The history, development, and performance of asphalt rubber at ADOT

              ARIZONA DEPARTMENT OF TRANSPORTATION
REPORT NUMBER: AZ-SP-8902
THE HISTORY,
DEVELOPMENT, AND
PERFORMANCE OF ASPHALT
RUBBER AT ADOT
Special Report
Final Report
Prepared by:
Lany A ScoReld
Arl;rona Transportation Research Center
206 South 1 7th Avenue
Phoenk Arizona 85007
Prepared for:
Arizona Department of Transportatbn
206 South 17th Avenue
Phoenk Arizona85007
in cooperation with
U.S. Department d Transpartation
Federal Highway Administretion
The contents of this report reflect the views of the authors who
are responsible for the facts and the accuracy of the data
presented herein. The contents do not necessarily reflect the
official views or policies of the Arizona Department of
Transportation or the Federal Highways Administration. This
report does not constitute a standard, specification, or
regulation. Trade or manufacturer’s names which may appear
herein are cited only because they are considered essential to
the objectives of the report. The U.S. Government and the State
of Arizona do not endorse products or manufacturers.
Technical Report Documentation Page
--
I. Report No. 1 2. Government Accession No. 1 3. Recipient's Catalog No.
1
4. Title and Subtitle
THE HISTORY, DEVELOPMENT, AND PERFORMANCE OF
ASPHALT RUBBER AT ADOT
5. Report bte~ecernber, 1989
6. Performing Organization Code
7. Author (s)
Larry A. ScofIeld
Prepared In cooperation with the U.S. Department of Transportation, Federal Highway Administration
8. Performing Organization Report No.
9. Performing Organization Name and Address
Arizona Transportation Research Center
206 South 17th Avenue
Phoenix, Arizona 85007
12. Sponsoring Agency Name and Address
ARIZONA DEPARTMENT OF TRANSPORTATION
206 SOUTH 17TH AVENUE
PHOENIX, ARIZONA 85007
16. Abstract (21 lines including this one =delete this comment) I
10. Work Unit No.
11. Contract or Grant No.
13. Type of Report & Period Covered
Final Report -
14. Sponsoring Agency Code
This report evaluates ADOT's two decades of experience with asphalt-rubber materials. it dlscusses the
chronological development of asphalt rubber by ADOT and the five principle uses of asphalt rubber. The performance
of asphalt-rubber materials are determined from historical records and pavement test sections.
The performance of asphalt rubber Is evaluated by utilizing hlstorical data from ADOT's pavement management
system database and by reviewing eight experimental projects which included 47 test sections. Pavement condition
distress surveys were performed on several of these projects to determine the terminal condition of the pavements.
Asphalt-rubber materials have been placed on over 700 miles of roadway on the State system. This is
approximately ten percent of ADOT's highway network. Although regularly used on the Interstate System, the principle
use has occurred on State and U.S. Routes. The major application has been in mitigating reflective cracking with over
90% of the appllcatlons consisting of SAMs and SAMls.
The average life of a SAM is approximately five years on the lnterstate and ten years and eight years for the State
and U.S. Routes, respectively. The coefficient of variation in service tife ranges between 30%-40% for aH three highway
classifications. The average life of a SAM1 is approximately nine years for both the lnterstate and State Routes while it is
only three years on the U.S. Routes.
Asphalt rubber has successfully been used as an encapsulating membrane to control pavement distortion due to
expansive soils and to reduce reflection cracklng in overlays on both rigM and flexible pavements.
During the twenty years of asphalt rubber use, ADOT has evolved from using slurry applied asphalt-rubber chip
seals to utilizing reacted asphalt rubber as a binder in open and dense graded asphalt concrete.
15. Supplementary Notes
Asphalt rubber. SAM, SAMI, Three
layer system. Stress absorblng
membrane, ACFC, ARACFC, Reacted
asphalt rubber, Overlay
17. Key Words
Document is availaMe to the U.S.
Public through the National
Technical Information Service,
Springfield, Virginia, 221 61 ,,,. ,
19 Secur~tyC lassification 20. Security Classification 21. No. of Pages 22. Price
(of this report) (of this page)
18. Distribution Statement
Unclassified I I I Unclassified
- ; ..
23. Regist~xplL% aJ-
TABLE OF CONTENTS
Section Page
INTRODUCTION
HISTORICAL DEVELOPMENT
Introduction
Specification Development
SAMI Development
Three Layer System Development
Moisture Membrane Development
Asphalt-Rubber as a Low Volume Road Design Strategy
Rubber Manufacturers
Asphalt-Rubber Asphalt Concrete Mixture Development
ADOT Use of Asphalt Rubber
NETWORK LEVEL HELD PERFORMANCE OF ASPHALT-RUBBER SECTIONS
Introduction
Performance of SAMs
Performance of SAMls
Rubber Membranes for Pavement Encapsulation
EXPERIMENTAL TEST SECTIONS
State Route 87(Beeline Highway)-1975
Rubber As A Mineral Filler Performance Results
Three Layer System Performance Results
Open Graded AC with Asphalt-Rubber Bindcr Performance Results
Modified Asphalt Performance Results
Control Section Performance Results
Cost Comparison
1-40 - (EAST FLAGSTAFF T.1.)-1977
Two Layer System Performance
Three Layer System Performance
Saw and Seal Performance
1-17 (Munds Park) - 19n
Effect of Asphalt-Rubber Membrane on Stripping
Effect of Aggregate Source
Effect of SAMI Location on Reflective Cracking
Performance of Test Sections After Rehabilitation
SR 169 (Dewey-Copper Canyon)-1977
Comparison of Thin Overlay To SAMI Plus ACFC
Comparison of Subgrade Treatments
SR 85 (Buckeye-Liberty)-1978
Evaluation Results
Multiple Comparisons of Design Strategies
Comparison of the Asphalt-Rubber Materials
Construction Effects and Experimental Design
1-17 (DURANGO CURVE-16TH STREET)-1979,1985
1979 Durango Curve Section
1985 16th Street Section
1-19 - 1988
CURRENT DEVELOPMENT
Asphalt Rubber as a Binder in Asphalt Concrete
RESULTS
REFERENCES
APPENDIX 1
Figure
LIST OF FIGURES
Page
1 - Distribution of Asphalt-Rubber Use by Application
2 - Typical Sections of Three Layer Projects
3 - Chronology of Asphalt-Rubber Development in ADOT
4 - Roadway Miles of SAM Placed By Year
5 - Roadway Miles of SAMI Placed By Year
6 - Roadway Miles of Threc Layer System Placed By Year
7 - Roadway Miles of AR-ACFC Placed By Year
8 - Asphalt Rubber Project Locations
9 - Pavement Survival Curves for SAM Surface Treatments
10 - Distribution of Existing Pavement Age At the Time of SAM Placement
11 - Pavement Crack Development in SAMs
12 - Pavement Survival Curves After SAMI Application
13 - Distribution of Pavement Ages of S M I Applications
14 - Pavement Crack Development in SAMIs
15 - Typical Sections for Membranes Placed on Pavements
16 - Mays Roughness With Time
17 - SR 87 (Beeline Highway) Typical Sections
18 - 1-40 (East Flagstaff T.I.) Typical Section
19 - 1-17 (Munds Park) Typical Section
20 - SR169 (Dewey-Copper Canyon) Typical Sections
21 - SR85 (Buckeye-Liberty) Typical Sections
Ll ST OF TABLES
Table
1 - Performance Data for SAM
2 - Rate of Roughness Increase for SAMs
3 - Performance Data for SAMI
4 - Rate of Roughness Increase for SAMI
5 - 1987 Distress Survey Results
6 - Performance Data for Membrane Protection Sections
7 - Mays Ride Roughness
8 - Beeline Highway Paver Survey Results
9 - Test Section Costs
10 - 140 Paver Survey Results
11 - Munds Park Paver Survey Results
12 - Results of Transverse Crack Survey
13 - Results of Crack Survey After Rehabilitation
14 - Dewey-Copper Canyon Pavement Survey Results
I5 - Buckeye-Liberty Paver Results
16 - Average P.C.I. for Design Strategies
17 -Aggregate Gradation for Asphalt Rubber M'udures
Page
INTRODUCTION
The A ~ O hTas used asphalt-rubber materials in the construction and preservation of pavements
since the mid 1960s. Through the last twenty years many tield experiments have been conducted and
changes made in materials, specifications, and design philosophies. This paper discusses ADOT's
experience with asphalt-rubber materials and the current use and state of the practice.
Asphalt-rubber materials have been placed on approximately 700 miles of roadway on the State
System. This is approximately ten percent of ADOTs highway network. Although regularly used on
the Interstate System, the principal use of asphalt rubber has occurred on State Routes and U. S.
Routes. The major application has been in mitigating reflective aacking with over 90% of the
applications consisting of stress absorbing membranes (SAMs) and stress absorbing membrane
interlayers (SAMls).
