible pavement empirical design equation. The following
subsections discuss:
● Assumptions
● Inputs
● Outputs
3.1.1 Assumptions
From the AASHO Road Test, equations were developed which related loss in serviceability, traffic, and
pavement thickness. Because they were developed for the specific conditions of the AASHO Road Test,
these equations have some significant limitations:
● The equations were developed based on the specific pavement materials and roadbed soil present
at the AASHO Road Test.
● The equations were developed based on the environment at the AASHO Road Test only.
● The equations are based on an accelerated two-year testing period rather than a longer, more
typical 20+ year pavement life. Therefore, environmental factors were difficult if not impossible to
extrapolate out to a longer period.
● The loads used to develop the equations were operating vehicles with identical axle loads and
configurations, as opposed to mixed traffic.
In order to apply the equations developed as a result of the AASHO Road Test, some basic assumptions are
needed:
● The characterization of subgrade support may be extended to other subgrade soils by an abstract
soil support scale.
● Material characterizations may be applied to other surfaces, bases, and subbases by assigning
appropriate layer coefficients.
● The accelerated testing done at the AASHO Road Test (2-year period) can be extended to a longer
design period.
When using the 1993 AASHTO Guide empirical equation or any other empirical equation, it is extremely
important to know the equation's limitations and basic assumptions. Otherwise, it is quite easy to use an
equation with conditions and materials for which it was never intended. This can lead to invalid results at
the least and incorrect results at the worst.
3.1.2 Inputs
The 1993 AASHTO Guide equation requires a number of inputs related to loads, pavement structure and
subgrade support. These inputs are:
● The predicted loading. The predicted loading is simply the predicted number of 80 kN (18,000
lb.) ESALs that the pavement will experience over its design lifetime.
● Reliability. The reliability of the pavement design-performance process is the probability that a
pavement section designed using the process will perform satisfactorily over the traffic and
environmental conditions for the design period (AASHTO, 1993). In other words, there must be
some assurance that a pavement will perform as intended given variability in such things as
construction, environment and materials. The ZR and So variables account for reliability.
● Pavement structure. The pavement structure is characterized by the Structural Number (SN). The
Structural Number is an abstract number expressing the structural strength of a pavement required
for given combinations of soil support (MR), total traffic expressed in ESALs, terminal
serviceability and environment. The Structural Number is converted to actual layer thicknesses (e.
g., 150 mm (6 inches) of HMA) using a layer coefficient (a) that represents the relative strength of
the construction materials in that layer. Additionally, all layers below the HMA layer are assigned
a drainage coefficient (m) that represents the relative loss of strength in a layer due to its drainage
characteristics and the total time it is exposed to near-saturation moisture conditions. Generally,
quick-draining layers that almost never become saturated can have coefficients as high as 1.4
while slow-draining layers that are often saturated can have drainage coefficients as low as 0.40.
Keep in mind that a drainage coefficient is basically a way of making a specific layer thicker. If a
fundamental drainage problem is suspected, thicker layers may only be of marginal benefit - a
better solution is to address the actual drainage problem by using very dense layers (to minimize
water infiltration) or designing a drainage system. Because of the peril associated with its use,
often times the drainage coefficient is neglected (i.e., set as m = 1.0).
● Serviceable life. The difference in present serviceability index (PSI) between construction and
end-of-life is the serviceability life. The equation compares this to default values of 4.2 for the
immediately-after-construction value and 1.5 for end-of-life (terminal serviceability). Typical
values used now are:
❍ End-of-life (called "terminal serviceability"): 1.5 - 3.0 depending upon road use (e.g.,
interstate highway, urban arterial, residential)
● Subgrade support. Subgrade support is characterized by the subgrade's resilient modulus (MR).
Intuitively, the amount of structural support offered by the subgrade should be a large factor in
determining the required pavement structure.
There cannot be a "fixed list" of design inputs for use in the AASHTO Guide which can be
applied to all WSDOT flexible pavement designs; however, some guidance is offered as a
starting point in the design process (typical values and associated ranges; required decisions).
Further, knowledge about these inputs will improve and undoubtedly change over time. Listed
below are the WSDOT suggested design inputs for the 1993 AASHTO Guide flexible
pavement empirical design equation:
● Future ESALs. Take the initial traffic and multiply it by a factor that is dependent
upon growth rate using the following equation to determine this factor:
● Design serviceability loss. Two decisions are required, selection of an initial PSI
(po) and terminal PSI (pt). A terminal PSI level of 3.0 is based, in part, on the
original pavement serviceability performance data reported by Carey and Irick
(1960). They found that about one-half of the panel of raters found a PSR of 3.0
acceptable and a PSR of 2.5 unacceptable. Thus, the following is suggested:
Western Washington
Eastern Washington
● Layer coefficients (a). For HMA Class A and B mixes as well as Superpave mixes,
typical values are a = 0.44 or less. For crushed surfacing base course, typical
values are a = 0.14 or less. For other materials, test results (such as MR, R-value or
CBR) should be used and correlated with a layer coefficient using the AASHTO
Guide.
1993 AASHTO Guide). Typically, WSDOT uses m = 1.0 and addresses drainage
issues separately.
3.1.3 Outputs
The 1993 AASHTO Guide equation can be solved for any one of the variables as long as all the others are
supplied. Typically, the output is either total ESALs or the required Structural Number (or the associated
pavement layer depths). To be most accurate, the flexible pavement equation described in this chapter
should be solved simultaneously with the flexible pavement ESAL equation. This solution method is an
iterative process that solves for ESALs in both equations by varying the Structural Number. It is iterative
because the Structural Number (SN) has two key influences:
1. The Structural Number determines the total number of ESALs that a particular pavement can
support. This is evident in the flexible pavement design equation presented in this section.
2. The Structural Number also determines what the 80 kN (18,000 lb.) ESAL is for a given load.
Therefore, the Structural Number is required to determine the number of ESALs to design for before the
pavement is ever designed. The iterative design process usually proceeds as follows:
1. Determine and gather flexible pavement design inputs (ZR, So, ∆PSI and MR).
2. Determine and gather flexible pavement ESAL equation inputs (Lx, L2x, G).
4. Determine the equivalency factor for each load type by solving the ESAL equation using the
assumed SN for each load type.
5. Estimate the traffic count for each load type for the entire design life of the pavement and multiply
it by the calculated ESAL to obtain the total number of ESALs expected over the design life of the
pavement.
6. Insert the assumed SN into the design equation and calculate the total number of ESALs that the
pavement will support over its design life.
7. Compare the ESAL values in #5 and #6. If they are reasonably close (say within 5 percent) use the
assumed SN. If they are not reasonably close, assume a different SN, go to step #4 and repeat the
process.
In practice, the flexible pavement design equation is usually solved independently of the ESAL equation by
using an ESAL value that is assumed independent of structural number. Although this assumption is not
true, pavement structure depths calculated using it are reasonably accurate. This design process usually
proceeds as follows:
1. Assume a structural number (SN) for ESAL calculations. Although often not overtly stated, a
structural number must be assumed in order to calculate ESALs.
2. Determine the load equivalency factor (LEF) for each load type by solving the ESAL equation
using the assumed SN for each load type. Typically, a standard set of load types is used (e.g.,
single unit trucks, tractor-trailer trucks and buses).
3. Estimate the traffic count for each load type for the entire design life of the pavement and multiply
it by the calculated LEF to obtain the total number of ESALs expected over the design life of the
pavement.
4. Determine and gather flexible pavement design inputs (ZR, So, ∆PSI and MR).
6. Check to see that the computed SN value is reasonably close to that assumed for ESAL
calculations. This step of often neglected.
4 Flexible - Mechanistic-Empirical
Method
Mechanics is the science of motion and the action of forces on Major Topics on this Page
bodies. Thus, a mechanistic approach seeks to explain
phenomena only by reference to physical causes. In pavement 4.1 Mechanistic Model
design, the phenomena are the stresses, strains and deflections
within a pavement structure, and the physical causes are the loads 4.2 Failure Criteria
and material properties of the pavement structure. The 4.3 A Mechanistic Computer Program
relationship between these phenomena and their physical causes
is typically described using a mathematical model. Various mathematical models can be (and are) used; the most
common is a layered elastic model.
Along with this mechanistic approach, empirical elements are used when defining what value of the calculated
stresses, strains and deflections result in pavement failure. The relationship between physical phenomena and
pavement failure is described by empirically derived equations that compute the number of loading cycles to
failure.
The basic advantages of a mechanistic-empirical pavement design method over a purely empirical one are:
● It can be used for both existing pavement rehabilitation and new pavement construction
The benefit of a mechanistic-empirical approach is its ability to accurately characterize in situ material (including
subgrade and existing pavement structures). This is typically done by using a portable device (like a FWD) to
make actual field deflection measurements on a pavement structure to be overlaid. These measurements can then
be input into equations to determine existing pavement structural support (often called "backcalculation") and the
approximate remaining pavement life. This allows for a more realistic design for the given conditions.
This section describes the basics behind flexible pavement mechanistic-empirical design to include:
● The mechanistic model. The layered elastic model and the finite element models are used as examples.
● Empirical failure definitions and equations. Equations from Finn et al. (1977), the AASHO Road Test
and the Roads and Transportation Association of Canada (RTAC) are used as working examples.
● A mechanistic computer program. The Everseries programs from the Washington State DOT are used as
examples.
4.1.1.1 Assumptions
The layered elastic approach works with relatively simple mathematical models and thus, requires some basic
assumptions. These assumptions are:
4.1.1.2 Inputs
A layered elastic model requires a minimum number of inputs to adequately characterize a pavement structure and
its response to loading. These inputs are:
❍ Modulus of elasticity
❍ Poisson's ratio
● Loading conditions
❍ Geometry. Usually specified as being a circle of a given radius (r or a), or the radius computed
knowing the contact pressure of the load (p) and the magnitude of the load (P). Although most
actual loads more closely represent an ellipse, the effect of the differences in geometry become
negligible at a very shallow depth in the pavement.
❍ Repetitions. Multiple loads on a pavement surface can be accommodated by summing the effects
of individual loads. This can be done because we are assuming that the materials are not being
stressed beyond their elastic ranges.
Figure 6.4 shows how these inputs relate to a layered elastic model of a pavement system.
4.1.1.3 Output
The outputs of a layered elastic model are the stresses, strains, and deflections in the pavement:
● Stress. The intensity of internally distributed forces experienced within the pavement structure at various
points. Stress has units of force per unit area (N/m2, Pa or psi).
● Strain. The unit displacement due to stress, usually expressed as a ratio of the change in dimension to the
original dimension (mm/mm or in/in). Since the strains in pavements are very small, they are normally
expressed in terms of microstrain (10-6).
● Deflection. The linear change in a dimension. Deflection is expressed in units of length (mm or µm or
inches or mils).
The use of a layered elastic analysis computer program will allow one to calculate the theoretical stresses, strains,
and deflections anywhere in a pavement structure. However, there are a few critical locations that are often used in
pavement analysis (see Table 6.1 and Figure 6.5).
Bottom of HMA layer Horizontal Tensile Strain Used to predict fatigue failure in the HMA
Top of Subgrade Vertical Compressive Strain Used to predict rutting failure in the subgrade
In the FEM analysis of a flexible pavement, the region of interest (the pavement and subgrade) is discretized into a
number of elements with the wheel loads are at the top of the region of interest (see Figure 6.6). The finite
elements extend horizontally and vertically from the wheel to include all areas of interest within the influence of the
wheel.
The drawing shows the discrete elements, wheel loads (tire patch loads), a modeled crack and a slip interface (where on layer
can slip - move independently - from another).
4.1.2.1 Assumptions
The FEM approach works with a more complex mathematical model than the layered elastic approach so it makes
fewer assumptions. Generally, FEM must assume some constraining values at the boundaries of the region of
interest. For instance, the computer program developed by Hongyu Wu and George Turkiyyah at the University of
Washington (Wu, 2001), called EverFlex, uses a 6-noded foundation element to model the Winkler Foundation.
This program also uses free boundaries on the four sides of the flexible pavement model. Additionally, the choice
of element geometry (size and shape) as well as interpolation functions will influence overall model performance.
4.1.2.2 Inputs
The typical finite elements method approach involves the following seven steps (Huebner et al., 2001):
1. Discretize the Continuum. The region of interest is divided into small discrete shapes.
2. Select Interpolation Functions. Nodes are assigned to each element and then a function is chosen to
interpolate the variation of the variable over the discrete element.
3. Find the Element Properties. Using the established finite element model (the elements and their
interpolation functions) to determine matrix equations that express the properties of the individual
elements.
4. Assemble the Element Properties to Obtain the System Equations. Combine the matrix equations
expressing the behavior of the elements and form the matrix equations expressing the behavior of the
entire system.
5. Impose the Boundary Conditions. Impose values for certain variables at key boundary positions (e.g., the
bottom and sides of the chosen region of analysis).
6. Solve the System Equations. The above process results in a set of simultaneous equations that can then
be solved.
7. Make Additional Computations If Desired. The unknowns are displacement components. From these
displacements element strains and stresses can be calculated.
4.1.2.3 Outputs
The outputs of a FEM analysis are the same as for a layered elastic model:
● Stress. The intensity of internally distributed forces experienced within the pavement structure at various
points. Stress has units of force per unit area (N/m2, Pa or psi).
● Strain. The unit displacement due to stress, usually expressed as a ratio of the change in dimension to the
original dimension (mm/mm or in/in). Since the strains in pavements are very small, they are normally
expressed in terms of microstrain (10-6).
● Deflection. The linear change in a dimension. Deflection is expressed in units of length (mm or µm or
inches or mils).
In addition, the finite elements method allows for extremely powerful graphical displays of these values (see
Figures 6.7 through 6.10).
Figure 6.7: 3-D Strain Figure 6.8: Surface Strain Figure 6.9: Section View Figure 6.10: Sample Load
Diagram Diagram Strain Diagram Profiles
The above equation defines failure as fatigue cracking over 10 percent of the wheelpath area. Figure 6.11 shows
the relationship between tensile strain in the asphalt concrete and the number of cycles to failure for two levels of
asphalt concrete elastic modulus. This relationship assumes bottom-up cracking rather than top-down cracking.
Figure 6.11: Limiting Horizontal Strain Criterion for Asphalt Concrete Fatigue Cracking
failure criteria are intended for rutting that can be attributed mostly to a weak pavement structure. This is typically
expressed in terms of the vertical compressive strain (εv) at the top of the subgrade layer:
The above equation defines failure as 12.5 mm (0.5-inch) depressions in the wheelpaths of the pavement. Figure
6.12 illustrates how the vertical compressive strain relates to the number of cycles to failure.
AASHO Road Test and Roads and Transportation Association of Canada (RTAC) criteria are shown here. Both
these criteria were developed based on spring seasonal deflections.
The AASHO Road Test results were used to develop the following relations (Highway Research Board, 1962b):
where: W2.5 = number of applications of axle load L sustained by the pavement to a terminal
1
serviceability index of 2.5
dsn = Benkelman Beam springtime measured pavement surface deflection (0.001 in.) measured at
the AASHO Road Test (Spring 1959) after "disappearance of frost."
This criterion was based on data from Loops 2 through 6 and single axle loads of 6, 12, 18, 22.4, and 30 kips (1 kip
= 1,000 lbs.). The following equation is obtained if L = 18,000 lbs. (a standard ESAL):
1
The RTAC criterion can be calculated as follows (after RTAC (1977) and Haas et al. (1994)):
where: BB = maximum rebound deflection (in.) (defined as the mean rebound deflection plus two standard
deviations) at a standard temperature of 21°C (70°F)
To install the Everseries programs on your computer, click the install icon below:
NOTES:
● These programs must be installed on to your computer before you can use them. During the installation you
will be prompted to specify a location to which they can be installed. Once installed, the programs and their
supporting files take about 3.34 MB of disk space.
The Everseries Pavement Analysis Programs can also be downloaded from the Washington State DOT Materials
Lab at: http://www.wsdot.wa.gov/biz/mats. Volume 3 of the WSDOT Pavement Guide (WSDOT, 1998) is
available at the same site for download and contains detailed instructions on how to run the Everseries programs.
5.2 Shrinkage/Expansion
● Curling stress. Differences in temperature between the top
and bottom surfaces of a PCC slab will cause the slab to curl. Since slab weight and contact with the base
restrict its movement, stresses are created.
● Load stress. Loads on a PCC slab will create both compressive and tensile stresses within the slab and any
adjacent one (as long as load transfer efficiency is > 0).
● Shrinkage/Expansion. In addition to curling, environmental temperatures will cause PCC slabs to expand
(when hot) and contract (when cool), which causes joint movement.
These three principal responses typically determine PCC slab geometry (typically described by slab thickness and joint
design). As slabs get longer, wider and thinner, these responses, or a combination of them, will eventually exceed the
slab's capacity and cause failure in the form of slab cracking, joint widening or blowup. Note that additional issues,
notably load transfer stresses and deflections, must also be accounted for in design.
There are a variety of ways to calculate or at least account for these responses in design. The empirical approach uses
the AASHO Road Test results to correlate measurable parameters (such as slab depth and PCC modulus of rupture) and
derived indices (such as the load transfer coefficient and pavement serviceability index) to pavement performance. The
mechanistic-empirical approach relates calculated pavement stresses to empirically derived failure conditions.
5.1 Stress
The stresses of primary concern are associated with slab bending either due to temperature gradients, loading or a
combination thereof.
5.1.1 Curling
Since PCC is much stronger in compression than tension, tensile stresses tend to control PCC pavement design.
Therefore, slab curling calculations seek to find the points of maximum tensile stress as the slab curls due to
temperature gradients within (see Figure 6.13). In 1935, measurements reported by Teller and Southerland of the
Bureau of Public Roads showed that the maximum temperature differential (hence, maximum curling and maximum
tensile stresses) is much larger during the day than during the night. Therefore, the daytime curling stresses are usually
most limiting.
To evaluate the tensile warping stresses which develop in the slab, the temperatures at the top and bottom of the slab
must be estimated. The first formulas used to estimate warping stresses were developed by Bradbury (1938):
C = coefficient which is a function of slab length and the radius of relative stiffness (shown
in Figure 6.13)
Figure 6.14: Curve Showing Variation in the Differential Temperature Stress Coefficient C for Different
Values of the Ratio B/l (redrawn from Bradbury, 1938)
Bradbury also developed an approximate formula for slab corner warping stresses
The radius of relative stiffness (the relative stiffness of the slab relative to that of the foundation) is required for the
h = slab thickness
5.1.2 Load
The original equations developed by Westergaard (1926) for three critical load locations will be presented. The critical
load locations are (after Bradbury, 1938 and Westergaard, 1926):
1. Interior loading. Occurs when a load is applied on the interior of a slab surface which is "remote" from all
edges.
2. Edge loading. Occurs when a load is applied on a slab edge "remote" from a slab corner.
3. Corner loading. Occurs when the center of a load is located on the bisector of the corner angle.
where: σi, σe, σc = maximum stress (psi) for in interior, edge and corner loadings, respectively
Note that all three equations involved the depth of slab (h) squared. This suggests that slab thickness is very critical in
reducing load stresses to acceptable levels.
5.2 Shrinkage/Expansion
Although slab shrinkage and expansion causes internal stress, especially as the PCC sets and hardens, the long term
concern centers on the joint movement that this shrinkage/expansion can cause. The following formula can be used to
estimate joint movement in PCC slabs (FHWA, 1989):
∆T = the maximum temperature range (for some cases it is the temperature of the PCC at
the time of placement minus the average daily minimum temperature in January) (°F)
(Note: δ should be omitted for rehabilitation projects as shrinkage (assuming no new slab PCC) is not
a factor.)
Many pavement design procedures use an empirical approach. This means that the relationship between design inputs (e.g., loads,
materials, layer configurations and environment) and pavement failure were arrived at through experience, experimentation or a
combination of both. Empirical design methods can range from extremely simple to quite complex. The simplest approaches specify
pavement structural designs based on what has worked in the past. For example, local governments often specify city streets to be
designed using a given cross section (e.g., 200 mm (8 inches) of PCC over 150 mm (6 inches) of crushed stone) because they have
found that this cross section has produced adequate pavements in the past. More complex approaches usually develop empirical
equations based on the results of experimentation. Some of this experimentation can be quite elaborate. For example, the empirical
equations used in the 1993 AASHTO Guide are largely a result of the original AASHO Road Test.
● The empirical equation – using the 1993 AASHTO Guide rigid pavement equation as an example
● An empirical computer program - using the 1993 AASHTO Guide equation for rigid pavements
∆PSI = difference between the initial design serviceability index, po, and the
design terminal serviceability index, pt
Cd = drainage coefficient
J = load transfer coefficient (value depends upon the load transfer efficiency)
This equation is not the only empirical equation available but it does give a good sense of what an empirical equation looks like,
what factors it considers and how empirical observations are incorporated into an equation. The rest of this section will discuss the
specific assumptions, inputs and outputs associated with the 1993 AASHTO Guide flexible pavement empirical design equation.
The following subsections discuss:
● assumptions
● inputs
● outputs
6.1.1 Assumptions
From the AASHO Road Test, equations were developed which related loss in serviceability, traffic, and pavement thickness. These
equations were developed for the specific conditions of the AASHO Road Test and therefore involved some significant limitations:
● The equations were developed based on the specific pavement materials and roadbed soil present at the AASHO Road Test.
● The equations were developed based on the environment at the AASHO Road Test only.
● The equations are based on an accelerated two-year testing period rather than a longer, more typical 20+ year pavement
life. Therefore, environmental factors were difficult if not impossible to extrapolate out to a longer period.
● The loads used to develop the equations were operating vehicles with identical axle loads and configurations, as opposed
to mixed traffic.
● For JPCP and JRCP, all transverse joints were the same spacing. JPCP was 4.6 m (15 ft) and JRCP was 12.2 m (40 ft). All
transverse joints used dowel bars.
● All PCC was of the same mix design and used the same aggregate and portland cement.
In order to apply the equations developed as a result of the AASHO Road Test, some basic assumptions are needed:
● The characterization of subgrade support may be extended to other subgrade soils by an abstract soil support scale.
● Material characterizations may be applied to other surfaces, bases, and subbases by assigning appropriate values.
● The accelerated testing done at the AASHO Road Test (2-year period) can be extended to a longer design period.
When using the 1993 AASHTO Guide empirical equation or any other empirical equation, it is extremely important to know the
equation's limitations and basic assumptions. Otherwise, it is quite easy to use an equation with conditions and materials for which it
was never intended. This can lead to invalid results at the least and incorrect results at the worst.
6.1.2 Inputs
The 1993 AASHTO Guide equation requires a number of inputs related to loads, pavement structure and subgrade support. These
inputs are:
● The predicted loading. The predicted loading is simply the predicted number of 80 kN (18,000 lb.) ESALs that the
pavement will experience over its design lifetime.
● Reliability. The reliability of the pavement design-performance process is the probability that a pavement section designed
using the process will perform satisfactorily over the traffic and environmental conditions for the design period (AASHTO,
1993). In other words, there must be some assurance that a pavement will perform as intended given variability in such
things as construction, environment and materials. The ZR and So variables account for reliability.
● PCC elastic modulus. If no value is known, the PCC elastic modulus (Ec) can be estimated from relationships such as the
following:
If no compressive strength data are available (or cannot be assumed), assume Ec = 27,500 MPa
(4,000,000 psi), which corresponds to a compressive strength of 34.5 MPa (5000 psi).
● PCC modulus of rupture (flexural strength). The modulus of rupture (S'c) is typically obtained from a flexural strength
test.
● Slab depth. The pavement structure is best characterized by slab depth (D). The number of ESALs a rigid pavement can
carry over its lifetime is very sensitive to slab depth. As a general rule, beyond about 200 mm (8 inches) the load carrying
capacity of a rigid pavement doubles for each additional 25 mm (1 inch) of slab thickness.
● Drainage coefficient. Rigid pavement is assigned a drainage coefficient (Cd) that represents the relative loss of strength
due to its drainage characteristics and the total time it is exposed to near-saturation moisture conditions. Generally, quick-
draining layers that almost never become saturated can have coefficients as high as 1.2 while slow-draining layers that are
often saturated can have drainage coefficients as low as 0.80. If subsurface drainage is expected to be a problem, positive
drainage measures should be taken. In general, the use of drainage coefficients to overcome poor drainage conditions is
not recommended (i.e. more slab thickness does not necessarily solve water-related problems). Because of the peril
associated with its use, often times the drainage coefficient is neglected (i.e., set as Cd = 1.0).
● Serviceable life. The difference in present serviceability index (PSI) between construction and end-of-life is the
serviceability life. The equation compares this to default values of 4.2 for the immediately-after-construction value and 1.5
for end-of-life (terminal serviceability). Typical values used now are:
❍ End-of-life (called "terminal serviceability" and designated "pt"): 1.5 - 3.0 depending upon road use (e.g.,
interstate highway, urban arterial, residential)
● Load transfer coefficient (J Factor). This accounts for load transfer efficiency. Essentially, the lower the J Factor the
better the load transfer. The J Factor for the AASHO Road Test was estimated to be 3.2. Typical J factor values are as
shown below.
Condition J Factor
● Modulus of subgrade reaction. The modulus of subgrade reaction (k) is used to estimate the "support" of the PCC slab by
the layers below. Usually, an "effective" k (keff) is calculated which reflects base, subbase and subgrade contributions as
well as the loss of support that occurs over time due to erosion and stripping of the base, subbase and subgrade. Typically,
large changes in keff have only a modest impact on PCC slab thickness.
There cannot be a "fixed list" of design inputs for use in the AASHTO Guide which can be applied to all WSDOT
rigid pavement designs; however, some guidance is offered as a starting point in the design process (typical values
and associated ranges; required decisions). Further, knowledge about these inputs will improve and undoubtedly
change over time. Listed below are the WSDOT suggested design inputs for the 1993 AASHTO Guide rigid
pavement empirical design equation:
● Future ESALs. Take the initial traffic and multiply it by a factor that is dependent upon growth rate using
the following equation to determine this factor:
● Overall standard deviation. Unless available project specific information suggests otherwise, use So =
0.40.
● Design serviceability loss. Two decisions are required, selection of an initial PSI (po) and terminal PSI
(pt). A terminal PSI level of 3.0 is based, in part, on the original pavement serviceability performance
data reported by Carey and Irick (1960). They found that about one-half of the panel of raters found a
PSR of 3.0 acceptable and a PSR of 2.5 unacceptable. Thus, the following are suggested:
● PCC elastic modulus. For design purposes (unless project specific information suggests otherwise), use
Ec = 27,500 MPa (4,000,000 psi).
● PCC modulus of rupture. In the absence of local data, this value typically ranges between 4.5 - 5.5 MPa
(650 - 800 psi).
● Load transfer coefficient ("J Factor"). Unless performance information indicates otherwise, the
following J factors are suggested:
Condition J Factor
Undoweled PCC on an HMA base with a widened outside lane and drainable
3.4
shoulder material
Undoweled PCC on crushed aggregate surfacing with widened outside lane and
3.4
drainable shoulder material.
Doweled PCC on ATPB with widened outside lane and drainable HMA surfaced
2.7
shoulders
Doweled PCC on ATPB (without widened outside lane) and tied PCC shoulders 2.7
Note: The J Factors in the Seattle area range from 2.9 to 3.8. J Factors in the
Snoqualmie Pass area ranged from 3.6 to 3.9 and in the Vancouver area about 3.8.
The joint spacing at all five locations examined was 4.8 m (15 ft.) and the joints are
not skewed.
● Drainage coefficient. If no other project specific information is available, use a Cd = 1.0 for PCC
sections without asphalt treated permeable base (ATPB) and Cd = 1.2 for PCC sections with ATPB.
Other values are certainly possible.
● Effective modulus of subgrade reaction (keff). No specific guidance for local conditions. A typical value
to start with is 54 MPa/m (200 pci). The following table shows some typical keff based on different base
and subgrade conditions:
Loss
Base Material Subgrade MR keff
of Support1
1A factor used to correct the modulus of subgrade reaction (k) based on the potential erosion of base material.
Values range from 0 to 3 in increments of 1. High values indicate more erosion.
6.1.3 Outputs
The 1993 AASHTO Guide equation can be solved for any one of the variables as long as all the others are supplied. Typically, the
output is either total ESALs or the required slab depth (D). In design, the rigid pavement equation described in this chapter is
typically solved simultaneously with the rigid pavement ESAL equation. The solution is an iterative process that solves for ESALs
in both equations by varying the slab depth (D). The solution is iterative because the slab depth (D) has two key influences:
1. The slab depth (D) determines the total number of ESALs that a particular pavement can support. This is evident in the
rigid pavement design equation presented in this section.
2. The slab depth also determines what the equivalent 80 kN (18,000 lb.) single axle load is for a given load.
Therefore, the slab depth (D) is required to determine the number of ESALs to design for before the pavement is ever designed. The
iterative design process usually proceeds as follows:
1. Determine and gather rigid pavement design inputs (ZR, So, ∆PSI, pt, Ec, S'c, J, Cd and keff).
2. Determine and gather rigid pavement ESAL equation inputs (Lx, L2x, G)
4. Determine the equivalency factor for each load type by solving the ESAL equation using the assumed slab depth (D) for
each load type.
5. Estimate the traffic count for each load type for the entire design life of the pavement and multiply it by the calculated
ESAL to obtain the total number of ESALs expected over the design life of the pavement.
6. Insert the assumed slab depth (D) into the design equation and calculate the total number of ESALs that the pavement will
support over its design life.
7. Compare the ESAL values in #5 and #6. If they are reasonably close (say within 5 percent) use the assumed slab depth
(D). If they are not reasonably close, assume a different slab depth (D), go to step #4 and repeat the process.
7 Rigid - Mechanistic-Empirical
Method
Mechanics is the science of motion and the action of Major Topics on this Page
forces on bodies. Thus, a mechanistic approach seeks to
explain phenomena only by reference to physical 7.1 Mechanistic Model
causes. In pavement design the phenomena are the
stresses, strains and deflections within a pavement 7.2 Failure Criteria
structure, and the physical causes are the loads and 7.3 A Mechanistic Computer Program
material properties of the pavement structure. The
relationship between these phenomena and their physical causes is typically a mathematical model.
Various mathematical models can be (and are) used; the most common is a layered elastic model.
The empirical portion of a mechanistic-empirical approach comes about when defining what value of the
calculated stresses, strains and deflections result in pavement failure (the point at which the pavement is
no longer serviceable). This relationship between physical phenomena and pavement failure is described
by empirically derived equations that compute the number of loading cycles to failure.
The basic advantages of a mechanistic-empirical pavement design method over a purely empirical one
are:
● It can be used for both existing pavement rehabilitation and new pavement construction
This section describes the basics behind rigid pavement mechanistic-empirical design.
Much of the discussion in this section is identical to that of Section 4.1.2, Finite Elements Model (the
flexible pavement FEM). However, this section uses EverFE, a FEM developed at the University of
Washington for the Washington State DOT and U.S. Army Corps of Engineers by Davids, Turkiyyah and
Mahoney (1998). As in flexible pavements, the FEM analysis of a rigid pavement discritizes the region
of interest (the pavement and subgrade) into a number of elements with the loads at the top (see Figure
6.15).
Figure 6.15: EverFE Sample Deflection Display Showing Discretized Region of Interest
7.1.2.1 Assumptions
The FEM approach works with a more complex mathematical model than the layered elastic approach so
it makes fewer assumptions. Generally, FEM must assume some constraining values at the boundaries of
the region of interest. Additionally, the choice of element geometry (size and shape) as well as
interpolation functions will influence overall model performance.
7.1.2.2 Inputs
The typical finite elements method approach involves the following seven steps (Huebner et al., 2001):
1. Discretize the Continuum. The region of interest is divided into small discrete shapes.
2. Select Interpolation Functions. Nodes are assigned to each element and then a function is
chosen to interpolate the variation of the variable over the discrete element.
3. Find the Element Properties. Using the established finite element model (the elements and
their interpolation functions) to determine matrix equations that express the properties of the
individual elements.
4. Assemble the Element Properties to Obtain the System Equations. Combine the matrix
equations expressing the behavior of the elements and form the matrix equations expressing the
behavior of the entire system.
5. Impose the Boundary Conditions. Impose values for certain variables at key boundary
positions (e.g., the bottom and sides of the chosen region of analysis).
6. Solve the System Equations. The above process results in a set of simultaneous equations that
can then be solved.
