Long-Term Assessment of Asphalt Trackbed Component

Materials’ Properties and Performance

by

Jerry G. Rose, PE
Professor of Civil Engineering
161 OH Raymond Bldg
University of Kentucky
Lexington, KY 40506-0281
859/257-4278, [email protected]

and

Henry M. Lees, Jr., PE

Sr. Engineer-Track & Structures
BNSF Railway Company
920 SE Quincy Street
Topeka, KS 66612-1116
785/435-6459, [email protected]

Submitted for Presentation at the 2008

AREMA Annual Conference
Salt Lake City, September 2008
and Publication in the Proceedings

Word Count: 5,794

June 1, 2008
 ABSTRACT

The uses of Hot Mix Asphalt as subballast layers within railroad track structures for new

trackbed construction and trackbed maintenance applications have grown steadily in the

United States during the past 25 years. The asphalt layer (termed underlayment) is

used in lieu of an all-granular subballast layer. This paper documents the results of a

characterization and evaluation study to ascertain the effects of long-term exposure in

various trackbed environments on the material properties of the trackbed materials –

asphalt and underlying (roadbed) subgrade. The primary purpose of the testing program

was to determine if any weathering or physical/chemical deterioration of the materials

were occurring that could adversely affect long-term performance of the trackbeds. Six

asphalt trackbeds, ranging in age from 12 to 25 years; on heavy traffic revenue lines in

three states were recently core drilled. Test data on the trackbed materials were

compared to data obtained previously. The expected benefits and trackbed life

projections are discussed relative to current basic design and construction practices.

Keywords: hot-mix asphalt, railway trackbeds, trackbed performance, subgrades,

subballast
INTRODUCTION

From its beginnings in 1830, the railroads have been a primary mode of freight transport

in this country. Its dominance is becoming significant in recent years as train speeds,

gross ton-miles, and axle loads have increased. The latest Association of American

Railroads statistics (1) indicate that in 2005 an all-time record 1.7 trillion ton-miles of

freight was carried over the nation’s nearly 141,000-mile (227,000 km) railroad network.

The average freight car weight has increased to 129 tons (117 metric tons) with most

new cars having gross weights of 143 tons (130 metric tons). The importance of

developing and specifying premium track structures and components to adequately

carry the increased tonnage is a current reality of the industry. Failure of the track

structure and components results in difficulty maintaining track geometric features

necessary for efficient and safe train operations. Maintenance costs and track outages

increase due to frequent maintenance and renewal cycles.

The inability of the track structure to adequately carry the imposed loadings can

be categorized into two primary failure types. The first one is failure of the subgrade

when the pressure transmitted to the subgrade is higher than the inherent hearing

capacity of the particular subgrade. The subgrade soil’s ability to accommodate loading

pressures is a function of its shear strength, cohesion, plasticity, density, and moisture

content. A well-compacted subgrade soil that is confined and maintained reasonably dry

will normally perform adequately for an indefinite period of time. A possible exception is

a highly compressible soil such as peat. Subgrade failures adversely affect track

geometry and are normally difficult and expensive to correct.

 The second type of trackbed failure occurs when one or more of the trackbed

structural components fail to perform satisfactorily for a reasonable period of time.

This is commonly manifested by the subballast, and particularly the ballast, becoming

clogged (fouled) with excessive quantities of fine size material. This lowers the shear

strength of the ballast and bearing capacity of the subballast. Fouling is normally due to

degradation of the ballast, infiltration of subgrade soil particles, extraneous droppings

from hopper cars, or an accumulation of wind-blown fine particles. Track geometry is

adversely affected to varying degrees. It is difficult to rectify track geometry in fouled

ballast with typical trackbed maintenance surfacing equipment.

Periodic replacement of the track components (rails, ties, fasteners, and special

trackworks) cannot be avoided (2). It is desirable to increase the service life of the

components. The adequacy of the trackbed structural components supporting the track

can have a significant effect on the life of the track components by reducing impact

stresses and minimizing deflections of the track.

The solution for minimizing subgrade failures involves a combination of reducing

the pressure on the top of the subgrade, improving drainage (effectively improving the

properties of the subgrade), adding thickness to the trackbed structural components, or

utilizing higher quality/load bearing trackbed components. The solution for minimizing

structural component failure is designing and selecting reasonable fasteners and

track components so that an optimum track structural support stiffness will be achieved.

In order to design optimum track structural support stiffness, it is necessary to

determine the applied pressures at different levels in the track support structure and

select a combination of materials and thicknesses to withstand the applied pressures.

ASPHALT TRACKBEDS

The most common trackbed is composed of all-granular materials consisting of layers of

ballast and subballast over a prepared subgrade, as noted in Figure 1a. During the past

twenty-five years, the use of Hot Mix Asphalt as a subballast layer within the track

structure has steadily increased until it is becoming standard practice in many areas of

the United States. The asphalt-bound impermeable layer, typically 5 to 6 in. (125 to 150

mm) thick, provides a “hardpan” to protect the underlying roadbed and to support the

overlying ballast and track. Various tests and performance evaluations have shown

numerous advantages over traditional all-granular (ballast) trackbeds, particularly on

heavy tonnage lines traversing areas of marginal geotechnical engineering

characteristics (3, 4, 5).

The most common asphalt trackbed, termed asphalt underlayment as depicted in

Figure 1b, incorporates the layer of asphalt in lieu of the granular subballast. Ballast is

used above the asphalt layer in a similar manner as conventional all-granular trackbeds.

The ballast provides a protective cover for the asphalt by blocking the sunlight,

protecting the surface from air and water, and maintaining a relatively constant

temperature and environment. The ballast provides a means to adjust the track

geometry, when necessary, with typical maintenance equipment and procedures.

Recent studies involve instrumenting asphalt trackbeds with earth pressure cells

and displacement transducers to measure pressure levels and distributions within the

track structure and rail deflections under moving trains. These tests, conducted in real

time domain train operations with 286,000 lb (130 metric ton) cars, confirm the positive
attributes of the asphalt layer (6, 7). Peak dynamic pressures range from 13 to 17 psi

(90 to 120 kPa) on top of the asphalt layer. These are further reduced to 7 to 8 psi (50

to 55 kPa) under the asphalt layer at the subgrade interface. Dynamic track deflections

average 0.25 in. (6.4 mm) for wood tie track and 0.05 in. (1.3 mm) for concrete tie track.

These are considered optimum for quality trackbeds. Dynamic track modulus values

consistently average 2,900 lb/in/in (20 MPa) for wood tie track and 7,200 lb/in/in (50

MPa) for concrete tie track, also considered optimum stiffness levels.

BASIC ASPHALT TRACKBED DESIGN AND CONSTRUCTION PRACTICES

The asphalt mix is similar to that used for highway applications, but can be slightly

modified for optimum performance in the trackbed environment. It is placed as a layer or

mat of specified thickness and the common term is “underlayment” since the layer is

placed under the ballast and above the subgrade or old roadbed. It basically serves as

a subballast. A lesser used technique, known as “full-depth or overlayment” is

applicable for special situations and involves placing the track directly on the asphalt

layer with no ballast between the ties or slab and the asphalt. This technique is primarily

used in Europe and Japan (8, 9, 10).