ADOT has successfully utilized asphalt rubber for five different applications: as a pavement
membrane to prevent reficctive aacking in overlays placed upon both flexible and rigid pavements; as
a subgrade/pavement membrane to mitigate differential movements induced by expansive soik; as a
sealant for cracks in asphalt conaete pavements and joints in concrete pavements; as an economical
design strategy for low volume roads; and as a membrane to prevent moisture intrusion into bridge
deck5.l The relative percentage of use for each of these applications is shown in Figure 1. The
percentages are based upon roadway miles at placement. No information is presented for crack and
joint sealants since this information was not readily available at the time of this paper. The use of
asphalt rubber for crack/joint sealing and as a membrane for bridge decks is not discussed in this
paper-
HISTORICAL DEVELOPMENT
Introduction
The development of asphalt rubber by ADOT paralleled the work of Charles McDonald of the
City of Phoenix. The first McDonald "band aid' test patch utilized by ADOT was applied on U. S. 666
in 1%4.~ Except for one experimentation in 1%7, no other field testing was conducted by ADOT until
1968. During that time the industry evolved from hand placement of McDonald "band aid' sedions to
mass application of asphalt rubber using a slurry machine. Although considerably more eficient, layer
thickness control was difficult due to surface irregularities such as wheel ruts where an excess amount
of asphalt rubber resulted.
In 1968, ADOT placed its fust asphalt-rubber surface application with a diseriiutor truck. That
marked the beginning of a new era for ADOT. With the ability to use distributor truck to dispense
asphalt rubber, the industry aquired a means for producing a quality product at an economical cost.
Over the nex! four years, ADOT developed its asphalt-rubber spccScations and construction
techniques througb field testing and utilization of industry standards.
The only application for asphalt rubber was as a bider for chip seals (SAM) and there was only
one supplier, Sahuaro Petroleum and Asphalt Company. Rubber particles were typically vulcanized
rubber supplied by Atlos Rubber Company ranging in size between the 116 and #25 sieve size.
Three Layer System / ' Low Volume Road Construct~on
Note: Membrane Placed In Conjunction with SAM1
Figure 1 - Distribution of Asphalt-Rubber Use by Application
SAM Development
Between 1968-1972 ADOT incorporated asphalt rubber on six construction projects. During this
time several improvements were developed. Most notable was the use of kerosene to lower the
viscosity of the asphalt-rubber mixture just prior to dispensing (1972), the need to overlap longitudinal
joints and utilize sand blotter (1972), and the change from penetration grade asphalts to AR graded
asphalts (1972-73).
Although numerous field trials were made to vary bider content, mineral aggregate gradation,
mineral apgegate voidage, and layer position within the pavement structure, ADOT has generally used
the same specifications for all asphalt-rubber mixtures. They typically consisted of asphalt rubber
proportioned 75% asphall to 25% rubber, mixed at elevated temperatures (3W+ F), diluted with
kerosene just prior to application (5%-7%), and spray applied at 0.6 gallons/yard2. CM-11 cover
material was rolled L?to the membrane after placement.
In 1975, Arizona Refining Company (ARCO) developed an asphalt-rubber mixture to compete
with the Sahuaro process. This new mixture diired in that it used devulcanized rubber, was
proportioned 80:20, and utilized extender oil in lieu of kerosene. Once suitable specifications were
developed for the ARCO process (approximately 1976), they remained essentially unchanged.
SAMI Development
ADOT placed its first stress absorbing membrane interlayer (SAMI) in 1972 on 1-40 when 18 test
sections were constructed in conjundion with the Federal NEEP Project No. 10 - Reducing Reflective
Cracking in Bituminous On r l a y ~ . ~Th e interlayer was situated above the overlay and beneath the
ACFC wearing course. The neld SAMI installation occurred in 1975, also on 1-40, when an
encapsulating membrane was placed over the existing pavement and cut ditches prior to overlay
placement. Although situating the SAMI beneath the overlay is believed to have been a result of
accommodating the cut ditch sealing, this was a fortunate installation. Research in the near future
would indicate that positioning an interlayer upon the existing pavement would maximize its benefit?
Most future SAM1 applications were placed upon a leveling course and beneath the overlay. The
leveling course was used to provide a smooth platform upon which to place the SAMI and subsequent
overlay. Tbe leveling course also prevented puddling of the asphalt rubber in depressions in the
existing pavement.
Three Layer System Development
ADOT first began experimenting with ways of overlaying old concrete pavements with asphalt-rubber
membranes in 1973. They participated in a joint effort with the City of Phoenix to construct a
test section on Madison Avenue. Although only a two layer system, this was the beginning of the three
layer development (See Figure 2)5. The next experiment occurred in 1974 on 1-17 near Bell Road.
Again this was a two layer system with a CM-11 chip applied to the final surface. Due to the high
traffic volumes and speeds encountered on 1-17, considerable windshield damage occurred. This
problem led to the ~ e e fdo r a third layer such as an ACFC to prevent windshield damage. However,
concern did exist as to whether the third layer was constructible and whether it would shove and rut
under traffic. To verify its constructibility and performance, a test section was placed in 1975 on State
Route 87 on an asphalt concrete pavement. The success of this project led to additional installations as
shown in Figure 2. During 1976-77 additional experimental sections were placed to study the effects of
application rates, void size in leveling course, and vulcanized versus devulcanized rubber. In 1985, the
Tust use of the three layer system as a non experimental design occurred when it again was placed on I-
17 to restore ride and prevent reflection cracking.
Moisture Membrane Development
The fir^ use of an asphalt-rubber membrane to control differential movement due to expansive
soils was in 1973. A catalylically-blown asphalt membrane was placed upon the subgrade prior to
placement of the select material and asphalt concrete. Although this treatment worked well it was
recognized that most of the roadways within the distribution of expansive so& had already been
constructed. Therefore, the membrane could only be placed at the subgrade level during
reconstruction. To alleviate this problem ADOT constructed four projects between 1974 and 1976 to
verify the effectiveness of a membrane placed over the existing pavements and shoulders. The
favorable performance of these tests resulted in ADOT utiliig the 'encapsulating membrane concept"
as a standard design strategy. Approximately 10-20 projects have utilized this technique.
Asphalt Rubber as a Low Volume Road Design Strategy
Only one experimental project has evaluated this strategy. The experimental project was
constructed in 1977 on State Route 169 and consisted of four different design strategies incorporating
asphzlt rubber. Although two of the six sections placed performed satisfactorily, little or no additional
use of this technique has occurred.
Rubber Manufacturers
Since the beginning of ADOT's involvement with asphalt rubber, two manufacturers dominated
the rubber supply market. Most rubber for the McDonald/Sahuaro process consisted of vulcanized
rubber (TPOQ4) from Atlos Rubber Company. Once the ARC0 process became competitive,
devulcanizcd rubber was utilized in this process which was generally obtained from the U. S. Rubber
Reclaiming Company.
-YEAR
1973
LOCATION AND TYPICAL SECTION
MADISON AVENUE
SLBS'YO' SAND ELOnER
FLUSH IN ASPHALT P$lBBER
@O5 04.03GALVD
3 4-
TACK COAT@ I GACVD~.SSI.H
1-17 BELL ROAD
/ FLUS^ IN ASPHALT RUBBER@ o 5 o 8 GAL yoi PRECOATEOCM I 1 @3 r e AR- 30LBsYD2
1 SR87 - BEELINE HIGHWAY
ASPHALTRUBBER@O ~ G A L Y D ~
CM3CHlPS@20 LBSYO~
CM 10@4%ARlOOO 5 8-
1 ASPHALT RUBBER @ 0 6 GAL^
C MC~HI PS @ 20 LBS VO'
5 8- ACFC @W/OAR~OOO
SR87 MESA UNDER PASS
A C F C lW/oPASSIffi38-@65'..
1-40 SIX TEST SECTIONS
7 ASPHALT RUBBER @O 6
ACFC C1M2 5 @ G 30A LLBYSDV ~~' .: , :.: .: . . . ' <::.-. - .. ..:.; , :; -.:;.;:, .-:.
' I ASPHALT CONCRETE
EXIST NGCWRETE PAVEMENT T T -
M7 DURANGO CURVE
ASPHALT RUPER @
ACFC @ 6). AlU000 0C M55. 1G1 A@JLO'V D LBS,YO~
ACFC 8 5% ARUXX) sa-
A 1-17 1 ~ ~ - 7SlT"R EET - -
ASPHA$T RUBBER @4 6
ACFC @ 6.4 AC 20
ACFC @ 5 5% AC-20 1 1 2 -
Figure 2 - Typical Sections of Three Layer Projects
4
Asphatt-Rubber Asphalt Concrcle Mixture Development
Reacted asphalt rubber was first used as a bider in an open-graded asphalt concrete mixture in
1975. Two test sections were incorporated into an experimental section on State Route 87 to evaluate
the effectiveness of this treatment. In 1978, three additional test sections were constructed on the
Buckeye-Liberty experimental project. The open-graded mixtures with asphalt-rubber binder were
used as a design strategy on a limited basis between 1977 and 1981. The first use of a dense graded
asphalt-rubber asphalt concrete occurred in 1986 when a detour was constructed on 1-40 using this
material. More recently (1987), a 3/4 inch lift of open graded asphalt-rubber asphalt concrete was
placed on 1-19 as an overlay to restore ride to a concrete pavement. In October of 1989 a one inch
thick asphalt-rubber ACFC was placed as an overlay on 1-17 near Camp Verde, Arizona.