7.1.2.3 Outputs
The outputs of a FEM analysis are the same as for a layered elastic model:
● Stress (see Figure 6.18). The intensity of internally distributed forces experienced within the
pavement structure at various points. Stress has units of force per unit area (N/m2, Pa or psi).
● Strain. The unit displacement due to stress, usually expressed as a ratio of the change in
dimension to the original dimension (mm/mm or in/in). Since the strains in pavements are very
small, they are normally expressed in terms of microstrain (10-6).
● Deflection (see Figure 6.19). The linear change in a dimension. Deflection is expressed in
In addition, the finite elements method allows for extremely powerful graphical displays of these values
(see Figures 6.12 through 6.14).
Screen Shot Thumbnails from EverFE (Davids, Turkiyyah and Mahoney, 1998). EverFE is a
project under development at the University of Washington. The project is supported by the
Washington State Department of Transportation. Additional support is provided by the US Army
Corps of Engineers. Click on each thumbnail to see a larger version of the picture.
● Joint spacing
● Joint orientation
● Joint size
Joint spacing is highly dependent on the local environment, materials and subgrade. First, expected
temperature changes will influence slab curling stresses. In general, the greater the temperature changes,
the shorter the joint spacing should be. Second, the materials within the PCC slab (the coarse aggregate
is of overriding concern) will influence the slab's thermal coefficient. The higher the thermal
coefficient, the more a slab will shrink and expand for a given temperature change. Generally, slabs
made with limestone coarse aggregate have lower thermal coefficients, while slabs made with quartz or
sandstone have higher thermal coefficients. Third, as the slab expands and contracts, the frictional
resistance offered by the base material will also influence slab stresses. In general, the more frictional
resistance, the higher the slab stresses.
Joint spacing is also related to slab thickness. In general, the thinner a slab is, the higher the curling
stresses and thus, the shorter the joint spacing should be. As a general rule-of-thumb, joint spacing
should be less than about 24 x slab thickness. Thus, a 230 mm slab (9 inches) should have joints spaced
no more than about 5.5 m (18 ft.) apart. Also, as a general guide, the ratio of longer side slab length to
the shorter side slab length should be kept less than about 1.25.
The FHWA (1990) recommends that the L/l ratio (slab length divided by radius of relative stiffness) not
exceed 5.0 when determining the maximum slab length. Table 6.3 shows some slab lengths resulting
from using L/l = 5.0 for a range of normal slab thicknesses.
Table 6.3: Slabs Lengths Resulting from Using an L/l Ratio = 5.0
k = 27 MPa/m (100 pci) k = 54 MPa/m (200 pci) k = 216 MPa/m (800 pci)
Slab Thickness
l L l L l L
The contraction joint spacing used by WSDOT is based on dowel bar use for load
transfer. A reasonable joint spacing when dowels are used is 3.7 m (12 ft.); however,
contraction joint spacings up to 4.5 m (15 ft.) can be used and are specified in the WSDOT
Standard Plans.
These contraction joint spacings are, in part, based on prior rigid pavement performance in
Washington State and elsewhere and slab stress calculations. For example:
● Contraction joint spacings of 3.7 - 4.5 m (12 to 15 ft.) result in lower slab
stresses due to thermal gradients.
● A contraction joint spacing of about 3.7 m (12 ft). conforms to the FHWA L/l
= 5.0 criterion for "thinner" slabs of about 228 mm (9 in.) on stiff subbases. A
spacing of about 4.5 m (15 ft.) conforms to the same criterion for "thicker"
slabs of about 330 mm (13 in.) on stiff subbases.
● In general, annual joint openings should be limited to no more than 0.6 - 0.9
mm (0.025 - 0.035 in.) to insure long term joint performance. Using the slab
shrinkage/expansion equation and PCC slabs on stabilized base for annual
temperature ranges estimated for eastern and western Washington, the resulting
joint movements are:
1. Estimate the total joint movement using the slab shrinkage/expansion equation.
❍ Hot pour liquid sealant / silicone sealant. Dependent upon the estimated joint
opening, the allowable sealant strain and a sealant shape factor. The shape factor is
used to determine the required depth of sealant. For example, if the required joint
width is 12.5 mm (0.5 inches), and the shape factor is 1:1, then the depth is 12. 5
mm (0.5 inches).
❍ Compression sealant. The uncompressed seal width (USW) should be selected based
upon the anticipated joint openings and the maximum and minimum recommended
compression of the seal (generally 0.5 and 0.2, respectively). The sawcut width is
determined based on the anticipated state of compression of the seal at the time of
compression, which is based largely on the expected temperature range and the
installation temperature.
transverse joints associated with medium to high traffic loading (FHWA, 1990).
The FHWA (1990) recommends the use of dowel bars. Further it recommends that they have a
minimum diameter of 1/8 the pavement thickness, but not less than 32 mm (1.25 inches). Typical
designs use 460 mm (18 inch) long dowel bars at 305 mm (12 inch) on center spacing, placed at slab
mid-depth.
WSDOT uses one standard dowel bar for all new construction, reconstruction
and dowel bar retrofits:
1. Slab length. This has a large effect on the maximum PCC tensile stresses developed within
the slab. As the slab length increases, the contact area with the base material increases, which
increases the total resisting frictional force, resulting in higher tensile stresses as the slab
contracts and/or loses moisture.
2. Steel working stress. This is usually taken to be 75% of the steel yield stress. The steel
working stress must be great enough to resist the frictional forces developed during slab
contraction.
3. Friction factor. This represents the frictional resistance between the bottom of the slab and
the top of the base material. It is like a coefficient of friction. Table 6.4 shows the 1993
AASHTO Guide recommended frictional factors.
Friction
Type of Material Beneath the Slab
Factor (F)
Sandstone 1.2
Taking the above three factors into account, the following equation is used to determine the amount of
reinforcing steel as a percentage of slab cross-sectional area:
F = friction factor
This JRCP design procedure is also used to design CRCP transverse reinforcing steel.
● Crack spacing. To minimize crack spalling, the maximum spacing between cracks should be
less than 2.5 m (8 ft.). To minimize the potential for punchouts, the minimum spacing
between cracks should be 1.07 m (3.5 ft.).
● Crack width. To minimize spalling and water penetration, the allowable crack width should
not exceed 1 mm (0.04 inches). Small crack widths are essential to CRCP performance.
● Steel stress. This is usually taken to be 75% of the steel yield stress to prevent any plastic
deformation, although studies have shown that many CRCP pavements have performed
adequately even though their steel stress was calculated to be above yield stress (Majidzadeh,
1978 as referenced in AASHTO, 1993).
One longitudinal steel design procedure is given by the 1993 AASHTO Guide:
1. Solve the following three limiting criteria equations for the percentage of steel required (yes,
they appear difficult, but the 1993 AASHTO Guide contains nomograph solutions). Note that
crack spacing (x) should be solved using input values of x = 2.5 m (8 ft.) to determine a
minimum amount of steel required to keep the maximum crack spacing less than 2.5 m (8 ft.),
and x = 1.07 m (3.5 ft.) to determine a maximum amount of steel required to keep the
minimum crack spacing greater than 1.07 m (3.5 ft.). Crack width and steel working stress
solutions will give a minimum amount of required steel.
2. The solutions to step 1 will provide the minimum (Pmin) and maximum (Pmax) required
percentage of reinforcing steel. If Pmax > Pmin then the design is feasible and can continue. If
not, the design inputs need to be modified and the above equations recalculated.
D = slab thickness
4. Determine the design number of reinforcing bars (Ndesign) such that it is a whole number
between Nmin and Nmax.
Transverse steel can then be designed using the JRCP procedure to define the amount of steel required
and the following equation to determine the reinforcing bar spacing:
D = slab thickness
1 Introduction
Construction quality is crucial to the long-term pavement performance. Construction factors such as
surface preparation, placement, joint construction and compaction/consolidation have an overwhelming
effect on pavement performance, which cannot be ignored or compensated for in mix or structural
design. Other construction considerations such as plant operations, mix transport and quality control
procedures can also directly influence pavement performance.
Pavement construction is somewhat of a combination of science and art. Although the mix design and
structural design are determined through carefully controlled experiments and equations, properly
constructing a pavement can be done in many different ways, each of which may be appropriate for a
specific combination of factors such as temperature, pavement thickness, material properties, and
subgrade to name a few. Essentially, there are so many variables involved in construction that it is
virtually impossible to reduce it down to a simple set of rules and equations; therefore, there is a
significant amount of “art” to it. However, there are equipment and methods common to almost all
pavement construction and there are accepted best practices. This Module outlines the basics of flexible
and rigid pavement construction in the following sections:
slipform paving
joints
http://training.ce.washington.edu/WSDOT/Modules/07_construction/07-1_body.htm4/2/2008 6:37:10 PM
7.2 Pavement Construction - Surface Preparation
2 Surface Preparation
Before a pavement is actually placed at the construction site the surface to be Major Topics on this Page
paved must be prepared. Adequate surface preparation is essential to long-
term pavement performance. Pavements constructed without adequate 2.1 Subgrade Preparation for New Pavements
surface preparation may not meet smoothness specifications, may not bond
to the existing pavement (in the case of overlays) or may fail because of 2.2 Existing Surface Preparation for Overlays
inadequate subgrade support. Surface preparation generally takes one of two 2.3 Summary
forms:
1. Preparing the subgrade and granular base course for new pavement. This can involve such activities as subgrade
stabilization (e.g., with lime, cement or emulsified asphalt), over-excavation of poor subgrade, applying a prime coat
or compacting the subgrade.
2. Preparing an existing pavement surface for overlay. This can involve such activities as removing a top layer
through milling, applying a leveling course, applying a tack coat, rubblizing or cracking and seating an underlying
rigid pavement, and replacing localized areas of extreme damage.
Specific actions for each method depend upon the pavement type and purpose, environmental conditions, subgrade conditions,
local experience and specifications.
● Subgrade elevation
1. Stabilization. The binding characteristics of these materials generally increase subgrade load-bearing capacity.
Typically, lime is used with highly plastic soils (plasticity index greater than 10), cement is used with less plastic
soils (plasticity index less than 10) and emulsified asphalt can be used with sandy soils. For flexible pavements, a
primecoat is not effective on silty clay or clay soils because the material cannot be absorbed into such a fine soil
(TRB, 2000).
2. Over-excavation. The general principle is to replace poor load-bearing in situ subgrade with better load-bearing
fill. Typically, 0.3 - 0.6 m (1 - 2 ft.) of poor soil may be excavated and replaced with better load-bearing fill such as
gravel borrow.
3. Add a base course and perhaps a subbase course over the subgrade. A base course offers additional load-bearing
capacity. New pavement structural designs often use some sort of granular base course unless subgrade structural
support is extremely good and expected loads are extremely low. Base courses are subjected to the same
compaction and elevation requirements as subgrade soils.
1. Fill the surface voids and protect the subbase from weather.
Generally, if a flexible pavement is to be less than 100 mm (4 inches) thick and placed over an unbound material, a primecoat
is recommended (Asphalt Institute, 2001).
1. Ensure the compacted subgrade is able to support construction traffic. If the subgrade ruts excessively under
construction traffic it should be repaired before being paved over. Left unrepaired, subgrade ruts may reflectively
cause premature pavement rutting.
2. Remove all debris, large rocks, vegetation and topsoil from the area to be paved. These items either do not compact
well or cause non-uniform compaction and mat thickness.
3. Treat the subgrade under the area to be paved with an approved herbicide. This will prevent or at least retard
future vegetation growth, which could affect subgrade support or lead directly to pavement failure.
In summary, subgrade preparation should result in a material (1) capable of supporting loads without excessive deformation
and (2) graded to specified elevations and slopes.
● Repair
● Tack coats
2.2.1 Repair
Generally, pavement overlays are used to restore surface course (both flexible and rigid) characteristics (such as smoothness,
friction and aesthetics) or add structural support to an existing pavement. However, even a structural overlay needs to be
placed on a structurally sound base. If an existing pavement is cracked or provides inadequate structural support these defects
will often reflect through even the best-constructed overlay and cause premature pavement failure in the form of cracks and
deformations. To maximize an overlay’s useful life, failed sections of the existing pavements should be patched or replaced
and existing pavement cracks should be filled.
At most, overlays are designed to add only some structural support; the remaining structural support must reside in the existing
pavement. Therefore, small areas of localized structural failure in the existing pavement should be repaired or replaced to
provide this structural support (see Figure 7.3). Often, existing pavement failure may be caused by inadequate subgrade
support or poor subgrade drainage. In these cases, the existing pavement over the failed area should be removed and the
subgrade should be prepared as it would be for a new pavement.
Existing pavement crack repair methods depend upon the type and severity of cracks. Badly cracked pavement sections,
especially those with pattern cracking (e.g., fatigue cracking) or severe slab cracks, must be patched or replaced because these
distresses are often symptoms of more extensive pavement or subgrade structural failure (TRB, 2000). Existing cracks other
than those symptomatic of structural failure should be cleaned out (blown out with pressurized air and/or swept) and filled
with a crack-sealing material when the cracks are clean and dry (TRB, 2000). Cracks less than about 10 mm (0.375 inches) in
width may be too narrow for crack-sealing material to enter. These narrow cracks can be widened with a mechanical router
before sealing. If the existing pavement has an excessive amount of fine cracks but is still structurally adequate, it may be
more economical to apply a general bituminous surface treatment (BST) or slurry seal instead of filling each individual crack.
1. Applying a leveling course (flexible pavements). The first lift applied to the existing pavement is used to fill in ruts
and make up elevation differences. The top of this lift, which is relatively smooth, is used as the base for the
wearing course.
2. Milling (flexible pavements). A top layer is milled off the existing pavement to provide a relatively smooth surface
on which to pave. Milling is also commonly used to remove a distressed surface layer from an existing pavement.
3. Diamond grinding (rigid pavements). A thin top layer can be milled off of an existing pavement to smooth out
relatively small surface distortions prior to flexible or rigid overlay.
Leveling courses (or prelevel) are initial lifts placed directly on to the existing pavement to fill low spots in the pavement (see
Figure 7.4). Typically, pavers use an automatic screed control, which keeps the screed tow point constant regardless of the
tractor unit’s vertical position. This allows the paver to drive over a rough, uneven pavement yet place a relatively smooth lift
with extra HMA making up for low spots in the existing pavement.
Leveling course lifts need to be as thick as the deepest low spot but not so thick that they are difficult to compact. Because it
is not the final wearing course, leveling course elevation and grade are sometimes not tightly specified or controlled.
However, contractors and inspectors alike should pay close attention to leveling course thickness because an excessively thick
leveling course can lead to large overruns in HMA and thus large overruns in project budget.
Although leveling courses can help produce a smoother pavement, they suffer from the previously discussed differential
compaction and therefore may not entirely solve the smoothness problem.
Milling (also called grinding or cold planing) can be used to smooth an existing flexible pavement prior to flexible or rigid
overlays. Rather than filling in low spots, as a leveling course does, milling removes the high points in an existing pavement
to produce a relatively smooth surface. For flexible pavements, milling can help eliminate differential compaction problems.
Milling machines are the primary method for removing old flexible pavement surface material prior to overlay (Roberts et al.,
1996). They can be fitted with automatic grade control to restore both longitudinal and transverse grade and can remove most
existing pavement distortions such as rutting, bumps, deteriorated surface material or stripping. The primary advantages of
milling are (Roberts et al., 1996):
1. Eliminates the need for complicated leveling courses and problems with quantity estimates for irregular leveling
course thicknesses used to fill existing pavement depressions.
3. Allows efficient removal of deteriorated flexible pavement material that is unsuitable for retention in the pavement
structure.
4. Provides a highly skid resistant surface suitable for temporary use by traffic until the final surface can be placed.
5. Allows curb and gutter lines to be maintained or reestablished before flexible overlays.
6. Provides an efficient removal technique for material near overhead structures in order to maintain clearances for
bridge structures, traffic signals and overhead utilities.
The basic components of a milling machine are a cutting drum to mill the existing pavement, a vacuum to collect the milled
particles and a conveyance system to transport the milled particles to a dump truck for hauling (see Figure 7.5, 7.6 and 7.7).
Table 7.1 shows ranges for some key milling machine parameters, Figures 7.8 and 7.9 show two milling machine examples,
Figures 7.10 and 7.11 show milled pavements and Video 7.1 shows the basic milling process.
Cut Width 75 mm (3 inches) to Drums come in specific widths. Varying widths can be made with
4.5 m (14 feet) multiple passes.
Cut Depth up to 250 mm It is easier to make several shallow passes than one deep pass.
(10 inches) per pass
Figure 7.6: Milling Machine Cutting Drum Figure 7.7: Milling Machine Cutting Teeth
Figure 7.8: Small Milling Machine Figure 7.9: Large Milling Machine
Figure 7.10: Milled road showing complete removal of the Figure 7.11: Milled road in preparation for HMA overlay.
HMA overlay, which exposes the PCC slabs beneath. Notice some areas of the previous HMA overlay remain.
After a pavement has been milled the resulting surface is quite dirty and dusty. The surface should be cleaned off by sweeping
or washing before any overlay is placed otherwise the dirt and dust will decrease the bond between the new overlay and the
existing pavement (see Figure 7.12 and 7.13). When sweeping, more than one pass is typically needed to remove all the dirt
and dust. If the milled surface is washed, the pavement must be allowed to dry prior to paving.
Milling also produces a rough, grooved surface, which will increase the existing pavement’s surface area when compared to an
ungrooved surface. The surface area increase is dependent on the type, number, condition and spacing of cutting drum teeth
but is typically in the range of 20 to 30 percent, which requires a corresponding increase in tack coat (20 to 30 percent more)
when compared to an unmilled surface (TRB, 2000).
For many situations, milling may be a superior alternative to a leveling course. Leveling course quantities are difficult to
accurately estimate and leveling course thicknesses are usually small, precluding the use of nuclear gauge density testing.
Thus, adequate mix density is difficult to achieve and measure. In some overlay projects a combination of milling and
leveling course application may be best.
Although typically used for rigid pavement surface restoration, diamond grinding can be used to eliminate relatively small
surface distortions in existing rigid pavement prior to flexible or rigid overlays. Because it roughens the existing rigid
pavement surface, diamond grinding also improves the bond between the existing pavement and the overlay. Non-overlay
applications of diamond grinding are covered in Module 10, Section 4, Rigid - Maintenance.
movement at slab and crack interfaces is great enough to crack a flexible overlay (called reflection cracking). There are
several techniques to prevent (or at least delay the onset of) reflection cracking:
● Prevent the slabs or sections from moving by stabilizing the material beneath them. This involves drilling holes in
an unstable PCC slab or section and injecting an asphaltic or cementitious material to fill any underlying voids.
Typically, this method is only an option for isolated instances of instability. It does not work well as a general
roadway treatment.
● Make the flexible structure strong enough to resist cracking. This usually involves extra granular base layers
between the flexible overlay and the existing rigid pavement or extremely thick flexible layers, both of which are
often not cost effective. Even if these types of preventative measures are used, they still cannot be guaranteed to
prevent reflective cracking.
WSDOT's experience is that reflection cracking has generally not been a problem if the
flexible overlay is at least 100 mm (4 inches) thick. Thinner overlays have exhibited
reflection cracking.
● Crack/break and seat the underlying rigid pavement. This involves breaking the underlying rigid pavement into
relatively small pieces (on the order of about 0.3 m2 to 0.6 m2 (1 ft2 to 2 ft2) by repeatedly dropping a large weight
(see Figure 7.14). The pieces are then seated by 2 to 3 passes of a large rubber tired roller. The result is a pavement
made of small firmly-seated pieces (see Figure 7.15). Video 7.2 briefly shows the process.
● Rubblize the underlying rigid pavement. This involves reducing the underlying rigid pavement to rubble. This
rubble is then used as a high quality base course to support the flexible overlay. Rubblizing is typically done with
one of the following two pieces of equipment:
❍ Resonant pavement breaker (see Figure 7.16 and Video 7.3). This equipment strikes the rigid pavement
at low amplitude with a small plate at the resonant frequency of the slab (usually about 44 Hz) causing the
slab to break apart (see Figure 7.17) (Roberts et al., 1996). Usually it takes about 14 to 18 passes for a
resonant pavement breaker to rubblize an entire 3.6 m (12 ft.) lane (NCAT, 2001).
❍ Multi-head breaker (MHB) (see Figures 7.18, 7.19 and Video 7.4). This equipment uses a series of
independently controlled high amplitude drop hammers to smash the slab. Typically, there are between
12 and 16 hammers, each weighing between 450 - 680 kg (1000 - 1500 lbs.). Hammers can be dropped
from variable heights (0.3 - 1.5 m (1 - 5 ft.)) to create impact energies between 2,700 - 16,300 N-m
(2,000 - 12,000 ft.-lbs.). Hammers cycle at a rate of 30 - 35 impacts per minute. MHBs can
rubblize an entire lane (up to 4 m (13 ft.)) in a single pass (Antigo Construction, 2001).
Figure 7.14: Drop Hammer Used for Cracking and Seating Figure 7.15: PCC Pavement After Cracking and Seating with
PCC Drop Hammer
Video 7.2: Drop Hammer Used for Breaking and Seating PCC
(video has no sound)
Figure 7.16: Resonant Pavement Breaker Used to Rubblize Figure 7.17: PCC Pavement After Rubblization With a
PCC Pavement Resonant Pavement Breaker
A 38-state survey published in 1999 (Ksaibati, Miley and Armaghani, 1999) revealed the following about rigid pavement
rubblizing:
● Distresses in the subsequent flexible overlay such as fatigue cracking and rutting are most often traced to a weak
subgrade. This subgrade is also the most likely cause of the original rigid pavement distress. Rubblization is risky
when subgrade support conditions are not well known.
● A majority of rubblized particles are in the 25.4 - 76.2 mm (1 - 3 inch) range, although particles near pavement
edged or under existing reinforcing steel can be as large as 380 mm (15 inches).
● Rubblizing is generally better than cracking and seating for reducing reflective cracking.
Given the expense of these techniques, some agencies just choose to live with joint reflection cracking rather than prevent it.
This is especially true on low volume, low speed roads where ride smoothness and structural integrity may not be given the
high priority they are on high volume, high speed roads like interstates.
Unbonded rigid overlays do not require much surface preparation, which is one of the principal reasons they are used.
Bonded rigid overlays of flexible pavement require several additional considerations. First, the success of a bonded overlay is
contingent on a good bond between the rigid overlay and the underlying flexible pavement. In order to develop this bond, the
underlying flexible pavement must have a clean, rough surface. Preferably, the flexible pavement should be milled, however,
as a minimum, water or abrasive blasting should be used to clean the HMA surface. If water blasting is used, the surface must
be allowed to air dry before the PCC is placed.
Once the flexible pavement surface is clean, it must be kept clean until the bonded overlay is placed. Dust,
dirt and debris that falls or blows onto the asphalt surface must be removed. If the surface is cleaned on
the day prior to paving, air cleaning may be required on the day of paving to remove dirt and dust. If traffic
is allowed on the milled surface, the surface must be cleaned again prior to paving.
2.3 Summary
Pavements should be placed only on properly prepared surfaces to ensure they perform properly. Pavements constructed on
inadequately prepared surfaces may be excessively rough, may not bond to the existing pavement (in the case of overlays) or
may fail because of inadequate subgrade support. For a new pavement, surface preparation involves compacting, grading and
possibly stabilizing the underlying subgrade. For an overlay, surface preparation involves repairing, leveling and cleaning the
existing pavement.
● Transportation Research Board (TRB). (2000). Hot-Mix Asphalt Paving Handbook 2000.
Transportation Research Board, National Research Council. Washington, D.C.
● Roberts, F.L., Kandhal, P.S., Brown, E.R., Lee, D.Y., and Kennedy, T.W. (1996). Hot Mix
Asphalt Materials, Mixture Design, and Construction. National Asphalt Pavement
Association Research and Education Foundation. Lanham, MD.
Drum plants, which produce HMA in a continuous manner, generally offer higher production rates than
batch plants for comparable cost. Typical production rates for drum plants vary between about 100 tons/
hr up to over 900 tons/hr depending upon drum design. Figure 7.19 shows the basic components of a
drum plant and their functions.
4 Flexible - Transport
Mix transport involves all actions and equipment required to convey HMA Major Topics on this Page
from a production facility to a paving site including truck loading, weighing
and ticketing, hauling to the paving site, dumping of the mix into the paver or 4.1 Truck Types
material transfer vehicle hopper, and truck return to the HMA production
facility (Roberts et al., 1996). Ideally, the goal of mix transport should be to 4.2 Operational Considerations
maintain mix characteristics between the production facility and the paving 4.3 Summary
site. Transport practices can have a profound effect on mix temperature at the
paving site, aggregate and/or temperature segregation of the mix and mat quality. This section will discuss the types of trucks
used for mix transport and the various considerations involved with mix transport.
● End dump
A windrow elevator is used to pick up HMA from the windrow and feed it into the paver hopper. Windrow elevators do not
have any method of regulating material flow, which makes it necessary to place the correct amount of HMA in the windrow to
match the paving width and depth being placed without allowing the paver hopper to run out of mix or become overloaded
(TRB, 2000).
Figure 7.24: Bottom Dump Truck in an Urban Setting Figure 7.25: Windrow made by emptying a bottom dump truck
Each truck type is capable of adequately delivering HMA from a production facility to a paving site. However, certain situations
such as the ones listed in Table 7.2 below, may make one truck type advantageous over another.
● Operation synchronization
1. Truck bed cleanliness and lubrication. Truck beds should be clean and lubricated to prevent the introduction of
foreign substances into the HMA and to prevent the HMA from sticking to the truck bed. Non-petroleum based
products should be used for lubrication such as lime water, soapy water or other suitable commercial products
(Roberts et al., 1996). Petroleum based products, such as diesel fuel, should not be used because of environmental
issues and because they tend to break down the asphalt binder.
2. Aggregate segregation. HMA should be discharged into the truck bed so as to minimize segregation. Dropping HMA
from the storage silo or batcher (for batch plants) in one large mass creates a single pile of HMA in the truck bed (see
Figure 7.27 and Video 7.8). Large-sized aggregate tends to roll off this pile and collect around the base. Dropping
HMA in several smaller masses (three is typical) at different points in the truck bed will largely prevent the collection
of large aggregate in one area and thus minimize aggregate segregation.
Figure 7.27: Truck Loading Under a Storage Silo Video 7.8: Truck Loading Close-up
1. Minimize haul distance. This can be accomplished by choosing an HMA production facility as close as possible to the
paving site. Closer production facilities create shorter haul times and result in less HMA cooling during transport.
Unfortunately, many paving locations may not be near any existing production facilities and economics may prohibit
the use of a mobile production facility.
2. Insulate truck beds. This can decrease HMA heat loss during transport. Insulation as simple as a sheet of plywood
has been used.
3. Place a tarpaulin over the truck bed. A tarp over the truck bed (see Figure 7.31) provides additional insulation,
protects the HMA from rain and decreases heat loss. A study by the Quality Improvement Committee of the National
Asphalt Pavement Association (NAPA) studied truck tarping and found that the HMA surface temperatures of tarped
loads dropped more slowly than untarped loads but temperatures 100 mm (4 inches) below the surface between tarped
and untarped loads were not significantly different (Minor, 1980).
Figure 7.28: Infrared picture of an HMA storage silo loading a Figure 7.29: Infrared picture of a truck dumping HMA showing
truck showing the hot uniform temperature of the mix the cold surface layer crust (blue) and the hot inner mass (red)
Figure 7.30: Infrared picture of a truck dumping HMA showing Figure 7.31: Driver covering his truck bed with a tarpaulin
the cold surface layer crust (blue) and the hot inner mass (red)
In most cases, truck transport appears to cool only the surface of the transported HMA mass, however this cool surface crust can
have detrimental effects on overall mat quality if not properly dealt with. Actions such as reducing transport time, insulating
truck beds or tarping trucks can decrease HMA surface cooling rate. Additionally, since the majority of the HMA mass is still at
or near its original temperature at loading, mixing the crust and interior mass together at the paving site (“remixing”) will
produce a uniform mix near the original temperature at loading.
1. HMA should be unloaded quickly when it arrives at the paving site. This will minimize mix cooling before it is
placed.
2. Before HMA is loaded into the paver, the inspector and/or foreman should be certain it is the correct mix.
Occasionally, paving jobs require several different mix designs (i.e., one for the leveling course and one for the
wearing course) and these mixes should not be interchanged.
● National Asphalt Pavement Association (NAPA). (1996). Balancing Production Rates in Hot Mix Asphalt
Operations, IS 120. National Asphalt Pavement Association. Landham, MD.
Ideally, all operations are designed to meet optimal mat laydown rates. However, these rates can vary based on paving width
and lift thickness. Also, complicated paving locations such as intersections or near manholes and utility vaults can temporarily
increase or decrease the laydown rate.
Truck transport should be planned such that the HMA transport rate (expressed in tons/hr) closely matches plant production rate
and laydown rate. Some factors to consider are:
● Truck type.
Traffic plays a large role in HMA delivery rates because it affects truck speed. Especially in congested urban areas, heavy and/
or unpredictable traffic may substantially increase, or at least vary, truck travel time. As truck travel time increases, more trucks
are needed to provide a given HMA delivery rate. Therefore, as traffic gets worse, trucking costs increase. Additionally, the
unpredictability of traffic may result in either long paver idle times as it waits for the next truckload of HMA or large truck
backups as several trucks all reach the paving site or production facility at the same time.
Finally, production facility output is typically controlled to match haul or laydown rate. However, this can result in suboptimal
plant efficiency or HMA uniformity, which may increase plant exhaust output, shorten emission control device lifetimes, and
affect contractual payment if payment is tied to HMA uniformity. It may often be more economical to run the production
facility at maximum rate and store excess material in storage silos for discharge into trucks as they arrive. Storage silo
insulation has progressed to a state where dense-graded HMA can be stored in them for up to a week at a time without
significantly affecting HMA characteristics. However, gap graded mixes such as SMA or OGFC should still not be stored for
more than about 2 to 3 hours.
In sum, synchronization should be the goal but it is often difficult to achieve (based on varying laydown rates, haul time and
traffic) and may result in plant inefficiency and HMA quality degradation. If a production facility has modern well-insulated,
airtight storage silos and is producing a dense-graded HMA, it may be beneficial to run the plant at maximum production rate
and store the mix until needed rather than try and match haul or laydown rate.
4.3 Summary
Mix transport can have a large impact on flexible pavement construction quality and efficiency. Mix characteristics such as
laydown temperature, aggregate segregation and temperature differentials are largely determined by transport practices. In
general, there are three types of HMA transport trucks: the end dump, bottom dump and live bottom dump (flo boy). End dump
trucks are most common, however bottom dumps and live bottom dumps are well-suited for certain situations. Key
considerations in mix transport are:
If properly managed, mix transport can successfully move HMA from the production facility to the paving site with little or no
change in mix characteristics.
5 Flexible - Placement
Mix placement and compaction are the two most important elements in HMA Major Topics on this Page
pavement construction. Mix placement involves any equipment or procedures
used to place the delivered HMA on the desired surface at the desired thickness. 5.1 Placement Considerations
Mix placement can involve complicated asphalt paver operations or simple
manual shoveling. This section provides a basic description of HMA placement 5.2 Asphalt Paver
operations. The Hot Mix Asphalt Paving Handbook (TRB, 2000) and the Asphalt 5.3 Material Transfer Vehicles (MTV)
Institute's HMA Construction manual (2001) contains detailed information on
asphalt paver components. 5.4 Summary
● Lift thickness. A "lift" refers to a layer of pavement as placed by the asphalt paver. In order to avoid mat tearing (which
generally shows up as a series of longitudinal streaks) a good rule-of-thumb is that the depth of the compacted lift should
be at least twice the maximum aggregate size or three times the nominal maximum aggregate size (TRB, 2000).
● Longitudinal joints. The interface between two adjacent and parallel HMA mats. Improperly constructed longitudinal
joints can cause premature deterioration of multilane HMA pavements in the form of cracking and raveling.
● Handwork. HMA can be placed by hand in situations where the paver cannot place it adequately. This can often occur
around utilities, around intersection corners and in other tight spaces. Hand-placing should be minimized because it is
prone to aggregate segregation and results in a slightly rough surface texture. If hand placement is necessary the
following precautions should be taken (Asphalt Institute, 2001):
❍ Place the HMA in a pile far enough away from the placement area that the whole pile must be moved. If the
pile is located in the placement area its appearance, density or aggregate distribution may be slightly different
than the surrounding handworked mat.
❍ Carefully deposit the material with shovels and then spread with lutes. Do not broadcast (scoop and pitch) the
HMA with shovels - this is likely to cause aggregate segregation.