The most common asphalt mix is produced as a hot mix asphalt, thus the

acronym – HMA. Cold mix asphalt mixtures and in-place stabilization of roadbeds with

liquid asphalts have been used sparingly. Normally the asphalt mix is produced in a

local mixing plant, at a temperature around 275°F (135°C), hauled to the site in dump

trucks, spread to the desired thickness, and compacted while being maintained at an

elevated temperature.
 The asphalt underlayment system is equally applicable for heavy tonnage freight

lines, high-speed passenger lines, commuter and transit lines, freight and intermodal

yards, ballast loadout facilities, and practically all types of special trackworks including

crossing diamonds, turnouts, tunnel floors, bridge approaches, and highway crossings.

The majority of the asphalt trackbed applications are on existing lines. The applications

number in the thousands and most have been used on in-service lines in conjunction

with rehabilitation or renewal of special trackworks, particularly when existing subgrade

support and drainage conditions are inferior. Current installation practices, which require

removal of the track, are not applicable for long sections of in-service lines since the

time required to remove and replace the track is not commensurate with typical work

windows. Studies are underway to develop equipment to place asphalt under a raised

track on in-service lines without removing the track.

New construction, particularly double-tracking and yard installations, account for

the largest projects. At these selected locations, conventional trackbed designs were

considered to be inadequate or uneconomical to provide the required level of long-term

performance because of inherent poor qualities of the roadbed support materials and

drainage conditions. The roadbed/subgrade is readily available for regular highway

paving practices prior to track placement (11, 12).

Recommended asphalt mixture specifications and trackbed section designs have

evolved over the years. Following is a summary of prevailing practices. Detailed

information is available elsewhere (5, 13).

Normally a local dense-graded asphalt highway base mix is specified, slightly

modified with an additional 0.5% asphalt (binder) cement content. The ideal design air
void content for the compacted asphalt layer is 2 to 3%. Typical asphalt layer width is 12

ft (3.7 m) and thickness ranges from 5 to 6 in. (125 to 150 mm). Ballast thickness above

the asphalt is from 8 to 12 in. (200 to 300 mm).

The roadbed should be reasonably well-compacted, well-drained, and capable of

accommodating the hauling and spreading equipment without excessive rutting or

deformation. A slight crown or side slope is desirable. The need to purposefully improve

sub-surface drainage, or improve support with additional granular material prior to

placing the asphalt, will depend on an analysis of the conditions at the specific site.

ASPHALT TRACKBED MATERIALS TESTS AND EVAULATIONS

Eight asphalt trackbeds, located in five different states, ranging from 12 to 29 years old

and having various asphalt thicknesses and trackbed support materials, were selected

for materials characterization studies. Pertinent classification and descriptive data for

the projects are presented in Table 1. Samples were obtained during summer 2007.

Previous characterization studies, primarily conducted in 1998 (14, 15), were available

for selected projects and are included herein for comparison purposes.

Samples normally were taken at three randomly selected locations at each

project. Samples were removed from the field side crib area (Figure 2). The following

sequence was followed at each location:

• Remove and sample ballast from crib area down to top of asphalt layer

• Measure ballast thickness and observe condition

• Obtain 6 in. (150 mm) diameter core sample with core drill
 • Protect samples from core drilling water so as to not contaminate the

underlying roadbed

• Measure asphalt core thickness, observe condition, and place in sealed

plastic bag

• Auger out roadbed samples, note distance below asphalt, separate if layered

conditions existed, place in sealed plastic bags

• Repeat drilling sequence, normally three cores were taken at each location

• Fill core holes with cold mix patch and replace ballast

Geotechnical Tests and Evaluations

The following geotechnical laboratory tests and evaluations using standard ASTM

procedures were conducted on the subgrade/roadbed samples:

• Moisture Content; in-situ condition – as sampled

• Grain Size Analysis; sieve and hydrometer

• Atterberg Limits; liquid limit, plastic limit, plasticity index

• Soil Classification Determinations; unified system

• Standard Proctor Moisture-Density

• California Bearing Ratio; unsoaked and soaked

The samples were recorded by depth below the asphalt and placed in separate

containers when differences in size, color, texture, or moisture content were observed.

The sealed containers were transported to the geotechnical laboratory at the Kentucky

Transportation Cabinet for subsequent tests.

 Table 2 contains the geotechnical evaluations for the subgrade/roadbed

samples. Data from the 1998 sampling is included for comparison with the recent 2007

data. Subgrade samples were obtained from four projects. The subballast and subgrade

were sampled separately at the Hoover site. This was the only project where granular

subballast was used below the asphalt. The Quinlan site had two distinctly different

subgrades due to differing topography. Thus, six different samples were analyzed for

the four projects.

The initial testing phase involved in-situ moisture content tests, grain-size

analysis, and Atterberg limits tests followed by soil classifications by the Unified

procedure. Based on the classifications, similar materials from a site were combined to

accumulate samples of sufficient size for the subsequent standard Proctor moisture-

density test to determine optimum moisture content for maximum dry density and for the

California bearing ratio (CBR) test.

In-Situ Moisture Contents

There was significant interest in determining the existing moisture contents of the

subgrade materials directly under the asphalt layer and subsequently comparing these

with previous measurements with the optimum moisture contents for the respective

materials. Every effort was made to remove core drilling water to protect subgrade

samples. No significant water penetrated the soil (particularly clay) subgrades. No

sample appeared to be overly wet or wet of optimum based on initial observations.

In-situ moisture contents are provided in Table 2 for both the 1998 and 2007

sampling operations. The values varied relative to the type of subgrade soil, but were
very site specific comparable with values obtained during the 1998 sampling. These

data are shown in Figure 3. There was an average net decrease of 0.1% change in

moisture contents over the span of nine years.

Two of the projects had in-situ moisture tests taken during similar coring

operations on several previous occasions, dating to the early 1980s. This data is

presented in Figure 4. The Oklahoma City trackbed has a highly plastic clay under the

asphalt. The range in moisture values is minimal. The Conway trackbed has the existing

old roadbed under the asphalt that is highly variable mixture of large-size ballast, small-

size ballast, cinder, coal, soil, etc. The significance of the data is that the average

moisture contents of the materials underlying the asphalt have remained essentially

unchanged at each respective site over the years from the time the asphalt was placed.

Previous concerns about pore water pressure, and its effects on lowering subgrade soil

strengths, are not founded.

Unified Soil Classifications

The soil classifications, based on grain size analyses and Atterberg limits tests, are

provided in Table 2. The test projects were selected to include a wide variety of

subgrade materials, ranging from reasonably high plastic clays to more silty/sandy

materials having little or no plasticity. As expected, little difference in soil classifications

was noticed at individual sites for the samples taken in 1998 and 2007.
Standard Proctor Moisture-Density

The standard Proctor moisture-density test was conducted to determine the optimum

moisture content for achieving maximum density. The minus 0.50 in. (12.5 mm) size

material was removed. The optimum moisture content data is included in Table 2.

Figure 5 shows the change in optimum moisture contents for the six samples between

1998 and 2007 sampling. The changes were typically less than 1 percent, indicating

similar materials.