ADOT Use of Asphalt Rubber
The chronological development of asphalt-rubber applications by ADOT is shown in Figure 3. It
should be noted that the development of asphalt-rubber applications began with chip seals (SAM),
then evolved into SAMIs, and finally into the three layer system. At the time ADOT placed the first
true three layer system (1975) they also placed the first open-graded asphalt concrete with an asphalt-rubber
bider. This marked the beginning of the development of an open-graded asphalt concrete
utilizing asphalt rubber as a bider. This development has continued on a sporadic basis until the
present time.
Development of the various applications occurred through field testing only. The laboratory
investigations into rhe mixture properties of asphalt rubber did not occur on a major scale until all the
design applications had been field tested.
The number of roadway miles of asphalt rubber placed in each year for each application is shown
in Figures 4 through 7. It is evident that the greatest application of SAMs was durring 1975-1976 while
the greatest application of SAMIs occurred during 1976-1977. The mid 1970s was a milestone in the
use of asphalt rubber by ADOT. During this era, the most field testing and construction applications
occurred.
The asphalt-rubber project locations are shown in Figure 8 along with the locations of the
experimental projects.
NETWORK LEVEL FIELD PERFORMANCE OF ASPHALT-RUBBER SECTIONS
Introduction
ADOT has been collecting network level pavement distress data since 19'72. A pavement
management system has been operational since 1980 and indudes distress measurements for each
milepost location on ADOTs 7369 centerline milts of roadway. Pavement distress measurements are
currently recorded for roughness, skid, cracking, rutting, flushin& patching, and faulting. Historical
records do not include all these distresses.
The performance of the asphalt-rubber applications can be assessed utilizing network level
Pavement Management System (PMS) data. This was satisfactorily demonstrated in a previous study.6
Utilizing PMS data it is possible to determine the useful service life of the various strategies and also to
gain insight as to the agencies philosophies for the use of asphalt rubber. Spedfkally, the types of
roadways the strategy is employed on, i.e. Interstate, State Route, etc, and the condition and age of the
existing pavements prior to the rehabilitation.
When utilizing PMS data for network level pavement performance evaluations, several important
pints should be remembered. The performance of any given section is not solely a function of the
treatment itself, It is also a function of the existing condition of the roadway prior to rehabilitation, the
judicious selection of the design engineer in choosing the most appropriate treatpent, the contractors
and inspectors abilities to produce a quality product, and the maintenance authority's abilities to
properly maintain the roadway. All these aspects should be considered when assessing performance in
terms of PMS data.
Furthermore, the Interstate System is maintained at a higher serviceability (PS1~3.0)t han the
State and U.S. Routes. Therefore, the time until rehabilitation would be different between these
functional classifications even if the pavement deterioration were the same.
Y) BUIO LID P*lC*!S i Developmenl of AR Spec~l~caf~oannsd Applrcat~onT echnrques For SAM - Developmenl of SAM1 Des~gnA ppllcatron
I
Begtnning of AR Usage For Membrane Seals 8 Crack Sealanls
I
Developmenl 01 Three Layer System To Overlay Concrere Pavemenls
I
Major Fmld Exper~menfal~oAnn d Grealesl use 01 AR
I
Vulcanrzed 8 Devulcanrzed Rubber Opf~onsE olabl~shedA s Alternar~ves m
Laboratory ~nvest~galoonins lo AR M ~ t ~Prroepe rf16 - Development ol 0 G A R
I
Figure 3 - Chronology of Asphalt-Rubber Development in ADOT
The use of asphalt rubber did not evolve as a preventative maintenance treatment, instead its use
was limited to badly deteriorated pavements where reflective cracking was a concern. In 1975, ADOT
initiated a policy which required the use of a SAM1 on all projects where overlays of less than four
inches were to be placed over badly cracked pavements (> 10% cracking)'. Therefore, comparisons to
other strategies which did not include rubber are difficult since they would not have hen utilized on as
badly deteriorated pavements.
Performance of SAMs
Pavement survival curves were developed utilizing the PMS data through 1988. These curves
represent the pavement life from the time the SAM was placed until the next major treatment occurred
such as a seal coat or overlay. Figure 9 displays these curves for three roadway classifications;
Interstate, State Route, and U. S. Route. Figure 10 displays the pavement age curves which indicate
the existing pavement age distribution for each of the roadway classifications at the time of SAM
placement.
It is evident that the performance of SAMs on the Interstate is significantly different than either
the State or U. S. Routes. This is not surprising since the Interstate receives approximately ten times
the loadings as shown in Table 1. The marked difference in the position of the survival curves for the
U. S. and State Routes is surprising since both have similar loading levels (see Table 1). However,
upon inspection of the age distribution curves it becomes clear that the U. S. routes were significantly
older at the time of SAM placement. Approximately 40% of these pavements were older than 22 years
at the time of the SAM treatment.
W 100
0 a Note. Data Through 1988 e C a LEDGEND:
Z 2 80 0 Inremate Data
2 (? State Route Data
P A U.S Route Data
P
Z
60
m
P-L
>
a
3 40
U)
I
ii
P a
LL ; W Y L I I I I I t
W I
P 2 4 6 8 10 12 14
PAVEMENT AGE ( Y a m )
Figure 9 - Pavement Sumval Curves for SAM Surface Treatments
TABLE 1 - PERFORMANCE DATA FOR SAM
* Through 1985~
Route
Interstate
State Route
U.S. Route
In Figure 10, it is interesting to note that a distinct linear range exists in each of the curves for the
three roadway classifications.
Note Data Tnrwgh 1985
SAM LIFE
(Years)
-
X ff C.V. R
5.3 1.7 31% 2- 7
10.0 3.8 38% 3-15
8.2 3.2 38% 4-13
LEGEND
c? Interscale
Slale Roules
9 US Routes
ix
I I I I I I I
5
1
10 15 20 25 30 35
I
40 45
PAVEMENT AGE (Ywrs)
Pavement Age @
SAM Placement (Years)
-
X a C.V. R
11.6 2.7 23% 8-17
17.9 8.0 45% 2-29
23.4 9.7 41% 16-48
Figure 10 - Distribution of Existing Pavement Age At the Time of SAM Placement
Mean
18K ESALS*
Since SAM
Placement
3944
401
496
The average time from when a SAM is placed until the fust major surface treatment occurs is
approximately five years on the Interstate and almost twice that on the State Routes and U. S. Routes.
The coefficient of variation is similar for all three classifications, ranging between 30% and 40%.
The development of cracking on the Interstate also appears to occur at a different rate than for
the other routes as shown in Figure 11. While cracking is randomly distributed around the mean for
both the State and U.S. Routes, there is a slight rate f increase to the Interstate data. The resulting
linear equation for the Interstate data only had an R2=0.32. As would be expected the U.S. Routes
exhibited higher mean cracking than the State Routes. This is presumably due to the higher age of the
underlying pavements.
Note Data Thrwgh 1985
LEGEND
E Interslare
0 Stare Roule
5 U S Route
YEARS AFlER SAM PLACEMENT
Figure 11 - Pavement Crack Development in SAMs
The rate of roughness increase was previously investigated and the findings restated in TabIe 2.6
The rate of roughness increase is similar for both the Interstate and State Routes. The U.S. Routes
exhibited a 13% higher rate of increase. This could possibly be a result of less maintenance on these
routes and/or due to the oIder underlying pavements.
TABLE 2 - RATE OF ROUGHNESS INCREASE FOR SAMS~
Route Mays Roughness (inches/mile/year)
Interstate 125
State Route 12.7
U.S. Route 143
Performance of SAMIs
Pavement survival curves were developed utilizing the PMS database for data through 1988.
These curves represent the,pavement Life from the time the SAMI was placed until the next major
treatment occurred such as a seal coat or overlay. Figure 12 displays the curves for the three road
classifications; Interstate, State Routes, and U. S. Routes. Figure U displays the pavement age curvcs
which indicate the existing pavement age distribution for each of the roadway classifications at the timc
of SAMI placement.
Surprisingly, SAMIs appear to last longer on the Interstate than on the State Routes or U. S.
Routes. This is probably due to the fact that Interstate pavements were in better condition at the timc
of SAM1 placement and received approximately twice the overlay thickness (4 inches). Although,
Table 3 indicates that State Routes have a slightly higher mean (9.5 years), examination of Figure 12
indicates that two of the seven data points for the State Routes appear to be outliers. For this reason,
no curve is shown to represent the "survival curve." The paucity of data precludes satisfactory
interpretation. The lnterstate pavements have received approximately 10 times the loadings t h a ~th c
State or U. S. Routes received. The coefficient of variation for SAMI life for all classifications \itas
bctween 41% and 45%, a significant performance variability.
TABLE 3 - PERFORMANCE DATA FOR SAMI
Figure 14 indicates the extent of cracking for each cldcation with time. It should be noted that
all the data are not shown since most of the data would have plotted on the abscissa. The data wcre
randomly distributed about the mean for each of the classiications. The mean was approximately 1%
cracking or less for up to 10 years of service for all classifications.
The rate of inaease in roughness was previously investigated and the results restated in Table 4.G
The data analyted for the analysis had considerable variance and therefore sections with R?> =0.70
were utilized to determine the values shown in Table 4. The Tidings suggest that the State Routes
have sisnif~cantly lower rates of roughness inaease. However, only 30% of the data were used to
establish this value due to the variability in the remaining data. These findings should thereforc bc
viewed with caution.