❍ All material should be thoroughly loosened and evenly distributed. Chunks of HMA that do not easily break
apart should be removed and discarded.
❍ Check the handworked surface with a straightedge or template before rolling to ensure uniformity.
● SMA. SMA mixes behave differently than dense-graded mixes during placement and compaction. Experience and
understanding of dense-graded mix placement should be augmented with specific training and precautions before
attempting to place an SMA mix for the first time. SMAs are generally stickier and more difficult to work with than
dense-graded mixes because (1) they have more asphalt binder, (2) the asphalt binder is modified, and (3) the binder and
filler combination creates a viscous mastic. Also, it is not uncommon for large amounts of mastic (the combination of
asphalt binder and mineral filler) to collect on paving equipment. If not carefully monitored, this mastic will release from
the equipment into the mat leaving an over-asphalted area - commonly referred to as a "fat spot". These considerations
only scratch the surface of SMA construction. A more thorough treatment can be found in:
❍ National Asphalt Pavement Association (NAPA). (1999). Designing and Constructing SMA Mixtures - State-
of-the-Practice, Quality Improvement Series 122. National Asphalt Pavement Association. Landham, MD.
● Mat problems. The asphalt paver, MTV, rollers, mix design and manufacturing introduce many variables into flexible
pavement construction. A familiarity with common causes of the more typical mat problems can help improve
construction quality. Some common mat problems are microcracking, fat spots, joint problems, non-uniform texture,
roller marks, shoving, surface waves, tearing (streaks) and transverse screed marks.
● The screed
● Push Roller and Truck Hitch. The push roller is the portion of the paver that contacts the transport vehicle and the truck
hitch holds the transport vehicle in contact with the paver (see Figures 7.32 and 7.33). They are located on the front of
the hopper.
Figure 7.32: Push Roller and Truck Hitch Figure 7.33: Truck Hitch Engaged
● Hopper. The hopper is used as a temporary storage area for HMA delivered by the transport vehicle. Therefore, the
paver can accept more material than is immediately needed and can use the volume in the hopper to compensate for
fluctuating material demands created by such things as paving over irregular grades, utility access openings or irregular
intersection shapes. Hopper sides (or “wings”) can be tilted up (or “folded”) to force material to the middle where it is
carried to the rear by the conveyor system (see Figure 7.34). Hoppers can also be fit with inserts to allow them to carry
more HMA (see Figure 7.35). These inserts are typically used in conjunction with a material transfer vehicle (MTV).
Figure 7.34: Hopper with Wings Figure 7.35: Caterpillar AP-1055B with
Folded Up Hopper Insert
● Conveyor. The conveyor mechanism carries the HMA from the hopper, under the chassis and engine, then to the augers
(see Figure 7.36 and 7.37). The amount of HMA carried back by the conveyors is regulated by either variable speed
conveyors and augers or flow gates, which can be raised or lowered by the operator or, more often, by an automatic feed
control system.
● Auger. The auger receives HMA from the conveyor and spreads it out evenly over the width to be paved (see Figure
7.38). There is one auger for each side of the paver and they can be operated independently. Some pavers allow the
augers to be operated in reverse direction so that one can be operated forward and the other in reverse to send all the
received HMA to one side of the paver. The auger gearbox can either be located in the middle (between the augers as
shown in Figure 7.39) or on the outside edge of each auger. If an inadequate amount of HMA is distributed under a
middle-located gearbox the result can be a thin longitudinal strip of mat aligned with the gearbox that exhibits lower
densities from aggregate segregation and/or temperature differentials (see Figures 7.40 and 7.41).
Figure 7.40: Gearbox Streak Figure 7.41: Gearbox Streak in a New Mat
Operation of the tractor, and specifically the material feed system, can have significant effects on overall construction quality and
thus long-term pavement performance. Although there are many detailed operational concerns, the two broad statements below
encompass most of the detailed concerns:
1. HMA must be delivered to maintain a relatively constant head of material in front of the screed. This involves
maintaining a minimum amount of HMA in the hopper, regulating HMA feed rate by controlling conveyor/auger speed
and flow gate openings (if present), and maintaining a constant paving speed. As the next section will discuss, a
fluctuating HMA head in front of the screed will affect the screed angle of attack and produce bumps and waves in the
finished mat.
2. The hopper should never be allowed to empty during paving. This results in the leftover cold, large aggregate in the
hopper sliding onto the conveyor in a concentrated mass and then being placed on the mat without mixing with any hot
or fine aggregate. This can produce aggregate segregation or temperature differentials, which will cause isolated low
mat densities. If there are no transport vehicles immediately available to refill the hopper it is better to stop the paving
machine than to continue operating and empty the hopper (TRB, 2000).
5.2.2 Screed
The most critical feature of the paver is the self-leveling screed unit, which determines the profile of the HMA being placed
(Roberts et al., 1996). The screed takes the head of HMA from the material delivery system, strikes it off at the correct thickness
and provides initial mat compaction. This section describes:
● Screed terminology
The following is a list of basic screed components and terms (see Figure 7.44):
1. Screed plate. The flat bottom portion of the screed assembly that flattens and compresses the HMA.
2. Screed angle (angle of attack). The angle the screed makes with the ground surface.
3. Strike-off plate. The vertical plate just above the leading edge of the screed used to strike off excess HMA and protect
the screed’s leading edge from excessive wear.
4. Screed arms. Long beams that attach the screed to the tractor unit (see Figure 7.42).
5. Tow point. Point at which the screed arm is attached to the tractor unit (see Figure 7.43).
6. Depth crank. The manual control device used to set screed angle and ultimately, mat thickness (see Figure 7.42).
7. Screed heater. Heaters used to preheat the screed to HMA temperature. HMA may stick to a cold screed and cause mat
tearing. After the screed has been in contact with the HMA for a short while (usually about 10 minutes) its temperature
can be maintained by the HMA passing beneath it and the heater can be turned off. If the screed is removed from contact
with HMA for an extended period of time, it may need to be pre-heated again before resuming paving.
8. Screed vibrator. Device located within the screed used to increase the screed’s compactive effort. Screed compaction
depends upon screed weight, vibration frequency and vibration amplitude.
9. Screed extensions. Fixed or adjustable additions to the screed to make it longer (see Figures 7.44 and 7.45). Basic
screed widths are between 2.4 m (8 ft.) and 3.0 m (10 ft.). However, often it is economical to use wider screeds or
adjustable width screeds. Therefore, several manufacturers offer rigid extensions that can be attached to a basic screed or
hydraulically extendable screeds that can be adjusted on the fly.
Figure 7.42: Screed Close-Up Showing the Screed Arm and Depth Figure 7.43: Tow Point
Crank
There are six basic forces (see Figure 7.46) acting on the screed that determine its position and angle (Roberts et al., 1996):
1. Towing force. This is provided by the tractor and exerted at the tow point. Thus, towing force is controlled by paver
speed.
2. Force from the HMA head resisting the towing force. This is provided by the HMA in front of the screed and is
controlled by the material feed rate and HMA characteristics.
3. Weight of the screed acting vertically downward. This is obviously controlled by screed weight.
4. Resistive upward vertical force from the material being compacted under the screed. This is also a function of HMA
characteristics and screed weight.
5. Additional downward force applied by the screed’s tamping bars or vibrators. This is controlled by vibratory amplitude
and frequency or tamping bar force.
6. Frictional force between the screed and the HMA under the screed. This is controlled by HMA and screed
characteristics.
Since the screed is free floating it will slide across the HMA at an angle and height that will place these six forces in equilibrium.
When any one of these forces is changed, the screed angle and elevation will change (which will change the mat thickness) to bring
these forces back into equilibrium. Therefore, changing anything on the paver that affects these forces (such as paver speed,
material feed rate or screed tow point) will affect mat thickness. Furthermore, since mat thickness needs to be closely controlled,
pavers have controls to manually set screed angle rather than rely on a natural equilibrium to determine mat thickness. In typical
paving operations the screed angle is adjusted to control mat thickness. In order to understand how a manually controlled screed
angle affects mat thickness, a brief discussion of how the paver parameters of speed, material feed rate and tow point elevation
affect screed angle, screed height and therefore mat thickness is provided.
Speed
Paver speed affects mat thickness by changing the screed angle. If a paver speeds up and all other forces on the screed remain
constant, the screed angle decreases to restore equilibrium, which decreases mat thickness. Similarly, as paver speed decreases,
screed angle increases, which increases mat thickness.
Therefore, in order to maintain a constant mat thickness for a change in paver speed or material head in front of the screed, the
natural equilibrium of forces on the screed cannot be relied upon and the screed angle must be manually adjusted using a thickness
control screw or depth crank. Screed angle adjustments do not immediately change mat thickness but rather require a finite amount
of time and tow distance to take effect. Figure 7.47 shows that it typically takes five tow lengths (the length between the tow point
and the screed) after a desired level is input for a screed to arrive at the new level.
Because of this screed reaction time, a screed operator who constantly adjusts screed level to produce a desired mat thickness will
actually produce an excessively wavy, unsmooth pavement.
The interaction of paver speed, material feed rate and tow point elevation determine the screed position without the need for direct
manual input. This is why screeds are sometimes referred to as "floating" screeds.
As discussed previously, the screed angle can be manipulated manually to control mat thickness. However, tow point elevation is
not practical to manually control. Therefore, pavers usually operate using an automatic screed control, which controls tow point
elevation using a reference other than the tractor body. Since these references assist in controlling HMA pavement grade, they are
called “grade reference systems” and are listed below (Roberts et al., 1996):
1. Erected stringline. This consists of stringline erected to specified elevations that are independent of existing ground
elevation. Most often this is done using a survey crew and a detailed elevation/grade plan. Although the stringline
method provides the correct elevation (to within surveying and erecting tolerances), stringlines are fragile and easily
broken, knocked over or inadvertently misaligned. Lasers can be used to overcome the difficulties associated with
stringlines because they do not require any fragile material near the pavement construction area. Lasers can establish
multiple elevation or grade planes even in dusty or high-electronic and light-noise areas and are therefore sometimes
used to construct near-constant elevation airport runways. The laser method becomes quite complicated, however, when
frequent pavement grade changes are required.
2. Mobile reference. This consists of a reference system that travels with the paver such as a long beam or tube attached to
the paver (called a "contact" device since it actually touches the road - see Figure 7.48) or an ultrasonic device (called a
"non-contact" device since it relies on ultrasonic pulses and not physical contact to determine road elevation). The
mobile reference system averages the effect of deviations in the existing pavement surface over a distance greater that the
wheelbase of the tractor unit. Minimum ski length for a contact device is normally about 7.5 m (25 ft.) with a typical ski
lengths being on the order of 12 to 18 m (40 to 60 ft.) (Asphalt Institute, 2001).
3. Joint matching shoe. This usually consists of a small shoe or ski attached to the paver that slides on an existing surface
(such as a curb) near the paver. Ultra sonic sensors accomplish the same task without touching the existing surface by
using sound pulses to determine elevation. This type of grade control results in the paver duplicating the reference
surface on which the shoe or ski is placed or ultra sonic sensor is aimed.
In addition to grade control, the screed can also be set to control pavement slope and/or crown. A slope controller uses a slope
sensor mounted on a transverse beam attached to the screed to determine screed slope, then adjusts screed slope to the desired
amount. Generally, one side of the screed is set up to control grade and the opposite side is set up to control slope based on that
grade. The usual practice is to run grade control on the side of the screed nearest the pavement centerline and run slope control on
the screed side nearest the pavement edge because it is easier to match the centerline joint if grade control is used on that side of the
paver (TRB, 2000).
Screed crown (the elevation of the middle in relation to the edges) can also be controlled. Typically screeds offer separate front and
rear crown controls. If crown control is used, the front control is usually set to a slightly more severe crown than the rear control to
allow for easier passage of HMA under the screed.
The floating screeds used by today’s pavers are acted upon by six basic forces, which when left undisturbed result in an equilibrium
screed angle and elevation that determines mat thickness. Adjusting paver speed, material feed rate or tow point elevation will
change these forces and result in a new equilibrium screed angle and elevation and eventually a new mat thickness. In order to
achieve the most consistent thickness and smoothest possible surface, pavers attempt to maintain a constant speed, use automatic
feed controls to maintain a consistent head of material in front of the paver, and use automatic screed control to maintain a
consistent tow point. Although the screed angle can be adjusted manually to change mat thickness, excessive adjustments will
result in a wavy, unsmooth mat. In addition to grade, screeds can also control mat slope and crown to provide almost complete
control over mat elevation at any location.
● Windrow elevators
Other MTVs are used to provide an additional surge volume for the paver (see Figures 7.51 through 7.54). This surge volume
allows for continuous paver operations because with an MTV the paver no longer has to stop while one truck leaves and the next
truck backs up. Additionally, the MTV serves as a buffer between the paver and the haul trucks, which eliminates most truck
bumping problems. Finally, most MTVs offer some sort of remixing capability that remixes the cool HMA crust formed during
transport with the hot interior HMA to produce a more uniform mix entering the paver. This remixing can essentially eliminate
aggregate segregation and temperature differentials. Some states have actually implemented specifications that require a remixing
MTV for paving contracts where segregation and temperature differentials are of concern.
Figure 7.51: Blaw Knox MC-30 MTV Figure 7.52: Cedarapids MS-3 MTV
Figure 7.53: Roadtec Shuttle Buggy MTV Figure 7.54: Roadtec Shuttle Buggy Front View Showing
Loading Hopper for End Dump and Live Bottom Trucks
In summary, MTV’s assist with the transfer of HMA from the transport trucks to the paver. There are two basic types of MTVs:
windrow elevator MTVs and surge volume / remix MTVs. Windrow elevator MTVs are used to pick up HMA from a windrow
and place it into the paver hopper while surge volume / remix MTVs provide an additional material surge volume that allows for
continuous paving and/or a remix capability that can reduce aggregate/temperature segregation. MTV use costs money and will
increase the per ton cost of HMA paving but can help provide superior mat quality. However, MTVs should not be used as a
substitute for good production and laydown practices, which are fundamental to good mat quality.
5.4 Summary
Mix placement is one of the most important elements in HMA pavement construction. The key piece of equipment in mix
placement, the asphalt paver, consists of two major systems: the tractor system and the screed system. Proper operation of these
systems results in an independent machine that is capable of placing a smooth, continuous HMA pavement. MTVs can be used to
augment mix transfer to the asphalt paver. Proper employment of MTVs can increase laydown rates, streamline mix transfer and
help reduce segregation problems.
6 Flexible - Compaction
It has been said that the top three factors in real estate are “location, Major Topics on this Page
location, location”. It can also be said that the top three factors in HMA
pavement construction are “compaction, compaction, compaction”. 6.1 Compaction Measurement and Reporting
Compaction is the process by which the volume of air in an HMA mixture
is reduced by using external forces to reorient the constituent aggregate 6.2 Compaction Importance
particles into a more closely spaced arrangement. This reduction of air 6.3 Factors Affecting Compaction
volume in a mixture produces a corresponding increase in HMA unit
weight, or density (Roberts et al., 1996). Numerous researchers have 6.4 Compaction Equipment
stated that compaction is the greatest determining factor in dense graded
pavement performance (Scherocman and Martenson, 1984; Scherocman, 6.5 Roller Variables
1984; Geller, 1984; Brown, 1984; Bell et. al., 1984; Hughes, 1984; 6.6 Summary
Hughes, 1989). Inadequate compaction results in a pavement with
decreased stiffness, reduced fatigue life, accelerated aging/decreased durability, rutting, raveling, and moisture damage
(Hughes, 1984; Hughes, 1989).
Percent air voids is typically calculated by using AASHTO T 269, ASTM D 3203 or an equivalent procedure (AASHTO,
2000). These procedures all use laboratory-determined bulk specific gravity and theoretical maximum specific gravity in
the following equation:
where: Gmm = theoretical maximum specific gravity of the particular HMA in question
These procedures require a small pavement core (usually 100 - 150 mm (4 - 6 inches) in diameter), which is extracted from
the compacted HMA (see Figure 7.56 and 7.57). This type of air voids testing is generally considered the most accurate but
is also the most time consuming and expensive.
Figure 7.56: Core Extraction Figure 7.57: Two Cores – The Core on the Right has
Significantly Higher Air Voids
Since core extraction is time consuming and expensive, air voids are often measured indirectly using a portable density-
measuring device such as a nuclear density gauge (see Figure 7.58) or electrical density gauge (see Figure 7.59).
Figure 7.58: Nuclear Density Gauge Figure 7.59: Electrical Density Gauge (Trans Tech PQI
pictured)
Each contracting agency usually specifies the compaction measurement methods and equipment to be used on contracts
under their jurisdiction. Most agencies stipulate some sort of extracted core density testing and usually allow testing by
nuclear gauge. Electric density gauges are relatively new on the market (in the last five years). Accurate calibration of
these devices is essential for their proper use.
WSDOT uses the nuclear density gauge for acceptance density testing. The gauge
is calibrated to densities measured in the laboratory on core samples from the same
material.
Although percent air voids is the HMA characteristic of interest, measurements are usually reported as a measured density
in relation to a reference density. This is done by reporting density as a (1) percent of TMD (sometimes called Rice
density), (2) percent of a laboratory density or (3) percent of a control strip density (a control strip is a short pavement strip
that is compacted to the desired value under close scrutiny then used as the compaction standard for a particular job).
In sum, percent air voids is the critical HMA characteristic with which compaction is concerned. It can be measured using
pavement cores or portable nuclear or electric gauges; measurement specifications vary from one contracting agency to the
next. Percent air voids is usually reported as a density in one of three forms: (1) percent TMD, (2) percent of laboratory
density or (3) percent of control strip density. Regardless of the measurement device or reporting method, the key
characteristic is percent air voids.
WSDOT's generally requires a minimum density of 91.0 percent of TMD for its surface course mixes (Class
A, B, E, F and Superpave mixes used in traffic lanes). On projects exceeding 2,500 tons of a specific class
of HMA (e.g., Class A or 19 mm Superpave) the specified level of density attained is determined by a
statistical evaluation of five random nuclear density gauge tests taken from a 400 ton lot.
1. Decreased stiffness and strength. Kennedy et al. (1984) concluded that tensile strength, static and resilient
moduli, and stability are reduced at high air void content.
2. Reduced Fatigue Life. Several researchers have reported the relationship between increased air voids and
reduced fatigue life (Pell and Taylor, 1969; Epps and Monismith, 1969; Linden et. al., 1989). Finn et al. (1973)
concluded “...fatigue properties can be reduced by 30 to 40 percent for each one percent increase in air void
content.” Another study concluded that a reduction in air voids from eight percent to three percent could more
than double pavement fatigue life (Scherocman, 1984a).
3. Accelerated Aging/Decreased Durability. In his Highway Research Board paper, McLeod (1967) concluded
“compacting a well-designed paving mixture to low air voids retards the rate of hardening of the asphalt binder,
and results in longer pavement life, lower pavement maintenance, and better all-around pavement performance.”
4. Raveling. Kandhal and Koehler (1984) found that raveling becomes a significant problem above about eight
percent air voids and becomes a severe problem above approximately 15 percent air voids.
5. Rutting. The amount of rutting which occurs in an asphalt pavement is inversely proportional to the air void
content (Scherocman, 1984a). Rutting can be caused by two different mechanisms: vertical consolidation and
lateral distortion. Vertical consolidation results from continued pavement compaction (reduction of air voids) by
traffic after construction. Lateral distortion – shoving of the pavement material sideways and a humping-up of
the asphalt concrete mixture outside the wheelpaths – is usually due to a mix design problem. Both types of
rutting can occur more quickly if the HMA air void content is too low (Scherocman, 1984a).
6. Moisture Damage. Air voids in insufficiently compacted HMA are high and tend to be interconnected with each
other. Numerous and interconnected air voids allow for easy water entry (Kandhal and Koehler, 1984; Cooley et
al., 2002) which increases the likelihood of significant moisture damage. The relationship between permeability,
nominal maximum aggregate size and lift thickness is quite important and can change significantly as these
parameters change.
Air voids that are either too great or too low can cause a significant reduction in pavement life. For dense graded HMA, air
voids between 3 and 8 percent generally produce the best compromise of pavement strength, fatigue life, durability,
raveling, rutting and moisture damage susceptibility.
Environmental factors are determined by when and where paving occurs. Paving operations may have some float time,
which allows a limited choice of “when” but paving location is determined by road location so there is essentially no choice
of “where”. Mix and structural design factors are determined before construction and although they should account for
construction practices and the anticipated environment, they often must compromise ease of construction and compaction to
achieve design objectives. Obviously construction factors are the most controllable and adaptable of all the factors
affecting compaction. Although some factors like haul distance/time, HMA production temperature, lift thickness and type/
number of rollers may be somewhat predetermined, other factors associated with roller timing, speed, pattern and number
of passes can be manipulated as necessary to produce an adequately compacted mat. This subsection discusses:
Section 6.4, Compaction Equipment and 6.5, Roller Variables discuss construction factors.
6.3.1 Temperature
HMA temperature has a direct effect on the viscosity of the asphalt cement binder and thus compaction. As HMA
temperature decreases, its asphalt cement binder becomes more viscous and resistant to deformation, which results in a
smaller reduction in air voids for a given compactive effort. As the mix cools, the asphalt binder eventually becomes stiff
enough to effectively prevent any further reduction in air voids regardless of the applied compactive effort. The
temperature at which this occurs, commonly referred to as cessation temperature, is a function of the mix property factors
in Table 7.4. In some literature it is reported to be about 79oC (175°F) for dense-graded HMA (Scherocman, 1984b;
Hughes, 1989). Below cessation temperature rollers can still be operated on the mat to improve smoothness and surface
texture but further compaction will generally not occur. Conversely, if the binder is too fluid and the aggregate structure is
weak (e.g., at high temperatures), roller loads will simply displace, or “shove” the mat rather than compact it. In general,
the combination of asphalt cement binder and aggregate needs to be viscous enough to allow compaction but stiff enough to
prevent excessive shoving.
Mat temperature then, is crucial to both the actual amount of air void reduction for a given compactive effort, and the
overall time available for compaction. If the initial temperature and cool-down rate are known, the temperature of the mat
at any time after laydown can be calculated. Based on this calculation rolling equipment and patterns can be employed to:
1. Take maximum advantage of available roller compactive effort. Rollers can be used where the mat is most
receptive to compaction and avoided where the mat is susceptible to excessive shoving.
2. Ensure the mat is compacted to the desired air void content before cessation temperature is reached. This can be
done by calculating the time it takes the mat to cool from initial temperature to cessation temperature. All
compaction must be accomplished within this “time available for compaction”.
The major factors affecting time available for compaction are (Roberts et al., 1996):
● Initial mat temperature. Higher initial mat temperatures require more time to cool down to cessation
temperature, thus increasing the time available for compaction. However, overheating the HMA will damage the
asphalt binder and cause emissions.
● Mat or lift thickness. Thicker lifts have a smaller surface-to-volume ratio and thus lose heat more slowly, which
increases the time available for compaction.
● Temperature of the surface on which the mat is placed. Hotter surfaces will remove heat from the mat at a slower
rate, increasing the time available for compaction.
● Ambient temperature. Hotter air temperatures will remove heat from the mat at a slower rate, increasing the time
available for compaction.
● Wind speed. Lower wind speeds will decrease mat heat loss by convection, which will increase the time
available for compaction.
Jordan and Thomas (1976) point out additional factors affecting mat cool-down rate that include mat density, pavement
layer thermal conductivity, specific heat, convection coefficient, incident solar radiation and coefficients of emission and
absorption of solar radiation for the pavement surface.
David Timm, Vaughan Voller and David Newcomb have developed a software tool at the University of Minnesota called
MultiCool that automatically calculates pavement cool-down rate and time available for compaction (see Program 7.1).
Table 7.5 is a sampling of MultiCool output for some representative values of pavement thickness and ambient temperature.
Approximate Time
25 mm 149 °C 16 °C 9 minutes
(1 inch) (300 °F) (60 °F)
25 mm 149 °C 7 minutes
-4 °C
(1 inch) (300 °F) (25 °F)
50 mm 121 °C 16 minutes
16 °C
(2 inches) (250 °F) (60 °F)
50 mm 121 °C 12 minutes
-4 °C
(2 inches) (250 °F) (25 °F)
9. Single lift
MultiCool is quick and powerful. It can easily be installed on a laptop and used by contractors or inspectors to give a
general idea of the time available for compaction on a given job site, which can be quite helpful in determining roller use
and patterns. Figure 7.60 relates HMA temperature with typical aspects of compaction.
HMA temperature affects its binder viscosity, which affects compaction in two ways: (1) the colder and more viscous the
binder, the less actual amount of air void reduction for a given compactive effort, and (2) HMA can only be compacted until
it reaches cessation temperature, therefore initial HMA temperature and mat cool-down rate establish a fundamental
compaction parameter – the overall time available for compaction. Many factors influence HMA temperature and cool-
down rate including initial mat temperature, mat thickness, temperature of the surface on which the mat is placed, ambient
temperature and wind speed. Using these factors as inputs, MultiCool, a program developed at the University of Minnesota,
can easily produce a mat cool-down curve and calculate the time available for compaction.
Gradation affects the way aggregate interlocks and thus the ease with which aggregate can be rearranged under roller loads.
In general, aggregate effects on compaction can be broken down by aggregate size (TRB, 2000):
1. Coarse aggregate. Surface texture, particle shape and the number of fractured faces can affect compaction.
Rough surface textures, cubical or block shaped aggregate (as opposed to round aggregate) and highly angular
particles (high percentage of fractured faces) will all increase the required compactive effort to achieve a specific
density.
2. Midsize fine aggregate (between the 0.60 and 0.30-mm (No. 30 and No. 50) sieves). High amounts of midsize
fine, rounded aggregate (natural sand) cause a mix to displace laterally or shove under roller loads. This occurs
because the excess midsize fine, rounded aggregate results in a mix with insufficient voids in the mineral
aggregate (VMA). This gives only a small void volume available for the asphalt cement to fill. Therefore, if the
binder content is just a bit high it completely fills the voids and the excess serves to (1) resist compaction by
forcing the aggregate apart and (2) lubricate the aggregate making it easy for the mix to laterally displace.
3. Fines or dust (aggregate passing the 0.075-mm (No. 200) sieve). Generally, a mix with a high fines content will
be more difficult to compact than a mix with a low fines content.
The asphalt binder grade affects compaction through its viscosity. A binder that is higher in viscosity will generally result
in a mix that is more resistant to compaction. Additionally, the more a binder hardens (or ages) during production, the more
resistant the mix is to compaction.
Asphalt binder content also affects compaction. Asphalt binder lubricates the aggregate during compaction and therefore,
mixes with low asphalt content are generally difficult to compact because of inadequate lubrication, whereas mixes with
high asphalt content will compact easily but may shove under roller loads (TRB, 2000).
Sometimes, a combination of mix design factors produces what is known as a tender mix. Tender mixes are internally
unstable mixes that tend to displace laterally and shove rather than compact under roller loads.
1. By applying its weight to the HMA surface and compressing the material underneath the ground contact area.
Since this compression will be greater for longer periods of contact, lower equipment speeds will produce more
compression. Obviously, higher equipment weight will also increase compression.
2. By creating a shear stress between the compressed material underneath the ground contact area and the adjacent
uncompressed material. When combined with equipment speed, this produces a shear rate. Lowering equipment
speed can decrease the shear rate, which increases the shearing stress. Higher shearing stresses are more capable
of rearranging aggregate into more dense configurations.
These two means of densifying HMA are often referred to collectively as “compactive effort”. This section discusses the
paver screed, the steel wheeled roller (both static and vibratory) and the pneumatic tire roller as they apply to HMA
compaction. Section 6.5.1, Compaction Sequence discusses how each one of these pieces of compaction equipment work
together in a typical construction scenario. This subsection covers:
Figure 7.61: Small Static Steel Wheel Roller (1.32 tonnes Figure 7.62: Large Vibratory Steel Wheel Roller (17 tonnes
(1.45 tons), 86 cm (34-inch) wide drum) (18.7 tons), 213 cm (84-inch) wide drum)
In addition to their own weight, some steel wheel rollers can be ballasted with either sand or water to increase their weight
and thus, compactive effort. Although this ballasting is a fairly simple process (see Figure 7.63), it is usually done before
rolling operations start and rarely during rolling operations. Since asphalt cement binder sticks to steel wheels, most steel
wheel rollers spray water on the drums to prevent HMA from sticking, and are equipped with a transverse bar on each drum
to wipe off HMA (see Video 7.10). Note, however, that this water will cool the HMA and can reduce the time available for
compaction.
Figure 7.63: Filling up with water Video 7.10: Steel wheel roller showing transverse
bar and water spray
Some steel wheel rollers are equipped with vibratory drums. Drum vibration adds a dynamic load to the static roller weight
to create a greater total compactive effort. Drum vibration also reduces friction and aggregate interlock during compaction,
which allows aggregate particles to move into final positions that produce greater friction and interlock than could be
achieved without vibration. Roller drum vibration is produced using a rotating eccentric weight located in the vibrating
drum (or drums) and the force it creates is proportional to the eccentric moment of the rotating weight and the speed of
rotation (TRB, 2000). Operators can turn the vibrations on or off and can also control amplitude (eccentric moment) and
frequency (speed of rotation). Vibration frequency and amplitude have a direct effect on the dynamic force (and thus the
compactive force) as shown in Table 7.6.
Table 7.6: Vibratory Steel Wheel Roller Parameters (after TRB, 2000)
The ideal vibratory frequency and amplitude settings are a compromise based on desired mat smoothness, HMA
characteristics and lift thickness. Low vibration frequencies combined with high roller speeds will increase the distance
between surface impacts and create a rippled, unsmooth surface. In general, higher frequencies and lower roller speeds are
preferred because they decrease the distance between surface impacts, which (1) increases the compactive effort (more
impacts per unit of length) and (2) provides a smoother mat. The recommended impact spacing is 3 - 4 impacts per meter
(10 - 12 impacts per foot). Table 7.7 shows basic guidance for vibratory settings.
Operate in static mode. Under vibratory mode, as the pavement increases in density the
Thin Lifts
drums may begin to bounce, which may cause the HMA to shove and become less
(< about 30mm (1.25 inches))
dense. Also, some of the aggregates may be crushed.
Lifts beyond 65 mm
High frequency Higher amplitude
(2.5 inches)
As a general rule-of-thumb, a combination of speed and frequency that results in 3 - 3.5 impacts per meter (10 - 12 impacts
per foot) is good. At 3000 vibrations/minute that gives a speed of 4.5 - 5.5 km/hr (2.8 - 3.4 mph).
When density is difficult to quickly achieve with a vibratory steel wheel roller, the tendency may be to increase vibratory
amplitude to increase compactive effort. However, high amplitude is only advisable on stiff mixes or very thick lifts that
can support the increased amplitude without fracturing the constituent aggregate particles. For typical mix types and lift
thicknesses a better solution is usually to maintain low amplitude vibrations and increase the number of roller passes at low
amplitude.
Vibratory steel wheel rollers offer potential compaction advantages over static steel wheel rollers but they also require the
operator to control more compaction variables (amplitude, frequency and vibratory mode use) and there are certain
situations under which they must be used with caution (e.g., over shallow underground utilities, in residential areas, thin
overlays).
In general, steel wheel rollers provide the smoothest mat finish of all compaction equipment. When operated in the
vibratory mode, they also provide substantial compactive effort.
Asphalt binder tends to stick to cold pneumatic tires but not to hot pneumatic tires. A release agent (like water) can be used
to minimize this sticking, however if asphalt binder pickup (the asphalt binder sticking to the tires) is not permanently
damaging the mat it is better to run the roller on the hot mat and let the tires heat up to near mat temperature. Tires near
mat temperature will not pick up an appreciable amount of asphalt binder. Insulating the tire area with rubber matting or
plywood helps maintain the tires near mat temperature while rolling (see Figure 7.64).
Figure 7.64: Pneumatic Tire Roller (notice rubber matting insulation around tire area as well as tire marks left in
the new mat in front of the roller)
In addition to a static compressive force, pneumatic tire rollers also develop a kneading action between the tires that tends
to realign aggregate within the HMA. This results in both advantages and disadvantages when compared to steel wheel
rollers:
1. They provide a more uniform degree of compaction than steel wheel rollers.
2. They provide a tighter, denser surface thus decreasing permeability of the layer.
3. They provide increased density that many times cannot be obtained with steel wheeled rollers.
4. They compact the mixture without causing checking (hairline surface cracks) and they help to remove any
checking that is caused with steel wheeled rollers.