Figure 6 is a graphical comparison of the measured in-situ moisture contents and

the Proctor optimum moisture values. The linearity of the relationship is shown in Figure

7. Note that the R2 value is in excess of 0.9 indicating very good correlation. The in-situ

moisture contents were very close to optimum values. These findings indicate that the

subgrade materials under the asphalt layer can be considered, for design purposes, to

have a prevailing moisture contents very near optimum for maximum compactability and

strength.

In addition, strength or bearing capacity values used in design calculations

should be reflective of optimum moisture content values. It is common practice, when

designing conventional all-granular trackbeds, to assume the subgrade is in a soaked

condition, which for most soils is a weaker condition than when the soil is at optimum

moisture.

California Bearing Ratio

California Bearing Ratio (CBR) specimens were prepared at moisture contents

determined from previous Proctor tests to be optimum for maximum density. Specimens
were tested immediately in the unsoaked condition. Companion specimens were

soaked in water for 96 hours prior to testing. Tests were conducted at 0.1 in. (2.5 mm)

penetration.

The CBR data is presented in Table 2. The values were typical for the types of

materials tested. For example, the highest CBR value was in the 50 range, which was a

select river gravel used as a subballast (locally known as “Tex-Flex” base), for the

Hoover project. A select crushed stone product is considered to have a CBR value of

100. The other subgrade materials have CBR values significantly lower, as expected,

even for the unsoaked condition.

A comparison of unsoaked and soaked CBR test values is presented graphically

in Figure 8. CBR values were significantly lower for the soaked samples, particularly

those containing clay size material, which had values in the low single digits. Test

results for the 1998 and 2007 sampling were reasonably close considering that

materials sufficient for only one unsoaked and one soaked specimen per site were

available for tests. Likely the 1998 and 2007 test comparisons would have been less

variable had additional tests been conducted to obtain averages based on several

replicable tests.

As noted previously, the in-situ moisture contents for individual samples were

very close to the those determined from the Proctor test to be near optimum. This

relationship is shown graphically in Figure 7. Since the unsoaked CBR values are

derived from tests on samples at optimum moisture contents, and the test results from

samples under asphalt trackbeds were determined to be at or very near optimum

moisture contents, it is obvious that the unsoaked CBR bearing capacity values are
appropriate to use for structural design calculations. The soaked (lower) CBR values

result in a conservative overdesign. The preceding statements are not necessarily

applicable to the open all-granular trackbeds, which are prone to variable moisture

contents depending on the amount of rainfall and surface drainage conditions, and

corresponding variations in support strength. The subgrade/roadbed materials

underlying the asphalt layers were at moisture contents near optimum, and based on

long-term monitoring at two sites, maintain optimum moisture conditions for indefinite

periods.

Asphalt Mixture and Core Tests and Analysis

The following laboratory tests were conducted on the asphalt mixtures and cores at the

National Center for Asphalt Technology (NCAT) at Auburn University:

• Density and Voids Analysis

• Asphalt (binder) Content

• Extracted Aggregate Gradation

• Resilient Modulus @ 5°C (41°F) and 25°C (77°F) @ 1 loading cycle per

second

• Dynamic Modulus @ 5°C (41°F) and 25°C (77°F) @ 1 hertz load frequency

• Recovered Asphalt Binder Properties

Penetration @ 25°C (77°F)

Absolute Viscosity @ 60°C (140°F)

Kinematic Viscosity @ 135°C (275°F)

Dynamic Shear Rheometer @ 25°C (77°F)

 Figure 9 depicts typical asphalt cores as obtained from the trackbeds. Table 3

contains results for the Mix Extraction Tests and Core Analysis. Table 4 contains test

results on the Recovered Asphalt Binders. The most recent test results are listed in the

far right columns. This represents 2007 data for six of the projects. The significance of

the prior tests is so that the changes in the properties and weathering characteristics of

the asphalt layers can be evaluated over a period of time.

Mix Extraction Tests and Core Analysis

The extraction test results (Table 3) are indicative of dense-graded base mixes with 1.0

in. (25 mm) maximum size aggregate and about 6 percent passing the No. 200 sieve.

These are basically in conformance with guidelines previously described (5, 15).

Asphalt binder contents vary somewhat, ranging from 4.5 to 7.0 percent. No particular

changes are evident in aggregate gradations or asphalt binder contents over the period

of years.

Tests on the asphalt cores included density and voids analyses and dynamic and

resilient modulus tests. The air voids were typically higher than desirable for five of the

sites ranging from 5 to 9 percent. The air voids were purposefully maintained at 2 to 3

percent range at three of the sites. This range is considered to be optimum to resist

premature oxidation of the binder. Average air voids for each site were less than the 8%

maximum normally believed to represent the upper limit to provide an impermeable

layer.

The industry standard dynamic and resilient modulus tests were used to measure

the modulus of elasticity of the asphalt cores. In both tests, repeated loads were applied
to a cylindrical specimen and the displacements were measured. The values, reported

in Table 3, were measured under uniaxial compression loading for the dynamic modulus

and under indirect tensile loading for the resilient modulus. Tests were conducted at two

standard temperatures which represent the nominal lowest, 5°C (41°F) and highest,

25°C (77°F), temperature asphalt experiences in the insulated trackbed environment.

Recent tests were limited to resilient modulus since it is now considered as more

representative of the actual stiffness of the asphalt core.

Values were typically several orders of magnitude higher at the lower

temperature, which is normal for a viscoelastic, thermoplastic material – and is

characteristic of the asphalt binder in the mix. At lower temperatures, the asphalt

becomes stiffer, as reflected in higher modulus (or stiffness) values. At higher

temperatures, the asphalt becomes less stiff. Obviously, for asphalt highway

environments, where the asphalt is exposed to greater temperature extremes, the

stiffness differences from winter to summer are significantly greater than those existing

in the insulated trackbed environment.

Figure 10 is a plot of Resilient Modulus versus Age of the asphalt mixes. The

circled symbols represent data for cores (obtained from the trackbed in 1998) that cured

the final nine years in the laboratory environment. They are plotted directly above the

railroad cured data for similar ages. Note that the modulus values for the cores cured

the last nine years in the laboratory were higher than the cores in the railroad

environment.

The measured modulus values are reasonably consistent for the various sites.

There is no particular trend or changes in modulus as a function of time. The mixes vary
in asphalt contents, densities, aggregate gradations, and binder properties from site-to-

site, which can be expected to produce variations in modulus values. However, these

variations are minimal. The significant factor is that the values are reasonably typical for

new, unweathered mixes not exemplifying fatigue and cracking – thus low values, or

exemplifying hardening/weathering of the binder – thus high values. The values are

basically intermediate in magnitude, even after many years of loading and weathering in

the trackbed. The asphalt appears to be undergoing little, if any, weathering or

deterioration in the trackbed environment.

Recovered Asphalt Binder Tests

Test results for Penetration, Absolute and Kinematic Viscosities, and Dynamic Shear

Rheometer on the recovered asphalt binders are contained in Table 4. Plots of

Penetration and Absolute Viscosity versus Age of the Asphalt Underlayments are

contained in Figures 11a and 11b. The data points circled at the ends of the trend lines

represent the 2007 values. The preceding data points are nine years prior, or 1998

values.