Mean
18K ESALS *
Since SAM I
Application
2676
241
227
Pavement
Age @
SAM1
(years)
- x u
14 8-29
19 9-32
28 10-44
Route
Interstate
State Route
U.S. Route
Mean SAM1 Life
(Years)
-
a C.V. x R
3.9 44 9 5-15
3.9 41 95 3-13
3.6 45 7.8 612
Mean Overlay
Thickness
@ SAM1
(inches)
4.0
2.0
25
TABLE 5 - 1987 DlSTRESS SURVEY RESULTS~
Route U.S. 666 Route U.S. 60 Route U.S. 89
SAM! CONTROL SAM1 CONTROL SAM1 CONTROL
2.5 inch 4.5 inch 1.5 inch 15 inch 1.5 incq 4.0 inch
overlay overlay overlay overlay overlay overlay
- - - - - -
X 0- X u X 0- X cr X 0- X u
75.1 95 75.2 8.7 80.7 4.6 75.3 5.1 100 0 100 0
A 2 inch leveling course was placed prior to the SAMI.
Rubber Membranes for Pavement Encapsulation
Approximately 300 miles of highway are located over expansive soils in northeastern ~ r i z o n a . ~
The highways are situated within outcrops of the "Chinle Formation" which exhibit considerable
volume change due to moisture change in highly expansive soils. The differential movements incurred
by pavements overlying these soils results in increased roughness, cracking, and premature failure of
the roadway. To address this problem, ADOT constructed its lirst experimental project utilizing
asphall rubber to form a subgrade seal in 1973. A cut section on U. S. 180, located near the southern
boundary of tbe Petrifted Forest National Park, had a cataly~ically-blowna sphalt membrane placed
upon the subgrade prior to placement of the select material and asphalt concrete. All three cut-sections
u t i l i d paved shoulders to 1 foot up the back slope.
Due to the success of the 1973 experimental section, ADOT continued to pursue the development
of the encapsulating membrane seal. However, most of the highways situated within the Chinle
Formation had already been constructed and therefore a subgrade seal could only be considered in a
reconstruction situation. To soIve the problem ADOT elected to construct experimental projects
which utilized a membrane seal over the existing pavement and cut ditches. Three projects were
constructed between 1975 and 1976, two on Interstate 40 and one on U. S. 89. The first project
utilizing the membrane protection extended 11.8 miles along 1-40 east of ~olbrook? It was
constructed adjacent to an 11 mile project constructed one year earlier which did not utilize the
membrane protection. The project without membrane protection was considered the control section
for comparison with the three experimental membrane projects. The projed data and typical sections
are shown in Table 6 and Figure 15, respectfully. The details of construction and project evaluations
have previously been reported.B*9y10
The asphalt rubber on all three experimental sections consisted of 25% vulcanized rubber to 75%
AR1000. On the two 1-40 projects, the asphalt rubber was applied at 0.65 gallons/yard2 on the
pavement and 0.75 gallons/yard2 on the earthen shoulders. On the U. S. 89 project the asphalt rubber
was applied at 0.65 gallons/yard2 on the pavement and 0.70 gallons/yard2 on the earthen shoulders.
All the shoulders had an emulsion tack coat (0.08 gallonslyart.?) applied prior to the membrane skal.
After the asphalt-rubber seal was constructed, six inches of soil was placed on the shoulders as a
protective cover.
1-40 TY PlCAL SECTIONS
FILL SECTION
OVERLAY: 0.5 ACFC 1 . 5 AC
ASPHALT-RUBBER MEMBRANE
7 6" SOIL COVER 2.75" AC LEVELING COURSE
CUT SECTION
OVERtAY: 0.5 ACFC 1.5' AC
ASPHALT-RUBBER MEMBRANE
6' SOIL COVER 2.75' AC LEVELING COURSE
-1 I ASPHALT WBBER MEMBRANE I
US 89 TYPICAL SECTION
OVERLAY: 2" AC OVERLAY
ASPHALT.RUBBER MEMBRANE /
12'PAVEDTAC AC LEVELING COURSE
-1 24' NEW CUT DITCH 2' AC I ASPHALT RUBBER MEMBRANE ----Figure 15 - Typical Sections for Membranes Placed on Pavements
TABLE 6 - PERFORMANCE DATA FOR MEMBRANE PROTECTION SECl"I0NS
Rate of Roughness Increase
Project Year -
Number Constructed X t~ C.V. Range
1-40-5(38) 1974 15.6 3.8 24% 9.6-21.9
1-40-5(44) 1975 9.7 3.2 33% 4.6-16.2
1-40-5(45) 1976 5.5 1.4 26% 3.4- 8.0
F-037-2-502 1976 11.1 3.2 29% 7.3-13.1
Construction Roughness Significantly Different.
In 1979, approximately 3 to 4 years after construction, ADOT reported the results of an on-going
evaluation which evaluated changes in roughness, pavement distortion (changes in elevations),
cracking, and moisture changes in the subgradeB Several of these previous findings are reproduced
here:
(1) '...the untreated sections are showing larger heaves."
(2) The overall variance would indicate that the membrane seal is working particularly well
in the fills and the cut to fill transition."
(3) "The membrane sections in cut and fill show less moisture variance over time than the
untreated control sections for all depths except the two foot treated cut."
In addition to these conclusions, the authors produced a table of information shown as 'I-able 7.
The original information is shown in normal typeface, while the information displayed in italics is the
actual performance data through 1988. It is interesting to compare the actual rate of roughness
increase to the previously predicted rate, and the rate experienced prior to the overlay.
Utilizing the previous rate of increase shown in the table, the actual roughness level prior to
overlay, and the average actual rate of increase through 1988 (or the service life), the performance can
be presented graphically as in Figure 16. It is very obvious that the membrane sections have performed
substantially better than the control sections in terms of roughness.
The actual Mays roughness prior to overlaying is indicated on the ordinate axis. The as-constructed
roughness is indicated at the zero value of the abscissa. It is evident from this plot that the
roughness was typically reduced #)O inches/year by the overlay. The project which utilizes no
membrane protection, labeled (38), exhibits a 60% higher average rate of roughness increase than the
project constructed adjacent to it, labeled (44). The (44) membrane project also displayed a 30%
longer service life than the project with no membrane protection.
It is also evident in Figure 16 that only the (44) project reached 1988 without further
rehabilitation. The life of the membrane projects ranged between 7 and 13 or more years while the
"control" project lasted 10 years.
Since only the (44) and (38) projects were constructed adjacent to each otber, it is of questionablc
validity to extend the "control" secticin to projects 35 miles and 160 miles away. When comparing thc
average rate of increase before membrane placement to the average rate of increase after membrane
placement, the beneficial effect of the membranes is very evident. The membranes reduced the rate of
roughness increase between 26% and 69%.
If roughness was the only determining factor for the membrane projects, they would have lastcd
between 14 and 36 years. The dashed lines in F i r e 16 indicated the predicted service lives until
objectionable roughness would be attained. Obviously, roughness is not the only pavement distrcss
factor.
TABLE 7 - MAYS RlDE ROUGHNESS
PROJECT RIDE HISTORY BEFORE OVERLAY
Acntal Rcd~rctio~i ~t r
Rate of Roughness Predicied Years to Rorrgltness Ralc
Project Increase (inches/mile) Objectionable Ride Dliriitg Service Life
140-5(38) 15 14 + 4%
140-5(44) 16 14 - 39%
140-5(45) 18 12 - 69%
F-037-2-502 15 15 - 26%
-
OVERLAY PREDICTED ROUGHNESS FROM RlDE HISTORY AFTER OVERLAY
Rate of Acfual Years to Revised Years to Acniol
Increase Rate Objectionable Objectiotrable Paventent Lifc
Project (inches/mile) of Inceare Ride Ride O'can)
No Membrane:
140-5(38) 14 15.6 16 13 I0
With Membrane:
140-5(44) 6 9.7 23 13 +
I40-5(45) 6 5.5 36 9
F-037-2-502 6 11. I I3 7
-
Awmge 88 Awmge 9.7
Construction Roughness Significantly Diflerent
LEGEND:
Control Section (No Membrane): (38)
Membrane Projects = (502) (44) (45)
Pre Overlay
Roughness YEARS SINCE CONSTRUCTlON
Figure 16 - Mays Roughness With T i e
EXPERIMENTAL TEST SECTIONS
Eight experimental projects, containing a total of 47 sections, are reviewed in this paper. The
projects were constructed between 1975 and 1986, and represent ADOTs major efforts in the
development of asphalt rubber technologies. Many of these sections have reached their terminal
service life and the actual performance can therefore be analyzed without tbe need for "predictions.'
Before discussing each of the experimental projects, mention should be made of the historical
development of experimental projects in ADOT. UsualIy only projects which were severely distressed
were considered for experimental purposes. The experimental project designs never used replication
or randomization, and frequently inadequate pavement Wes smys were performed prior to the
experimental treatment. These problems make definite conclusions of field perfonnancc diff~cult.
State Route 87(Bteline Highway)-I975
In 1975, ADOT conducted its most extensive development program in the field of asphalt rubber.