Disadvantages
1. The individual tire arrangement may cause deformations in the mat that are difficult or impossible to remove with
further rolling. Thus, they should not be used for finish rolling.
2. If the HMA binder contains a rubber modifier, HMA pickup (mix sticking to the tires) may be so severe as to
warrant discontinuing use of the roller.
In summary, pneumatic tire rollers offer a slightly different type of compaction than steel wheel rollers. The arrangement
of multiple tires on both axles serves to both compress and kneed the mat, which may or may not be advantageous over
steel wheel rollers.
● Roller speed
Not all these variables are infinitely adjustable, but by adjusting a combination of them a rolling plan can be developed that
will optimize mat compaction.
1. Screed. The screed is the first device used to compact the mat and may be operated in the vibratory mode.
2. Breakdown Roller. The breakdown roller is the first roller behind the screed and therefore, generally effects the
most density gain of any roller in the sequence. Breakdown rollers can be of any type but are most often
vibratory steel wheel and sometimes pneumatic tire.
3. Intermediate Roller. The intermediate roller is used behind the breakdown roller if additional compaction is
needed. Pneumatic tire rollers are sometimes used as intermediate rollers because they provide a different type of
compaction (kneading action) than a breakdown steel wheel vibratory roller. This can help further compact the
mat or at the very least, rearrange the aggregate within the mat to make it receptive to further compaction.
4. Finish Roller. The finish roller is last in the sequence and is used to provide a smooth mat surface. Although the
finish roller does apply compactive effort, by the time it comes in contact with the mat, the mat may have cooled
below cessation temperature. Static steel wheel rollers are almost always used as finishing rollers because they
can produce the smoothest surface of any roller type.
5. Traffic. After the rollers have compacted the mat to the desired density and produced the desired smoothness, the
new pavement is opened to traffic. Traffic loading will provide further compaction in the wheel paths of a
finished mat. For instance, a mat compacted to eight percent air voids and then opened to heavy traffic (e.g., an
interstate freeway) may further compact to about three to five percent air voids in the wheelpaths over time.
Each position in the roller train (breakdown, intermediate and finish) may be performed by one roller or several rollers in
parallel. For instance, a large paving project may use two vibratory steel wheel rollers for breakdown rolling, one
pneumatic tire roller for intermediate rolling and two static steel wheel rollers for finish rolling. The determination of the
best rolling sequence and the number of rollers is generally made on a case by case basis and depends upon the desired final
air voids, available rollers and their operating parameters, rolling patterns, mix properties, and environmental conditions.
Static Steel Wheel 3.2 – 5.6 km/h 4.0 – 6.5 km/h 4.8 – 8.0 km/h
(2.0 – 3.5 mph) (2.5 – 4.0 mph) (3.0 – 5.0 mph)
Pneumatic 3.2 – 5.6 km/h 4.0 – 6.5 km/h 6.4 – 11.2 km/h
(2.0 – 3.5 mph) (2.5 – 4.0 mph) (4.0 – 7.0 mph)
Vibratory Steel Wheel 3.2 – 4.8 km/h 4.0 – 5.6km/h not used
(2.0 – 3.0 mph) (2.5 – 3.5 mph)
As mentioned previously, roller compactive effort comes in two forms: (1) material compression under the ground contact
area and (2) shear stress between the compressed area and adjacent uncompressed areas. Operating at lower speeds allows
the roller to remain in contact with a particular mat location longer than it would at higher speeds. This results in more
compression per roller pass and therefore increases compactive effort. Speed also affects the magnitude of shear stress
developed. Lower speeds result in the shearing force between compressed and uncompressed areas being applied for a
longer period of time for a particular area (giving a lower shear rate), which results in a higher shear stress. The higher the
shear stress, the better able it is to rearrange aggregate into a denser configuration. Therefore, as roller speed decreases,
shear stress increases and compactive effort increases.
Because speed affects compactive effort, varying roller speed will vary compactive effort resulting in uneven compaction.
Varying roller speed typically occurs when operators are not closely monitoring their speed or when they speed up to roll an
area more quickly so that they can catch up to the paver. If the mat is being laid down at a faster rate than it can be rolled,
the solution should not be to speed up the rollers but rather should involve one of the following options (TRB, 2000):
1. Slow down the paver. This may involve adjusting production and material delivery rate as well.
2. Use more rollers. Adding rollers can increase the number of roller passes in a given time without reducing the
compactive effort per pass.
3. Use larger, wider rollers. Wider rollers allow greater coverage per pass.
Finally, rollers should not be stopped on a fresh mat because they can cause large indentations that are difficult, if not
impossible, to remove.
Roller speed directly affects compactive effort. The best compactive effort and most uniform densities are achieved by
slow, consistent roller speeds. If rollers cannot keep up with the pace of the paving operation, they should not be operated
at higher speeds because this reduces compactive effort. Rather, the paving operation should be slowed or more/larger
rollers should be used.
Test Strip
Contractors will often (and are sometimes required to) construct a “test strip” to help determine the necessary number of
passes. A test strip is a small section of mat laid out at the beginning of a project with the purpose of determining the best
roller type, sequence, number of passes and rolling pattern to use.
In general, the greatest compaction per roller pass can be achieved right behind the paver because the mat is the hottest and
least viscous in that position. Therefore, the breakdown roller(s) should operate as close to the paver as possible to achieve
the most compaction per roller pass. Likewise, the intermediate roller(s) and finish roller(s) should be placed on the mat at
a safe distance from the roller in front of them and begin rolling as soon as possible. Sometimes when a tender mix is
placed, these general rules do not apply.
Uniform compaction depends on getting the same number of roller passes over each area of the mat. This means that a
pattern must be developed that covers the entire mat with an equal number of roller passes from each type of roller. For
example, if two vibratory steel wheel rollers are operating as the breakdown rollers they must work together so each portion
of the mat receives the same number of passes, but since they are the same type of roller it is not necessary for each roller to
cover the entire mat. If two different rollers such as a vibratory steel wheel roller and a pneumatic tire roller are performing
breakdown rolling, each roller should cover the entire mat an equal number of times, otherwise compaction may be non-
uniform. Although roller patterns can vary widely, some general rules-of-thumb are:
● Overlap between two successive passes should be at least 15 cm (6 inches) (Roberts et al., 1996; Ingersoll-Rand,
2001). This ensures that small steering inaccuracies do not leave gaps between successive passes.
● The roller should be turned slightly to the side when reversing directions or stopping. Rollers tend to create a
slight bow hump (see Video 7.11) when moving and will leave this hump in place when reversing directions or
stopping. Often, it is difficult to flatten out this hump on subsequent passes if it is perpendicular to the direction
of roller travel. By turning the roller slightly before changing direction or stopping, the resulting hump will be
diagonal to the direction of roller travel and easier to flatten out with subsequent passes. However, hard steering
should be avoided because it can tear or shove the mat.
● Roller passes should end at different points to prevent developing a hump (caused by the direction change) that
spans the entire transverse length of the mat.
● Where there is an unconfined edge on the mat, the first roller pass should stay about 0.15 - 0.30 m (0.5 - 1 ft.)
away from the mat edge. The small resultant strip of uncompacted mat helps confine the rest of the mat and
minimize lateral displacement near mat edges. This strip should then be compacted on the next roller pass
(Ingersoll-Rand, 2001).
● Do not roll over a designed crown in the road. Rolling over a crown will flatten it out.
● When compacting a longitudinal joint, the first roller pass should be entirely on the hot mat about 0.15 - 0.30 m
(0.5 - 1 ft.) away from the joint. On subsequent passes, the roller should travel mostly on the newly constructed
mat and only overlap the older mat by about 0.15 m (0.5 ft.) (Roberts et al., 1996; Ingersoll-Rand, 2001).
● Joints should be compacted with the roller operating parallel to the joint. Although transverse joints cannot
always be compacted this way, perpendicular rolling does not compact the hot/new side as well.
● For steel wheeled rollers, operate the powered wheel on the paver side. This will minimize humps that can be
caused by the drive wheel.
The above thumb-rules are just general guidance; other methods may work. However, without a clear roller pattern, the
center of a lane typically receives more roller passes than the outsides. This is of particular concern because most wheel
loads occur nearer the edges of any particular lane in the wheelpaths. In summary, any method that achieves uniform
coverage, acceptable density and acceptable smoothness without damaging the mat can be considered a good method.
6.6 Summary
Although compaction looks like a simple job, it is far from it. Variables such as sequence, speed, number of passes,
location, pattern and mat temperature make it quite complex. All these variables have a profound effect on air voids and
thus pavement performance. Simply put, good compaction is essential to quality pavement.
● National Ready Mixed Concrete Association (NRMCA). (2002). NRMCA web site, Concrete Basics
home page. National Ready Mixed Concrete Association. Silver Spring, MD. http://www.nrmca.org.
Figure 7.67: Typical Charging Order for Truck Mixed PCC (redrawn from ACPA, 1995)
Truck mixed PCC (see figure 7.68) is appropriate for all types of pavement construction, but is particularly well-
suited for instances that can take advantage of a concrete mixing truck's monitored, continuously agitated storage
volume (the mixing drum) and its ability to precisely deliver PCC through its chute (see Figure 7.69). These
situations might include: intersection paving, street paving, pavement repair, urban environments, high traffic areas
and staged/phased construction. Disadvantages of truck mixed PCC are longer load and unload times (due to the
nature and opening size of the mixing drum) and the higher operating cost of a concrete mixing truck when compared
to an end or bottom dump truck. Truck mixer specifics are covered in Section 8, Mix Transport.
● High production volume. A high production volume would be on the order of 3000 m3/day (3950 yd3/
day). Most central mix drums can mix about 9 m3 (12 yd3) in a single batch and produce fully mixed PCC
in excess of 150 m3/hr (200 yd3/hr). Mixing times are on the order of 30 to 90 seconds (ACPA, 1995).
● Improved quality control. Since mixing is controlled by a central facility and not by individual truck, PCC
quality can be more closely monitored and controlled.
● Reduced wear on truck mixer drums. If used, concrete mixing trucks are used as an agitating haul unit
rather than a mixing unit. Agitating rotation speeds are much slower than mixing speeds and thus, produce
less wear on drum components.
8 Rigid - Transport
Mix transport involves all actions and equipment required to Major Topics on this Page
convey PCC from a batching facility to a paving site including
truck loading, weighing and ticketing, hauling to the paving site, 8.1 Truck Types
mixing (if required), agitation, placing, truck washing and truck
return to the batching facility. The goal of mix transport is to 8.2 Operational Considerations
delivery PCC to the paving job site that conforms to the specified 8.3 Summary
mix design. Transport practices can affect such mix characteristics
as (1) homogeneity, (2) workability, (3) water content and (4) air content. This section will discuss the types of
trucks used for mix transport and the various considerations involved with mix transport.
● Truck mixer. Truck mixers, use a truck-mounted rotating drum that is capable of mixing (if necessary) and
agitating the ready mixed PCC.
● Non-agitating truck. Non-agitating trucks are not able to mix or agitate their payload and usually consist
of end dump, bottom dump or side-dump trucks.
accurate to about 12.5 mm (0.5 inches)) and digital water meters to monitor water usage (ACPA, 1995).
Figure 7.75: Typical Truck Mixer Figure 7.76: Large Truck Mixer
Drum rotation (see Video 7.12) is used for two purposes: mixing and agitation. High speed rotation (on the order of
12 - 15 rpm) is used to mix PCC ingredients into a homogenous material. This type of mixing typically takes
between 50 and 100 revolutions depending upon PCC characteristics and environmental factors. After this period of
mixing, the PCC is usually required to meet at least 5 of the 6 homogeneity specifications listed in Table 7.9.
Samples for these specifications should be taken from widely separated portions but should also come from the
middle 15 - 85 percent of the load so as not to be influenced by beginning and end of load abnormalities.
Table 7.9: Ready-Mix Concrete Homogeneity Test Requirements from AASHTO M 157 and ASTM C 94
Metric English
Slump
Low speed rotation (about 2 rpm) is used to agitate the PCC to (1) maintain its homogeneity and (2) prevent slump
loss while in transit. Truck mixers are equipped with a revolution counter to help maintain tight control over the total
number of drum revolutions. Mixing, which is generally short in duration, is usually planned for a specific time or
place. If not mixing, truck mixers usually operate in the low speed agitation mode. Mixing is typically done using
one of the following three methods (NRMCA, 2002):
● Mixing at the batching facility. The drum is turned at high speed (12 - 15 rpm) for about 50 revolutions
while at the production facility, which allows for a quick check of batch characteristics. The PCC is then
agitated (< 2 rpm) while in transit to the paving site.
● Mixing in transit. The drum is turned at medium speed (about 8 rpm) for 70 revolutions while driving to
the job site. The PCC is agitated (< 2 rpm) until discharge.
● Mixing at the paving site. The PCC is agitated (< 2 rpm) while in transit to the paving site. Upon arrival,
the PCC is mixed (12 - 15 rpm) for 70 to 100 revolutions, or about five minutes.
In general, short times between mixing and placement can better avoid the problems of premature hardening and
slump loss that result from potential delays in transit. Regardless of the mixing mode, PCC is a perishable
construction material. First, if it begins to set before being placed and consolidated it is of little use. Second, if it is
mixed and agitated excessively it can loose its air entrainment or the effects of certain admixtures can diminish.
Therefore, a typical specification will require that ready mixed PCC delivered to the paving site meet the following
criteria:
● A minimum time (often 1 to 2 hours) between the time at which when the mixing water was introduced to
the portland cement and aggregates and discharge at the site.
● A maximum number of revolutions (typically around 300) between the time at which when the mixing
water was introduced to the portland cement and aggregates and discharge at the site.
haul distances are short and mixing requirements are simple. Their chief advantages are:
● Quick Loading and discharge. Unlike truck mixers, which have a relatively small loading hopper and
discharge chute, the loading and discharge areas for a dump truck are quite large.
● Quick cycle times. Because of their quicker loading and unloading times, dump truck cycle times are
shorter than mixing truck cycle times, thus requiring fewer trucks to maintain a particular delivery rate.
● Lower maintenance/cost. Fewer moving parts and greater accessibility make dump trucks less expensive
to maintain and use than truck mixers.
Because they are not designed to transport PCC, dump trucks also have some severe disadvantages, which limit their
use:
● No mixing/agitating ability. Dump trucks can only be used with central mixed PCC. Further, the lack of
agitation over long haul distances may allow segregate and excessive slump loss.
● No integral cover. Although most dump trucks can be covered with a tarp, the tarp is not integral to the
bed and usually allows some water into the bed. Thus, when operating in the rain, rainwater can
unintentionally increase the water-cement ratio of the transported PCC. Additionally, hot weather may
cause excessive water evaporation, which can also change the water-cement ratio.
● Limited placing ability. Dump trucks can only place PCC at their discharge point. They are not practical
on small jobs (such as sidewalks) nor can they place material in confined areas.
Figure 7.77: End Dump Truck Discharging PCC in Front of Figure 7.78: End Dump Trucks Discharging PCC into
a Paver Placer/Spreaders
There are several mix transport considerations or best practices that are essential to maintaining PCC characteristics
between the production facility and the paving site. These considerations can generally be placed into four
categories:
● Water management
● Operation synchronization
1. Inadequate or improper material mixing. If the order of material batching is not carefully controlled, it is
possible that they will be introduced in such an order that they do not adequately mix. Typically this
happens when liquid admixtures are not mixed in with the water and fine aggregate, or when the portland
cement and mixing water are introduced simultaneously. Also, certain admixtures should not come into
contact with one another until they are in the mixing drum.
2. Head packs. A head pack occurs when fine aggregate and portland cement become lodged in the drum
entrance and are not mixed with the rest of the PCC. Head packs need to be detected during charging
because they can frequently break away during discharge and end up in the finished pavement without
being noticed.
3. Cementitious balls. These are 60 - 75 mm (2.4 - 3 inch) balls of fine aggregate, portland cement and
perhaps some coarse aggregate that do not get thoroughly mixed with the rest of the PCC. They are more
prevalent in batches mixed for a small number of revolutions and can generally be avoided by a proper
material loading sequence (one that starts loading water before the other ingredients and one that starts
loading coarse aggregates before fine aggregates).
incorporated into the pavement causing a potential weak spot (ACPA, 1995). Water is used to maintain cleanliness
at three different stages of mix transport (ACPA, 1995):
● At the production facility after loading (called "wash off"). Applicable to mixing trucks, the driver should
"wash off" the back of the mixer to prevent buildup of materials. The driver must take care to minimize
water entry into the drum. Generally, a consistent amount of water (such as 20 liters or 5 gallons) is used
to wash off the back hopper and fins because this water will run off into the drum. If the volume is known,
it can be counted as part of the mixing water.
● After unloading at the paving site (called "wash down"). Applicable to mixing trucks, the chute and
discharge hopper should be washed to prevent buildup of materials. Wash down locations need to be
determined in advance because wash water should not be discharged into catch basins, road ditches or
environmentally sensitive areas. PCC left in the drum after discharge can be either washed out or recycled.
● At the production facility at the end of the day (called "washout"). The drum/bed should be washed out at
the end of the day to prevent material buildup. Prior to loading the next day, the drum should be run
discharged or the bed dumped to eliminate any remaining water. See Figure 7.79.
PCC performance is closely tied to its water-cement ratio and the amount of mixing water. Therefore, adding water
to the mix between batching at the plant and final placement and consolidation must be a conscious decision that
carefully considers all the potential consequences. All water inputs to the mix should be closely controlled, or
managed. As an example, truck mixers are equipped with water tanks for water addition and cleaning, so it may be
tempting to use this water to adjust PCC slump on the fly to account for a long haul time. If this type of addition is
not in accordance with the water management plan (and it most often is not), avoid it - excess water can result in a
multitude of untraceable problems such as excessive slump, mix segregation, poor finishing characteristics, bleeding
and scaling. They key to water management is to use a consistent process and keep careful control of all water
sources that go into the mix (ACPA, 1995).
1. PCC should be unloaded quickly when it arrives at the paving site. This will minimize the time available
for water evaporation and excessive agitation.
2. PCC should be unloaded as close as possible to its final placement spot. This will prevent mix segregation
that can occur if PCC is moved excessively. Paving machines are not designed to move PCC but rather to
finish it.
Ideally, all operations are designed to meet optimal PCC placement rates because a consistently moving PCC paver
can, in general, produce a smoother pavement than one that must constantly start and stop while trying to match its
paving rate to the material delivery rate. Truck transport should be planned such that the PCC transport rate
(expressed in m3/hr or yd3/hr) closely matches plant production rate and paving rate. Some factors to consider are:
● Truck type.
Traffic plays a large role in PCC delivery rates because it affects truck speed. Especially in congested urban areas,
heavy and/or unpredictable traffic may substantially increase or at least vary truck travel time. As truck travel time
increases more trucks are needed to provide a given PCC delivery rate. Additionally, PCC usually must be delivered
within a specified amount of time. Therefore, as traffic gets worse, trucking costs increase. Additionally, the
unpredictability of traffic may result in either long paver idle times as it waits for the next truckload of PCC or large
truck backups as several trucks all reach the paving site or production facility at the same time.
In sum, synchronization should be the goal but it is often difficult to achieve (based on varying paving rates, haul
time and traffic) and may result in paving inefficiency and degraded PCC quality.
8.3 Summary
Mix transport can have a large impact on rigid pavement construction quality and efficiency. Mix characteristics
such as homogeneity, workability, water content and air content are all affected by mix transport practices. In
general, there are two types of mix transport trucks: the truck mixer and the non-agitated truck. The most common
one, the truck mixer, hauls the mix in a large rotating drum giving it the capability to actually perform PCC mixing as
well as the ability to agitate the PCC while in transport. Non-agitating trucks, which are typically dump trucks, are
much simpler in design and are often used to transport central mixed PCC over short haul distances. Key
considerations in mix transport are proper mixing and batch at the production facility, truck cleanliness, water
management, unloading timeliness and location, and operation synchronization. If properly managed, mix transport
can successfully move PCC over long distances with no impact to final pavement quality.
● Dowel baskets
When using dowel baskets, the dowels must be properly aligned and the dowel basket firmly anchored to the base
course. The FHWA recommends that the dowel baskets be secured with steel stakes with a minimum diameter of 8
mm (0.3 inches) embedded at least 100 mm (4 inches) in stabilized bases, 150 mm (6 inches) in treated permeable
bases and 250 mm (10 inches) for untreated bases or subgrade. Further, a minimum of 8 stakes per basket is
recommended.
Figure 7.84: Dowel Basket Figure 7.85: Dowel Baskets Placed on a Lean Concrete Base in
Placed on an HMA Base Advance of PCC Placement
Figure 7.88: Tie Bar Insertion Equipment for Side Tie Bars
Figure 7.89: Inserter for Mid-Slab Tie Bars (the area in Figure 7.90: Tie Bar Insertion Cartridge Protruding
which the tie bars are inserted will later be cut as a Upward from the Center of a
longitudinal joint) Slipform Paver
● Manual method
● Mechanical method
method (see Figure 7.91), involves hand-placing the reinforcing steel before the PCC is placed. Since the steel is
located at mid-depth or higher in the finished slab, the reinforcing steel must be supported by small metal or plastic
"chairs" in order to achieve this elevation before PCC placement. These chairs must be strong enough to hold the
reinforcing steel in place during PCC placement, consolidation and finishing.
The typical placement process involves (1) placing the transverse bars (which function only as placement aids) on
chairs (see Figure 7.92), (2) arranging the longitudinal bars on top, and then (3) tying the longitudinal bars to the
transverse bars. Typically, they are tied or clipped to the transverse bars every 1.2 - 1.8 m (4 - 6 ft.) (Burke, 1983).
Figure 7.93 shows reinforcing bars in their final position before PCC placement.
The chief advantage of the manual method is that it allows for easy checking of bar placement, height and lap
distance. However, the manual method is slower and more labor intensive than mechanical methods.
Additionally, CRCP pavements (and JRCP pavements when they were more routinely constructed) can be placed in
two lifts. First, the bottom lift is placed, followed by reinforcing bar placement, and then finished with a second PCC
lift. Although feasible, this method can be costly as it usually requires two passes of the PCC paver. However, some
equipment companies offer pavers capable of placing two lifts simultaneously.
● American Concrete Pavement Association (ACPA). (1995). Construction of Portland Cement Concrete
Pavements. National Highway Institute Course No. 13133. AASHTO/FHWA/Industry joint training.
Federal Highway Administration, Department of Transportation. Washington, D.C.
10.1 Placement
PCC can be placed directly in the desired location by truck or truck attachments (see Figures 7.95 and 7.96), or can be
fed into a placement machine for more accurate and even placement. PCC that is moved excessively once it has been
unloaded from the transport truck will tend to segregate (become less homogeneous).
10.3 Consolidation
Consolidation is the process of making the freshly placed PCC into a more uniform and compact mass by eliminating
undesirable air voids and causing it to move around potential obstructions (such as reinforcing steel). Consolidation is
usually accomplished using long, slender vibration rods called vibrators. Vibrators work by rotating an eccentric
weight which causes the entire vibrator to move back and forth. This movement excites particles within the PCC mass,
causing them to move closer together and better flow around obstructions.
Vibrators can be defined by the amount of energy the impart to the surrounding PCC mass. This energy transmission
is defined by two processes. First, the amount of energy generated by the vibrator is proportional to the size and speed
of the rotating weight. Usually, the size is fixed and the speed is variable. Second, the energy transmitted from the
vibrator to the surrounding PCC mass is related to paver speed (the faster the paver runs, the less time the vibrator has
in a particular volume of PCC) and vibrator location within the PCC mass. All of these factors together comprise and
control the size and shape of an "influence zone" - the volume of PCC mass around a vibrator that receives its energy
(see Figure 7.97). This influence zone is usually conical in shape and varies in size depending on the previously
mentioned factors.
● Too much vibration, either by allowing vibrators to operate too long in one area or by using to high a
vibration rate, can result in (1) non-uniform distribution of coarse aggregate particles, (2) loss of entrained
air, and (3) bleeding (water accumulation on the surface). All of these results can greatly reduce PCC
durability.
● Too little vibration, either by not allowing vibrators enough time to operate in one area or by using to low a
vibration rate, can result in (1) non-uniform distribution of coarse aggregate particles, and/or (2) large air
voids within the PCC mass. Again, either result can greatly reduce PCC durability.
● Vibrator static head (amount of PCC above the vibrator) influences efficiency. Higher static heads will help
push coarse aggregate particles together behind the vibrator as it travels along.
10.4 Finishing
Finishing involves all processes and equipment used to create the final surface finish and texture of fresh PCC.
Generally, finishing can be divided into floating and texturing:
● Floating. A flat surface is run across the PCC in order to eliminate high and low spots, embed larger
aggregate particles beneath the surface, remove slight imperfections and to compact the mortar at the surface
in preparation for texturing (PCA, 1988). Floating can involve a number of different tools and may involve
multiple passes over the same surface.
● Texturing. After floating, fresh PCC is usually quite smooth. In order to create a slip resistant surface for
traffic, a rough pattern is usually imparted by dragging a broom, rough-textured item, or tined instrument
across the surface. Typically, texturing is divided into the following two categories (FHWA, 1999):
❍ Microtexture (Figure 7.98). This is achieved by dragging a section of burlap or artificial turf
behind the paver. Microtexture enhances friction between vehicle tires and the pavement surface,
and enhances safety at low speeds.
❍ Macrotexture (Figure 7.99). This is generally achieved by tining the pavement surface.
Macrotexture permits water to escape from between tires and the pavement surface and enhances
safety at high speeds. Typically, an average texture depth of 0.7 mm (0.03 in) will substantially
reduce both total and wet weather accident rates. Tining practices vary by agency, but many
states require transverse grooves on the order of 3 - 5 mm (0.12 - 0.20 inches) deep, 3 mm (0.12
inches) wide and spaced 12 - 20 mm (0.47 - 0.79 inches) apart (ACPA, 1995). Sometimes the
area over the future joint locations is not textured in order to provide a good sawing and sealing
surface. Some agencies consider microtexturing sufficient and do not macrotexture their rigid
pavements.
WSDOT Macrotexturing
10.5 Curing
Curing refers to the maintenance of satisfactory moisture and temperature within a PCC mass as it sets and hardens
such that the desired properties of strength, durability and density can develop (PCA, 1988). The desired properties of
strength, durability and density are related to the extent of hydration within the PCC mass; the more complete the
hydration, the better a PCC's properties. The extent and rate of hydration depend on two critical construction-
controlled parameters: moisture and temperature. This subsection covers:
● Curing methods
1. Curing compound
3. Wet curing
10.5.1 Moisture
Hydration requires portland cement and water. The extent of hydration is controlled by the limiting ingredient, which
is usually portland cement. However, if any substantial portion of water is lost to evaporation, hydration may be
limited by a lack of water, causing it to slow or virtually stop. Thus, inadequate moisture will inhibit hydration, which
results in a weaker, less durable PCC. Rapid moisture loss will also cause excessive shrinking and cracking.
Therefore, a high relative humidity around a hydrating PCC mass will ensure an adequate water supply for hydration
and limit shrinkage cracking. Generally, some method of curing is specified in order to maintain the relative humidity
within the hydrating PCC at an adequate level.
10.5.2 Temperature
Hydration rate is also dependent upon temperature. Higher temperatures speed up hydration's chemical reactions,
while lower temperatures slow them down. Therefore, temperature will affect PCC strength gain. Often, minimum
ambient temperatures for PCC construction are specified to ensure an adequate hydration rate and thus, strength gain.
Maturity
Since hydration progresses over time, and the rate of this progression is dependent on temperature, it should be
possible to estimate the extent of hydration by tracking time and temperature. "Maturity" is the term used to describe
this concept. Most maturity measures are expressed as a function of the product of curing time and temperature (see
Figure 7.100). For example, the Nurse-Saul expression is:
∆t = time interval
Often, maturity is correlated to PCC strength gain by laboratory testing prior to PCC placement. A non-destructive
maturity measurement can then be used to estimate strength and avoid destructive strength tests during construction.
ASTM C 1074 defines the maturity method as "...a technique for estimating concrete strength that is based on the
assumption that samples of a given concrete mixture attain equal strengths if they attain equal values of maturity
index." The maturity method is useful because it can provide strength estimates of in-place PCC subject to actual
environmental temperatures rather than relying solely on controlled-environment laboratory tests. There are also a
number of significant limitations when using maturity to estimate strength (Mindess and Young, 1981):
● The maturity method requires establishment of strength-maturity relationship in the laboratory prior to any
field measurements. Because different PCC mixes mature at different rates, maturity meters are typically
calibrated to actual compressive strength using laboratory test cylinders. Thus, any change in mix
proportions from the laboratory design used for calibration will require a new calibration.
● Other characteristics affecting PCC strength. Items such as moisture content, portland cement chemical
composition and fineness, and construction practices (e.g., consolidation, finishing, air content) are not
accounted for.
● Maturity only accounts for ambient temperature. In large concrete volumes, the heat of hydration
contributes significantly to the PCC mass temperature, and thus, strength gain. In typical PCC pavements,
which are relatively thin, this heat is quickly lost to the environment and can be ignored.
● Maturity functions are not accurate at low maturities. This is probably because the point at which time
should be measured from is poorly defined. Probably, the best time is not the time of mixing or casting, but
rather the time that the PCC actually begins to gain strength.
● Maturity does correlate well with strength when there are large temperature variations during curing.
Typically, a low initial curing temperature followed by a high temperature will lead to higher strengths,
while the opposite (high followed by low) leads to lower strengths.
In sum, the maturity method is not a physical law, but rather a convenient way to estimate strength gain. In PCC
pavement applications, maturity meters (see Figures 7.101 and 7.102) can be used to estimate the appropriate time for
form removal, joint cutting or opening a pavement to traffic, but should not be entirely substituted for basic laboratory
strength tests.
1. Water curing. Methods that prevent moisture loss and supply additional water to the PCC surface. These
methods usually involve ponding water on top of a slab, continuously spraying a slab with a fine mist or
covering a slab with a water-retaining material such as burlap. These methods are labor intensive and are
generally not used on PCC pavements any more.
2. Sealed curing. Methods that prevent moisture loss but do not supply any additional water. These methods
usually involve placing a waterproof covering over a slab (such as plastic) or using a liquid membrane-
forming chemical compound. Curing compounds are typically formed using resins, waxes or synthetic
rubbers with a dissolved volatile solvent. Once the solvent evaporates, the curing compound forms a near-
impermeable membrane over the PCC. Pigments are often added to curing compounds in order to reduce
(white pigment) or increase (dark pigment) heat absorption. Additionally, pigments allow workers to see
where the curing compound has been applied, which helps to ensure complete coverage.
10.6 Joints
All PCC pavement types use all types of joints, however, CRCP uses longitudinal reinforcing steel in order to limit the
number of transverse contraction joints. This subsection discusses the basics of transverse contraction joint
construction including:
● Joint location
● Joint sealing
10.6.1 Location
Typical joint locations are covered in Module 2, Section 6: Rigid Pavement Types, and are not repeated here.
However, it is important to note that joint locations should be indicated on the construction plans and planned in
advance (see Figure 7.103). Intersection joint locations can be quite complex and should be marked out on the base in
advance (see Figure 7.104).
Figure 7.103: Joint Layout on Base Material Figure 7.104: Joint Layout in an Intersection
● Shrinkage cracking. Since contraction joints are used to control shrinkage cracking, they should be sawed
before slab shrinkage stresses become great enough to cause uncontrolled cracking. See figure 7.105.
● PCC support strength and joint raveling. Sawing must be delayed until the PCC is strong enough to both
support the sawing equipment and to prevent raveling during the sawing operations. See Figure 7.106.
Thus, as the PCC hydrates and strengthens, there is a short window of time in which sawing can occur as illustrated by
Figure 7.107.
● Hot-pour liquid sealants. These sealants are heated up to decrease their viscosity and then poured. Joints
are ready for traffic as soon as the sealant has cooled. About 25 percent of roadway agencies use hot-pour
sealants in transverse contraction joints. Most hot-pour sealants are used in longitudinal joints and low-
traffic PCC pavements. Figure 7.109 shows joints filled with hot-pour sealant.
● Compression seals. These are preformed rubber compounds placed into a joint under compression. After
they are placed, they form a seal by pushing against each side of the joint and are immediately ready for
traffic. Compression seals, commonly called neoprene seals after their primary constituent, are used by
about 21 percent of roadway agencies in transverse contraction joints.