Penetration values will tend to decrease and viscosity values will tend to increase

with time due to expected oxidizing and hardening of the asphalt binders. There is

indication of this phenomenon when comparing the 1998 and 2007 test values.

However, the Abson method (ASTM D1856) was used for the 1998 and prior asphalt

recoveries; whereas, the Rotary Evaporator method (ASTM D5404) was used for the

2007 recoveries. The Rotovapor method is considered more effective at removing the

solvent. Therefore, the 2007 penetration values would be expected to be lower and the
2007 absolute viscosity values would be expected to be higher than their respective

1998 values. These trends are evident from Figures 12a and 12b respectively.

It is likely that the original asphalt binders were PAC 60-70 penetration or AC-20

viscosity graded. The effects of short-term aging (elevated temperatures) during the

pavement construction process and long-term aging for several years will reduce the

binder penetration to the 25 to 40 range and the absolute viscosity at 60°C (140°F) will

be maintained to less than 15,000 poises (17). These samples meet these criteria,

indicating minimal oxidation and weathering.

The Dynamic Shear Rheometer (DSR) procedure for evaluating asphalt binders

was developed in the mid-1990s. Fortunately this test was conducted in 1998 on

samples from 5 of the 6 sites and this data is compared to the 2007 data in Figure 13.

The standard for performance grade asphalt binders, after short- and long-term aging, is

that the DSR at 25°C (77°F) should be less than 5,000 kPa. Note in Figure 13 that all of

the samples are well below 5,000 kPa, another indication that the asphalt binders in the

trackbed cores are not oxidizing and hardening excessively (17).

Discussion

It is not surprising that the asphalt binder in the trackbed cores are not oxidizing and

hardening to the extent normally observed for asphalt highway pavements. This is

largely due to two factors. The surface of the asphalt is typically submerged 20 in. (500

mm) from the surface (atmosphere) by the ballast/tie cribs and the depth of ballast

below the ties. The lack of sunlight and reduced oxygen largely negates normal

weathering which occurs in highway pavements exposed to sunlight.

 Secondly, the range in temperature extremes which the HMA mat undergoes

from summer to winter is significantly less in the insulated trackbed environment than

for exposed highway pavements. This information was developed initially during 1982

and 1995 tests in Kentucky from buried thermistors, and reported previously (14) and

reproduced in Table 5. Additional tests during 2000 at the AAR Pueblo test site

confirmed the previous tests (6).

SUMMARY AND CONCLUSIONS

The primary purpose of this investigation was to determine, based on test results,

current materials properties of the asphalt and underlying materials in order to assess if

any weathering or deterioration of the materials was occurring in the trackbed

environment which could adversely affect long-term performance.

Material characterization evaluations were conducted on asphalt cores and

subgrade/roadbed samples from eight asphalt trackbeds. The trackbeds were from 12

to 29 years old when tested and were distributed over five states. The inherent

conditions varied significantly from site-to-site. These included asphalt thickness and

composition, ballast thickness, trackbed support, and traffic. Previous characterization

evaluations were available for the projects and the results were included for

comparisons with recent evaluations.

The significant finding relative to the materials (old roadbed/subgrade) directly

under the asphalt layer, is that the in-situ moisture contents are very close to laboratory

determined optimum values for maximum density of the respective materials. The

asphalt layer is not performing as a membrane to collect and trap moisture, thus
weakening support. Actually, since the in-situ moisture contents are at or near optimum

for maximum density, the strengths and load carrying capacities of the underlying

materials are also at or near optimum. Furthermore, average moisture contents remain

essentially unchanged, at or near optimum, for the two projects from which previous

data was available. For design purposes, it is reasonable to base strength or bearing

capacity values at optimum conditions (moisture content and density) for the material

under the asphalt layer. Using strength or bearing capacity values determined for the

soaked condition, common for highway designs, is inappropriate for asphalt trackbed

designs. The unsoaked, optimum moisture content condition is consistent with in-

service trackbed conditions.

An equally significant finding, relative to the asphalt cores characterizations, is

that the asphalt binders and asphalt mixes do not exhibit any indication of excessive

hardening (brittleness), weathering, or deterioration even after many years in the

trackbed environment. This is considered to be primarily due to the insulative effects of

the overlying ballast which protects the asphalt from excessive temperature extremes

and oxidation and hardening of the asphalt binder. These factors will contribute to a

long fatigue life for the asphalt layer. There is no indication that the asphalt layers are

experiencing any loss of fatigue life based on resilient modulus test on the extracted

cores.

The typical failure modes experienced by asphalt highway pavements are 1)

rutting at high temperatures, 2) cracking and fatigue at low temperatures, 3)

stripping/raveling under the suction of high tire pressures on wet pavements, and 4)

progressive fatigue cracking due to inadequate subgrade support, generally augmented

by high moisture and improper drainage. These conditions do not exist in asphalt

railroad trackbeds. For example, the temperatures are not sufficiently high to promote

rutting. Conversely, the temperatures are not sufficiently low to promote low

temperature cracking and decreased fatigue life, nor does the asphalt binder weather or

harden excessively in the insulated trackbed environment which would have further

negative influence on cracking and fatigue life. Obviously the tendency to strip/ravel is

essentially eliminated in the trackbed environment since there is no rubber suction

action. Also, the moisture contents of the underlying subgrade/roadbed support

materials are maintained at or near optimum for maximum density and support strength.

In addition, peak dynamic vertical pressures on top of the asphalt layer are

typically less the 20 psi (138 kPa) under 286,000 lb (130 metric ton) locomotives and

heavily loaded cars. (16) This is only two to three times larger than the pressure exerted

by an average-size person standing on an asphalt pavement, and much less than

pressures exerted by heavily loaded highway tracks, which can be in excess of 100 psi

(690 kPa). These peak dynamic pressures are further reduced to less than 10 psi (69

kPa) under the asphalt layer at the subgrade interface (6).

Based on the findings and analyses of the research reported herein, asphalt

underlayments installed in conformance with the basic design and construction

practices also reported herein, should have an extremely long service life as a premium

subballast to properly support railroad tracks. There is no indication of any deterioration

or cracks of the asphalt after many years of heavy traffic under widely varying

conditions.
 Ancillary benefits of a long-lasting premium subballast support material for

railroad tracks include the following: increased strength, decreased abrasion, and

increased life of the ballast; decreased wear and improved fatigue life of the ties, rail,

and premium-cost track components such as special trackworks; a consistent level of

track stiffness (modulus) designed for optimum levels; reduced maintenance activities

and associated track closures; and improved adherence to track geometric parameters.

All of these benefits impact favorably on achieving efficient operation of the rail

transportation system.

ACKNOWLEDGEMENTS

The research was primarily supported by CSX Transportation and the BNSF Railway

Company. The geotechnical laboratory testing was performed by the Geotechnical

Branch of the Kentucky Department of Transportation. The asphalt laboratory testing

was performed by the National Center for Asphalt Technology at Auburn University.

William (Zack) Dombrow, BNSF Summer Intern from the University of Illinois, assisted

with the sample collections and tests.