During that year, ADOT constructed its fust true three layer system, first SAM1 beneath an overlay,
utilized the ARC0 devulcanized process, and constructed its fust open graded asphalt concrete
utilizing asphalt rubber as a bider. ADOT also constructed one of its most extensive test sections. On
State Route 87, locally known as the Beeline Highway, ADOT constructed eleven test sections. Four
major pavement treatments were evaluated; opet graded asphalt concrete with low percentages of
rubber as a mineral filler, three layer system, open graded asphalt concrete with asphalt rubber as a
binder, and a section with asphalt modification. In addition to these major experiments, vulcanized and
devulcanizd rubber sections were constructed for comparison as well as long shard rubber versus
granulated rubber as a mineral fder.
The experimental project, constructed between mileposts 193.7 and 205.4, consisted of overlaying
the entire 38 foot width of the northbound roadway under construction project F-053-1-926. This
section of State Route 87 has incurred approximately 1.2 million 18K ESALS during the 13 years thc
test sections have been in service. The area receives approximately 10-15 inches of annual rainfall.
The existing pavement consisted of nine inches of select material, two inches of aggregate basc,
and one inch of bituminous surfacing with several chip seals. It was approximately 20 years old at the
time the experimental project was constructed. The northbound roadway consisted of two tralfic lancs
with two seven foot shoulders. The experimental project consisted of placing ten sections each
approximately 1/2 mile in length between MP 193.9 and MP 199.0. The remaining section of project to
the north was considered the control section.
The pavtmei~is ections, shown in Figure 17, were constructed in 1975 and are still in service altcr
14 years (1989). A pavement distress survey was conducted in the travel lane using the Corp of
Engineers' Paver Method in 1987 prior to reha~ilitation.~' The results of that survey are shown in
Table 8. No information is shown for sections D and Dl because these sections failed in 1979 and wei-e
replaced. Section E had extensive patching and sealing which prevented an accurate evaluation.
Section I was not evaluated because it appeared to be overlaid at the time of the survey. However, at a
later date it was discovered that Section I had had no maintenance.
Rubber As A Mineral Filler Performance Results
From Table 8 it is evident that sections A through E, which represented rubber as a mineral fillcr,
pcrformed the poorest of the four major experiments. Only section C which utilized granutalcd,
devulcanized rubber performed well. The long shard rubber performed substantially worse than the
granulated rubber. The devulcanized, granulated rubber performed better than the vulcanized rubbcr.
Thm Lapr System Performance Results
The three layer section had the highest average P.C.I. of all the sections and the lowest standard
deviation. The high P.C.I. rating and low standard deviation suggests that this section may have
significantly better properties as a treatment.
Open Graded AC with Asphalt-Rubber Binder Performance Results
The average P.C.I. of the two sections which make up this experiment are the 2nd and 3rd highcsl
overall. Although the devulcanized rubber had a higher average P.C.I. than the vulcanized section it
cxhibited greater variability. It appears this design strategy is relatively insensitive to rubber type (it.
vulcanized or devulcanized).
TABLE 8 - BEELINE HIGHWAY PAVER SURVEY RESULTS~
WOR n ~ r a d c dA $
EXPERIMENT Rubher as a Mineral Filler t $ g p / A R Binder No Rubber
Rubber Granulated Long Shard Granulated
TYPC Rubber Rubber Rubber
Rubber Vulcanized Dcvulcanized Dcvul. 6%AR4000
Composition Rubber Rubber Vulcanized Rubber Rubber 5% Dulrcx 6% AR
1000
SECTION A B C D l D E F G H I CONTROL '
Avenge PC1 43.9 40.1 60.2 N/A N/A N/A 61.9 60.9 61.3 N/A 60.1
PC1 Std. Dcv. 9.4 7.4 3.2 N/A N/A N/A 03 1.7 3.0 N/A 2.4
Modified Asphalt Performance Results
Although quantitative data is not available on section I, it was reviewed at the same time all the
other sections were evaluated. It performed so well that the observers mistakenly thought it had been
overlaid. It is the authors' conclusion that this section performed the best of all sections.
Control Section Performance Results
The designated control section performed as well as any other major experiment in terms of
average P.C.1.. This section was located beyond an intersection to a recreational area and therefore
incurred somewhat less traffic than the other sections.
Cost Comparison
Since the Beeline test sections were constructed as a change order, it provided a unique
opportunity to establish relative cost comparisons between the different strategies. Table 9 indicates
the cost of construction and the increased cost relative to the control section for each strategy. As
evident in the table, Section I, which utilized modified asphalt, was only 8% greater in cost while the
three layer system was 185% greater in cost than the control section.
TABLE 9 - TEST SECTION COSTS
SECTION COST INCREASE IN COST
(dollars) OVER CONTROL SECTION (%)
A :f 10,897 55%
B S11,.529 64%
C 110,475 49%
D S9J61 36%
DI 59,983 42%
E $10,264 46%
F $20,036 185%
G $12,935 &r%
H $10,967 56%
1 $7,593 8%
Control $7,030
1-40 - (East FlagstaffT.1.)-1977
In 1977, ADOT constructed six test sections under project 1-4-4425 between milepost 202.8 and
204. Three major pavement treatments were evaluated for their effectiveness to restore ride quality to
a concrete pavement while preventing reflective cracking. Two two-layer systems, three three-layer
systems, and a saw and seal section were compared. Each section was approximately 1,000 feet in
length and covered the entire 38 feet of the WB roadway. In addition to the three major experiments,
vulcanized versus devulcanized rubber, and bottom layer type (i.e. dense graded ac or open graded)
were evaluated.
Since construction in 1977, this section of Interstate 40 has incurred approximately 7 million 18K
ESALS during its 12 year service life. The area receives approximately 16-20 inches of annual rainfall.
The existing pavement consists of seven inches of subgrade seal, four inches of C.T.B., and eight
inches of JPCP. It was approximately 8 years old at the time the experimental projects were
constructed. he westbound roadway consists of two 12 It. travel lanes and a 10 ft. and 4 ft. shoulder.
The pavement sections, shown in Figure 18, were evaluated in 1989. A PAVER distress survey
was performed in the travel lane on sections 3 and 6 which were the only sections which had not been
overlaid (See Table 10). Sections 4 and 5 failed first and were repaired in 1980. Sections 1 and 2 were
repaired at some unknown later date. Section 1 had 100% of the joints reflected through by 1979.
Section 1 also only had one of the two designated SAM applications constructed. During construction,
it was decided to eliminate the second SAM application after stability problems occurred with the first
application.
Based upon the PAVER results, section 3 scored significantly higher than section 6. This was
primarily a result of longitudinal cracking at the shoulder joint. Both sections had been sealed and
were performing satisfactorily.
Two Layer System Performance
Due to construction problems only one two-layer system was constructed. Unfortunately, it did
not sumve until the 1989 evaluation and the historical records are too inadequate to accurately assess
its performance.
Thm Layer System Performance
Both the vulcanized and devulcanized rubber systems placed upon the dense graded asphalt
concrete failed within 3 years of construction. The vulcanized rubber system placed upon an asphalt-rubber
leveling course provided satisfactory sentice for 12 years and had the highest average P.C.I.. It
should be noted, however, that the passing lane was in significantly worse condition than tbe travel lane
where the PAVER survey was performed.
Saw and Seal Performance
This section has been resealed during its semce life and is still performing satisfactorily.
Considering Be condition of both the travel lane and passing lane, this strategy has performed the best
in the author's opinion.
An interesting note is that the saw and seal section and three layer scdion exhibited significantly
higher load transfer across the JPCP joints than the adjoining ConCrete pavement which had not been
overIayed (i.e. 90% versus 35%). Since the concrete pavement was badly Daacked, it is possible that
all these sections reduced moisture inliltration and therefore the severity of D-aackiig.
SECTION
NO. 140 TYPICAL SECTIONS
ASPHALT RUBBER @O 6 GALNO'
(75%ARlOOO - 25% VULCANIZED RUBBER)
CM ('1 @ 35 LBSYD'
ASPHALT RUBBER @ 1 Z ~ G A W D ~
(75% ARIOOO - 25% VULCANIZED RUBBERl
CM ("1 @ 30 LBSWO'
ASPHALT RUBBER @ 1 ~ ~ G A L ' Y ~
(75% ARlOOO - 255 VULCANIZED RUBBER)
CM I"; @ 30 LBSIYD'
, , ;:,. .: : ..,,i-::,1.,. . .*{.';.'.;..e ..., *'.':,::a
ASPHALT CONCRETE LEVELING COURSE @' 74-
1 ASPHALT RUBBER @ 0 6 GAWD'
1R- (7511 ARlMX) - 25% VULCANIZED RUBBER)
T CM 1") @ 3 0 ~ 0 S Y d
3 LAYER
4 SYSTEMS (
I / - ASPHUT RUBBER @ 0 75 G U M
ACFC @ (W ARUXK) - 20% DEVULCANIZED RUBBER1
L SAW 8 SEA!. OVER PCCP .!€IlNTS
A SEAL
f
ASRULT CONCRETE SURFACE COURSE
6 OVERLAY
Figure 18 - 1-40 (East Flagstaff T.I.) Typical Section
TABLE 10 - 1-40 PAVER SURVEY RESULTS
MAJOR TWO LAYER THREE LAYER SAW &
EXPERIMENT SYSTEMS SYSTEMS SEAL
Rubber Type Vulcanized Rubber Dewlcan. No Rubber
Bottom Layer Type SAM ARLC' ARLC' ACSC* ' ACSC'. ACSC"
Section 1 2 3 4 5 6
Average PC1 80 72.9
PC1 Standard
Deviation 6.0 6.7
I 2
ARLC = Asphalt Rubber Leveling Course
ACSC = Asphalt Concrete Surface Course
1-17 (Munds Park) - 1977
Stripping of a three layer system placed upon a concrete pavement in 1975 produced concern that
an asphalt-rubber membrane facilitated stripping of the cover material in high rainfall areas. To
investigate this problem, an experimental section was constructed on 1-17 between mileposts 313 and
323 near Munds Park. Three experimental sections and a control section were constructed on project
I-IR-17-2(78). Three major factors were evaluated; the effect of an asphalt-rubber membrane on the
stripping potential, the effect of the aggregate source on the stripping potential, and the effect of the
position of the asphalt-rubber membrane in mitigating reflective cracking. Each experimental section
and the control section were approximately 1/2 mile in length and extended across the full roadway
width of the northbound alignment. Figure 19 indicates the cross-sections for the sections.