● Silicone sealants. These sealants are silicone polymer compounds that are poured into joints at ambient
temperatures. It generally takes about 30 minutes for them to harden and make the joint ready for traffic.
About 52 percent of roadway agencies use silicone sealants in transverse contraction joints.
Figure 7.109: Joints Sealed with Hot-Pour Liquid Sealant on a Freeway On-Ramp
(normally, joints should coincide with lane divisions as they do near the horizon of this photograph)
10.7 Summary
This section has provided an overview of the basic elements of rigid pavement surface course construction: placement,
consolidation, finishing, curing and jointing. These basic elements are common to both fixed form and slipform
paving; the differences are in the equipment and methods.
● Custom geometry. Forms can be placed in just about any pavement geometry, which allows for multiple changes in
pavement width, smooth curves, blockouts and other abnormalities.
● Better construction staging. Forms can be placed such that staged construction can be used to maintain traffic flow
or intersection use (see Figure 7.112).
● Less expensive equipment and mobilization. Forms and equipment are less expensive than slipform paving
equipment. If paving operations are small enough, this cost savings can more than offset the higher production rates
of slipform paving.
This section presents PCC forms, placement, consolidation, finishing and curing as they are typically done in fixed form
paving. Often more than one of these steps can be performed by the same piece of equipment, such as a vibrating screed,
which serves to strike off and consolidate the fresh PCC, or a traveling carriage paver which can perform all three steps.
● Form depth should be equal to PCC slab thickness. This way PCC slab elevation can be tightly controlled out to
and including the edges.
● Forms should have some base width to provide stability. If they do not, they may fall over when subject to the
loading imposed by the fresh PCC.
● The top of the form should be straight and true. Specifications vary, but generally anything more than 3 mm (0.12
inches) every 3 m (10 ft.) is considered excessive.
● Form ends should be able to lock together. If not, they may come apart when subject to the loading imposed by the
fresh PCC.
● Forms should be attached to the ground (via stake) every 1 - 2 m (3 - 6 ft.). Forms that are not staked may move
and forms that are infrequently staked may bow out between stakes.
● Forms should be cleaned and oiled before use. Dirty, unlubricated forms will cause surface defects in the slab sides
and may stick to the slab during removal.
● Curves less than about 30 m (100 ft.) in radius should be done with flexible forms (e.g., wood) or curved metal
forms. Above 30 m (100 ft.) straight 3 m (10 ft.) long metal forms can be laid in a smooth enough curve.
Form placement (or "setting") should be a careful process that is entirely completed before PCC placement. Forms are quite
easy to adjust before PCC is placed and near impossible to adjust afterwards. Form placement is also crucial to rigid pavement
quality because finishing equipment generally rides on the forms making final pavement smoothness dependent on form
elevation uniformity. Forms can often be removed as soon as 6 to 8 hours after placement (ACPA, 1995).
11.2 Screed/Paver
Fixed form screeds vary in complexity from a simple hand screed (sometimes even as simple as a small piece of lumber) to an
automatic self-propelled combination screed-finisher. Often, these combination machines are referred to as "pavers" because
of their multiple functions. Some typical screed/paver types are listed below:
● Manual screed (Figure 7.116 and 7.117). Consists of just about anything that is straight and flat enough to strike
off PCC at a desired elevation. Figure 7.117 shows a manual screed consisting of a long-handled metal straightedge.
● Roller screed/paver (Figures 7.118 through 7.120). Consists of a steel tube that rotates up to 250 rpm in the
opposite direction of movement. The tube strikes off PCC and pushes excess PCC forward by the rolling action (see
Figure 7.119). The roller interior can be outfitted with a small vibrating system. Larger roller screeds can be
mounted on a single unit that travels on fixed form tracks (see Figure 7.120).
Figure 7.119: Roller Screed in Action Figure 7.120: Self-Propelled Roller Screed
● Truss screed (Figure 7.121). Consists of an aluminum or steel blade mounted to a truss frame. Typical truss
screeds can span up to about 22.9 m (75 ft.). Usually, truss screeds are vibrated by either a small gasoline engine or
air pressure.
● Traversing roller tube paver (Figure 7.122 and 7.123). Consists of a traveling carriage mounted on a truss frame.
The carriage travels along the truss frame from one end to the other, with the machine moving forward after each
carriage pass. Typically, the traveling carriage can screed and finish PCC in one pass. The truss frame upon which
the carriage travels can be set to provide flat, parabolic, crowned, super-elevated and tapered surfaces.
Figure 7.122: Traveling Carriage Screed Figure 7.123: Traveling Carriage Close-up
11.3 Consolidation
There are three general consolidation options in fixed form paving:
1. Omit (not recommended). Sometimes, on very small jobs such as the one
pictured in Figure 7.113, vibration is omitted. However, this increases the risk
of leaving large voids in the PCC mass, which creates a honeycomb-like
structure.
3. Screed vibration. Equipment like the vibrating truss screed and the traveling
Figure 7.124: Hand-Operated
carriage screed will vibrate the placed PCC while striking it off. Vibration is
Vibrator
usually accomplished by mechanical or air power.
11.4 Finishing
Fixed form finishing is usually accomplished by hand tools or form riding equipment. Typically, after the PCC has been
screeded it is floated using hand floats (see Figure 7.126 and 7.128) and straightedges (see Figure 7.126) in order to eliminate
any remaining high/low spots and to embed large aggregate particles. Finally, texturing is added using a hand tining tool (see
Figure 7.129) or a traveling carriage attachment.
11.5 Curing
Fixed form construction uses both wet and sealed curing. Wet curing is typically limited to small jobs, while sealed curing,
which is more prevalent, can be used on any job size. Curing seals can be distributed from machines that straddle the finished
pavement or from hand-operated sprayers (see Figure 7.130).
12.2 Paver
Slipform paving is most appropriate for larger jobs that require high
production rates. Particular advantages of slipform paving are (ACPA, 12.3 Texturing and Curing Machine
1995):
● Uses low-slump PCC. Low-slump PCC (on the order of 0 - 75 mm (0 - 3 inches)) is necessary so that the fresh
PCC is able to hold its shape once the slipform paver has passed. Low slump PCC can be made with less water
and usually has higher compression and flexural strengths than comparable high slump mixes.
● High productivity. Large jobs generally require high production rates in order to be profitable. Slipform
paving production rates are typically in the range of 65 - 100 m3/hr (85 - 131 yd3/hr) for mainline paving. That
translates into between 70 - 90 m/hr (230 - 300 ft./hr) of 3.66 m (12 ft.) wide, 250 mm (10 inch) thick PCC
surface course.
● Smooth riding surface. Automation and computer control allow slipform pavers to produce very smooth riding
surfaces (IRI on the order of 0.90 m/km or less).
This section presents PCC placement, consolidation, finishing and curing as it is typically done in slipform paving. Most
often, these steps are accomplished by three pieces of equipment: the placer/spreader (used for rough placement), the
concrete paver (used for final placement, consolidation and initial finishing), and the texturing and curing machine.
These machines usually travel together in series down the length of the project.
12.1 Placer/Spreader
Although not always used, placer/spreaders are quite common. They place a metered supply of PCC in front of the paver
using a series of conveyor belts, augers, plows and strikeoff devices (see Figures 7.131 and 7.132). Using a placer/
spreader allows the contractor to receive material from transport vehicles and place a uniform amount of PCC in front of
the entire paver width, while minimizing segregation.
12.2 Paver
The paver usually performs screeding, consolidation and initial finishing. A typical track-mounted, self-propelled paver
operates at speeds between 1 and 2.5 m/minute (3 and 8 ft./min) (ACPA, 1995). Some pavers are equipped to place
reinforcing steel (if needed), dowel bars and tie rods as well. Figure 7.133 shows the basic slipform paving process as it
occurs underneath the paver. First, an auger spreads the PCC in front of the strike off plate. Second, the strike off plate
(screed) removes excess portions of the auger-placed PCC and brings the slab near its final elevation. Third, the PCC is
consolidated by a group of vibrators. Fourth, a tamper (typically operating between 0 and 150 strokes per minute), if
present, pushes large aggregate particles below the slab surface. Finally, the profile pans level off the slab at the right
elevation and provide initial finishing. The remainder of this section describes this process in more detail.
12.2.1 Screeding
Slipform pavers first use an auger to perform any final material spreading and then strike off the PCC at the correct
elevation using a simple strike off plate, or screed.
12.2.2 Consolidation
After screeding, the paver consolidates the fresh PCC using a series of vibrators (see Figure 7.134). Typically, the most
effective vibrator position is after the strike-off mechanism and at the final slab elevation. Depending upon mix design
and slab depth, vibrators are usually set in the 7,000 - 9,000 vibrations per minute (VPM) range. Vibrators are positioned
next to one another such that their influence zones overlap by about 50 - 75 mm (2 - 3 inches) at normal paver speed
(ACPA, 1995). Gaps between the influence zones (caused by incorrect vibrator settings or excessively fast paver
operation) can cause segregation (ACPA, 1995). Most pavers use fully adjustable vibrator spacing to account for
different conditions and mix types, while still providing adequate influence zone overlap.
Curing is typically done once finishing of an area is complete and the original wet sheen has nearly disappeared. On
tined pavements, curing is usually specified to occur in two passes, one forward and one in reverse, to ensure both sides
of the texture ridges are coated with curing membrane.
13 Rigid - Joints
Joints are common to all rigid pavements and both rigid pavement Major Topics on this Page
construction methods. Even CRCP uses longitudinal joints and periodic
transverse joints. Joints can be formed in two ways. Contraction joints 13.1 Sawed Joints
are most often sawed in after PCC placement. Others such as expansion,
13.2 Other Joints
isolation and construction joints, are created by formwork before the PCC
is placed. Each one of these methods of joint construction has its own method and set of considerations.
● Saw blades
● Sawing equipment
Dry sawing uses abrasive blades that are usually made from fiber reinforced silicone carbide or carborundum since these
types of blades do not require water for cooling (ACPA, 1995). These abrasive blades have less cutting ability than
diamond blades and are usually used on PCC with softer aggregate (ACPA, 1995). When cutting, these blades will wear
down over time making it important to periodically check the sawed depth and saw blade diameter to ensure adequate joint
depth (ACPA, 1995). Additionally, unless water is used to prevent it, these types of blades will generate sawing dust.
13.1.3 Procedure
Joint sawing is generally accomplished in two passes. First, an initial thin cut is made to control shrinkage cracking. The
timing of this joint cut is critical in order to avoid permanent slab damage. Later on, a second, wider cut is made over the
first cut in order to make the joint wide enough to accommodate joint filler material. Some joint cutting considerations are:
● Sawing window. This prevents permanent slab damage due to uncontrolled shrinkage cracking, loading an
inadequately set slab, or spalling the cut joint of an inadequately set slab.
● Proper joint location. Contraction joints must be cut over the internal dowel bar locations. If these locations are
not marked, joints may not be cut at the dowel midpoints, which can result in excessive slab and dowel bar
stresses.
● Joint cleaning. Joint cutting will result in a fair amount of debris generation. After sawing, this debris should be
washed out to prevent it from contributing to faulting or joint stresses. Before sealing, joints should be abrasively
cleaned to provide a good textured surface with which the joint sealant can bond. Finally, the residue from the
abrasive cleaning should be blown out to provide a clean bonding surface.
● Ensure adequate PCC is available to finish the last slab of the day. The construction joint is set before all the
PCC is placed. If truck delivery is stopped to soon, the temptation may be to use the lower-quality PCC that has
been pushed in front of the screed to fill the remaining volume. This low-quality PCC may contain little portland
cement, excess water, low air and/or segregated aggregate.
● Ensure proper consolidation. Typically, the paver does not traverse the construction joint. Therefore, the
construction joint is not consolidated by the paver vibrators and it should be manually consolidated.
● Dowel placement should not segregate or cause air voids in the adjacent PCC. Sometimes, dowel bars are placed
after the PCC has been placed up to the header. If dowel bars are pounded in by hammer, the resultant vibrations
may cause air pockets or segregated aggregate.
1 INTRODUCTION
Quality assurance and specifications are two topics vital to pavement design, construction and
performance. Quality assurance, loosely defined as the actions undertaken to ensure a product will
perform satisfactorily in service, is a part of many activities including planning, design, plans and
specifications, contracts, construction and maintenance. Specifications are used by a buyer to
communicate product requirements to a seller or potential seller. In pavement construction, they
delineate what is to be built and how it is to be paid for.
● Quality control
● Acceptance sampling
● Independent assurance
Specifications are then covered in a general sense, while the last section examines statistical acceptance
specifications in some detail.
http://training.ce.washington.edu/WSDOT/Modules/08_specifications_qa/08-1_body.htm4/2/2008 6:40:34 PM
8.2 QA & Specifications - Quality Assurance (QA)
2 Quality Assurance
Quality has become one of the most important Major Topics on this Page
consumer decision factors in the selection among
competing products and services (Montgomery, 2.1 Quality Control
1997). This is true not only for individual consumers 2.2 Acceptance
but also for large corporations, government
organizations and the taxpaying public as a group. 2.3 Independent Assurance
Thus, quality is a key factor in pavement construction. 2.4 Summary
But what is “quality”? In its broadest sense, quality is
a degree of excellence: the extent to which something is fit for its purpose. In the narrow sense, product
or service quality is defined as conformance with requirement, freedom from defects or contamination,
or simply a degree of customer satisfaction. In quality management, quality is defined as the totality of
characteristics of a product or service that bears on its ability to satisfy stated and implied needs (IQA,
2001). This section discusses a few of the more visible aspects of quality: quality assurance, quality
control, acceptance and independent assurance. A deeper discussion of quality as a management
philosophy is beyond the scope of this Guide but has been addressed by countless textbooks and
seminars throughout the world.
AASHTO and the FHWA subscribe to definitions that designate “quality assurance” as an all-
encompassing term, to include “quality control”, “independent assurance” and “acceptance” as its three
key components (see Figure 8.1). This Guide uses the same definitions (listed below) for consistency
(TRB, 1999):
● Quality assurance. All those planned and systematic actions necessary to provide confidence
that a product or facility will perform satisfactorily in service. Quality assurance addresses
the overall problem of obtaining the quality of a service, product, or facility in the most
efficient, economical, and satisfactory manner possible. Within this broad context, quality
assurance involves continued evaluation of the activities of planning, design, development of
plans and specifications, advertising and awarding of contracts, construction, and
maintenance, and the interactions of these activities.
● Quality control. Those quality assurance actions and considerations necessary to assess
production and construction processes so as to control the level of quality being produced in
the end product. This concept of quality control includes sampling and testing to monitor the
process but usually does not include acceptance sampling and testing. Also called process
control.
● Acceptance. Sampling, testing, and the assessment of test results to determine whether or not
the quality of produced material or construction is acceptable in terms of the specifications.
● Independent assurance. A management tool that requires a third party, not directly
responsible for process control or acceptance, to provide an independent assessment of the
product and/or the reliability of test results obtained from process control and acceptance
testing. The results of independent assurance tests are not to be used as a basis of product
acceptance.
As defined here, quality control, acceptance and independent assurance are wholly separate components
of quality assurance. It is difficult, if not impossible, to influence one component, solely through actions
within another component. For instance, acceptance procedures are essentially monitoring methods
used to determine whether or not a particular process is meeting quality standards. As such, they can be
used to accept or reject material based on its quality but they should never be used as a method to
control or improve quality; quality will not necessarily improve based on increased or stricter
monitoring alone. In short, no amount of inspection changes the quality of a product or service (IQA,
2001). Quality control (process control) is used to control and systematically improve quality.
Furthermore, independent assurance test results should not be used for acceptance or quality control. If
they are, the tests are no longer independent and should not be used as if they were.
In the narrowest sense, quality control seeks to control the level of quality being produced in the end
product. This level of quality consists of two key components:
1. Target value. This is the goal set for a certain material characteristic. As a minimum it
should conform to standards and be achievable. For example, on a specific contract the
specified in-place HMA density might be 92-percent of TMD. Therefore, a contractor may
set an in-place density target value at 93-percent of TMD. This meets (in fact, it exceeds) the
standard.
2. Variability. This describes how much a process varies from item-to-item (or location-to-
location). A process that meets the target value on average but is highly variable will result in
pavement locations where the material characteristic is unacceptable.
Thus, a quality control program consists of (1) the actions and considerations necessary to assess
production and construction processes and (2) setting the end product target value and controlling
variability. In order for a quality control program to be effective it should (1) base actions and decisions
on measurable results and (2) be statistically valid.
Quality control actions and considerations should be based on objective evidence and not subjective
opinion. Consider the difference between changing a roller pattern based on (1) an inspector’s opinion
that compaction is inadequate or (2) consistently low-density test results. While the first option is
subjective, the second option results in unequivocal evidence from which confident actions can be
taken. This does not mean that experience and expertise are not valuable but rather that they should be
used to determine what measurements to take and how to improve the process. A training manual for
Komatsu Ltd. in Japan says it this way (Walton, 1986):
“The first step in quality control is to judge and act on the basis of facts. Facts are data such as
length, time, fraction defective and sales amount. Views not backed by data are more likely to
include personal opinions, exaggeration and mistaken impressions. Data volume has nothing to
do with accuracy of judgment. Data without context or incorrect data are not only invalid but
sometimes harmful as well. It is necessary to know the nature of that data and that proper data be
picked as well.”
Since variability can only be accurately described and evaluated through statistical methods, quality
control involves statistics. Fortunately, these statistics are relatively straightforward. For instance,
control charts are frequently used to analyze production results (e.g., production, density, strength).
2.2 Acceptance
"Acceptance" is a formal procedure used to decide whether work should be accepted, rejected, or
accepted at a reduced payment (Freeman and Grogan, 1998). Specifically, it is the monitoring method
used to determine whether or not a particular process is meeting quality standards; it is not, however, a
form of quality control. Acceptance procedures simply accept or reject things based on their quality;
they do not ensure proper quality standards. Thus, they should never be used as a method to control or
improve quality; process controls are used to control and systematically improve quality (Montgomery,
1997).
Acceptance procedures can take one of the following three broad forms: (Montgomery, 1997):
1. Accept with no inspection is generally used when there is no economic justification to look for
defective units or material.
2. 100 percent inspection is generally used where components or material are extremely critical
and passing any defective components or material would result in an unacceptably high failure
cost.
3. Acceptance sampling is generally used when there is some economic justification to look for
defective material and either (1) some small finite percentage of defective material is
acceptable or (2) it is not economical or practical to use 100 percent inspection. Acceptance
sampling uses statistics to estimate information about a large amount of material from a small
random sample.
Pavement construction typically uses acceptance sampling and thus, many specifications are statistically
based. Section 3, Statistical Acceptance Specifications takes a more in-depth look at these types of
specifications.
Quality control and acceptance tests are often performed by a multitude of laboratories (agency,
contractor and independently owned) within a particular county, region or state. Even if all tests are
performed at the same central laboratory, they are at least conducted at different times. Thus,
independent assurance insures that sampling variability (the variation in sample characteristics from
sample-to-sample that is attributable to variations in sampling technique), testing variability (the lack of
repeatability of test results - operators, equipment condition, calibration, and test procedure all
contribute to testing variability) and the differences in these variabilities between laboratories and over
time do not unduly influence testing measurements. For example, the same sample tested by different
laboratories should give roughly the same result.
2.4 Summary
In simple terms, quality is the extent to which something is fit for its purpose. Quality assurance
involves those actions necessary to provide confidence that a product or facility is fit for its purpose.
These actions involve quality control, acceptance or independent assurance. Quality control (or process
control) involves assessing and adjusting production/construction processes so as to control the level of
quality being produced. Acceptance is a formal procedure instituted by the specifying agency used to
decide whether work should be accepted, rejected, or accepted at a reduced payment. Finally,
independent assurance provides an assessment of the product and/or the reliability of test results
obtained from process control and acceptance testing. Together, these three elements form the basis of
quality assurance.
3 Specifications
Specifications are used for three primary purposes: Major Topics on this Page
There are four types of specifications generally recognized in the construction industry: proprietary
product, method, end-result and performance.
This section is largely taken from a series of three articles written for HMAT Magazine and a
Washington State DOT research report as listed below:
● Newcomb, D.E. and Epps, J.A. (Jan/Feb 2001). Statistical Specifications for Hot Mix
Asphalt: What Do We Need to Know? HMAT, vol. 6, no. 1. National Asphalt Pavement
Association (NAPA). Landham, MD. (first in a series of 3 articles)
● Newcomb, D.E. and Epps, J.A. (Mar/April 2001). Statistical Specifications for Hot Mix
Asphalt: What Do We Need to Know? HMAT, vol. 6, no. 2. National Asphalt Pavement
Association (NAPA). Landham, MD. (second in a series of 3 articles)
● Muench, S.T. and Mahoney, J.P. (2001). A Quantification and Evaluation of WSDOT’s Hot
Mix Asphalt Concrete Statistical Acceptance Specification. WA-RD 517.1. Washington State
Department of Transportation, Transportation Center (TRAC). Seattle, WA. (http://www.
wsdot.wa.gov/ppsc/research/CompleteReports/WARD517_1HotMixAsphalt.pdf).
Although widely used, method specifications have several key disadvantages. First, they tend to stifle
contractor innovation. Since a contractor’s only motivation is instructional compliance, there is virtually
no incentive to develop better, more efficient construction methods. Second, since they are not
statistically based and 100 percent compliance is usually not possible, method specifications usually
required “substantial compliance,” a purposely vague and undefined term that can lead to disputes.
Finally, spot checks of material quality, which are often used in method specifications, do not reflect
overall material quality because they are taken from subjectively determined non-random locations.
Since they are not random, these spot checks have no statistical validity and therefore do not reflect
overall material quality.
Despite their flaws, method specifications are still widely used on the local agency level (e.g., counties,
small cities, towns, etc.). In general, this is because they are familiar, straightforward to write and can
be implemented with minimal agency involvement. Local agencies often lack the expertise and
resources required to use statistical specifications or warranties.
An end-result specification is one in which the final characteristics of the product are stipulated, and the
contractor is given considerable freedom in achieving those characteristics. In their roughest form, they
specify minimum, maximum or a range of values for any given characteristic and base acceptance on
conformance to these specifications. For instance, they may state a minimum layer thickness or a range
of in-place air voids. However, since it is impractical to measure every square foot of constructed
pavement, end-result specifications use statistical methods to estimate overall material quality based on
a limited number of random samples. Therefore, end result specifications improve on methods
specifications in two key areas: (1) they shift the focus away from methods and on to final product
quality and (2) they do not rely on the nebulous "substantial compliance" because they clearly define
acceptable quality.
Today, most large state and Federal pavement contracts use statistically based end-result specifications
that incorporate some elements of method specifications (usually used to guard against early failure of
the product). These end-result specifications are often referred to as a "quality assurance specifications",
"QA/QC specifications" or "QC/QA specifications". Essentially, these specifications specify the end
results and also specify certain minimum construction method requirements (e.g., temperatures below
which paving is not allowed, descriptions of initial test sections, minimum number of rollers, conditions
under which the agency may halt paving operations, etc.).
End result specifications assign pavement construction quality to the contractor, they define the desired
final product, and they allow the contractor significant latitude in achieving that final product. This
leads to innovation, efficiency, and lower costs. However, these specifications and their statistical
sampling requirements are often too complex and resource intensive to be used at the local agency level.
contractor warrants the pavement for performance over a specific amount of time (usually 2 to 7 years
although some have been done up to 20 years). During the warranty period, any defects attributable to
construction are repaired at the contractor’s expense.
There are two basic types of pavement construction warranties, the second of which, performance
warranties, is what is typically meant when referring to a "warranty specification" (Huber, 2002):
1. Materials and workmanship. Almost all HMA construction is covered by a short duration
(usually 1 year) materials and workmanship warranty. This type of warranty assigns risk to
the contractor for following agency specifications in regards to materials and workmanship. If
a problem or defect is detected within the warranty period, the agency usually uses a forensic
analysis to determine the cause. If it is determined that specification non-compliance caused
the problem, it is repaired at the contractor's expense. Otherwise, the agency assumes repair
costs. This type of warranty is almost universal, rarely collected on and is usually covered by
sureties at no additional charge to the contractor.
2. Performance. This type of warranty assigns a large portion of the pavement performance risk
to the contractor. During the warranty period the agency monitors pavement performance and
any unacceptable performance attributable to construction is remedied at the contractor's
expense. Because the contractor assumes greater risk he/she is allowed to control most
construction aspects.
For specifying agencies, warranties represent an advancement in specifications over end result
specifications because they can specify actual pavement performance rather than material characteristics
that are only indicative of pavement performance (Table 8.1 gives an example of performance standards
used by the Indiana DOT). Thus, warranty specifications are best able to align the sometimes competing
influences of economic incentive, innovation, customer requirements and pavement quality. This
alignment, when achieved, allows market forces and economics, rather than specifications alone, to
drive pavement quality.
Table 8.1: Indiana DOT Pavement Performance Thresholds for a Five Year
Warranty Specification (from Andrewski, 2002)
Threshold Value
Parameter
(contractor must take action above this value)
Longitudinal Cracking 5.5 m (18 ft.) per 152.5 m (500 ft.) section
Although warranty specifications are being used in other countries, most notably in Western Europe,
they are used somewhat sparingly in the United States for several reasons. First, the industry has been
somewhat reluctant to change. Second, the Federal Government places severe legal restrictions on
warranty use. Third, performance tests need further development so they can accurately and fairly
invoke warranty clauses. Finally, the surety industry may have the largest say. Contracting agencies
usually limit their risk by requiring a bonded contractor. Bonding agencies may or may not accept the
risk associated with a 2 to 7 year performance warranty. They are especially wary since contractors
typically have no say in pre-construction pavement design, and no control over post-construction
pavement use (Hancher, 1994).
3.5 Summary
Each of the three basic types of specifications (method, end result and warranty) is appropriate for
certain types and sizes of jobs. While method specifications are still adequate for small jobs and local
agency work, end result specifications are typically used on larger state and Federal contracts and
warranty specifications are beginning to be used on selected large jobs.
4 Statistical Acceptance
Specifications
As of 2002, about 35 States are using some form of Major Topics on this Page
statistical acceptance. This section provides a practical
background for statistical acceptance specifications by 4.1 Acceptance Plan Basics
establishing what statistical acceptance specifications are,
why they are used, and what their applications and 4.2 Acceptance Plan Components
limitations are. A statistical acceptance specification is 4.3 Acceptance Plan Categories
only one way of accepting or rejecting work; other methods
such as accepting with no inspection and full inspection are 4.4 summary
also commonly used, however each method has its specific
areas of applicability. Regardless of the details of accept / reject methods, they are all only designed to
monitor or audit a process. They do not provide any means of process or product improvement and should
not be used to achieve such goals.
This section is largely taken from the Washington State DOT research report listed below:
● Muench, S.T. and Mahoney, J.P. (2001). A Quantification and Evaluation of WSDOT’s Hot Mix
Asphalt Concrete Statistical Acceptance Specification. WA-RD 517.1. Washington State
Department of Transportation, Transportation Center (TRAC). Seattle, WA. (http://www.wsdot.
wa.gov/ppsc/research/CompleteReports/WARD517_1HotMixAsphalt.pdf).
● Random sampling
1. Accept with no inspection is generally used when there is no economic justification to look for
defective units or material.
2. 100 percent inspection is generally used where components or material are extremely critical and
passing any defective components or material would result in an unacceptably high failure cost.
3. Acceptance sampling is generally used when there is some economic justification to look for
defective material and either (1) some small finite percentage of defective material is acceptable or
(2) it is not economical or practical to use 100 percent inspection. Acceptance sampling uses
statistics to estimate information about an entire lot from a small random sample.
Of these three approaches, pavement construction typically uses acceptance sampling because excessive out-
of-specification (defective) material will substantially affect long-term pavement performance but it is
neither practical nor economical to inspect everything. Basically, acceptance sampling uses random
sampling to make quality and material property estimates about a large amount of material. This highlights
two key concepts involved in the effective use of acceptance sampling: (1) acceptance sampling only
estimates material properties, and (2) acceptance sampling depends on random sampling.
and drawing conclusions about an entire lot is invalid. Thus, any exercise in judgment as to whether or not a
sample will produce a good, failing, or average test result nullifies the random sampling assumption and
therefore the assumptions on which a statistically oriented specification is based (Bowery and Hudson,
1976).
Pavement construction acceptance sampling uses a modified version of random sampling that satisfies the
random sampling assumption. In true random sampling any location or item within a lot must have an equal
probability of being sampled. In rare instances, this results in all samples being clustered together through
random chance (Freeman and Grogan, 1998). Although this sample clustering is statistically valid, pavement
specifications usually strive to ensure samples are spread more evenly throughout the lot. Therefore,
stratified random sampling, which involves dividing lots into several equal-sized sublots, is generally used
(Weed, 1982 as cited in Freeman and Grogan, 1998). Each individual sublot is still randomly sampled, but
stratification ensures that samples are more evenly spread throughout the lot. Stratified random sampling
conforms to the requirements of random sampling as long as three rules are obeyed (Weed, 1982 as cited in
Freeman and Grogan, 1998):
WSDOT uses stratified random samples for its HMA and PCC acceptance sampling. For instance, HMA
in-place density lots are divided into five equal-sized sublots and one random sample is taken from each
sublot (WSDOT, 2002).
In summary, acceptance sampling is only one of several acceptance procedure options. As such, it does not
provide any direct form of quality control; it simply accepts or rejects lots. Acceptance sampling only makes
estimates of actual lot properties and is dependent on random sampling to make these estimates. In order to
ensure a uniform distribution of samples, the pavement construction industry typically uses stratified random
sampling.
viable overall plan. Correct application involves proper implementation of the following acceptance
sampling components:
● Quality characteristics
● Specification limits
● Statistical model
● Risk
● Pay factors
Decisions regarding these components will significantly impact final acceptance plan performance.
In attribute sampling, each sample is inspected for the presence or absence of one or several attributes (often
called quality characteristics). Measurements used to detect these quality characteristics are not retained.
Rather, they are compared to a standard then recorded as either passing or failing. An aggregate fracture test
is an example of attribute sampling. Aggregate is accepted or rejected based on a minimum quality
characteristic of one fractured face on a specified percentage of the material. The actual percentage of
fractured face is not recorded; instead, a simple pass-fail record is used.
In variable sampling, measured quality characteristics are used as continuous variables, which means that,
unlike attribute sampling, measurement values are retained. Because these values are retained rather than
converted into a discrete pass-fail criterion, variable sampling plans retain more information per sample than
do attribute sampling plans (Freeman and Grogan, 1998). This means that compared to attribute sampling, it
takes fewer variable samples to get the same information. Because of this, most statistical acceptance plans
use variable sampling.
However, variable sampling does have disadvantages. Foremost, variable sampling plans are predicated on a
known distribution of the measured property. Therefore, most variable sample plans assume a normal
distribution of the measured property. For instance, acceptance testing for HMA compaction often assumes
that in-place HMA densities (the measured property) are normally distributed. If this normal distribution
assumption is not true then the resulting estimates of lot quality will be wrong. Fortunately, construction-
related lot characteristics are usually normally distributed (Markey, Mahoney, and Gietz, 1994; Aurilio and
Raymond, 1995; Cadicamo, 1999).
Therefore, although both attribute and variable sampling are used in pavement construction, variable
sampling is more prevalent because it provides more information per sample and its necessary assumption of
a normal distribution of the quality characteristic is usually satisfied.
Construction contracts, including pavement contracts, generally require full payment at substantial
completion. However, since the constructed pavement performs for many years after construction,
contracting agencies usually use some predictive method to relate construction quality to long-term pavement
performance. Statistical acceptance plans typically accomplish this by choosing construction quality
characteristics that are most predictive of pavement performance. These quality characteristics typically
include mix properties (such as aggregate gradation, HMA asphalt content and PCC slump), HMA in-place
density, PCC strength, and pavement smoothness (Schmitt, et al., 1998).
Quality characteristics must also be chosen to avoid correlation with one another. If not carefully selected, a
change in one quality characteristic (such as aggregate gradation) could result in a change in another quality
characteristic (such as HMA VMA or PCC cement content). Lin, Solaimanian, and Kennedy (2001) point out
that this correlation will always cause biases in pay factor determination. In the gradation-VMA instance
mentioned previously, the bias occurs because a poorly graded aggregate would be penalized not only by
lower pay for poor gradation but also by lower pay for the correlated poor VMA. Bias in the opposite
direction (higher pay for well-graded aggregate) is equally likely. Therefore, biased pay factors will unfairly
penalize either the agency or the contractor.