REFERENCES

1. Association of American Railroads (2006) Railroad Facts, 2006 Edition, 84 pages.

2. Lopresti, J., Davis, D., and Kalay, S. (2002) Strengthening the Track Structure for
Heavy Axle Loads, Railway Track & Structures, September, pp. 21-26.

3. Rose, J. and Anderson, J. (2006) Long-Term Performance of Asphalt

Underlayment Trackbeds for Special Trackbed Applications, American Railway
Engineering and Maintenance-of-Way Assoc. 2006 Annual Conference
Proceedings, Louisville, KY, September, 27 pages.
4. Rose, J., Li, D., and Walker, L. (2002) Tests and Evaluations of In-Service Asphalt
Trackbeds, Proceedings of the American Railway Engineering and Maintenance-
of-Way Association, 2002 Annual Conference & Exposition, September, 30 pages.

5. Rose, J. (2006) Hot-Mix Asphalt in Railway Trackbeds, ASPHALT, Asphalt Institute

Magazine, Vol. 21, No. 3, pp. 22-25.

6. Li, D., Rose, J., and LoPresti, J. (2001) Test of Hot-Mix Asphalt Trackbed over Soft
Subgrade Under Heavy Axle Loads, Technology Digest-01-009, Assoc. of
American Railroads, April, 4 pages.

7. Rose, J., Su, B., and Twehues, F. (2004) Comparisons of Railroad Track and
Substructure Computer Model Predictive Stress Values and In-Situ Stress
Measurements, American Railway Engineering and Maintenance-of-Way Assoc.
2004 Annual Conference Proceedings, Nashville, TN, September, 17 pages.

8. European Asphalt Pavement Association (2003) Asphalt in Railway Tracks,

www.eapa.org, October, 11 pages.

9. Teixeira, P., Pita, A., Ubalde, L. and Gallego, I. (2005) New Possibilities to Reduce
Track Maintenance Costs on High-Speed Lines by Using a Bituminous Sub-ballast
Layer, Proceedings of Railway Engineering 2005, London, June, 11 pages.

10. Momoya, Y., Horiike, T., and Ando, K. (2002) Development of Solid Bed Track on
Asphalt Pavement, Quarterly Report, Railway Technical Research Institute, Vol.
43, No. 3, September, pp. 113-118,

11. Frailey, F. (2004) BNSF Reborn, TRAINS, Vol. 64, No. 10, October, pp. 34-49.

12. Lustig, D. (2007) Paving a Way for a Railroad Line, TRAINS, Vol. 67, No. 3, March,
pp. 26-27,

13. Rose, J. and Hensley, J. (2000) Design, Construction, and Maintenance Practices
for Asphalt Trackbeds, Proceedings, Transportation Systems 2000 Workshop, San
Antonio, February, pp. 275-281.

14. Rose, J., Brown, E., and Osborne, M. (2000) Asphalt Trackbed Technology
Development; The First 20 Years. Transportation Research Record 1713,
Transportation Research Board, pp. 1-9.

15. Rose, J. (1998) Long-Term Performances, Tests, and Evaluations of Asphalt

Trackbeds. Proceedings of the American Railway Engineering and Maintenance-
of-Way Association 1998 Conference, September, 27 pages.
16. Rose, J. (2008) Test Measurements and Performance Evaluations of In-Service
Railway Asphalt Trackbeds, Proceedings of the Transportation Systems 2008
Workshop, Phoenix, April, 24 pages.

17. American Society for Testing and Materials (2007) Standard Specification for
Performance-Graded Asphalt Binder, ASTM D6373, Book of Standards Volume:
0403, 5 pages.
LIST OF TABLES

Table 1. Asphalt Test Trackbeds

Table 2. Subgrade/Roadbed Geotechnical Evaluations

Table 3. Mix Extraction Tests and Core Analyses from Asphalt Trackbeds

Table 4. Tests on Recovered Asphalt from Asphalt Trackbeds

Table 5. Temperature Range from Winter to Summer in Trackbed Environment

LIST OF FIGURES

Figure 1. Cross-Sectional Views of Typical All-Granular and Hot Mix Asphalt

Trackbeds.

Figure 2. Core Drilling Operation to Obtain Asphalt Cores and Underlying

Roadbed/Subgrade Samples.

Figure 3. Changes in In-Situ Subgrade Moisture Contents Between 1998 and 2007.

Figure 4. Subgrade/Roadbed In-Situ Moisture Tests After Coring.

Figure 5. Changes in Optimum Subgrade Moisture Contents Between 1998 and 2007.

Figure 6. Comparison of 1998 and 2007 Measured In-Situ Moisture Contents and
Optimum Moisture Contents for the Roadbed/Subgrade Samples.

Figure 7. Relationships for Roadbed/Subgrade In-Situ and Optimum Moisture

Contents.

Figure 8. Comparison of 1998 and 2007 Unsoaked and Soaked CBR Test Values for
the Roadbed/Subgrade Samples.

Figure 9. Typical Asphalt Cores of Various Compositions and Thicknesses.

Figure 10. Resilient Modulus versus Age of Asphalt.

Figure 11. Penetration and Absolute Viscosity versus Age of Asphalt.

Figure 12. Penetration and Absolute Viscosity Values for Railroad and Laboratory
Cured Asphalt Cores.

Figure 13. Dynamic Shear Rheometer Values for 1998 and 2007 Tests.
Table 1. Asphalt Test Trackbeds
Location Conway, KY Deepwater, WV Cynthiana, KY Guthrie, OK Oklahoma City, Quinlan, OK Hoover, TX Raton, NM
(Railroad) (CSXT) (CSXT) (CSXT) (BNSF) OK (BNSF) (BNSF) (BNSF) (BNSF)

High Speed
High Speed High Speed High Speed High Speed High Speed Slow Speed
Type of Mainline Open Slow Speed
Mainline Mainline Bridge Mainline Bridge Mainline Mainline Branch Line
Facility Track/ Road Yard Lead
Open Track Approaches Approaches Open Track Open Track Open Track
Crossings
Unit Coal Unit Coal Unit Coal Unit Coal Unit Coal Unit Coal
Traffic Type
Intermodal Mixed Intermodal Mixed Intermodal Mixed Intermodal Mixed Mixed Freight Intermodal Mixed Intermodal Mixed Unit Coal
(million gross
freight Freight freight freight (10) freight freight (3)
tons per year)
(40+) (40+) (40+) (40+) (40+) (40+)
Year 1994
≈1900 ≈1900 ≈1900 1989 1980 1995 1969
Roadbed (new double
(original) (original) (original) (new alignment) (new yard) (new double track) (new coal spur)
Constructed track)
Year HMA
1983 1984 1984 1989 1982 1995 1994 1969
Placed
(24 years) (23 years) (23 years) (18 years) (25 years) (12 years) (13 years) (38 years)
(Age of HMA)
HMA Section 1000 ft, 8 in. 200 ft, 8 in. 1300 ft, 6 in. 3100 ft, 4 in. 532 ft, 8 in. 7.9 miles, 6 in. 4.4 miles, 4 in. 700 ft, 2 ½ in.
Length and (305 m, 200 mm) (61 m, 200 mm) (396 m, 150 mm) (945 m, 100 mm) (162 m, 200 mm) (12.7 km, 150 mm) (7.1 km, 100 mm) (213 m, 65 mm)
Thickness
1000 ft, 5 in. 280 ft, 4 in. 700 ft, 5 in.
(305 m, 125 mm) (85m, 100 mm) (213 m, 125 mm)

700 ft, 7 ½ in.