The local aggregate source for this project consisted of quarried basalt. This is typical to the
Flagstaff and surrounding area and it was the source of concern for the stripping problem. ADOT
personnel suspected the basalt aggregate contributed to stripping in asphalt concrete mixtures. To
verify this hypothesis, Salt River aggregate from the Phoenix area was hauled over one hundred miles
to build one of the test sections. Since the Salt River aggregate was widely used and exhibited minimal
stripping potential in the Phoenix area, it was selected for experimentation on the Munds Park project.
To evaluate the effect of the asphalt-rubber membrane location within the pavement structure, an
"inverted" SAMI section was constructed. The inverted SAMI section placed the SAMI above the
overlay and beneath the ACFC wearing course. The standard practice was to place the SAM1 beneath
the overlay on the leveling course or existing pavement.
This project received approximately four million 18K ESAk during its eleven year service life.
The northbound alignment consisted of two 12-foot travel lanes and a ten and four foot shoulder. The
average annual rainfall is approximately 20-25 inches.
A PAVER distress swey was conducted in both the travel and passing lanes in 1988 prior to
rehabilitation. The results of this survey and a survey of transverse cracks alone is shown in Tables 11
and 12, respectively. The PAVER survey did not include rutting which would have significantly
lowered the results. At a significance level of 0.1, the results of the passing lanes and travel lane are
statistically different. Similarly, the control section results were statistically different when compared to
the SAMI sections.
SECTION 1-17 TYPICAL SECTIONS
ASPHALT RUBBER @ 0.6 GALND'
A.C.F.C @ CM-11 (BASALT AGGREGATE)
INVERTED :;..' ..' .,.,..; ::: .:d- .,:;; :':;.P'.
BASALT
SAM ASPHALT CONCRETE @
CONTROL 112- ACFC @
SECTION ASPHALT CONCRETE @
1
1.112'
(NO SAMI) -r
L
112' A C F.C.@
BASALT ASPHALT CONCRETE @ 1-112'
SAM1
ASPHALT RUBBER B 0.6 G A W ~
L
112' A.C.F.C. @ 1 t
SALT RIVER I ASPHALT CONCRETE @
SAM1
ASPHALT RUBBER @ 0.6 G A W ~
Figure 19 - 1-17 (Munds Park) Typical Section
During construction of this project the travel lane was over asphalted due to hot plant problems.
This reduces the reliability of the experimental comparisons. Therefore, discussions pertaining to
comparisons between the sections are based only on tht passing lane P.C.I..
TABLE 11 - MUNDS PARK PAVER SURVEY RESULTS
h4AJOR EFFECT OF AR MEMBRANE
EXPERIMENT ON STRIPPING & REFLECTIVE CRACKING
EFFECT OF AGGREGATE .
SOURCE ON STRIPPING
EFFEa OF AR MEMBRANE
LOCATION ON REFLECTIVE
CRACKING
SECTION SALT RIVER AGG. BASALT AGG. lNVERTED CONTROL
SAM1 SAM1 SAM1 (NO SAMI)
T=TRAVEL LANE
P = PASSING L4NE T P T P T P T P
t
AVERAGE P.C.I. 66 98 63 93 66 94 65 80
TABLE 12 - RESULTS OF CRACK SURVEY
Section
Inverted Basalt
SAM1
Control
No SAM1
Basalt Aggregate
SAMI
Salt River Aggregal
SAM1
Full Width Transverse
Cracks (Cracks/Mile)
NA.
Quantity of Cracking
Lineal Feet % of Control
(Cracks/Milc)
Effect ol Asphalt-Rubber Membrane on Stripping
The results of the pavement distress surveys suggests that the presence of the asphalt-rubber
membrane in the SAMI sections did not contribute to stripping. In fact, when the existing overlay \\,as
milled out in 1988, field inspections confirmed that this had not occurred.
Effect of Aggregate Source
No apparent differences were evident in the P.C.I. ratings of the Salt River aggregate SAMI and
basalt aggregate SAMls. However, a marked reduction in lineal feet of aacking is evident in Table 12,
with the basalt aggregate SAM1 exhibiting approximately half the cracking of the Salt River aggregate
SAMI. During milling operations, however, no stripping was evident in either overlay.
EfYect of SAMl Location on Reflective Cracklag
The two basalt SAMls directly compare the effect of membrane position on the performance of
the strategy. The P.C.I. rating shown in Table 11 indicates no significant difference betwecn thc
sections. Table 12 suggests that the standard SAMI design perlormed slightly better than the invcrtcd
design in regards to reflection cracking. However, when considering the results of the Salt Rivcr
SP-MI, it is questionable whether the results are statistically different. Insufficient data are availablc to
provide statistical inference.
Performance of Test Sections After Rehabilitation
In 1988 the Munds Park test sections were removed as part of a construction rehabilition project.
This project consisted of milling off the existing ACFC and 1.5 inch overlay and replacing in-kind.
Except for the inverted SAMI section, the SAMIs were not removed and continued to function with the
newly placed overlay. The inverted SAMI was milled off with the existing overlay due to its proximity
to the ACFC. This produced two sections which no longer had a SAMI beneath the new ovcrlay.
A "windshield" reconnaissance was performed on the test sections in October 1989 approximntcly
one year after construction. The results of this survey are shown in Table 13. The survey recordcd
only full-width transverse cracks. Since the test sections are of unequal lengths, the results are reportcd
as cracks per mile.
TABLE 13 - RESULTS OF CRACK SURVEY AFTER REHABILITATION
ORIGINAL TEST SECTION TRANSVERSE CRACKS PER MILE
SALT RIVER SAM1 36
BASALT SAM1 3
CONTROL 91
INVERTED SAM1 72
As evident in the Table, the two sections which still have the SAMI seaion in place (i.e. Salt River
and Basalt SAMIs) exhibit less than half the refledion aacking of the control and inverted sections.
These results indicate that a SAMl continues to provide benefit to the pavement structure even after
subsequent rehabilitation strategies. It also points out another advantage of placing a SAMI beneath
an overlay. On subsequent rehabititation it will still fundion even if the existing overlay is milled all.
SR 169 (Dewey-Copper Canyon)-1977
In 1977 ADOT constructed its only test section to evaluate the use of asphalt rubber for
construction of an economical low volume road.12 Eight test sections were constructed under Project
F-058-1-501 to compare four major features; performance of a thin overlay to a SAMI with an ACFC,
thc effect of subgrade stabilization, the effect of a compaction aid, and the effect of an encapsulating
membrane.
*
Figure 20 indicates the pavement sections which were constructed. Unfortunately, during
construction borrow material was imported for subgrade construction which possessed greatcr
plasticity than had been designed for. This resulted in many of the sections exhibiting distress hithin
approximately one year of construction. Table 14 reports the results of patch surveys conducted in
1978 and 1980. As indicated in the Table, four of the six SAMI sections exhibited significant distress.
In fact, of the eight test sections, only the two sections which utilized subgrade stabilization performed
satisfactorily. In 1981 all but the two sections with subgrade stabilization and Section 1 had been
overlaid. Section 1w as overlaid in 1985. The remaining two s~ctionws ere chip sealed by maintenance.
In 1989 the two subgrade stabilized sections are still performing satisfactorily, although alligator
cracking is evident.
Comparison of Thin Overlay to SAM! With ACFC
No valid comparisons can be made between these strategies because different borrow sourccs
were used for these sections during construction. Since testing of the borrow during construction was
not performed, the quality of the subgrade is unknown.
Comparison of Subgrade Treatments
Without a doubt, the two sections with stabilized subgrades out performed all the other sections.
The effect of subgrade treatment was the dominant factor in this experiment. As of 1989, no significant
difference is discernable between the two stabilized sections. These sections have performed
satisfactorily for 12 years with only a chip seal. During this time the project has received approximately
118,000 18K ESALS.