Acceptance sampling determines overall construction quality by measuring quality characteristics. Proper
selection of these characteristics ensures that (1) their quality accurately reflects construction quality, which
should in turn reflect long-term pavement performance and (2) they are relatively independent of one another
so that final pay is not biased in either direction.
Specification limits must be based on sound engineering judgment and sound statistical analysis.
Specifically, engineering judgment is used to establish a target value for each quality characteristic and
statistical analysis is used to establish an acceptable range around the target value. This range is used to
account for the various sources of variability inherent in producing and testing HMA. Specifically, there are
four types of variability to consider: (Hughes, 1996):
1. The material’s inherent variability is the true random variation of the material and is a function of
material characteristics alone. A contractor’s manufacturing and construction process cannot
control this variability.
3. Testing variability is the lack of repeatability of test results. Operators, equipment condition,
calibration, and test procedure all contribute to testing variability. A contractor’s manufacturing
and construction process cannot control this variability.
The total variability is then the sum of the material, sampling, testing and manufacturing/construction
variability:
Since contractors can only control manufacturing and construction variability, if the sum of inherent material,
sampling and testing variability is greater than the allowable specification band, a potentially large amount of
material will be judged out-of-specification for no contractor-correctible reason. For instance, an asphalt
content specification of the JMF ± 0.1 percent does not make statistical sense because the combination of
inherent asphalt content variability, sampling variability, and testing variability will typically cause test
results to vary by more than ±0.1 percent from the JMF (Hughes, 1996). A more practical approach, which
adequately accounts for material, sampling, and testing variability might specify the JMF asphalt content
±0.5 percent. In sum, specification limits should be tight enough to detect manufacturing and construction
variability, but loose enough to allow a reasonable amount of testing, sampling, and inherent material
variability.
Statistical models all rely on random samples, which provide two pieces of data: (1) the average of the
sample measurements and (2) the variation in sample measurements. Both pieces of data are needed to
estimate the distribution of the measured quality characteristic within a lot (see Figure 8.2).
Note: This distribution represents hypothetical quality characteristic measurement results if an entire lot
were broken down into infinitesimally small sections and the quality characteristic associated with each
section was measured. As stated earlier, this distribution can never be known for certain unless a 100
percent inspection method is used.
1. Use the average of sample measurements only. This method calculates the sample average and
uses this to estimate lot average. It does not calculate sample variation, thus it is unable to
estimate the overall distribution of the quality characteristic within the lot.
2. Use the average of sample measurements and assume typical lot variation. This method estimates
lot average as the calculated sample average and assumes a typical lot variation based on historical
data of the measured quality characteristic. By assuming a typical lot variation, this method can
use the standard normal distribution (a relatively well-understood distribution) to estimate the
overall distribution of the quality characteristic within the lot. This estimate is only accurate if the
actual variation of the quality characteristic within the lot is close to the assumed variation
(Freeman and Grogan, 1998).
3. Use the average of sample measurements and variation in sample measurements. This method
estimates lot average as the calculated sample average and estimates lot variation as the calculated
sample variation. It estimates the overall distribution of the quality characteristic within the lot by
applying the non-central t distribution (Johnson and Welch, 1940).
Methods like #3 are typically preferable because they fully describe the distribution of the quality
characteristic within a lot and make the fewest assumptions. However, methods such as #1 and #2 are still
often used.
“Quality” is then defined as the fraction of the overall quality characteristic distribution that falls within
specification limits. It is usually expressed as either (TRB, 1999):
● Percent defective (PD) – also called percent nonconforming. The percentage of the lot falling
outside specification limits.
● Percent within limits (PWL) – also called percent conforming. The percentage of the lot falling
above a lower specification limit, below an upper specification limit, or between upper and lower
specification limits. PWL is related to PD by the following: PWL = 100% - PD.
To summarize, the statistical model determines how and to what extent the overall quality characteristic
distribution is estimated. Some models are quite simple and only estimate an average quality characteristic
value while other models are more complete and estimate both average and variation, which then provides
the ability to estimate lot quality. Lot quality, expressed as PWL, is simply the fraction of the lot that falls
within specifications. Figure 8.3 presents a summary of the PWL concept and the common approaches to
increase PWL.
The appropriate levels of AQL and RQL are matters of judgment. It would be nice but unrealistic to expect
all material within a lot to meet specifications (PWL = 100). However, some small fraction of defective
material must be permitted due to the unavoidable variability that accompanies any material or production
process (Comisky, 1974 as cited in Freeman and Grogan, 1998). To account for this, AQL should be some
value less than 100 PWL. Additionally, AQL should also be set at a value equal to the maximum amount of
defective material present within the pavement that will not substantially degrade overall road quality
(Freeman and Grogan, 1998). These considerations result in typical AQL values of 90 or 95 PWL.
RQL is generally set much lower than AQL because it should represent a PWL below which the pavement is
essentially worthless to the contracting agency. Typical values of RQL range from 60 PWL down to 30
PWL and often depend upon sample size. If the actual material quality level is between AQL and RQL then
it is often accepted at reduced pay because although defects in the material will degrade overall road
performance they will not degrade it to a point where the pavement has no value.
AQL and RQL are difficult to accurately set. Typically there is not enough data to accurately relate material
quality to final pavement worth. Although current research is addressing this issue (Weed, 1998; Deacon et
al, 2001), most AQL and RQL values seem to be set using a combination of historical data, experience, and
statistical tradition.
4.2.6 Risk
Using samples to make estimates about the quality of a large amount of construction material involves risk;
there is some probability that a random sample will not be representative of the material as a whole, and will
thus be an incorrect estimate of material quality. Therefore, risk is an inherent part of statistical acceptance
plans. An incorrect estimate, or error, and its associated risk can be either of two types:
● Type I error (α risk). Acceptable construction quality will be rejected as unsatisfactory. This is the
contractor’s (seller’s) risk and can result in unnecessary removal and reconstruction of large
pavement sections. There are two types:
❍ Primary type I error (primary α risk). The contractor’s risk that material produced at
AQL will be either rejected or subject to reduced pay.
❍ Secondary type I error (secondary α risk). The contractor’s risk that material produced
at AQL will be rejected.
● Type II error (β risk). Unacceptable construction quality will be accepted as satisfactory. This is
the contracting agency’s (buyer’s) risk and can result in additional maintenance costs, and
premature pavement failure. There are two types:
❍ Primary type II error (primary β risk). The contracting agency’s risk that material
produced at RQL will be accepted at bonus pay.
❍ Secondary type II error (secondary β risk). The contracting agency’s risk that material
produced at RQL will be accepted.
These risks can be calculated and must be balanced. For a given sample size, reducing the likelihood of
accepting poor material usually means increasing the likelihood of rejecting good material and vice versa
(Freeman and Grogan, 1998). To simultaneously reduce both of these risks, the sample plan must make
more accurate estimates. This usually means increasing the sample size, which means higher inspection and
testing costs to the contracting agency. Therefore, the contracting agency will try and achieve an acceptable
balance between sample size (accuracy) and inspection and testing costs.
Selecting the appropriate contractor risk and contracting agency risk is a matter of judgment. However, these
risks should be related to the criticality of the quality characteristic as well as economic considerations
(Freeman and Grogan, 1998). If the failure of a certain material characteristic will render an entire project
useless, then it is a critical material characteristic. Therefore, the probability of accepting poor material (β
risk) should be set quite small. Conversely, if a material characteristic is not critical, then the probability of
accepting poor material (β risk) can be set higher (Freeman and Grogan, 1998). For pavement construction,
the primary α risk is often set near 5 percent and the primary β risk is often set near 10 percent (Cominsky,
1974 as cited in Freeman and Grogan, 1998). As long as these risks are quantified and known in advance,
both parties can account for them in their respective budgets and bids.
The risks involved in a particular acceptance plan are often expressed using an operating characteristic (OC)
curve. An OC curve describes the relationship between a lot’s quality and its probability of acceptance for a
given sample size. Each sample size has a different OC curve. Figure 8.4 shows a WSDOT OC curve for a
sample size of five (n = 5). The better the sampling plan is at estimating actual lot quality, the steeper the OC
Curve. Figure 8.5 shows a much steeper OC curve for a sample size of 50.
Figure 8.4: Example Operating Characteristic (OC) Curve for a Sample Size of 5
Figure 8.5: Example Operating Characteristic (OC) Curve for a Sample Size of 50
First, the pay a contractor can expect for consistently producing material at a particular quality level is not
necessarily the same as the pay factor shown in the specification for that quality level (referred to as the
contractual pay factor).
The WSDOT Standard Specifications (2002) show that for five samples (n = 5), material estimated at AQL (95 PWL)
shall receive a 1.04 pay factor. However, a contractor consistently producing AQL material should expect to receive,
over time, an average pay factor near 1.03 (see note below). The figure below shows this difference between the
specified, or contractual, pay factor and the expected pay factor for the WSDOT specification.
Note: For this report expected pay is calculated assuming a normal distribution of quality (as measured by PWL) about
the quality level in question with a standard deviation of about 19 percent PWL. Results using this model compare
almost identically with results from the OCPLOT simulation software contained in the FHWA’s demonstration Project
89, Quality Assurance Software for the Personal Computer (1996b).
This difference occurs because sampling only estimates actual material quality. Therefore, material
produced at AQL may be estimated by sampling to be either above or below AQL. Over time, sample
estimates of quality will be normally distributed about a mean equal to the actual material quality. Figure 8.5
shows how this looks for material produced at AQL under a typical acceptance plan using an ideal normal
distribution of samples (the large number of lots with estimated quality at 100 PWL occur because 100 PWL
is the maximum achievable quality, therefore the entire portion of the normal distribution that falls above 100
PWL is represented by the 100 PWL value). Since each lot receives a contractual pay factor, Figure 8.7
shows the resulting pay factors associated with Figure 8.6. Figure 8.7 shows that material consistently
produced at AQL (95 PWL) will not receive the contractual pay factor (1.04) associated with AQL but rather
a lesser pay factor (1.0349 in this example). Simulations run by the FAA (FAA, 1999b) and Weed (1995,
1998) have also shown this type of behavior, which is a characteristic of almost all statistical acceptance
plans that use pay factors.
Figure 8.6: Typical Sample Distribution for Material Produced at AQL for a Hypothetical Project Consisting of
100 Lots
Figure 8.7: Typical Pay Factor Distribution for Material Produced at AQL for a Hypothetical Project Consisting of
100 Lots
Second, material produced at AQL does not always receive a 1.00 pay factor. In the example shown in
Figures 8.6 and 8.7, AQL material produced a 1.0349 pay factor. Therefore, material produced at the
contractually specified quality is paid at a higher rate than the contractually specified price. Conversely, in
acceptance plans that do not include pay factors above 1.00, AQL material could receive a pay factor
significantly less than 1.00. In these cases, material produced at the contractually specified quality is paid at
a lower rate than the contractually specified price.
Pay factors relate material quality to actual pay. An ideal pay factor system typically allows bonus pay for
material produced in excess of AQL, pays the contractual price for AQL material, applies a deduction for
material produced between AQL and RQL and rejects material produced at or below RQL. Meeting all four
of these goals is quite difficult because expected pay is often different than contractual pay and providing
bonus pay for material produced in excess of AQL may lead to expected pay above the contractual price for
AQL material.
● Single specification limit, single decision criterion. Single specification limits are used when a
material must be controlled above a minimum or below a maximum. An AQL is set and material is
either accepted or rejected based on it. There is no pay factor provision.
● Single specification limit, dual decision criteria. An AQL and RQL are set. Material at or above
AQL is accepted at full or bonus pay while material below RQL is rejected. Material with an
estimated quality level between AQL and RQL is usually accepted at reduced pay according to a
pay scale.
● Dual specification limits, single decision criterion. Dual specification limits are used when a
material must be controlled within a range of values. The percent of material between these values
is calculated as the PWL and compared to the AQL. Material is then either accepted or rejected.
There is no pay factor provision.
● Dual specification limits, dual decision criteria. An AQL and RQL are set. Material at or above
AQL is accepted at full or bonus pay while material below RQL is rejected. Material with an
estimated quality level between AQL and RQL is usually accepted at reduced pay according to a
pay scale.
● Single specification limit, dual decision criteria is used for HMA compaction and PCC
compressive strength.
● Dual specification limits, dual decision criteria is used for HMA asphalt content, HMA
aggregate gradation and PCC air content.
4.4 Summary
Statistical acceptance specifications use acceptance sampling to audit construction quality. Acceptance
sampling is a powerful audit tool because it allows reasonably accurate estimates of lot quality to be made
based on test results from a relatively small number of random samples within the lot.
In general, there are seven components of a statistical acceptance specification that define its performance:
(1) the sampling type, (2) the quality characteristics, (3) the specification limits, (4) the statistical model, (5)
the quality level goals, (6) the risk inherent in the plan, and (7) the pay factor. All of these components
interact with one another to determine how the specification pays for pavement construction.
1 INTRODUCTION
Pavement performance is a function of its relative ability to serve traffic over a period of time (Highway
Research Board, 1962). Originally, a pavement’s relative ability to serve traffic was determined quite
subjectively by visual inspection and experience. However, experience is difficult to transfer from one
person to another, and individual decisions made from similar data are often inconsistent. In the late
1950s, systems of objective measurement (such as roughness meters, deflection and skid test equipment)
began to appear that could quantify a pavement’s condition and performance. These systems, along with
visual distress surveys, were used to aid in making maintenance and rehabilitation decisions, which, over
the years have been refined and upgraded to provide rapid, objective means to (Hicks and Mahoney,
1981):
● Establish maintenance priorities. Condition data such as roughness, distress, and deflection
are used to establish the projects most in need of maintenance and rehabilitation. Once
identified, the projects in the poorest condition (low rating) will be more closely evaluated to
determine repair strategies.
● Determine maintenance and rehabilitation strategies. Data from visual distress surveys are
used to develop an action plan on a year-to-year basis; i.e., which strategy (patching, surface
treatments, overlays, recycling, etc.) is most appropriate for a given pavement condition.
● Predict pavement performance. Data, such as ride, skid resistance, distress, or a combined
rating, are projected into the future to assist in preparing long-range budgets or to estimate the
condition of the pavements in a network given a fixed budget.
● Surface distress
● Skid resistance
● Structural evaluation
● Discusses these categories to include (1) the definition of the category, (2) the most common
measurement scales and (3) how measurements are made.
● Presents two sections that enumerate the typical flexible and rigid pavement distresses.
2 Roughness
Pavement roughness is generally defined as an expression of irregularities in the Major Topics on this Page
pavement surface that adversely affect the ride quality of a vehicle (and thus the
user). Roughness is an important pavement characteristic because it affects not 2.1 Measurement
only ride quality but also vehicle delay costs, fuel consumption and maintenance
costs. The World Bank found road roughness to be a primary factor in the 2.2 Measurement Techniques
analyses and trade-offs involving road quality vs. user cost (UMTRI, 1998).
Roughness is also referred to as "smoothness" although both terms refer to the same pavement qualities.
For detailed information, the University of Michigan Transportation Research Institute (UMTRI) maintains an
excellent Web page on roughness and all related aspects at http://www.umtri.umich.edu/erd/roughness.
2.1 Measurement
Today, roughness is typically quantified using some form of either present serviceability rating (PSR), international roughness index
(IRI) or other index with IRI being most prevalent.
Figure 9.2: IRI Roughness Scale (replotted from Sayers et al., 1986)
Another correlation was reported in a 1992 Illinois funded study performed by Al-Omari and Darter (1992):
This study used data from the states of Indiana, Louisiana, Michigan, New Mexico, and Ohio for both flexible and rigid pavements.
The associated regression statistics are R2 = 0.73, SEE = 0.39, and n = 332 sections. Correlations are highly dependent upon the
data that are used.
The following discussion with a few modifications was taken directly from the "Pavement Condition Data
Collection Equipment" article in the FHWA Pavement Notebook (1989).
2.2.1 Survey
A survey (performed by a survey crew) can provide an accurate measurement of the pavement profile. The use
of surveys for large projects, however, is impractical and cost prohibitive.
2.2.3 Profilographs
Profilographs have been available for many years and exist in a variety of different forms, configurations, and
brands. Due to their design they are not practical for network condition surveys. Their most common use today
is for rigid pavement construction inspection, quality control, and acceptance. The major differences among the
various profilographs involve the configuration of the wheels and the operation and measurement procedures of
the various devices.
Profilographs have a sensing wheel, mounted to provide for free vertical movement at the center of the frame
(see Figure 9.5). The deviation against a reference plane, established from the profilograph frame, is recorded
(automatically on some models) on graph paper from the motion of the sensing wheel (see Figure 9.6).
Profilographs can detect very slight surface deviations or undulations up to about 6 m (20 ft) in length.
WSDOT uses the California Profilograph to check rigid pavement construction smoothness (WSDOT Test Method
807). The measured parameter, called a "profile index", was developed by the California Division of Highways
starting in 1956. The goal was to develop a relationship between a calculated index and a subjectively obtained
panel rating of road roughness. After a careful survey, the profile index was developed based on 0.2 inch "blanking
band" as illustrated below.
The blanking band significantly reduced the data analysis effort since only "scallops" (deviations or excursions of
roughness above or below zero) that exceeded the blanking band would be significant and have to be analyzed.
Scofield recognized that "…the blanking band precluded faulting up to 0.2 inches. Although considered annoying,
the vibration caused by this level of faulting did not create accelerations large enough to produce discomfort to the
passengers."
Thus, the profile index is a count of the inches per mile in excess of the 0.2 inch blanking band. WSDOT specifies
the following profile index parameters:
● A daily profile index < 7 inches per mile. This specification attempts to ensure an overall construction
smoothness.
● High points having deviations > 0.3 inches shall be ground down so that they do not exceed 0.1 inches.
This specification prevents a single large bump from being averaged out over a days' worth of data.
Profile index can be roughly related to other measurements of smoothness and condition. One study showed that
for every 2 inches per mile increase in profile index, PSI decreases by about 0.1.
Scofield, L.A. (1993). Profilograph Limitations, Correlations, and Calibration Criteria for Effective Performance-Based
Specifications. Final Report, Project 20-7, Task 53, National Cooperative Highway Research Program, Transportation Research
Board. Washington, D.C.
RTRRMs measure the vertical movements of the rear axle of an automobile or the axle of a trailer relative to the
vehicle frame. The meters are installed in vehicles with a displacement transducer on the body located between
the middle of the axle and the body of a passenger car or trailer. The transducer detects small increments of
axle movement relative to the vehicle body. The output data consists of a strip chart plot of the actual axle body
movement versus the time of travel.
The disadvantage of a RTRRM is that its measured axle body movement vs. time depends on the dynamics of the
particular measurement vehicle, which results in two unwanted effects (UMTRI, 1998):
● Roughness measuring methods have not been stable with time. Measures made today with road meters
cannot be compared with confidence to those made several years ago.
● Roughness measurements have not been transportable. Road meter measures made by one system are
seldom reproducible by another.
Because of these two effects, profiling devices are becoming more popular.
● Straight edge. The simplest profiling system is a straight edge. Modifications to the straight edge, such
as mounting it on a wheel, result in a profilograph.
● Low speed systems. Low speed systems such as the CHLOE profilometer are moving reference planes.
The CHLOE is a long trailer that is towed at low speeds of 3 to 8 kph (2 to 5 mph). The slow speed is
necessary to prevent any dynamic response measurement during the readings. A few agencies still use
the CHLOE to calibrate their RTRRMs.
● Inertial reference systems. Most sophisticated road profiling equipment uses the inertial reference
system. The profiling device measures and computes longitudinal profile through the creation of an
inertial reference by using accelerometers placed on the body of the measuring vehicle to measure the
vehicle body motion. The relative displacement between the accelerometer and the pavement profile is
measured with either a "contact" or a "non-contact" sensor system.
The earliest profiling devices used a measurement system in direct contact with the pavement to measure
profile. Several contact systems have been used, and are still in use today. The French Road Research
Laboratory developed the Longitudinal Profile Analyzer (APL) in 1968.
Systems used today in the United States are frequently installed in vans (see Figure 9.7) which contain
microcomputers and other data handling and processing instrumentation. Older profiling devices are usually
contact systems, while the more recently manufactured devices use non-contact sensors. The non-contact
systems use probes, either acoustic or light, to measure differences in the pavement surface. For instance, the
South Dakota road profiler simultaneously collects three ultrasonic profiles, one for each wheelpath and one for
the lane center. These profiles are used to calculate (by computer) a mathematical measure of roughness and
an estimate of rutting at specified intervals along the roadway. A hybridized South Dakota road profiler
combines the three ultrasonic sensors with two laser sensors, one for each wheelpath, for simultaneous
measurement of the same roadway by two different sensor types under identical conditions (Virginia
Transportation Research Council, 1996). Integrated analysis units, as pictured in the interactive picture below,
can continuously collect a wide variety of data at highway speeds such as:
● Pavement texture
● Pavement video
● Pavement condition or distress
● Feature location
Figure 9.7: South Dakota Road Profiler Integrated Analysis Vehicle Interactive Picture
(van-mounted) (Click Picture to Launch)
WSDOT makes roughness measurements using laser equipment mounted in their distress data collection van.
This van, purchased in 1999, records pavement profile (ride, faulting, and rutting) and digital images of the
pavement surface, ahead view, and shoulder view.
Prior to 1999, roughness measurements were made with a South Dakota Road Profiler equipped with ultrasonic
sensors. WSDOT's switch to the laser measurements caused roughness measurements throughout the State to
decrease substantially. These decreases, often in the 0.5 - 1.0 m/km range, were not consistent but did occur on
almost all pavements regardless of type. Therefore, extreme caution must be exercised when comparing pre-1999
and post-1999 IRI values in Washington State as a substantial percentage of their difference may be due to the
change in measurement equipment and not actual road roughness.
Relative
Roughness Relative
Relative Data Approximate Extent of Projected
Data Principle of Degree
Initial Collection Decade of Current Extent of
Collection Measurement of
Cost Cost Development Use Use
Device Accuracy
(Network)
Direct Limited,
Same as
Dipstick Differential Low Impractical Very High 1980s Used for
Current Use
Measurement Calibration
Extensive
Direct Profile Same as
Profilographs Low Impractical Medium 1960s for Const.
Recordation Current Use
Acceptance
BPR Device
Low Low Medium 1940s Limited None
Roughometer Response
Vehicle Decreasing
Mays Meter Low Low Medium 1960s Extensive
Response Continuously
South
Direct Profile Rapidly
Dakota Medium Low High 1980s Growing
Recordation Increasing
Road Profiler
Contact
Direct Profile
Profiling High Medium Very High 1970s Limited Decreasing
Recordation
Device
Non-Contact
Direct Profile Increasing
Profiling High Medium Very High 1980s Medium
Recordation Continuously
Device
3 Surface Distress
Surface distress is "Any indication of poor or Major Topics on this Page
unfavorable pavement performance or signs of
impending failure; any unsatisfactory performance 3.1 Measurement
of a pavement short of failure" (Highway Research
Board, 1970). Surface distress modes can be 3.2 Measurement Techniques
broadly classified into the following three groups:
1. Fracture. This could be in the form of cracking (in flexible and rigid pavements) or
spalling resulting from such things as excessive loading, fatigue, thermal changes,
moisture damage, slippage or contraction.
3. Disintegration. This is in the form of stripping. raveling or spalling, which can result
from such things as loss of bonding, chemical reactivity, traffic abrasion, aggregate
degradation, poor consolidation/compaction or binder aging.
Thus, surface distress will be somewhat related to roughness (the more cracks, distortion
and disintegration - the rougher the pavement will be) as well as structural integrity (surface
distress can be a sign of impending or current structural problems).
These galleries include all the major types of pavement damage/distress. Each distress discussion
includes (1) pictures if available, (2) a description of the distress, (3) why the distress is a problem and
(4) typical causes of the distress. The gallery is organized alphabetically and the pictures are not
included in the Module list of figures.
3.1 Measurement
Measures of distress can be either subjective or objective. A simple example of a subjective
measurement may be a rating of high, medium, or low based on a brief visual inspection.
Objective measurements, which are generally more expensive to obtain, use different types
of automated distress detection equipment.
● Safety. Data are collected at highway speed, eliminating the need for driving at slow speeds or on
the shoulder.
● Accurate and complete distress data. Each distress along with its extent, severity and location is
identified and stored in a database. The system is also less prone to rating errors.
● More effective quality control. A centralized evaluation location and less subjective data make
quality control much better.
● More efficient data collection. Surface distress, rut and roughness data are all collected at the
same time using the same data collecting vehicle.
● Video and digital images are available for other users. They are available to bridge and
maintenance personnel and can be made available on the Internet in the future.
Figure 9.8: Washington State DOT Pavement Figure 9.9: Inside a Pavement Condition
Condition Rating Van Rating Van
Figure 9.12: Screen Shot from a Computer-Based Automatic Crack Detection System
(Image from Roadware's WiseCrax System)
Integrated analysis units can collect pavement surface distress data in the previously
described manner as well as collect data on a variety of other characteristics at highway
speeds such as:
● Grade, cross-slope
● Pavement video
● Pavement texture
● Feature location
● GPS coordinates
4 Skid Resistance
Skid resistance is the force developed when a tire that is Major Topics on this Page
prevented from rotating slides along the pavement surface
(Highway Research Board, 1972). Skid resistance is an important 4.1 Measurement
pavement evaluation parameter because:
4.2 Measurement Techniques
● Inadequate skid resistance will lead to higher incidences
of skid related accidents.
● Most agencies have an obligation to provide users with a roadway that is "reasonably" safe.
● Skid resistance measurements can be used to evaluate various types of materials and construction
practices.
Skid resistance depends on a pavement surface's microtexture and macrotexture (Corley-Lay, 1998).
Microtexture refers to the small-scale texture of the pavement aggregate component (which controls contact
between the tire rubber and the pavement surface) while macrotexture refers to the large-scale texture of
the pavement as a whole due to the aggregate particle arrangement (which controls the escape of water
from under the tire and hence the loss of skid resistance with increased speed) (AASHTO, 1976). Skid
resistance changes over time. Typically it increases in the first two years following construction as the
roadway is worn away by traffic and rough aggregate surfaces become exposed, then decreases over the
remaining pavement life as aggregates become more polished. Skid resistance is also typically higher in the
fall and winter and lower in the spring and summer. This seasonal variation is quite significant and can
severely skew skid resistance data if not compensated for (Jayawickrama and Thomas, 1998).
4.1 Measurement
Skid resistance is generally quantified using some form of friction measurement such as a friction factor or
skid number.
It is not correct to say a pavement has a certain friction factor because friction involves two bodies, the tires
and the pavement, which are extremely variable due to pavement wetness, vehicle speed, temperature, tire
wear, tire type, etc. Typical friction tests specify standard tires and environmental conditions to overcome
this.
In general, the friction resistance of most dry pavements is relatively high; wet pavements are the problem.
The number of accidents on wet pavements are twice as high as dry pavements (but other factors such as
visibility are involved in addition to skid resistance). Table 9.3 shows some typical Skid Numbers (the higher
The most commonly used method in the U.S. for skid resistance testing uses some form of a lock wheel tester (see Figure
9.13). Basically, this method uses a locked wheel skidding along the tested surface to measure friction resistance. A typical
lock-wheel skid measurement system must have the following:
● A test vehicle with one or more test wheels incorporated into it or as part of a towed trailer.
● A standard tire for use on the test wheel. The standardized skid-test tire, a tubeless, bias-ply G78x15 tire with seven
circumferential grooves, is defined by AASHTO M 261 or ASTM E 501. A newer tire, one with no grooves, appears
to be gaining acceptance as well. By defining the standard test tire, the tire type and design are eliminated as
variables in the measurement of pavement skid resistance.
● A means to transport water (usually 750 to 1900 liters (200 to 500 gallons)) and the necessary apparatus to deliver it
in front of the test wheel at test speed
● A transducer associated with the test wheel that senses the force developed between the skidding test wheel and the
pavement
● Electronic signal conditioning equipment to receive the transducer output signal and modify it as required
● Suitable analog and/or digital readout equipment to record either the magnitude of the developed force or the
To take a measurement, the vehicle (or trailer) is brought to the desired testing speed (typically 64 km/hr (40 mph)) and water is
sprayed ahead of the test tire to create a wetted pavement surface. The test tire braking system is then actuated to lock the test
tire. Instrumentation measures the friction force acting between the test tire and the pavement and reports the result as a Skid
Number (SN).
1. No force measurement is necessary, the force can be computed by knowing the test wheel's moment of inertia and its
rotational acceleration (Wambold et al., 1990). Force measuring devices for the locked wheel tester cost a significant
amount of money.
2. Because the test tire is in contact with the pavement while locked for a much shorter time than the locked wheel
tester, it significantly reduces test tire wear.
Laser or advanced image processing equipment are capable of determining surface macrotexture from a vehicle moving at
normal travel speeds. One particular device, the Road Surface Analyzer (ROSAN), a series of non-contact pavement surface
texture measurement devices, has been developed by the FHWA's Turner Fairbanks Research Center Pavement Surface
Analysis Laboratory. The ROSAN (see Figure 9.14) can be used for measuring texture, aggregate segregation, grooves, tining,
joints, and faulting (FHWA, 2001). ROSAN systems have been used in a number of NCHRP and FHWA sponsored studies.
Some integrated analysis units can use surface texture measuring to estimate skid resistance.
The one drawback to this method is that a pavement's surface macrotexture does not entirely determine its skid resistance.
Therefore, correlation between surface macrotexture and skid resistance is often difficult to extrapolate into any general
guidance.
5 Deflection
Pavement surface deflection measurements are the Major Topics on this Page
primary means of evaluating a flexible pavement
structure and rigid pavement load transfer. Although 5.1 Measurement
other measurements can be made that reflect (to
some degree) a pavement's structural condition, 5.2 Measurement Techniques
surface deflection is an important pavement
evaluation method because the magnitude and shape of pavement deflection is a function of
traffic (type and volume), pavement structural section, temperature affecting the pavement
structure and moisture affecting the pavement structure. Deflection measurements can be used
in backcalculation methods to determine pavement structural layer stiffness and the subgrade
resilient modulus. Thus, many characteristics of a flexible pavement can be determined by
measuring its deflection in response to load. Furthermore, pavement deflection measurements
are non-destructive.
5.1 Measurement
Surface deflection is measured as a pavement surface's vertical deflected distance as a result of
an applied (either static or dynamic) load. The more advanced measurement devices record
this vertical deflection in multiple locations, which provides a more complete characterization of
pavement deflection. The area of pavement deflection under and near the load application is
collectively known as the "deflection basin".
● Static deflections
The general principal is to apply a load of known magnitude to the pavement surface and
analyze the shape and magnitude of the deflection basin to assess the strength of the pavement
structure (see Figure 9.15).
The Benkelman Beam (see Figure 9.16), developed at the Western Association of State Highway
Organizations (WASHO) Road Test in 1952, is a simple device that operates on the lever arm
principle. The Benkelman Beam is used with a loaded truck - typically 80 kN (18,000 lb) on a
single axle with dual tires inflated to 480 to 550 kPa (70 to 80 psi). Measurement is made by
placing the tip of the beam between the dual tires and measuring the pavement surface rebound
as the truck is moved away (see Figure 9.17). The Benkelman Beam is low cost but is also
slow, labor intensive and does not provide a deflection basin.
The stead state deflection equipment (see Figure 9.18) is stationary when measurements are
taken with force generator (counter rotating weights) started and deflection sensors
(transducers) lowered to the pavement surface. Figure 9.19 is a plot of a typical force output
and Figure 9.20 shows the location of the equipment's loading wheels and five transducers. The
equipment is most suitable for use on thinner pavements including low volume rural highways,
county roads, municipal streets, and parking lots (IMS, 2001).
The Road Rater (see Figure 9.21) is the other popular type of steady state deflection
equipment. It must also be stationary to start and operates in a similar fashion to the
Dynaflect.
● ASTM D 4694: Standard Test Method for Deflections with a Falling Weight Type
Impulse Load Device
(based on unpublished data collected by the Washington State DOT Materials Laboratory in
1982-1983)
BB = 20.63 (D)
R2 = 0.72
Comparing a Benkelman Beam load at 9,000 pounds on dual tires with 70-80 psi inflated tires
and Road Rater at 8,000 pound peak-to-peak load at 15 Hz on a 12 inch diameter plate on a
stabilized pavement:
BB = 2.57 + 1.27(RR)
R2 = 0.66
The Western Direct Federal Division, Federal Highway Administration, Vancouver, Washington
provides the following correlation for the Benkelman Beam to Road Rater Model 400:
● Trigger treatment. For instance, once a pavement's condition rating reaches a certain level, it
can be scheduled for maintenance or rehabilitation.