(213 m, 190 mm)
Ballast 5 – 7 in. 8 – 12 in. 10 in. 10 in. 8 in. 12 in. 12 in. 10 in.
Thickness (125 – 175 mm) (200 – 300 mm) (250 mm) (250 mm) (200 mm) (300 mm) (300 mm) (250 mm)

select subballast
Type of existing select clay soil clay & silt soil
existing mixture existing mixture soil mixture Select subgrade
Roadbed mixture subgrade subgrade subgrades
subgrade
 
 Table 2. Subgrade/Roadbed Geotechnical Evaluations*
Unified Soil
Grain Size Analysis Atterberg Limits Proctor California Bearing Ratio Values
Test Classification
Project % No. 4 to % Optimum Unsoaked Soaked
Location In-situ No. 200 Passing Group % CBR, % CBR, %
% Moisture % Retained Size No. 200 Symb Group Moisture 2.5 mm (0.1 2.5 mm (0.1
Content No. 4 Sieve Material Sieve LL PL PI ol Name Content in.) in.)
Guthrie, OK
Select 10.8 - 16.0 Silty
1 67 32 Non Plastic SM 11.5 16.0 6.0
Subgrade (13.2)** Sand
Select 10.1 - 13.5 Silty
0 67 33 Non Plastic SM 12.5 12.1 3.9
Subgrade (11.8) Sand
Oklahoma City, OK
Clay 16.7 - 20.4 Lean
1 6 93 38 20 18 CL 17.6 8.5 3.2
Subgrade (18.1) Clay
Clay 15.1 – 22.4 Lean
0 4 96 34 18 17 CL 18.0 8.2 2.8
Subgrade (17.6) Clay
Quinlan, OK
Clay 15.9 – 20.5 Lean
4 11 85 37 19 18 CL 17.0 10.0 3.8
Subgrade (18.0) Clay
Clay 15.4 – 19.8 Lean
0 12 88 30 17 13 CL 17.0 8.8 4.2
Subgrade (17.6) Clay

Silt 8.4 – 15.2 Sandy

12 38 50 Non Plastic ML 13.0 23.1 22.7
Subgrade (11.6) Silt
Silt 10.6 – 13.2 Sandy
0 50 50 20 18 2 ML 13.2 42.1 30.0
Subgrade (11.9) Silt
Hoover, TX
Poorly
graded
Subballast 5.1 – 11.4 GP-
51 43 6 Non Plastic gravel w/ 9.2 59.1 54.1
(river gravel) (6.9) GM
silt and
sand
Subballast 6.2 – 9.6 Silty
39 50 11 Non Plastic SM 9.1 51.7 38.6
(river gravel) (7.6) sand

Silty
7.6 - 14.2 GC- clayey
Subgrade 39 36 25 27 18 9 11.4 4.8 2.8
(10.7) GM gravel w/
sand
9.1 – 13.6 Clayey
Subgrade 36 34 30 21 14 7 SC 10.0 8.6 4.7
(11.0) sand
* 1998 data in normal print, 2007 data in bold print
** Test data in parenthesis represents averages
Table 3. Mix Extraction Tests and Core Analyses from Asphalt Trackbeds
Project Location
Conway, KY (1983) Hoover, TX (1994)
(Date Constructed)
Age After 1 Day After 2 Years After 11 Years After 15 Years* After 24 Years** After 4 Years* After 13 Years**

Exposure RR RR RR RR RR RR RR Lab (9 Years)

Extraction Results
Maximum Aggregate 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1)
Size, mm (in.)
Percent Passing 4.5 – 11.7 4.8 – 5.0
3.5 – 5.3 4.6 – 5.9 5.8 – 6.9 -- -- --
No. 200 Sieve Avg. 6.3 Avg. 4.9

Asphalt Binder % 5.0 – 5.4 4.4 – 4.7 5.7 – 6.4 6.6 – 6.8
by Weight of 4.8 – 4.9 4.5 – 4.8 5.1 – 5.3 6.8
Avg. 5.3 Avg. 4.6 Avg. 6.2 Avg. 6.7
Total Mix
Core Analysis
Thickness, mm 108 – 213 121 – 216 114 –210 102 – 210 114 – 216 64 – 102 51 - 102 64 – 102
(in.) (4 1/4 – 8 3/8) (4 ¾ - 8 ½) (4 ½ - 8 ¼) (4 – 8 ¼) (4 ½ - 8 ½) (2 ½ - 4) (2 – 4) (2 ½ - 4)

Density, kg/m3 2260 – 2340 2225 – 2340 2305 – 2420 2327 – 2427 2242 – 2391 2171 – 2286 2213 – 2325 2267 – 2286
3
(lb/ft ) (141 – 146) (139 – 146) (144 – 151) (145 – 151) (140 – 149) (136 – 143) (138 – 145) (141 – 143)

Gmb bulk 2.242 – 2.391 2.213 – 2.325 2.267 – 2.286

2.390 2.237
Avg. 2.316 Avg. 2.270 Avg. 2.277
Gmm max 2.511 – 2.534 2.350 – 2.405
2.492 2.372 2.383
Avg. 2.522 Avg. 2.383
Air Voids, % 2.4 – 6.4 4.8 – 11.5 3.3 – 8.5 2.5 – 7.9 4.1 – 5.1
7.0 – 10.1 6.9 – 13.2 3.5 – 10.9
Avg. 4.4 Avg. 8.2 Avg. 5.8 Avg. 5.1 Avg. 4.5
Dynamic Modulus
psi x 103 @ 1 Hz
517 – 914 -- 921 225 – 400 -- -- -- --
5ΕC (41ΕF)
Avg. 308

84 - 171 -- 230 101 – 186 -- 127 – 174 -- --

25ΕC (77ΕF)
Avg. 143 Avg. 154
Resilient Modulus
3
psi x 10 @ 1 Hz
-- -- -- 655 – 976 -- 328 – 728 -- --
5ΕC (41ΕF)
Avg. 796 Avg. 580

-- -- -- 267 – 508 206 – 508 140 – 232 228 – 384 386 – 465
25ΕC (77ΕF)
Avg. 363 Avg. 387 Avg. 185 Avg. 309 Avg. 423
*1998 data
**2007 data
 Table 3 (Cont.). Mix Extraction Tests and Core Analyses from Asphalt Trackbeds
Project Location
Cynthiana, KY (1984) Deepwater, WV (1984) Raton, NM (1969)
(Date Constructed)
Age After 29
After 1 Year After 10 Years After 14 Years* After 23 Years** After 1 Year After 14 Years* After 14 Years
Years*
Exposure RR RR RR RR Lab (9 Years) RR RR RR RR