TABLE 14 - DEWEY-COPPER CANYON PAVEMENT SURVEY RESULTS
Experimental Feature T AC I Asphalt Rubber SAMI + 1" ACFC
Subgrade Enzymatic Cutoff S t a b i t i o n Borrow
Treatment Compaction Walls
AR2000 ARlOOO Lime Cement
Fly Ash
Section 1 8 7 5 2 3 6 4
% of Section 15.2% 85% 6.1% 4.1% 0 0 3.6% 7.9%
Patched in 1978
% of Section 20% 93% 78% U% 0 0 48% 41%
Patched in 1980
SECTION
NO. STATIONS SR169 TYPICAL SECTIONS
2- 1 ASPHALT CONCRETE @ % AR-2000 (
ASPHALT RUBBER @ 0.6 GAVYD'
CM-ll(?]
520.555
ASPHALT RUBBER @ 0 6 G A W ~
1" A.C.F.C.
555-590 ' +& 4 5% CEMENT STABILIZED SUBGRADE # 6.
L, ASPHALT RUBBER @ 0.6 G A W ~
1 ' A.C.F.C 2 f C M - ~ I P )
ASPHALT RUBBER @ 0.6 G A W ~
1' A.C.F.C. CM-1 I(?)
640-670
INCLUDED CUT.OFF WALLS @
13 FT. LEFT 8 RIGHT Of CENTERLINE
1 I / - ASPHALT RUBBER @ 0.6 GAWYD'
1" A.C F C CM.ll(7)
1 ASPHALT RUBBER @ 0.6 G A W ~
1- A.c.F.c I /-cM.I~P)
F i t 20 - SR169 (DewcyCopper Canyon) Typical Sections
In the mid 1970s ADOT began laboratory investigations into asphalt-rubber material properties
and behavior. During this same time, the ARCO asphalt-rubber process became competitive with the
Sahuaro process. ADOT construction projects were bid with alternate rubber specifications applicable
to the ARCO or Sahuaro process. In an effort to coordinate the on-going laboratory asphalt-rubber
research with specfication development and field performance, ADOT constructed an ambitious
asphalt-rubber test section. This project, generally referred to as the Buckeye-Liberty project, was
constructed in 1978 on State Route 85 between mileposts 164.4 and 172.0. Fourteen test sections were
constructed within the 7.6 mile project. Five different design strategies and three asphalt-rubber
materials comprised the major experiment of this project. The five design strategies consisted of
SAMIs, three layer systems, thin overlays with ACFC wearing courses, open graded AC mixtures with
asphalt-rubber biders, and ACFC wearing courses. In addition to providing multiple comparisons
between these strategies, this project provided one of the few direct comparisons between a SAM1 and
a thin overlay, and an open graded asphalt concrete with asphalt-rubber binder and an ACFC.
Three asphalt-rubber products were utilized on this project; two Sahuaro products and one
ARCO product. The ARCO product was their standard material which consisted of 20% devulcanized
rubber and 80% AR4000 modified with extender oil. The Sahuaro material was their standard blend
which consisted of 25% vulcanized rubber and 75% ARlOOO diluted with kerosene, and a modified
blend which consisted of 2i3% vulcanized rubber with 80% AR4000 or AR8000 diluted with kerosene.
Stale Route 85 in the vicinity of the experimental project has received approximately 1.3 million
18K ESALs during the period 1979 through 1988. The project is located within the desert region at an
elevation of 870 feet and receives 7 inches of annual rainfall.
The existing roadway prior to the experimental section consisted of two as-built projects, one in
1947 and one in 1958. The first project constructed a 44 foot roadway between milepost 164.4 and
milepost 169 and a 40 foot roadway between milepost 169 and milepost 170.5. Both sections consisted
of 9 inches of aggregate base with 2.5 inches of bituminous treated surface. The 1958 project
constructed a 40 foot roadway between milepost 1705 and milepost 172.0 consisting of 12 inches of AB
with 25 inches of bituminous treated surface.
The pavement sections, shown in Figure 21, were constructed in 1978 and rehabilitated between
1987 and 1988. A PAVER distress survey was performed in the Eastbound travel lane in 1988. The
PAVER survey did not include rutting due to traffic control problems. At the time of the survey,
Section 8 had been overlaid and many of the sections had extensive patching.
Evaluation Results
Although the Buckeye-Liberty project was ADOT's most ambitious asphalt-rubber lest section, it
appears that fate and poor experimental design worked together to thwart its success. The
experimental sections were placed over two as-built projects which varied in age, section, and width.
The project was constructed between October and January when temperatures were often below
acceptable levels for proper construction. Coupled with inclement weather which periodically caused
construction problems and changes, the design shortcomings precluded obtaining useful experience
from this experiment.
Table 15 indicates the results of the PAVER sumy performed in 1988 in conjunction with the
experimental problems. Unfortunately, extensive patchiig had been performed by maintenance in
1986 on most of these stctions. This undermines the value of the distress survey significantly.
SECTION
NO. SR85 TYPICAL SECTIONS
L
112- A C.F.C. @ 4 5% AR4OOO
ASPHALT CONCRETE (DENSE) 1-
I
ASPHALT RUBBER @ 0.6 G A W d
CM.1 I
1 \ ASPHALT CONCRETE @ 5 3% AR4OOO ( 1-
L
1 12'
WITHMIITHOUT
ASPHALT-RUBBER BINDER
ASPHALT.RUBBER @ 0.5 G A W d
A.C.F.C. @ 4.5% AR40DO CM-11
A.C.F.C. @ 4.5% A R W
112' \ A.C.F.C. Q 4 5% A R W ?
L
1 R-ASPHALT
RUBBER @ 0.5 GAWd
CM-11
Figure .21- SR85 (Buckeye-Liberty) Typical Seaions
TABLE l5 - BUCKEYE-LIBERTY PAVER RESULTS
Multiple Comparisons 01 Design Strategies
DESIGN
STRATEGY
Rubber
CRS9 Tack
MC-250 Tack
Asbuilt 19 -
Road Width
Month Constr.
Section No.
AVG. P.C.1
The average P.C.I. for each of the design strategies is shown in Table 16. From these data it
appears that the SAM1 sections, three layer sections, and thin overlays with ACFCs performed
similarly. The open graded AR ACFC performed the best.
TABLE 16 -AVERAGE P.C.I. FOR DESIGN STRATEGIES
SAM1
Thin Overlay Three Layer OGAR
SAM1 and ACFC System ACFC ACFC
143 20 21 36.7 10
Comparison of the Asphalt-Rubber Materials
Three Layer
The original experiment attempted to evaluate the performance of vulcanized and devulcanized
rubber systems commonly used in the industry as well as an optional vulcanized system utilizing the
same rubber percentage and grade of asphalt commonly used in the devulcanized system.
At the time of the 1988 survey it was not possible to attest to any diierences between the
performance of tbe asphalt-rubber materials.
OG AR ACFC
Vulcanized
Construction Emects and Experimental Design
None
X X
47 47
44' 44'
3A 3B
10 10
The author believes that the problems in construction and the lack of rigorous experimental
design precludes definitive conclusions on Lhi projed. The only conclusive statement is that all the
seaions carried heavy truck traftic for eight years without rehabilitation. The original as-built project
was 41 years old in 1988 and had only 3 inches of surfacing material on 9 inches of AB in some areas.
None
X
47 58
44' 40'
Dec
5D 8
10 NA.
Thin OL
+ ACFC
X X X X X X
X X X X
47 47 47 47 47 47 47 47 47 47
44' 44' 44' 40' 40' 40' 40' 44' 44' 44'
Oct Oct Dec Dec Dec Dec
1 4 2 6B 7A 61% 7B 5B 5C 5A
10 10 22.9 21 21 NA.NA. 10 45.8 54.4
Dewlcanized
ACFC
Vulcanized Devul.
1-17 (Durango Curve-16th StmO-1979,1985
I979 Durango Curve Scction
Two three layer systems have been placed on 1-17 to restore ride to the concrete pavement and
prevent reflection cracking. The first experiment consisted of a 1500 ft. section placed on the
eastbound roadway of 1-17 at the Durango curve. This section of 1-17 is on a superelevated curve and
incurred approximately 82,000 ADT at the time of the three layer placement.
The project was initiated to verify the performance of the three layer system under extreme traffic
conditions. The three layer system had been developed by ADOT as a rehabilitation strategy for its 20
year old urban conaete freeway. Only vulcanized rubber was used and the mixture proportioned 25%
rubber to 75% ARUWX).
The section performed satisfactorily from 1979 to 1985 when 1000 feet were removed and
replaced with a subsequent three layer section. During the six years of service life the pavement
received approximately 11.5 million 18K ESALS. Reflection cracking was first observed after
approximately five years, at the transverse join~s ?~T he reflection cracking at the transverse joints
became progressively worse with time.
Both three layer projects are located within the Phoenix Metropolitan area. They receive 6-7
inches of annual rainfall.
1985 16th Street Section
The second three layer system was placed on the eastbound and westbound roadways contiguous
with the location of the previous project. This project utilized the three layer system as the first
"routine" rehabilitation strategy for concrete pavement. This section did, however, contain a 1500 It.
experimental section which utilized devulcanized rubber proportioned 20% rubber to 80% AR2000.