● Determine the extent and cost of repair. A pavement condition score is a numerical
representation of a pavement's overall condition and can thus be used to estimate the extent of
repair work and the likely cost.
● Determine a network condition index. By combining pavement condition scores for an entire
road network, a single score can be obtained that gives a general idea of the network condition
as a whole.
● Allow equal comparison of different pavements. Since a pavement condition score accounts
for all types of pavement performance measures it can be used to compare two or more
pavements with different problems on an equal footing.
A pavement condition index is simply the scale, or series of numbers, used to describe a pavement
condition. Typical pavement condition indices may be based on a scale of 0 to 5 or perhaps 0 to 100.
The proper pavement condition index depends upon the objectives of whatever system is used to manage
a particular pavement network (called a Pavement Management System or PMS). This section presents
two pavement condition index methods.
PSR was a ride quality rating that required a panel of observers to actually ride in an automobile over the
pavement in question. Since this type of rating is not practical for large-scale pavement networks, a
transition to a non-panel based system was needed.
To transition from a PSR serviceability measure (panel developed) to a PSI serviceability measure (no
panel required), a panel of raters during 1958 to 1960 rated various roads in the states of Illinois,
Minnesota, and Indiana for PSR. This information was then correlated to various pavement
measurements (such as slope variance (profile), cracking, etc.) to develop PSI equations. Further, the
raters were asked to provide an opinion as to whether a specific pavement assessed for PSR was
"acceptable" or "unacceptable" as a primary highway (see Figure 9.1). Thus, although PSI is based on
the same 5-point rating system as PSR it goes beyond a simple assessment of ride quality. About one-
half of the panel of raters found a PSR of 3.0 acceptable and a PSR of 2.5 unacceptable. Such
information was useful in selecting a "terminal" (or failure) serviceability (PSI) design input for
empirical structural design equations. It is interesting to note that the original AASHO Road
Test rater opinions are based on car ride dynamics; it is unclear whether such levels are
acceptable for trucks.
Pavement performance can then be defined as "The serviceability trend of a ... (pavement segment) with
increasing number of axle applications" (Highway Research Board, 1962). Figure 9.27 further
demonstrates this concept.
6.3 Summary
In order to manage a pavement network (be it for a town, city, county or state), there must be some
means of comparing one pavement to the next. Thus, pavement management systems usually
implement a some type of pavement condition index. These usually take into account the types of
pavement evaluation presented in this Module. Condition indices can be subjective or objective and can
vary in complexity, however they should be relevant, reliable, affordable and appropriate (Deighton,
1998).
1 INTRODUCTION
The combined effects of traffic loading and the environment will cause every pavement, no matter how
well-designed/constructed to deteriorate over time. Maintenance and rehabilitation are what we use to
slow down or reset this deterioration process. Maintenance actions, such as crack sealing, joint sealing,
fog seals and patching help slow the rate of deterioration by identifying and addressing specific
pavement deficiencies that contribute to overall deterioration. Rehabilitation is the act of repairing
portions of an existing pavement to reset the deterioration process. For instance, removing and replacing
the wearing course in a pavement provides new wearing course material on which the deterioration
process begins anew. Reconstructing an entire pavement, however, is not considered rehabilitation but
rather new construction because the methods used are generally those developed for new pavement
construction. This section will discuss the following:
● Maintenance options for flexible and rigid pavement. Basic preventive and corrective
maintenance options.
● Rehabilitation options for flexible and rigid pavement. Basic rehabilitation options to include
overlays and their design procedures.
http://training.ce.washington.edu/WSDOT/Modules/10_maintenance_rehab/10-1_body.htm4/2/2008 6:42:09 PM
10.2 Maintenance & Rehabilitation - Flexible Pavement Maintenance
2 Flexible - Maintenance
Pavement maintenance describes all the methods and techniques used to preserve Major Topics on this Page
pavement condition, safety, and ride quality, and therefore aid a pavement in
achieving its design life (Hall et al., 2001). The performance of a pavement is 2.1 Crack Seals
directly tied to the timing, type and quality of the maintenance it receives. This
section, taken largely from Roberts et al. (1996), describes the more common U.S. 2.2 Fog Seals
preventative and corrective maintenance options for HMA pavement. The timing 2.3 Rejuvenators
of these maintenance items is discussed in Module 11, Pavement Management.
2.4 Slurry Seals
2.7 Patching
Crack seal products are used to fill individual pavement cracks to prevent entry of
water or other non-compressible substances such as sand, dirt, rocks or weeds. 2.8 Summary
Crack sealant is typically used on early stage longitudinal cracks, transverse
cracks, reflection cracks and block cracks. Alligator cracks are most often too extensive to warrant filling with crack sealer; they
usually require an area treatment such as a patch or reconstruction. Crack filler material is typically some form of rubberized
asphalt or sand slurry.
Purpose: Preventive maintenance. Crack filling to prevent entry of water or other non-compressible substances into the pavement.
Other Before applying crack sealant, cracks need to be routed out and cleaned.
Info:
Crack sealing is best done in moderate temperatures (spring or fall) and is most effective if performed immediately after cracks
develop.
Purpose: Preventive maintenance. Fog seals are used to restore or rejuvenate an HMA surface. They may be able to postpone the need for
a BST or non-structural overlay for a year or two.
Mix None. A test patch may be needed to determine the proper application rate.
Design:
Other Fog seals are suitable for low-volume roads which can be closed to traffic for the 4 to 6 hours it takes for the slow-setting asphalt
Info: emulsion to break and set.
An excessive application rate may result in a thin asphalt layer on top of the original HMA pavement. This layer can be very
smooth and cause a loss of skid resistance. Sand should be kept in reserve to blot up areas of excess application.
2.3 Rejuvenators
Rejuvenators are products designed to restore original properties to aged (oxidized) asphalt binders by restoring the original ratio of
asphaltenes to maltenes. Many rejuvenators are proprietary, making it difficult to offer a good generic description. However, many
rejuvenators contain maltenes because their quantity is reduced by oxidation. Rejuvenators will retard the loss of surface fines and
reduce the formation of additional cracks, however they will also reduce pavement skid resistance for up to 1 year (Army and Air
Force, 1988). Because of this, rejuvenators are generally appropriate for low-volume, low-speed roads or parking lots.
Purpose: Preventive maintenance. Restore original properties to aged asphalt binder. Rejuvenators may be able to postpone the need for a
BST for a year or two.
Materials: Various compounds. Most rejuvenators are proprietary and thus a general description of their constituent materials is not
possible.
Mix None. A test patch may be needed to determine effectiveness and the proper application rate.
Design:
Other A rejuvenator should not be applied to a pavement having an excess of binder on the surface such as that found in slurry seal,
Info: OGFC, or BSTs. When excessive binder is on the surface, the rejuvenator will soften the binder and cause the surface to become
tacky and slick (Army and Air Force, 1988).
The amount of air voids in the HMA being rejuvenated should be at least 5 percent to ensure proper penetration of the
rejuvenator into the pavement. If the voids are less than 5 percent, the rejuvenator may fill the voids and thus cause an unstable
mix (Army and Air Force, 1988).
Rejuvenators should be applied in hot weather, above 20°C (70°F), so that the rejuvenator (1) will penetrate more deeply into the
asphalt pavement and (2) will cure sooner (Army and Air Force, 1988).
1. Type I (fine). This type has the finest aggregate gradation (most are smaller than the 2.36 mm (No. 8) sieve) and is used
to fill small surface cracks and provide a thin covering on the existing pavement. Type I aggregate slurries are sometimes
used as a preparatory treatment for HMA overlays or surface treatments. Type I aggregate slurries are generally limited
to low traffic areas (ISSA, 2001).
2. Type II (general). This type is coarser than a Type I aggregate slurry (it has a maximum aggregate size of 6.4 mm (0.25
inches)) and is used to (1) treat existing pavement that exhibits moderate to severe raveling due to aging or (2) to improve
skid resistance. Type II aggregate slurry is the most common type.
3. Type III (coarse). This type has the most coarse gradation and is used to treat severe surface defects. Because of its
aggregate size, it can be used to fill slight depressions to prevent water ponding and reduce the probability of vehicle
hydroplaning.
Microsurfacing
Microsurfacing is an advanced form of slurry seal that uses the same basic ingredients (emulsified asphalt, water, fine aggregate and
mineral filler) and combines them with advanced polymer additives. Figures 10.1 through 10.4 show a microsurfacing slurry seal
project.
Purpose: Preventive maintenance. Repair slight to moderate pavement surface defects, improve skid resistance.
Materials: Emulsified asphalt, water, well-graded fine aggregate and mineral filler.
Other As opposed to a fog seal, a slurry seal contains aggregate and can thus correct minor surface defects in a variably textured
Info: surface - filling cracks and voids, sealing weather-tight, and providing color and texture delineation in a single pass (ISSA,
2001b).
● An anti-glare surface during wet weather and an increased reflective surface for night driving.
1. Surface preparation. Surface defects, such as potholes, are repaired and the existing surface is cleaned (e.g., by a street
sweeper).
2. Asphalt material application. Typically, an asphalt emulsion is applied from a spray truck to the surface of the existing
pavement (see Figure 10.5).
3. Aggregate application. A thin aggregate cover (only one stone thick) is spread over the asphalt material before it has set
(see Figure 10.6). The aggregate usually has a uniform gradation.
4. Aggregate embedding. A roller (usually a pneumatic tire roller) is used to push the aggregate into the asphalt material
and seat it firmly against the underlying pavement (see Figure 10.7). Generally, about 50 percent of each aggregate
particle should be embedded in the asphalt material (see Figure 10.8) after final rolling. About 70 percent of each
aggregate particle will be embedded after several weeks of traffic. It is common to place an aggregate "chokestone" on
top of the uniformly graded larger aggregates after embedment. Chokestone is essentially a finer aggregate gradation (e.
g., less than 12.5 mm (0.25 inches)) used to make a more dense aggregate matrix at the level of embedment (see Figure
10.9). This more dense matrix helps prevent excessive aggregate loss due to traffic.
Multiple layer surface treatments are done by repeating the above process for each layer. Figure 10.10 shows a BST in Washington
State.
Figure 10.5: Placing the Asphalt Emulsion Figure 10.6: Placing the Aggregate
Figure 10.7: Embedding the Aggregate Figure 10.8: BST Before Chokestone Application (note
asphalt emulsion is
visible between aggregates)
Figure 10.9: BST After Chokestone Application (note small Figure 10.10: BST on SR 2 near
chokestone Coulee City, WA
between the larger aggregates)
Purpose: Preventive maintenance. Wearing course, waterproof covering for the existing pavement.
Materials: Asphalt (as asphalt binder, cutback asphalt or asphalt emulsion) and aggregate (uniformly graded).
Other Traditionally, BSTs were thought of as most applicable to low volume, low speed roads because they will eventually involve
Info: some amount of loose aggregate. On a high volume or high speed road, this loose aggregate can be picked up and thrown by
wheels, which can result in chipped paint and broken windshields. However, developments in asphalt cement modifiers and BST
construction procedures have made it possible to use them on high volume/speed roads including interstates.
The Minnesota Department of Transportation has an excellent resource on BSTs and other seal coats titled the Minnesota Seal
Coat Handbook and available at: http://mnroad.dot.state.mn.us/research/mnroad_project/restools/sealcoat.asp
WSDOT recommends BSTs be applied to roadways with 2,000 average daily traffic (ADT)
or less or less than 50,000 ESALs per year.
WSDOT defines a non-structural overlay as being any overlay less than 25 mm (1 inch).
This is slightly different than NAPA's definition.
The WSDOT Tech Note on Novachip is a discussion of one type of proprietary non-
structural overlay.
2.6.1 Categories
Non-structural overlays can vary widely in composition depending upon local practice, traffic and general purpose. A loose
classification of non-structural overlays follows (NAPA, 1995):
1. Light volume/residential traffic. The primary objective in light traffic areas is to retard asphalt binder aging of the
underlying pavement. Since heavy traffic loads are not of great concern, overlays are generally less stiff (resulting in a
more workable mix, increased durability and flexibility and a potential for the overlay to reheal under traffic) and use
smaller-sized aggregates.
2. Heavy, high-speed traffic. The primary objective in heavy, high-speed traffic areas is to prevent rutting and provide good
friction. Because of this, overlays typically use larger angular aggregate and more durable mixes such as SMA or OGFC.
● Thin lifts require less HMA per foot of road length than thick lifts. This can result in high paver speeds (in excess of 21 m
(70 ft.) per minute). Compaction may not be able to keep pace with these high speeds.
● Thin lifts will cool quicker than thick lifts. This can result in little time available for compaction before the thin overlay
reaches cessation temperature (sometimes as little as 3 to 5 minutes). Therefore, roller variables should be set to account
for this (e.g., enough rollers and an adequate roller pattern to compact the material before it reaches cessation
temperature).
● Thin lift construction produces greater screed wear. If the lift depth is less than about twice the maximum aggregate
size, the HMA may tear under the paver screed. Very thin lifts (less than 25 mm (1 inch)) can be damaged by the screed
dragging large particles.
● Thin lifts are more sensitive to vibratory rolling. Incorrectly chosen amplitude, frequency or roller speed can result in
aggregate degradation (i.e. breaking) and damage of the bond between the overlay and the existing pavement.
● Density control is difficult. Thin lifts provide fewer options for aggregate particles to rearrange under compaction. Thus,
mat densities will tend to be less uniform than those associated with a thicker lift. This should be recognized if pay is in
any way tied to mat density.
2.7 Patches
Patches are a common method of treating an area of localized distress. Patches can be either full-depth where they extend from the
pavement surface to the subgrade (see Figure 10.12) or partial where they do not extend through the full depth of existing pavement
(see Figure 10.13).
Full-depth patches are necessary where the entire depth of pavement is distressed. Often times, the underlying base, subbase or
subgrade material is the distresses root cause and will also need repair. Partial depth patches are used for pavement distresses like
raveling, rutting, delamination and cracking where the depth of crack does not extend through the entire pavement depth.
Patching material can be just about any HMA or cold mix asphalt material as well as certain types of slurries. Typically some form
of HMA is used for permanent patches, while cold mix is often used for temporary emergency repairs.
One form of patching, pothole patching, probably receives the greatest amount of public attention. Pothole patching procedures
cover a wide range of methods and intentions from permanent full-depth patches to temporary partial depth patches. Two general
patching procedures are described next.
Semi-Permanent Pothole Patch (see Figures 10.14 and 10.15) (from FHWA, 1998)
2. Square up the pothole sides so they are vertical and have in-tact pavement on all sides.
3. Place the patching material into the clean squared-up hole. The material should mound in the center and taper down to
the edges so that it meets flush with the surrounding pavement edges.
4. Compact the patching material starting in the center and working out toward the edges. Compaction can be accomplished
using a vibratory plate compactor or a single-drum vibratory roller. Check the compacted patching material for a slight
crown. This is done so that subsequent traffic loading will compact it down to the surrounding pavement height.
Figure 10.14: Pothole Patching Truck with a Hotbox Figure 10.15: Semi-permanent Pothole Repair
1. Place the patching material into the pothole without any preparation or water/debris removal.
2. Compact the patching material using the patching truck tires (usually 4 to 8 passes).
3. Check the compacted patch for a slight crown. If a depression is present add more patching material and compact.
Although it may seem that the semi-permanent technique would produce a higher quality patch than the throw-and-roll technique,
the FHWA's Long Term Pavement Performance (LTPP) Study found that the "throw-and-roll technique proved just as effective as
the semi-permanent procedure for those materials for which the two procedures were compared directly" (FHWA, 1998). Since the
semi-permanent technique is more labor and material intensive, the throw-and-roll technique will generally prove more cost
effective if quality materials are used.
2.8 Summary
Pavement maintenance prolongs pavement life by slowing its deterioration rate. This section has described some of the more
common maintenance options in the U.S. Each option's effectiveness is dependent upon a multitude of local conditions. For most
smaller agencies, the best advice when considering pavement maintenance options is to talk to local contractors and nearby agencies
about what types of maintenance options have worked best in your local area.
3 Flexible - Rehabilitation
The combined effects of traffic loading and the environment will Major Topics on this Page
cause pavements to deteriorate over time. Although maintenance can
slow the rate of deterioration, it cannot stop it. Therefore eventually 3.1 Structural HMA Overlays
the effects of deterioration need to be reversed by adding or replacing
material in the existing pavement structure. This is called 3.2 Structural PCC Overlays
rehabilitation. Formally, rehabilitation can be defined as (Hall et al., 3.3 Summary
2001):
"...a structural or functional enhancement of a pavement which produces a substantial extension in service life,
by substantially improving pavement condition and ride quality."
A wholesale replacement of the entire pavement structure is considered reconstruction rather than rehabilitation since it
follows new pavement construction methods. Flexible pavement rehabilitation options depend upon local conditions
and pavement distress types but typically include:
● Cold in-place recycling (CIR). Covered in Module 2, Section.4, Recycling Options. Full-depth CIR, known
as full-depth reclamation (FDR) is considered reconstruction.
● HMA overlays. Overlays can be placed on existing surfaces with or without preparatory milling. Overlays
are used for two primary purposes:
❍ Structural overlays are designed to add structural support to the existing pavement. Because of
this, they are structurally designed and are thicker than non-structural overlays.
❍ Non-structural overlays are designed to add to or replace the existing pavement wearing course
only. Because of this they contribute very little to the pavement structure and are generally
assumed to provide no additional structural support. Because most agencies consider non-
structural overlays to be maintenance items, they are discussed in Section 3, Maintenance.
● PCC overlays. Some agencies have used PCC overlays of flexible pavements (usually called
"whitetopping") in certain situations. PCC overlays can be divided into two types (Mack, Hawbaker and
Cole, 1998):
❍ Unbonded (termed "classical whitetopping"). The PCC overlay is not purposely bonded
to the surface of the underlying flexible pavement surface. The existing flexible
pavement serves as base for the new PCC overlay. These overlays are usually
greater than 100 mm (4 inches) thick.
❍ Bonded (termed "thin composite whitetopping"). The PCC overlay is purposely bonded to the
existing flexible pavement surface. Thus, the rigid overlay and existing flexible pavement act as a
composite structure. This allows for thinner PCC overlays.
NCHRP Web document 35 (Project C1-38): Rehabilitation Strategies for Highway Pavements (http://gulliver.trb.org/
publications/nchrp/nchrp_w35-a.pdf) provides some good guidelines for collecting data, evaluating pavement,
selecting rehabilitation techniques and forming rehabilitation strategies.
This section will concentrate on structural overlays by describing several typical structural overlay design methods.
Specific WSDOT structural design policy is contained in the WSDOT Pavement Guide, Volume 1. In
general, WSDOT uses the following structural design procedures:
❍ Flexible. The AASHTO Guide for Design of Pavement Structures (1986 or 1993
version). This is an empirical procedure.
❍ Rigid. The AASHTO Guide for Design of Pavement Structures (1986 or 1993
version). This is an empirical procedure.
● Rehabilitation.
❍ PCC overlays. The AASHTO Guide for Design of Pavement Structures for
unbonded PCC overlays. This is an empirical procedure. Generally, only
unbonded PCC overlays will be used if a PCC surfacing is selected. Bonded
PCC overlays are not considered as a structural solution and have a higher than
acceptable risk of premature failure.
● Engineering judgment
● Component analysis
● Mechanistic-empirical analysis
Mechanistic-empirical based design methods are useful in overlay design as well as new pavement design. Their
greatest advantage is the versatility provided in evaluating different materials under various environments and
pavement conditions. Mechanistic-empirical procedures provide a basis for rationally modeling pavement systems. As
these models improve, better correlations can be expected between design and performance parameters. In many
places these procedures have replaced limiting deflection overlay methods, since the latter do not account for
subsurface material properties. Mechanistic-empirical overlay design is essentially the same as mechanistic-empirical
structural design for new pavements but with the addition of more evaluation locations. Module 6, Section 4,
Mechanistic-Empirical covers this design method.
The chief advantage of classical whitetopping is that it requires minimal surface preparation.
However, minimum overlay thicknesses tend to be in the 125 - 175 mm (5 - 7 inch) range,
which is quite thick and possibly unsuitable in situations where a specific elevation must be
maintained such as in curbed areas or under bridges.
The design procedure contained in the 1993 AASHTO Guide is virtually identical to the AASHTO empirical design for
new rigid pavements with one exception: The effective modulus of subgrade reaction (k) is determined based on the
existing flexible pavement resilient modulus. Although perfectly acceptable, this method gives little credit to the
existing pavement's remaining strength.
The chief advantage of thin composite whitetopping is that it can be made thinner than
classical whitetopping because of the composite layer action. However, issues with slab size,
joint location and bonding effectiveness can complicate its use. This subsection covers:
● Structural design
● Joint design
● Other considerations
The 1993 AASHTO Guide design procedure does not account for the bonded composite action of the combined
pavement-plus-overlay. Therefore, it treats the bonded overlay design exactly the same as the unbonded one and does
not credit the existing flexible pavement with any structural capacity. In reality, if the bond between layers is adequate
then the structural support capacity of the underlying flexible pavement should be considered. Although multiple
studies have shown this bonding to be adequate (Mack, Hawbaker and Cole, 1998), the assumption of adequate bond
performance is still a significant risk. If, for some reason, the bond does not perform as intended then the pavement
will most likely fail prematurely. Surface preparation is critical.
The American Concrete Pavement Association (ACPA) has a web page that will calculate the load-
carrying capacity of an ultra-thin whitetopping (UTW) pavement during its service life. The
calculations are based on a comprehensive mechanistic analysis and correlation to UTW performance
data. This web page can be found at: http://www.pavement.com/pavtech/tech/utwcalc/main.asp.
Joints are typically design much closer than for typical new-construction rigid pavement. The closer joint spacing, on
the order of 1 - 4 m (3.3 - 13.1 ft.), does the following (Mack, Hawbaker and Cole, 1998):
● Reduces the moment arm of the applied wheel load and minimizes the stresses due to
bending.
● Reduces the curling and warping stresses by reducing the size of the slab that can curl or warp.
Because of the short joint spacing, the overlaid PCC slabs transfer load to the underlying
flexible pavement by deflecting downward as a unit rather than bending (Mack, Hawbaker and
Cole, 1998). Figures 10.17 and 10.18 show two different joint spacings.
Figure 10.17: 3.7 x 3.7 m (12 x 12 ft.) Figure 10.18: 1.2 x 1.2 m (4 x 4 ft.) UTW Slabs at the Mn/
UTW Slabs at the Mn/ROAD Test Facility ROAD Test Facility
Some criteria for deciding when to consider thin composite whitetopping as a rehabilitation alternative are
(Vandenbossche and Fagerness, 2001):
● The existing flexible pavement must be more than 100 mm (4 inches) thick. This provides a reasonably
strong structural layer to which the PCC overlay can bond. Mack, Hawbaker and Cole (1998) suggest a
minimum thickness of 75 mm (3 inches).
● No raveling on the existing flexible pavement. Raveling will adversely affect bonding.
● Little to no bottom-up fatigue cracking in the existing flexible pavement. Bottom up fatigue cracking will
continue to progress and weaken the flexible pavement structure even after the PCC overlay.
3.3 Summary
Rehabilitation essentially reverses the effects of deterioration by adding or replacing material in the existing pavement
structure. Although there are several common methods of rehabilitation (HIPR, CIR and overlays) this section has
concentrated on structural overlays - those used to increase a pavement's structural capacity. Non-structural overlays
are treated in Section 2, Flexible - Maintenance.
New road construction in the U.S. is not nearly as prolific as it has been in previous generations. Urban areas have
filled out greatly and the ratio of existing roads to new roads is now quite high. Consequently, rehabilitation (and not
new construction) has become the dominant force in today's pavement design and construction arenas.
4 Rigid - Maintenance
Pavement maintenance describes all the methods and techniques used Major Topics on this Page
to prolong pavement life by slowing its deterioration rate. Thus, the
performance of a pavement is directly tied to the timing, type and 4.1 Joint and Crack Sealing
quality of the maintenance it receives. This section, taken largely
from Roberts et al. (1996), describes the more common U.S. 4.2 Slab Stabilization
preventative and corrective maintenance options for rigid pavement. 4.3 Diamond Grinding
The timing of these maintenance items is discussed in Module 11,
Pavement Management. 4.4 Patches
4.5 Summary
"A typical hot-pour sealant provides an average of 3 to 5 years of life after proper installation.
Some low-modulus or PVC coal-tars can perform well past 8 years. Silicone sealants have
performed well for periods exceeding 8 to 10 years on roadways. This type of performance
hinges on joint preparation and installation. Of extreme importance is that the joint be clean and
dry. Compression seals provide service for periods often exceeding 15 years and sometimes 20
years."
Crack sealant is typically used on early stage, isolated panel cracks; extensive or advanced panel cracking is a symptom
of larger problem (e.g., lack of panel support, inadequate structural design or poor construction) that cannot be
addressed by simple crack sealing.
Purpose: Preventive maintenance. Joint and crack filling to prevent entry of water or other non-compressible substances into
the pavement.
Other Sealant performance is dependent upon proper joint design and cleanliness.
Info:
Purpose: Preventive and corrective maintenance. Restores proper base/subgrade slab support to prevent more serious
distresses such as faulting, corner breaks and cracking.
Other Slab stabilization only fills the voids under a slab, it should not be confused with slab jacking, which is used to raise
Info: the elevation of a depressed slab.
Diamond grinding can reduce the IRI of an older pavement to 1.0 - 2.0 m/km (63 - 126 inches/mile).
Purpose: Corrective maintenance. Restores pavement smoothness and frictional characteristics. Can also be used in new
construction to attain initial specified smoothness, although properly constructed PCC pavements should not require
any diamond grinding.
Other Diamond grinding addresses serviceability problems but not their root cause. For example, diamond grinding can
Info: substantially reduce the roughness on a faulted rigid pavement, but it will not address the cause of faulting, nor will
it prevent roughness in the future as a result of additional faulting.
Because the grinding head is cooled with water, a byproduct of diamond grinding is a slurry of ground PCC
particles and water. Most of this slurry is picked up by vacuums within the grinding machine and either deposited
along the highway shoulder (in some rural areas) or collected in trucks for disposal.
4.4 Patches
Rigid pavement patches are used to treat localized slab problems such as spalling, scaling (e.g., reactive aggregate
distress, over-finishing the surface), joint deterioration, corner breaks or punchouts. If the problem is limited in depth,
then a partial depth patch may be appropriate, otherwise a full depth patch is recommended. A high quality patch can
be considered a permanent repair, although all patches are treated as a form of pavement distress. Although HMA is
sometimes used for emergency patches, PCC should be used for permanent patches. Fast-setting PCC is often used to
minimize setting time.
usually accomplished by sawing and chipping. Figure 10.25: Partial Depth Patch Used to Repair
Small areas can be removed by sawing around the Spalling Damage
patch edges and then chipping out the interior.
The patch should be deep enough to remove all the damaged material.
3. Clean the area to be patched. Sandblasting or water blasting removes loose particles and creates a rough
texture to which the bonding agent can adhere.
4. Apply a bonding agent. A cementitious grout is used to help the patch material bond to the original slab
material.
5. Place, finish and cure the PCC. The PCC should be placed so that the patch is of the same elevation as the
surrounding slab. Finishing the patch from the center to the edges helps push the PCC patch material firmly
against the existing slab and increases the potential for a high strength bond.
Figure 10.26: Full-Depth Patch Preparation Figure 10.27: Pouring a Small Full Depth PCC Patch on a
Residential Street
1. Locate the area to be patched. If the area to be patched is too close to an existing joint or crack, the patch
area should be extended as follows:
❍ Patch boundary within 2 m (6 ft.) of an existing undoweled transverse joint. Extend the patch to
the transverse joint.
❍ Patch boundary on an existing doweled transverse joint. If the other side of the joint does not
require repair, extend the patch beyond the transverse joint by about 0.3 m (1 ft.) to remove the
existing dowels.
❍ The patch boundary falls on an existing crack (CRCP). Extend the patch beyond the crack by
about 0.15 m (0.5 ft.).
2. Remove the damaged material. Usually, full depth saw cuts are used to isolate the repair area from the rest of
the pavement. Then, the isolated section is lifted out as a whole or broken up and removed.
3. Prepare the patch area. The base material and subgrade is compacted, smoothed and dried (see Figure
10.28). Dowel bars holes are drilled into the adjacent slab transverse sections and dowel bars are inserted to
provide load transfer across the patch boundary. Slab replacements longer than about 4.5 m (15 ft.) require
longitudinal tie bars as well (see Figure 10.29).
4. Apply a bonding agent. A cementitious grout is used to help the patch material bond to the original slab
material.
5. Place, finish and cure the PCC. The PCC should be placed so that the patch is of the same elevation as the
surrounding slab. Vibratory screeds are often used to strike off and finish full depth patches.
4.5 Summary
Pavement maintenance prolongs pavement life by slowing its deterioration rate. This section has described some of the
more common maintenance options in the U.S. Each option's effectiveness is dependent upon a multitude of local
conditions. For most smaller agencies, the best advice when considering pavement maintenance options is to talk to
local contractors and nearby agencies about what types of maintenance options have worked best in your local area.
5 Rigid - Rehabilitation
The combined effects of traffic loading and the environment will Major Topics on this Page
cause pavements to deteriorate over time. Although maintenance
can slow the rate of deterioration, it cannot stop it. Therefore 5.1 Dowel Bar Retrofit
eventually the effects of deterioration need to be reversed by adding
5.2 Structural HMA Overlays
or replacing material in the existing pavement structure. This is
called rehabilitation. Formally, rehabilitation can be defined as 5.3 Structural PCC Overlays
(ITS, 2000):
5.4 Summary
"Measures to improve, strengthen or salvage existing
deficient pavements to continue service with only routine maintenance. Deficient pavements exhibit distress in
excess of what can be handled through routine maintenance."
A wholesale replacement of the entire pavement structure is considered reconstruction rather than rehabilitation since
it follows new pavement construction methods. Rigid pavement rehabilitation options depend upon local conditions
and pavement distress types but typically include:
● Dowel bar retrofit. Rigid pavements originally constructed without dowel bars may be subject to excessive
faulting and pumping due to poor load transfer. In these situations, retrofitting transverse cracks and/or
contraction joints with dowel bars can significantly increase load transfer and extend rigid pavement life.
● HMA overlays. HMA overlays are used for two primary purposes: structural and non-structural. Structural
overlays are thicker and are designed to add structural support to the existing rigid pavement. Non-
structural overlays are designed as a wearing course and are generally assumed to provide no additional
structural support.
● PCC overlays. PCC overlays are structural in nature and can be divided into two types (Mack, Hawbaker
and Cole, 1998):
❍ Unbonded. Bonding between the existing rigid pavement and the PCC overlay
is intentionally prevented by using a slurry seal, BST, or HMA bond breaking
interlayer. Unbonded PCC overlays are typically 125 - 305 mm (5 - 12 inches)
thick (AASHTO, 1993).
❍ Bonded. The PCC overlay is purposely bonded to the existing rigid pavement surface. Thus, the
PCC overlay and existing rigid pavement act as a composite structure. This allows for thinner
PCC overlays.
● Rubblization or crack-and-seat. This involves breaking up the existing distressed rigid pavement and
overlaying the resulting surface with a flexible or rigid pavement. Rubblization and crack-and-seat surface
preparation is discussed in Module 7, Section 2.2.4, Flexible Overlays on Rigid Pavement.
This section will concentrate on structural overlays by describing several typical overlay design methods. Discussion
of rubblization or break-and-seat surface preparation techniques used to prepare existing rigid pavements for either
rigid or flexible overlays can be found in Section 7.2.2.4, Flexible Overlays on Rigid Pavement.
Specific WSDOT structural design policy is contained in the WSDOT Pavement Guide, Volume 1. In
general, WSDOT uses the following structural design procedures:
❍ Rigid. The AASHTO Guide for Design of Pavement Structures (1986 or 1993
version). This is an empirical procedure.
● Rehabilitation.
❍ PCC overlays. The AASHTO Guide for Design of Pavement Structures for
unbonded PCC overlays. This is an empirical procedure. Generally, only
unbonded PCC overlays will be used if a PCC surfacing is selected. Bonded
PCC overlays are not considered as a structural solution and have a higher than
acceptable risk of premature failure.