Extraction Results
Maximum Aggregate 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1)
Size, mm (in.)
Percent Passing 5.1 - 9.3 1.5 – 1.9 8.8 – 10.4
6.1- 8.6 8.1 - 8.4 -- -- 1.8 – 2.0 9.3 – 10.1
No. 200 Sieve Avg. 7.0 Avg. 1.7 Avg. 9.5

Asphalt Binder % 4.5 – 5.2 4.9 – 5.3 4.7 – 5.1 6.6 – 7.4
by Weight of 4.7 – 5.0 4.9 – 5.3 5.1 4.0 – 4.4 6.9 – 7.3
Avg. 5.0 Avg. 5.1 Avg. 4.9 Avg. 7.1
Total Mix
Core Analysis
Thickness, mm 102 – 254 127 – 229 127 – 279 102 – 203 127 – 279 102 – 178 76 – 178 67 – 194 127 – 190
(in.) (4 –10) (5 – 9) (5 – 11) (4 – 8) (5 – 11) (4 – 7) (3 – 7) (2 5/8 – 7 5/8) (5 – 7 ½)

Density, kg/m3 2194 – 2339 2179 –2355 2196 – 2375 2217 – 2343 2236 2115 – 2243 2132 – 2317 2180 – 2225 2232 – 2278
(lb/ft3) (137 – 146) (136 – 147) (137 – 148) (138 – 146) (139) (132 – 140) (133 – 145) (136 – 139) (139 – 142)

Gmb bulk 2.217 – 2.343

-- -- 2.302 2.236 -- -- -- --
Avg. 2.272
Gmm max 2.438 – 2.506
-- -- 2.456 2.482 -- -- -- --
Avg. 2.483
Air Voids, % 4.0 – 11.2 4.9 – 11.2 6.7 – 13.0 0.9 – 4.2
6.2 – 12.6 6.9 – 8.1 9.9 9.4 –14.1 3.1 – 4.7
6.2 Avg. 8.5 Avg. 8.8 Avg. 2.2
Dynamic Modulus
psi x 103 @ 1 Hz
-- 678 -- -- -- 197 – 340 -- 895 – 1461 --
5ΕC (41ΕF)
--
25ΕC (77ΕF)
261 90 – 108 -- -- 52 - 60 46 – 66 95 - 127 51 – 90
Avg. 97 Avg. 56 Avg. 64
Resilient Modulus
psi x 103 @ 1 Hz
-- -- 500 – 600 -- -- -- 428 – 804 -- 479 – 748
5ΕC (41ΕF)
Avg. 555 Avg. 693 Avg. 609
25ΕC (77ΕF)
-- -- 185 – 236 199 – 391 417 – 511 -- 168 – 347 -- 172 – 576
Avg. 206 Avg. 270 Avg. 464 Avg. 287 Avg. 401
*1998 data
**2007 data
Table 3 (Cont.). Mix Extraction Tests and Core Analyses from Asphalt Trackbeds
Project Location
Guthrie, OK (1989) Quinlan, OK (1995)
(Date Constructed)
Age After 9 Years* After 18 Years** After 3 Years* After 12 Years**

Exposure RR RR RR RR Lab (9 Years)

Extraction Results
Maximum Aggregate 19 (3/4) 25 (1) 25 (1) 25 (1) 25 (1)
Size, mm (in.)
Percent Passing 4.0 – 5.8
7.3 -- -- --
No. 200 Sieve Avg. 4.9

Asphalt Binder % 5.3 – 5.6 4.5 – 4.5 3.8 –3.9

by Weight of 5.7 4.1
Avg. 5.5 Avg. 4.5 Avg. 3.9
Total Mix
Core Analysis
Thickness, mm 102 – 140 102 – 114 152 – 171 127 – 190 152 – 171
(in.) (4 – 5 ½) (4 – 4 ½) (6 – 6 ¾) (5 – 7 ½) (6 – 6 ¾)

Density, kg/m3 2458 – 2463 2399 – 2423 2292 – 2389 2293 – 2341 2293 – 2341
3
(lb/ft ) (153 – 154) (150 – 151) (143 – 149) (143 – 146) (143 – 146)

Gmb bulk 2.399 – 2.423 2.293 – 2.341 2.293 – 2.341

-- 2.355
Avg. 2.410 Avg. 2.322 Avg. 2.312
Gmm max 2.445 – 2.478 2.510 – 2.525 2.513 – 2.520
-- 2.539
(2.465) Avg. 2.516 Avg. 2.516
Air Voids, % 1.7 – 4.3 1.7 – 3.0 6.0 – 9.4 5.2 – 7.5 7.1 – 8.8
Avg. 2.8 Avg. 2.2 Avg. 8.3 Avg. 6.7 Avg. 8.1
Dynamic Modulus
psi x 103 @ 1 Hz 68 – 177
-- -- Avg. 132 -- --
5ΕC (41ΕF)
20 – 25 -- 46 – 68 -- --
25ΕC (77ΕF)
Avg. 21 Avg. 54
Resilient Modulus
psi x 103 @ 1 Hz
1431 – 2130 -- 1142 – 1441 -- --
5ΕC (41ΕF)
Avg. 1680 Avg. 1262
25ΕC (77ΕF)
351 – 415 232 – 315 419 – 571 366 –572 589 – 644
Avg. 383 Avg. 276 Avg. 502 Avg. 450 Avg. 628
*1998 data
**2007 data
Table 3 (Cont.). Mix Extraction Tests and Core Analyses from Asphalt Trackbeds
Project Location
Oklahoma City Yard (1982)
(Date Constructed)
Age After 2 Months After 3 Years After 7 Years After 16 Years* After 25 Years**

Exposure RR RR RR RR RR Lab (9 Years)

Extraction Results
Maximum Aggregate 25 (1) 25 (1) 25 (1) 25 (1) 25 (1) 25 (1)
Size, mm (in.)
Percent Passing 5.2 – 6.4
7.0 6.0 – 6.5 -- -- --
No. 200 Sieve Avg. 5.9

Asphalt Binder % 5.3 – 5.7 5.9 – 6.2

by Weight of 5.7 5.5 – 5.6 -- 5.7
Avg. 5.5 Avg. 6.1
Total Mix
Core Analysis
Thickness, mm 152 – 241 241 – 267 203 – 229 178 – 229 190 – 229 178 – 229
(in.) (6 – 9 ½) (9 ½ - 10 ½) (8 – 9) (7 – 9) (7 ½ - 9) (7 – 9)

Density, kg/m3 2385 – 2420 2385 – 2405

--
2352 – 2414 2381 – 2406 2368 – 2382
3
(lb/ft ) (149 – 151) (149 – 150) (147 –151) (149 – 150) (148 – 149)

Gmb bulk 2.381 – 2.406 2.368 – 2.382

-- -- -- 2.388
Avg. 2.393 Avg. 2.374
Gmm max -- -- -- 2.445 Avg. 2.422 Avg. 2.415