During construction of the 1985 three layer section considerable windshield damage occurred
whiie traflic travelled over the CM-11 chips placed in the interlayer prior to placement of the second
ACFC. This problem resulted in the Department reducing the "cure time" on the asphalt-rubber
membrane from 72 hours to placing the top ACFC as soon as possible. Originally, a 72 hour cure time
was required to allow the kerosene sufficient time to evaporate. Under heavy traffic conditions and
high speeds (45MPH), windshield damage resulted.
The 1985 three layer system has performed satisfactorily to date. It reduced Mays roughness #)O
inches per mile and has received approximately 7 million 18K ESALS. Reflection cracking was first
observed at the transverse joints approximately three years after construction. In 1988 4 inch diameter
cores were removed from the three layer section to expose a longitudinal and transverse joint in the
underlying JPCP. The longitudinal joint was 112" to 1' wide while the transverse joint was 112" to 5/8"
wide. In 1989 crack sealing operations were performed on most of the eastbound roadway and sections
of the westbound roadway. Aside form the crack sealing operation, this pavement is still performing
well.
Due to chip retention problems experienced on the 1985 three layer installation, ADOT
developed a new syslem for restoring ride to plain jointed conaete pavements. This application
consisted of an open-graded asphalt concrete with asphalt-rubber binder. Its first use was on concrete
pavement on 1-19 just south of Nogales, Arizona. One and one half miles of the southbound roadway
were overlaid with a one inch surface course. Since it is a plant mixture, the application is constructed
in a manner similar to conventional asphalt concrete. The experimental section was constructed
without problems and has performed satisfactorily for over one year.
CURRENT DEVELOPMENT
Asphalt Rubber as a Binder in Asphalt Concrete
The utilization of asphalt rubber as a binder in asphalt concrete began in 1975 with the placement
of two test sections on State Route 87. These sections utihxd 10.5% total binder content and were
placed approximately 1/2 inch in thickness. The binder consisted of 75%AR1000 and 25% rubber.
Both vulcanized and reclaimed rubber were utilized.
Currently ADOT utilizes two designs with asphalt rubber as a binder. The first is an ACFC with
approximately 8% total binder content consisting of 80% AClO and 20% vulcanized rubber. This
material is placed as a one inch overlay on both flexible and rigid pavements. The second design is a
dense graded mixture which utilizes approximately 6% total binder content consisting of 80% AClO
and 20% vulcanized rubber. The gradations for each of these mixtures are shown in Table 17. The
open graded mixture is typically placed 3/4" to 1" in thickness while the dense graded mixture is
typically placed 1.5" to T in thickness. These mixes have only been utilized since 1987 and are
experiencing increased use.
TABLE 17 -AGGREGATE GRADATIONS FOR ASPHALT RUBBER MIXTURES
OPEN GRADED ARAC DENSE GRADED ARAC
Sieve Size % Passing Sieve Size % Passing
3/8" 100% 1/2" 100%
#4 3060% 3/8" WWo
#8 610% 1/4" 4040%
#200 0-25% #8 %30%
#40 515%
d m 0-25%
RESULTS
Over the past two decades ADOT has cvolved from using sIurry applied asphalt-rubber chip seals
to an open graded asphalt concrete utilizing asphalt-rubber binder. Ninety percent of the asphalt
rubber use has been to mitigate refledion cracking through the utilization of SAMs and SAMls. These
design strategies reached their peak use between 1975 and 1977 and then gradually dedined in use until
1980 - 1981 when each strategy had virtually no use. The SAM1 strategy bas gradually increased in use
since 1981 but is still only about 11% of its peak use in 1976. SAM applications appear to not have
been used since 1982. Apparently, asphalt-rubber ACFCs have replaced the SAM strategy, although
only limited use of AR-ACFC has occurred.
Approximately 10% of ADOTs system bas utilized asphalt rubber in some form. Although a
signif~cant decline in its use occurred in the beginning of this decade, projects incorporating asphalt
rubber continue to provide additional benefit even after they have been rehabilitated. ADOT typically
leaves the asphalt-rubber material in place when rehabilitating and places addidonal pavement courses
above it. This compouxtd use makes it difficult to evaluate the life cycle value of the strategy.
SAM1 treatments have demonstrated at a network level and on experimental projects that they
mitigate the effects of reflection cracking. However, the service life of a SAM1 on ADOT's system is
not a function of crack distress alone. Other forms of distress such as roughness, bleeding, etc. ma-]
determine the in-service life of the treatment.
Although only a limited investigation was attempted, it appears that inadequate information is
available to determine the reasons for rehabilitation of these strategies on the ADOT system. One
thing is clear, however, the level of cracking on these projects was not adequate to trigger the
rehabilitations at the lime of occurrence.
ADOTs PMS database assesses pavement cracking as the percentage of total cracking within a
given area. It does not distinguish between the different types of cracking such as longitudinal,
transverse, alligator etc., or the severity of these distresses. It is therefore difficult to determine where
an asphalt-rubber membrane would perform most successfully. For example, previous research has
suggested that the block size of cracking can be correlated with the crack movement and pavement
structure.14 This infers that the performance of the overlay strategy is affected by the type and size of
cracking. If this information was available for the sections in the PMS database it may then be possible
to explain the 30%-45% variation in pavement service life for the SAM and SAM1 sections.
ADOTs current philosophy is to utilize reacted asphalt rubber as a binder in open graded and
dense graded asphalt concrete. These treatments are utilized for overlaying rigid and flexible
pavements and are typically placed in one inch and 15 to 2 inch compacted thicknesses for the open
graded and dense graded mixtures, respectively.
REFERENCES
Morris, G. R. and McDonald, C. H., "Asphalt-Rubber Stress Absorbing Membranes, Field
Performance and State-of-the-Art," Arizona Department of Transportation.
Gonsalvas, G. F., "Evaluation of Road Surfaces Utilizing Asphalt-Rubber 1978," Report
Number 1979-GG3, Arizona Department of Transportation, November 1979.
Way, G. B., "Prevention of Reflective Cracking Minnetonka-East (1979 Addendum Report)
Report Number 1979 GW1, Arizona Department of Transportation, August 1979.
Chang, H. S., Lytton, R. L., and Carpenter, S. H., "Prediction of Thermal Reflection Cracking
in West Texas," Interim Report, Texas Transportation Institute, March 1976
Sarsam, J. B. and Morris, G. R., "The Three Layer System on Arizona Highways Development
and Analysis," Prepared for 21st Idaho Asphalt Conference, Arizona Department of
Transportation, November 1982.
Zaniewski, J. P., "Summary of Arizona Department of Transportation Experience with
Asphalt-Rubber," FHWA-AZ88-818, Arizona Department of Transportation, August 1988.
Ford, W. 0. and Landson, H. G., "Development and Construction of Asphalt-Rubber Stress
Absorbing Membrane," Prepared for presentation at the 55th Annual WASHTO Conference,
June 1976.
Forstie, D., Walsh, H, and Way, G. B., "Membrane Technique for Control of Expansive Clays,"
Prepared for presentation at the 58th Annual Meeting of the Transportation Research Board,
January 1979.
Morris, G. R., "Control of Expansive Highway Subgrades with Asphalt-Rubber Membranes:
Arizona's Experience," National Seminar on Asphalt-Rubber, October 1981.
Landson, H. G., "Construction Techniques of Placement of Asphalt-Rubber Membranes,"
Thirteenth Paving Conference, University of New Mexico, January 1976.
U.S. Army CERL Technical Report M-294, "Pavement Maintenance Management for Roads
and Parking Lots,' Urbana Illinois, October 1981.
Morris, G. R., Chen, N. J. and Di Vita, I. A., .Application of Asphalt-Rubber on New
Highway Pavement Construction," Prepared for the 61st Annual Meeting of the
Transportation Research Board, January 1982.
McCullagh, F. R., and Foppe, L. E., "A Five Year Evaluation of Arizona's Three Layer System
on the Durango Curve In Phoenix, Prepared for 64 Transportation Research Board.
Viljoen, A.W., Freeme, C.R., Servas, V.P., Rust, F.C., 'Heavy Vehicle Simulator Aided
Evaluation of Overlays on Pavements with Active CracW, Sixth International Conference on
the Structural Design of Asphalt Pavements, July 1987.
APPENDIX 1
1 YEAR ITEM NO. DEVELOPMENT ROADWAY I
1 McDonald "band aid" Repair Applied U .S. 666
2 Experimental SAM SR 87
Slurry Applied SAM
Distributor Applied SAM
U.S. 60
1-17 (Frontage Road)
SAM1 Application Above Overlay
Kerosene Dilution of AR Mirdure
AR Membrane Seal of Subgrade
3 Major AR Treatment in Northern Arizona
AR Crack Sealing
Three Layer System on HMAC Pavement
AR Open Graded AC Mixture
Sahuaro versus ARCO Applications Comparisons
Rubber as a Mineral Filler
4 Three Layer System on Concrete Pavement
56 Encapsulating AR Membrane Seal
Over Existing Pavement and Shoulder
ADOT Implemented the Use of SAM1 as SOP
for Overlays c 4' where cracking is a concern
ARCO Application Technique Became Avahble
SAM1 Placed Beneath Overlay
Devulcanized Rubber Three layer
AR Open Graded AC Mixture
AR Open Graded AC Overlay on Conaete
U.S. 60
SR 71
SR 180
U.S. 89
U.S. 89A
1-40
1-17
1-40 (detour)
I- 19