1. Cut slots across the joint (see Figure 10.30). Typically, three or four slots are cut across the joint in each
wheel path. These slots are cut parallel to the direction of traffic flow and must also be parallel to one
another so that the retrofitted dowel bars do not restrict slab expansion and contraction.
2. Insert dowel bars into the slots (see Figure 10.31). Each dowel bar is placed on a small support to keep it
at the correct elevation. A Styrofoam joint reformer and plastic end caps are used to allow the slab to
expand without bearing on the grout.
3. Fill the slot with grout (see Figure 10.32). A small maximum aggregate size (e.g., 10 mm (0.4 inch)) is
used to ensure the grout fills in completely around the dowel.
4. Diamond grind the entire pavement area (see Figure 10.33). This removes any elevation differences due
to faulting or grout placement.
Figure 10.30: Dowel Bar Slots Figure 10.31: Dowel Bars in Slots
Figure 10.32: Filling Slots with Grout Figure 10.33: Slots After Diamond Grinding
Retrofitting dowel bars has grown in popularity over the last 10 years and has resulted in good
pavement performance.
5.3.1 Unbonded
An unbonded PCC overlay consists of a relatively thick PCC layer (typically 125 - 305 mm (5 -
12 inches) thick) over an existing rigid pavement. Bonding between the existing pavement
and overlay is intentionally prevented by using a slurry seal, BST, or HMA bond breaking
interlayer. This intentional separation allows the original pavement and overlay to act
independently of each other and helps prevent distresses in the existing pavement from
reflecting through into the overlay (ACPA, 2001). Unbonded overlays are generally used as
an alternative to rubblization when the existing rigid pavement is badly deteriorated. Their
primary advantages are that they (1) can be applied over a badly deteriorated pavement
without much surface preparation and (2) they do not require the existing pavement to be
removed. Their primary disadvantages are (1) because they are relatively thick and placed
directly over the existing pavement, they add substantially to roadway elevation, which could
pose overhead clearance problems, and (2) they are relatively expensive.
The design procedure contained in the 1993 AASHTO Guide is virtually identical to the AASHTO empirical design
for new rigid pavements with one exception: The effective modulus of subgrade reaction (k) is determined based on
the existing pavement resilient modulus. Although perfectly acceptable, this method gives little credit to the existing
pavement's remaining strength. The basic equation in the 1993 AASHTO Guide is:
1. Determine the slab thickness required to carry all future traffic (Df). This is done as it would be in the
normal AASHTO empirical design process for new rigid pavements. Material properties used in these
calculations should be those of the overlay material and not the existing rigid pavement. Other values, are
specified in the 1993 AASHTO Guide.
2. Determine the effective thickness of the existing slab (Deff) by one of the following two methods:
❍ From a condition survey. The following equation is used to adjust the actual existing slab
thickness based on condition survey results:
Fjcu = joints and cracks adjustment factor - accounts for the extra loss
of serviceable life caused by deteriorated transverse joints and
cracks in the existing pavement.
❍ From a remaining life calculation. This method does not account for any benefits from pre-overlay
repair. The following equation is used to determine the remaining life as a percentage of total life:
Joints are typically designed to be mismatched with underlying existing joints. The
assumption is that the thick overlay and bond breaking interlayer will prevent reflective
cracking, therefore mismatching the joints will improve load transfer. AASHTO recommends
that "the placement of joints in the overlay should be mismatched from existing joints and
working cracks by at least 0.9 m (3 ft.) where possible" (AASHTO, 1990).
5.3.2 Bonded
A bonded PCC overlay consists of a relatively thin PCC layer (typically less than 100 mm (4
inches) thick) over an existing rigid pavement. The overlay is intentionally bonded to the
existing pavement with a PCC slurry or grout in order to create a composite pavement section
(McGhee, 1994). Bonded overlays are generally used to add structural capacity to existing rigid
pavements that have little deterioration (e.g., no faulting or spalling and cracked slabs should be
replaced before overlay). Their primary advantages are that they (1) are thinner than
unbonded overlays and (2) their structural design accounts for the strength of the underlying
pavement. Their primary disadvantages are (1) they should not be applied over badly
distressed pavements because the distress may affect bond quality, and (2) they are
dependent upon good bond development - if for some reason this does not occur, the
pavement could be structurally inadequate.
The design procedure in the 1993 AASHTO Guide is somewhat similar to that for unbonded overlays, however, it is
more dependent on existing pavement characteristics. Specifically, there are three key differences:
1. The overlay thickness is a linear function of the total required slab thickness (Df) and effective slab
thickness (Deff) instead a quadratic function. This results in a thinner overlay for a given Df and Deff. In
essence, the bonded overlay design procedure gives more credit to the existing pavement's structural
contribution.
2. Design is based on the material properties of the existing pavement structure because of the expected
bonding.
3. Effective thickness (Deff) calculations incorporate more aspects of the existing pavement's structure.
Because the design relies more heavily on the existing pavement's structure, it's evaluation should be more
rigorous.
1. Determine the existing pavement structure and material properties. This includes determining the existing
slab thickness, type of load transfer, type of shoulder design (flexible or rigid), condition survey, effective
modulus of subgrade reaction (k), PCC parameters (elastic modulus, modulus of rupture), joint load
transfer, etc.
2. Determine the slab thickness required to carry all future traffic (Df). This is done as it would be in the
normal AASHTO empirical design process for new rigid pavements. Material properties used in these
calculations should be those of the existing material and not the overlay. Other values, are specified in the
1993 AASHTO Guide.
3. Determine the effective thickness of the existing slab (Deff) by one of the following two methods:
❍ From a condition survey. The following equation is used to adjust the actual existing slab
thickness based on condition survey results:
Fjc = joints and cracks adjustment factor - accounts for the extra loss of
serviceable life caused by deteriorated reflection cracks in the overlay that
will result from any unrepaired distresses in the existing pavement.
Fdur = durability adjustment factor - accounts for the extra loss in serviceability
caused by any durability problems (such as "D" cracking) in the existing
pavement.
Ffat = fatigue damage adjustment factor - accounts for past fatigue damage in the
existing pavement.
❍ From a remaining life calculation. This method does not account for any benefits from pre-overlay
repair. The following equation is used to determine the remaining life as a percentage of total life:
Joints are typically designed to coincide with underlying existing joints, otherwise
uncontrolled reflective cracking may occur and substantially affect overlay performance
(McGhee, 1994). Dowels and reinforcing steel are generally not placed in the overlay joints.
5.5 Summary
Rehabilitation essentially reverses the effects of deterioration by adding or replacing material in the existing
pavement structure. This section has concentrated on structural overlays of existing rigid pavements. These overlays,
which can be either flexible or rigid, are used to increase a pavement's structural capacity. In order to do this, they
must be structurally designed using one of several methods. Flexible overlays tend to be less expensive, thinner and
quicker to construct, while rigid overlays are more expensive, thicker and take longer to construct, but may offer
longer life. The choice of overlay type and method is highly dependent upon local practice and conditions.
Finally, new road construction in the U.S. is not nearly as prolific as it has been in previous generations. Urban areas
have filled out greatly and the ratio of existing roads to new roads is now quite high. Consequently, rehabilitation
(and not new construction) has become the dominant force in today's pavement design and construction arenas.
1 INTRODUCTION
"Pavement management" refers to a systematic process of maintaining, upgrading and operating a
network of pavements. Pavement management involves the following three major components:
● Pavement life-cycles. This includes how pavements are built, how their condition changes
over time, and how this process can be affected by different forms of maintenance,
rehabilitation and reconstruction.
● Costs associated with the pavement life-cycle. This includes the costs of initial construction,
maintenance and rehabilitation, assessing end-of-life pavement salvage value, and
determining user costs incurred throughout the life-cycle.
● Pavement management systems. This includes all the different systems used to determine the
most appropriate time to rehabilitate pavement, what the most cost-effective method is, and
how many dollars it will take to maintain a roadway system at a desirable condition level
(WSDOT, 1994).
http://training.ce.washington.edu/WSDOT/Modules/11_pavement_management/11-1_body.htm4/2/2008 6:42:54 PM
11.2 Pavement Management - Pavement Life Cycle
2 Pavement Life-Cycle
This section discusses the basic pavement life-cycle Major Topics on this Page
and how maintenance and rehabilitation affect this life-
cycle. As noted previously, all pavement deteriorates 2.1 Effects of Maintenance and Rehabilitation
over time. Typically, pavement deteriorates at an ever-
increasing rate: at first very few distresses are present 2.2 Effects of Maintenance and Rehabilitation
and the pavement stays in relatively good condition, Timing
but as it ages more distresses develop with each
2.3 Summary
distress making it easier for subsequent distresses to
develop. For instance, once a substantial crack occurs
it is then easier for water to (1) infiltrate the HMA layer and (2) penetrate and weaken the subgrade.
Additionally, freeze-thaw problems in the crack may develop and any expansive materials that get into
the crack (such as dirt, sand or weeds) may make the crack even wider thus compounding the previous
problems.
1. They immediately improve the pavement condition. For instance, a slurry seal can eliminate
most minor surface distresses.
2. They affect the future rate of deterioration. For instance, crack sealing prevents water from
entering the pavement structure and subgrade through open cracks, which slows future
deterioration.
In general, maintenance can slow the rate of deterioration by correcting small pavement defects before
they can worsen and contribute to further defects. Beyond a certain point, however, defects become too
large for correction by mere maintenance. At this point, rehabilitation can be used to effect a wholesale
correction of a large number of relatively severe defects, which provides a step increase in pavement
condition. Figure 11.1 illustrates this concept.
In general, there are several levels of treatment to correct pavement deterioration. These are (Deighton,
1997):
● Ancillary (leveling and spot repair, usually done in concert with other treatments)
Keep in mind that not all treatments will reset pavement condition to its original new-construction
value. Determining how a particular pavement should be treated, when and to what extent are all
decisions that can be made with the aid of a pavement management system (PMS).
Figure 11.2: Rehabilitation Time vs. Cost (based on an illustration in Stevens, 1985)
WSDOT experience has been that in order to restore pavement condition to a predetermined
level, it will cost 2 to 3 times as much if the pavement is allowed to deteriorate for 2 to 3
years beyond the optimum rehabilitation point.
2.3 Summary
Left alone pavements will typically deteriorates over time at an ever-increasing rate. Maintenance and
rehabilitation can slow or reverse this deterioration. The degree to which this occurs is dependent on the
type of maintenance or rehabilitation as well as the timing of such actions. In general, an early and
systematic maintenance and rehabilitation plan is the most cost effective and results in the greatest
extension of useful pavement life. This concept is further illustrated in Figure 11.3 below.
LCCA should be used as a decision support tool when selecting pavement type, determining structure and mix
type (for flexible pavements), construction methods, as well as maintenance and rehabilitation strategy.
Typically, LCCA involves the following basic steps:
● Make initial strategy and analysis decisions. Certain baseline decisions, estimates and assumptions are
needed in order to establish the parameters under which a LCCA can be carried out.
● Estimate costs. Costs associated with the owning agency and users are calculated for each alternative.
● Compare alternatives. Comparison usually involves expressing each alternative using a common
metric such as net present value (NPV) or benefit-cost ratio (B/C).
● Analyze the results and reevaluate alternatives. Results should be scrutinized for the most influential
costs, factors and assumptions. A sensitivity analysis is often used to do this. Original design strategy
alternatives should be reevaluated base on these results analysis in order to improve the cost-
effectiveness of each alternative.
The end result of a successful LCCA is not simply the selection of one alternative over the other but the selection
of the most cost-effective design strategy for a given situation and a greater understanding of the factors that
influence cost effectiveness. The following subsections describe the general steps involved in a LCCA.
Certain baseline decisions, estimates and assumptions are needed in order to establish the parameters under which
a LCCA can be carried out. These decisions, estimates and assumptions can be broadly categorized as follows:
● Alternative pavement design strategies. A "pavement design strategy" is the combination of initial
pavement design and necessary supporting maintenance and rehabilitation activities. LCCA is most
often used to evaluate two or more different pavement design strategies and determine their relative
value.
● Determine performance periods and activity timing. Because LCCA is performed in advance of actual
pavement construction, estimates of pavement performance (e.g., how will a pavement deteriorate over
time) and maintenance and rehabilitation effort timing (e.g., when should maintenance and
rehabilitation activities be scheduled) need to be made so that an appropriate analysis period can be
chosen and appropriate cost estimates made. Often, agencies can use past local experience to estimate
these parameters.
● Analysis period. The time period over which alternate design strategies are analyzed. The LCCA
analysis period should be sufficiently long to reflect long-term cost differences between alternatives.
For instance, if one pavement design alternative requires rehabilitation at the 15-year point and the
other requires rehabilitation at the 25-year point, an 20-year analysis period would not provide a fair
comparison between the two alternatives since it would include rehabilitation costs for one alternative
but not for the other. In general, the selected analysis period should include at least one rehabilitation
activity for each alternative. In the above case, a more appropriate analysis period might be 30 years or
even 50 years depending upon the rehabilitation activity timing. The FHWA recommends an analysis
period of at least a 35 years.
It is important to realize that the nature of these initial decisions, estimates and assumptions can be critical to the
outcome of a LCCA. As input parameters are changed, the cost-effectiveness of alternatives will change.
3.2 Costs
3.2.1 Agency Costs
Agency costs are all those costs incurred by the owning agency over the life of the project. Items common to all
alternatives need not be considered because their costs will cancel one another out. Agency costs include:
● Preliminary engineering. Costs associated with preliminary items such as feasibility studies of
alternative designs, permitting, engineering design and consultation for each alternative. For instance,
one alternative may involve significantly more wetland mitigation which could be reflected in
additional permitting and engineering costs.
● Initial construction. Construction costs associated with each alternative. For instance, each
alternative's different roadway sections and material quantities should be accounted for in the analysis.
Costs which are unique to each alternative should be included in the analysis.
● Maintenance costs. Costs associated with maintaining the pavement surface at some acceptable level.
Routine reactive-type maintenance cost data are often difficult to obtain. However, these costs are
generally small and do not vary greatly from alternative to alternative. They have a negligible effect on
NPV and can generally be ignored. When maintenance costs are available for the alternatives
considered, they should be incorporated into the life-cycle cost analysis.
● Rehabilitation costs. Costs associated with each rehabilitation alternative (typically they are
resurfacing costs). They are computed in a manner consistent with the initial construction costs.
● Administrative costs. Any other administrative or overhead costs unique to each alternative.
● Salvage value. The value of an investment alternative at the end of the analysis period. This is usually
included as a negative agency cost (an agency benefit) and is comprised of two major components:
❍ Residual value. The net value from recycling the pavement. The differential residual value
between pavement design strategies is generally not very large, and, when discounted over
long periods of time (e.g., 35 years) tends to have little effect on LCCA results.
❍ Serviceable life. The remaining life in a pavement alternative at the end of the analysis
period. Serviceable life accounts for the differences in remaining pavement life between
different alternatives. For instance, suppose alternative A reaches terminal serviceability at
year 30, while alternative B requires a rehabilitation at year 25 that will extend its life for
another 10 years. If a 30-year analysis period is used then alternative A has no remaining
serviceable life at the end of the analysis period while alternative B has 5 years of remaining
service life. This additional serviceable life must be accounted for in LCCA and is usually
done so under the "agency cost" category. In this example, alternative B would be credited
with a monetary value equivalent to 5 years of service life. Often this is done by calculating
this 5 years as a percentage of design life remaining (5 years remaining on a 10-year
rehabilitation design life would give 50%) and then multiplied by the cost of alternative B's
rehabilitation at year 25.
Sunk costs are a special category of costs. Sunk costs are those that are irrevocable and should not be
used to influence the alternative selection decision. Consider the following example:
A certain pavement design strategy, alternative A, with a NPV of $20 million was selected
two years ago and has already has its initial preliminary engineering completed at a cost of
$1 million. Currently, a new technology has been approved for use and a second pavement
design strategy, alternative B, with a NPV of $15 million is put forward for consideration.
The already completed preliminary engineering cost for alternative A ($1 million) is a sunk
cost - the money has been spent and is irrevocable. Therefore, in a life-cycle cost
comparison comes down to the remaining $19 million for alternative A versus the $15
million for alternative B. There should extra consideration for alternative A just because $1
million (in NPV) has already been spent it. In this case, the lowest life-cycle cost is still
alternative B even though $1 million (in NPV) has already been spent on alternative A.
1. Normal operation. Roadway user costs associated with using a facility during periods free of
construction, maintenance, and/or rehabilitation activities that restrict the capacity of the facility. These
costs are generally driven by pavement roughness.
2. Work zone. Roadway user costs associated with using a facility during periods of construction,
maintenance, and/or rehabilitation activities that generally restrict the capacity of the facility and
disrupt normal traffic flow. These costs are influenced by the level, duration, and character of capacity
restriction (e.g., number of closed lanes, length of closure, traffic during closure, amount of stopping
and starting, etc.).
Often times, normal operation user costs are assumed to be equal for all alternatives involved and only work zone
user costs are analyzed. In general, user costs are an aggregation of three separate cost
components:
● Vehicle operating costs (VOC). Includes all costs associated with operating a
vehicle including fuel, oil, part replacement, upkeep and maintenance. Vehicle
operating costs will vary depending upon roadway condition (Figure 11.4 shows the relationship
between VOC and roughness for a stretch of roadway in Washington State).
● User delay costs. The costs associated with highway users' time. User delay
costs help quantify costs associated with slow downs due to construction and maintenance
activities and denial-of-use. User delay costs are the most difficult and most controversial life-cycle
cost to accurately calculate because they involve assigning a dollar value to individuals' delay time.
Table 11.1 (from Walls and Smith, 1998) presents typical dollar ranges for different vehicles.
● Crash costs. The costs associated with highway accidents. Generally crash costs
are categorized into fatality, non-fatal injury and property damage only.
Figure 11.4: Percent Increase in Vehicle Operating Costs (VOC) for Various Vehicle Types
as a Function of Roughness
Data are taken from HDM-4 calculations using basic default input values and pavement
deterioration models calibrated for Interstate 5 (northbound lanes) pavements between Olympia
and Marysville, Washington.
Trucks
Passenger Cars
Single-Unit Combination
n = year of expenditure
n = analysis period (i.e., the number of years into the future over which
you wish to compare the projects)
3.4 Analysis
Once initial NPV's have been calculated for all alternatives they should be analyzed to determine the relative
effects of inputs, the distribution of likely input values and the probability distribution of resultant NPVs. This
analysis helps in determining which alternatives are better in which situations and also where improvements can
be made to each alternative to make it more cost effective. Generally, analysis should involve a sensitivity
analysis and a risk analysis.
1. Potential variability of input parameters. Typically, it is not possible to predict the exact value of an
input parameter. Therefore, it is better to describe input parameters by a range of probable values
rather than one single most likely value. LCCA results based on input parameters described this way
will give a range of life-cycle costs.
2. Likelihood of occurrence for an input parameter value. Although sensitivity analysis can show how
the final LCCA result varies as input parameters are varied, it does not account for the relative
likelihood of each one of these variations. Therefore, input parameters are best described as a
probability distribution, which accounts for a range of values and their relative likelihood. LCCA
results based on input parameters described in this way will give a probability distribution of life-cycle
costs.
Probability analysis is important to perform because it can a range of potential life-cycle costs and their associated
probabilities of occurring. With this level of information, an agency can assess the risks associated with a
particular probability distribution of life-cycle costs (e.g., is it acceptable to accept the 20% chance that the
project will cost more than $10 million?) and make the most informed decision possible. Furthermore, if
probability analysis is not performed and left to an evaluator's intuition, this type of subjective risk analysis can
be wrong for any number of reasons including incomplete data, incorrect data or a poor perception of the risk.
● The Asphalt Pavement Alliance (APA) Web site (www.asphaltallinace.com). The APA has software
conforming to FHWA procedures available for free download.
3.6 Summary
LCCA can best be summarized by two paragraphs from Walls and Smith (1998):
"LCCA results are just one of many factors that influence the ultimate selection of a pavement design
strategy. The final decision may include a number of additional factors outside the LCCA process, such as
local politics, availability of funding, industry capability to perform the required construction, and agency
experience with a particular pavement type, as well as the accuracy of the pavement design and
rehabilitation models. Chapter 3 of the 1993 AASHTO Pavement Design Guide further discusses such
other factors. When such other factors weigh heavily in the final pavement design selection, it is
imperative to document their influence on the final decision.
Many assumptions, estimates, and projections feed the LCCA process. The variability associated with
these inputs can have a major influence on the confidence the analyst can place in LCCA results. It all
depends on the accuracy of the inputs used. The accuracy of LCCA results depends directly on the
analyst’s ability to accurately forecast such variables as future costs, pavement performance, and traffic for
more than 30 years into the future. To effectively deal with the uncertainty associated with such forecasts,
a probabilistic risk analysis approach...as increasingly essential to quantitatively capture the uncertainty
associated with input parameters in LCCA results."
Originally described as “a systems approach to pavement design”, the term “pavement management
system (PMS)” came into popular use in the late 1960s and early 1970s to describe decision support
tools for the entire range of activities involved in providing and maintaining pavements (OECD, 1987
and Peterson, 1987). Hudson et al. (1979) describe a “total pavement management system” as
“…a coordinated set of activities, all directed toward achieving the best value possible for the
available public funds in providing and operating smooth, safe, and economical pavements.”
Haas and Hudson (1978) expand on this by defining “activities” as those actions associated with
pavement planning, design, construction, maintenance, evaluation and research.
There are numerous different pavement management systems (PMS) from which to choose, each one
with its own level of complexity. For a small town or rural county a simple system based on visual
inspection and maintained in a Microsoft Excel or Access database may be more than sufficient. For a
state road network a more complex PMS is usually appropriate.
Most formal definitions of a “pavement management system” agree on five key components (Peterson,
1987):
1. Pavement condition surveys. Pavement condition surveys were probably the first PMS
component to be adopted on a large scale by U.S. transportation agencies. For example,
WSDOT, a PMS early adopter, began pavement surveys in about 1965 (Nelson and LeClerc,
1982). Condition survey research is largely concerned with advancing or refining
measurement and data collection.
2. Database containing all related pavement information. Databases have evolved along with
the pavement condition survey data they are designed to house. Computer databases gained
prevalence in the 1970s and as adequate, cost effective computing power and storage became
available. Recent research has concentrated on implementing more robust databases (e.g.,
Microsoft SQL server, Oracle, etc.) and better user interfaces including GIS-based spatial
interfaces. These interfaces are as important as the data itself because they enable users to
view and manipulate data in a meaningful way.
3. Analysis scheme. Analysis schemes are those algorithms used to interpret data in meaningful
ways. The late 1960s and early 1970s saw the introduction of computer-based optimization
algorithms (Haas et al., 1979). Recent software can combine the database, analysis scheme
and decision criteria in one package. Recent research has focused on advancing or refining
life-cycle costing analysis, optimization algorithms and performance prediction.
4. Decision criteria. Decision criteria are those rules developed to guide pavement management
decisions. As pavement management systems have evolved, decision criteria have become
more complex and now account for items such as user delay, vehicle operating costs and, in
limited cases, environmental effects. Research is ongoing to develop and refine appropriate
decision criteria and the ability to automatically apply these criteria.
is generally concerned with high-level decisions relating to network-wide planning, policy and budget.
For example, managers at this level might compare the benefits and costs for several alternative
programs and then identify the program/budget that will have the greatest network benefit-cost ratio
over the analysis period.
The project level deals with smaller constituent sections within the network and is generally concerned
with lower-level decisions relating to condition; maintenance, reconstruction and rehabilitation (MR&R)
assignment; and unit costs. For example, at this level detailed consideration is given to alternative
design, construction, maintenance and rehabilitation activities for specific projects. This might be
accomplished by comparing benefit-cost ratios of several design alternatives and picking the design
alternative that provides the desired benefits for the least total cost over the projected life of the project.
Approaches to pavement management tend to attack this two-level system either from the top down by
dealing with network-level decisions first, or from the bottom up by dealing with project-level decisions
first. Either method can be quite detailed or relatively simple depending upon data amount and quality
and desired analytical capabilities.
As an example, the AASHTO 1990 Guidelines for Pavement Management Systems sets forth three basic
pavement management methodologies, two are bottom-up and one is top-down. The following lists
these methodologies and discusses their basic approach and why they would be considered project-level
approaches (bottom-up) or network-level approaches (top-down) (Zimmerman and ERES
Consultants, 1995):
1. Pavement condition analysis (project-level approach). This method, considered the simplest
of the three, aggregates pavement condition information at the project level and then selects
the most appropriate MR&R strategy. Each project is assigned a priority based on a number
of factors including location, traffic, safety, etc. Pavement MR&R projects are then selected
by priority based on a network-level budget. This system is simple but limited in
effectiveness because it does not consider future pavement condition. Because the first
decisions are made at the project-level, this would be considered a project-level approach.
2. Priority assessment models (project-level approach). This method improves upon pavement
condition analysis by incorporating predicted future pavement condition information. Thus,
priority assessment models consider future predicted conditions and can possibly perform
limited “what if” scenarios based on network level decisions. However, conditional (“what
if”) scenario capability is limited because alternate decisions can only be modeled by
changing the project-level data – generally a time-consuming task. This is also considered a
project-level approach because, although more complex, it still begins at the lowest decision
level (determining an individual pavement section’s MR&R strategy) and progresses to the
highest decision level (an overall pavement network strategy).
3. Network optimization models (network-level approach). This method, considered the most
sophisticated, simultaneously evaluates an entire pavement network to determine the optimum
network management strategy. Specific MR&R projects and locations are then selected to
meet this strategy. This method is considered a network-level approach because it begins at
the highest decision level and progresses to the lowest decision level.
The most appropriate pavement management approach depends largely on the situation. The top-down
network-level approach offers (1) better institutional control, (2) clear advantages in conditional scenario
capability and (3) is best able to accommodate the ever-changing political and social issues inherent in
our republican form of government. The bottom-up project-level approach provides only basic
conditional scenario capabilities but can provide much more detailed and accurate data that is capable of
supporting individual project decisions.
Network-level approaches are very powerful and can produce optimum solutions for the entire defined
system. Key elements in a network-level approach are:
● System definition. The network-level approach optimizes solutions for the defined system. If
the system is incorrectly defined, solutions will not be optimal. For instance, many consider
the appropriate system for pavement management to be the larger transportation infrastructure
rather than the more limited pavement network.
● Network model. Network-level decisions, and thus all decisions, are based on outputs from a
complex simulation model. Thus, these decisions are only as good as the model. Inputs,
accuracy, sensitivity, assumptions and calibration must be known and considered when
selecting an appropriate model.
The Highway Design and Maintenance Standards Model (HDM-4), developed by the World Bank, is a
good example of a network-level pavement management system. More information on HDM-4 can be
obtained at: http://hdm4.piarc.org/main/home-e.htm. One shortcoming of a purely network-level
approach is that data concerning individual roadway sections may not be available in enough detail to
make fully informed project decisions.
In sum, the network-level approach is powerful and sophisticated, but requires large amounts of data/
resources and attention to detail in order to function in a meaningful manner.
Project-level approaches can be very useful and constitute the majority of pavement management
systems in operation today. The key elements in a project-level approach are:
● Project-level vs. network-level goals. Since decisions are made first at the project level, the
project-level approach requires more effort to coordinate the anticipated or promulgated
network-level priorities down to the project level.
● Project ranking. This will determine included/excluded projects based on network-level goals.
Many State and local pavement management systems can be classified as project-level. In sum,
although less capable at producing optimum solutions and conditional scenarios, the project-level
approach is advantageous because it maintains the detailed project-level information needed to make
fully informed project-level decisions.
1. Optimize solutions for the entire network. By definition, this is what a network-level
approach does. For instance, a network-level approach can optimize the cost-benefit ratio for
the entire network. This would seem the most logical since the system, rather than an
individual project, is the overarching concern. Project-level approaches attempt to replicate
this ability by assigning project priorities that are commensurate with network-level programs,
decisions or budgets. However, because projects are already planned before high-level
decisions are made, project-level decisions and priorities may not be consistent with network-
level decisions and priorities. This can lead to a suboptimal system solution that is driven by
individual project-level decisions instead of network-level decisions.
2. Quickly and accurately produce conditional scenarios. Software models using a network-
level approach allow the user to adjust top-level budget and policy inputs and then quickly
calculate the resulting network-wide effects because these models are driven by top-level
(network-level) decisions. For example, a network-level model could calculate the economic
and pavement conditions effects of a proposed lower axle load limit law or the long-term
network performance under varying levels of funding. Conversely, project-level software is
generally driven by low-level inputs and thus, a change in top-level budget or policy would be
input into the system by adjusting the multitude of lower-level inputs – a more laborious
process.
3. Prioritize broad areas of MR&R. Since network-level analysis provides target MR&R
treatments and costs, these targets can be easily and consistently applied to individual
projects. In order to accomplish the same thing with a project-level approach, network-level
targets need to be provided in advance such that project-level decision can be made with
network-level targets in mind.
5. More easily obtain top management attention. At the 1997 New Orleans pavement
management workshop, many pavement management practitioners raised the following
issues: (1) pavement management had lost the attention of top management because they did
not understand its capabilities or importance, and (2) some managers haphazardly overrode
pavement management recommendations with little or no though to network implications
(Zimmerman, Botelho and Clark, no date given). With its conditional scenario capability, a
network-level approach could easily show the fiscal importance of pavement management as
well as the implications of various decisions.
1. Relies less on aggregate data. The network-level approach relies on aggregate data to drive
its models. There is an inherent risk involved when using aggregate data to drive network-
level decisions. First, the aggregate data, if not carefully chosen, may not be representative of
actual conditions, which could lead to incorrect decisions at the highest level. Second,
aggregate data is sometimes difficult to translate into specific project-level results (e.g.,
overlaying all pavement above a certain IRI on a particular route may be costly if these high-
IRI sections are small isolated segments).
2. Able to be used with little data. Project-level systems can be used in situations where data
availability/requirements are small making them ideal for smaller agencies. Network-level
systems require large amounts of data and resources (such as computers, trained personnel,
advanced algorithms), which smaller municipal agencies may not be able to afford.
Additionally, the networks managed by these smaller municipal agencies may be simple
enough that they do not require a network-level analysis. For instance, a small city with
perhaps 40 km (25 miles) of roadway could probably use a simple pavement condition
analysis method to meet all their information needs. Project-level systems can, however,
require a significant amount of data depending upon their complexity and how they are
modeled.
3. Better link between network-level and project-level management decisions. Because decisions
flow from the bottom up, the high-level network decisions, although somewhat limited in
scope, in essence must be based on low-level project decisions. However, in a network-level
approach it is sometimes difficult to translate broad network-level decisions into specific
project actions. For instance, a decision to improve all IRI > 3.0 m/km roadways by overlay
may be made at the network level, but these segments of roadway may be located in small
sections over a wide area making for inefficient construction.
4. Less dependent upon feedback for success. Political sentiment, budgets, pavement condition
and MR&R strategies are highly dependent upon the local environment. Because of this,
network-level models will invariably need calibration, which can only occur through continual
feedback and updating. If this feedback/update process is interrupted or halted (e.g., through
budget cuts, personnel transfers) a network-level model's utility can quickly degrade.
Additionally, uncalibrated network-level models that are used will produce erroneous results.
5. Easier to obtain buy-in from others. Project-level approaches can be simpler and more easily
understood. Network-level approaches typically use sophisticated models that make many
generalizations and assumptions. Those who are unfamiliar with the model may be unwilling
to use its results because (1) they do not understand how it works or (2) they do not agree with
its generalizations and assumptions.
4.6 Summary
Pavement management systems assist decision makers in determining cost effective strategies for
maintaining, upgrading and operating a network of pavements. There are two general levels to
pavement management:
1. The network level. Deals with the pavement network as a whole. Structure is built to support
high-level decisions relating to network-wide planning, policy and budget.
2. The project level. Deals with smaller constituent sections within the network. Structure is
based on lower-level decisions relating to condition; maintenance, reconstruction and
rehabilitation.
Both the network-level and the project-level approaches have significant advantages and disadvantages.
In general, the advantages offered by the network-level approach, such as network optimization,
conditional scenarios, consistency and versatility are most suitable to a federal, state or large municipal
pavement management system. These agencies have large enough networks that manual or ad hoc
methods of network optimization, conditional scenario generation and maintaining consistency are
prohibitively difficult. Additionally, these larger organizations are better able to handle the large data
and resource requirements of a more sophisticated system. Conversely, the advantages offered by the
project-level approach, such as simplicity, economy and familiarity are more suitable for smaller
agencies such as local municipalities who generally have small networks, few resources and minimal
information requirements.