Air Voids, % 0.9 – 4.0 0.7 – 1.7 1.4 – 1.8

0.9 – 2.7 0.9 – 2.3 --
Avg. 2.3 Avg. 1.2 Avg. 1.7
Dynamic Modulus
psi x 103 @ 1 Hz
5ΕC (41ΕF)
1090 – 1110 -- -- -- -- --
25ΕC (77ΕF)
145 - 163 151 125 - 175 69 – 244 -- --
Avg. 137
Resilient Modulus
psi x 103 @ 1 Hz
-- -- -- 703 – 940 -- --
5ΕC (41ΕF)
(827)
25ΕC (77ΕF)
-- -- -- 199 – 341 335 – 474 424 – 966
Avg. 262 Avg. 377 Avg. 695
*1998 data
**2007 data
Table 4. Tests on Recovered Asphalt from Asphalt Trackbeds
Project Location
(Date Constructed)
Conway, KY (1983) Hoover, TX (1994)

Age After 1 Day After 2 Years After 11 Years After 15 Years* After 24 Years** After 4 Years* After 13 Years**

Exposure RR RR RR RR RR RR RR Lab (9 Years)

Recovered Asphalt

Penetration, dcm 39 – 44 50 – 54
50 35 41 - 44 25 31 34
25ΕC (77ΕF) Avg. 42 Avg. 52
Viscosity, P 6334 – 9378 3020 – 3358
4400 - 4410 6250 - 14060 9780 - 12034 13214 6022 5107
60ΕC (140ΕF) Avg. 7983 Avg. 3210
Viscosity, cSt 650 – 932 596 – 627
530 - 540 610 - 840 750 - 760 1752 778 722
135ΕC (275ΕF) Avg. 731 Avg. 612
Dynamic Shear
Rheometer, G*/sin  72.6 68.6 66.6
1.00 kPa, ΕC
Dynamic Shear
1819 – 2249
Rheometer, kPa 1213 1080 2084 1790
Avg. 2043
25ºC (77ºF)
*1998 data
**2007 data
Table 4 (Cont.). Tests on Recovered Asphalt from Asphalt Trackbeds
Project Location
(Date Constructed)
Cynthiana, KY (1984) Deepwater, WV (1984) Raton, NM (1969)

After 14 After 29
Age After 1 Year After 10 Years After 23 Years** After 1 Year After 14 Years* After 14 Years
Years* Years*
Lab (9
Exposure RR RR RR RR RR RR RR RR
Years)
Recovered

Asphalt

Penetration, dcm 30 – 51 25 – 35 61 – 77
43 - 51 41 42 21 53 -60 62 - 82
25ΕC (77ΕF) Avg. 40 Avg. 29 Avg. 68
Viscosity, P 8440 – 15405 19201 – 33891 1314 – 1477
6440 - 9177 13880 - 14480 13290 4193 - 5699 1060 - 1610
60ΕC (140ΕF) Avg. 11855 Avg. 25129 Avg. 1361
Viscosity, cSt 760 – 1159 1003 – 1104 290 – 318
631 - 688 886 - 894 763 1347 496 - 543 270 - 310
135ΕC (275ΕF) Avg. 936 Avg. 1050 Avg. 301
Dynamic Shear
Rheometer, G*/sin  77.3 78.3
1.00 kPa, ΕC
Dynamic Shear
Rheometer, Kpa 1188 1111 3706
25ºC (77ºF)
*1998 data
**2007 data
Table 4 (Cont.). Tests on Recovered Asphalt from Asphalt Trackbeds
Project Location
(Date Constructed)
Guthrie, OK (1989) Quinlan, OK (1995)

Age After 9 Years* After 18 Years** After 3 Years* After 12 Years**

Exposure RR RR RR RR Lab (9 Years)

Recovered

Asphalt

Penetration, dcm 28 – 34
42 28 22 21
25ΕC (77ΕF) Avg. 31
Viscosity, P 5922 – 6136 7827 – 10751
8276 11745 12949
60ΕC (140ΕF) Avg. 6029 Avg. 8927
Viscosity, cSt 678 – 782 856 – 1085
826 1127 968
135ΕC (275ΕF) Avg. 730 Avg. 941

Dynamic Shear
Rheometer, G*/sin  70.0 75.9 73.5
1.00 kPa, ΕC

Dynamic Shear
1387 – 2378
Rheometer, kPa 2197 3308 3842
Avg. 1883
25ºC (77ºF)
*1998 data
**2007 data
Table 4 (Cont.). Tests on Recovered Asphalt from Asphalt Trackbeds
Project Location
(Date Constructed)
Oklahoma City Yard (1982)
After 2 After 3 After 7 After 16
Age After 25 Years**
Months Years Years Years*
Exposure RR RR RR RR RR Lab (9 Years)

Recovered

Asphalt

Penetration, dcm 45 – 67
58 57 59 28 24
25ΕC (77ΕF) Avg. 54

Viscosity, P 2197 – 4482

3870 3490 2495 8678 12735
60ΕC (140ΕF) Avg. 3368
Viscosity, cSt 482 – 834
580 700 - 730 471 1105 869
135ΕC (275ΕF) Avg. 620
Dynamic Shear
Rheometer, G*/sin  71.6 75.3
1.00 kPa, ΕC
Dynamic Shear
Rheometer, kPa 1172 2821 2478
25ºC (77ºF)
*1998 data
**2007 data
Table 5. Temperature Range from Winter to Summer in Trackbed Environment*
Year Tested 1982 1982 1995
Location and Average Within 4 in. (100 mm) 4 in. (100 mm)
System HMA Layer Below HMA Below HMA
Ravenna, Ky 35ºF - 80ºF 37ºF - 77ºF
—
Overlayment (2ºC - 27ºC) (3ºC - 25ºC)
Cynthiana, KY 39ºF – 67ºF
— —
Underlayment (4ºC - 19ºC)
Conway, KY 41ºF - 74ºF 44ºF - 70ºF 39ºF - 65ºF
Underlayment (5ºC - 23ºC) (7ºC - 21ºC) (4ºC - 18ºC)
*Typical range in Kentucky from winter to summer for highway pavement expected to be
from 14ºF to 122ºF (-10ºC to 50ºC).
Figure 1. Cross-Sectional Views of Typical All-Granular and Hot Mix Asphalt
Trackbeds.
Figure 2. Core Drilling Operation to Obtain Asphalt Cores and Underlying
Roadbed/Subgrade Samples.
Figure 3. Changes in In-Situ Subgrade Moisture Contents Between 1998 and 2007.
Figure 4. Subgrade/Roadbed In-Situ Moisture Tests After Coring.
Figure 5. Changes in Optimum Subgrade Moisture Contents Between 1998 and 2007.
Figure 6. Comparison of 1998 and 2007 Measured In-Situ Moisture Contents and Optimum Moisture Contents for the
Roadbed/Subgrade Samples.
Figure 7. Relationships for Roadbed/Subgrade In-Situ and Optimum Moisture
Contents.
Figure 8. Comparison of 1998 and 2007 Unsoaked and Soaked CBR Test Values for
the Roadbed/Subgrade Samples.
Figure 9. Typical Asphalt Cores of Various Compositions and Thicknesses.
Figure 10. Resilient Modulus versus Age of Asphalt.
Figure 11. Penetration and Absolute Viscosity versus Age of Asphalt.
Figure 12. Penetration and Absolute Viscosity Values for Railroad and
Laboratory-Cured Asphalt Cores.
Figure 13. Dynamic Shear Rheometer Values for 1998 and 2007 Tests.