Use of Scrap Rubber in

Asphalt Pavement Surfaces

Robert A. Eaton, Richard J. Roberts and
Robert R. Blackburn December 1991
For conversion of Sf metric units to U.S./British customary units
of measurement consult ASTM Standard E380, Metric Practice
Guide, published by the American Society for Testing and
Materials, 1916 Race St., Philadelphia, Po. 19103.

This report is printed on paper that contains a minimum of

50% recycled material.
Special Report 91-27

U.S. Army Corps

of Engineers
Cold Regions Research &
Engineering Laboratory

Use of Scrap Rubber in

Asphalt Pavement Surfaces
Robert A. Eaton, Richard J. Roberts and
·Robert R. Blackburn December 1991

Prepared for
STRATEGIC HIGHWAY RESEARCH PROGRAM

Approved for public release; distribution is unlimited.

PREFACE

This report was prepared by Robert A. Eaton, Research Civil Engineer, and Richard J. Roberts,
Civil Engineering Technician, of the Experimental Engineering Division, U.S. Army Cold Regions
Research and Engineering Laboratory, and Robert R. Blackburn, Head, Engineering and Statistical
Sciences Section, Midwest Research Institute, Kansas City, Missouri.
Funding for this study was provided through Michigan Technological University, Houghton,
Michigan, under Strategic Highway Research Program (SHRP) Contract H-204, Ice-Pavement Bond
Disbanding, Surface Modification and Disbanding, Subtask 1a: Investigate Passive Pavement
Modifications. Albert Wuori was the contract administrator and the SHRP contract monitor was David
Minsk. This report covers work conducted at both CRREL and the Midwest Research Institute from
July 1989 to September 1990.
The authors would like to thank their respective organizations and support personnel for their
assistance and input on this project. Special thanks are due R. McGilvary, S. Shoop, J. Bayer, T.
Marlar, B. O'Donnell, P. Bosworth, R. Melendy, and H. Sanborn, all of CRREL, for their technical
and administrative support. Dr. V. Janoo, Dr. R. Berg, and J. Stark ofCRREL are sincerely thanked
for their technical review of the report. R. Doty, R. Page, and J. VanKirk of the State of California,
Department of Transportation, Sacramento, are thanked for their assistance and input.
A short videotape summarizing this work is available from Robert Eaton, CRREL, 72 Lyme Road,
Hanover, New Hampshire 03755-1290.
The contents of this report are not to be used for advertising or promotional purposes. Citation of
brand names does not constitute an official endorsement or approval of the use of such commercial
products.

11
CONTENTS
Page
Preface..................................................................................................................................... ii
Introduction ............................................................................................................................ .
Research procedure .....................................................................................,. ........................... 2
Test materials .... .. .. .. ....... .. ... ...... .... ...... .. ... .. ... ..... ..... ...... ...... ...... ...... ...... ............ .. .... ...... ...... ..... 2
Development of rubber-aggregate asphalt concrete mix design based on the "Gap Graded"
PlusRide concept ................................................. ~........................................................ 3
Revised Chunk Rubber Asphalt Concrete mix design ..... ..... ...... ...... ...... .. .... ...... ...... ... ... .... ... . 6
Rolling resistance on rubber-aggregate asphalt pavements..................................................... 9
Laboratory testing of pavement modifications .... ..... .... .. ..... ..... ... ................ .. ...... ............ ........ 10
Development of testing equipment ..... ..... ...... ...... ...... ...... ...... .... .. .... ......... ... .. ...... ...... ...... ... 10
Preparation of pavement surface specimens ....... ....... ... ..... ...... ...... .. .... ...... .... .. ...... ....... ..... . 11
Preliminary loading experiments .................. :..................................................................... 11
Wheel passage tests .... .. ... ....... ...... ............. ..... ...... ...... .................. ...... ...... ....... .. ................. 13
Summary.................................................................................................................................. 13
Recommendations
Literature cited......................................................................................................................... 13
Abstract .... ...... ... ... ...... .... .. ... ... ...... ...... ...... ...... ...... ...... ...... .... ...... ........ ...... .. .. ............ .. .... .. ...... .. 15

ILLUSTRATIONS

Figure
1. Mean resilient modulus for a 0.05-second load time of fout mix designs based on
the PlusRide concept ............................................................................................. :..... 4
2. Mean resilient modulus for a 0.10-second load time of four mix designs based on
the PlusRide concept ............................................................ .'..................... .. ............... 4
3. Total creep of four mix designs based on the PlusRide concept................................... 5
4. Mean creep modulus of four mix designs based on the PlusRide concept .. .. .. .. .. .. .. .. .. . 5
5. Mean resilient modulus for a 0.05-second load time of various CRAC mix designs... 8
6. Mean resilient modulus for a 0.1 0-second load time of various CR~C mix designs ... 8
7. PlusRide: Resilient modulus vs aging time and test temperature.................................. 9

TABLES

Table
1. Recommended specifications for rubber-asphalt paving mixtures for different
levels of traffic .... ... ... .. .. .. ...... ...... ... ... ...... ...... ...... ... ... ...... ...... ..... ...... ....... ............ . 1
2. Mix designs using PlusRide concept .... ....... .... ... .. .. .. .... ...... ...... ... ... ... .. ..... .. ............ ...... . 2
3. Average Marshall stability results................................................................................. 3
4. Optimum Marshall values .... ....... ...... .. ... .. .. .. ...... ...................... ... .. ... ...... ....... .. ... .. .... .. ... 3
5. Mean resilient modulus................................................................................................. 4
6. Creep tests, 1000-second load time ......................................................................... ;..... 5
7. Chunk Rubber Asphalt Concrete mix design control-no rubber .. .. .. .. ... .. ...... .... ... ... ... 6
8. Average Marshall stability results .................................................................... ;............ 6
9. Chunk Rubber Asphalt Concrete mix design .. .. .. .. ... ...... .. .. .......... .... ... ... ... .... .. .. ... .... .. ... 7
10. Control mix designs....................................................................................................... 8
11. Mean resilient modulus Chunk Rubber Asphalt Concrete .. .. .. .. .. .. .. ... .. .... ... .. ....... ... .. ... 8
12. Effect" of aging: resilient modulus test results ............................ :............................. ..... 9
13. Rolling resistances of passenger cars....................................................................... ..... 10

iii
 Use of Scrap Rubber in
Asphalt Pavement Surfaces
ROBERT A. EATON, RICHARD J. ROBERTS AND ROBERT R. BLACKBURN

INTRODUCTION for some stone aggregate. Thus, the rubber particles are
relatively large compared to the particles used in the
Passive pavement modifiers have been used in the asphalt rubber concrete. In addition, th~ larger rubber
past either to prevent ice from bonding to the roadway particles are thought to act as elastic aggregates that flex
surface or to weaken the bond so that the ice can be on the pavement surface under traffic and contribute to
disbonded with wheel passages. These modifications ice disbondment.
have included various rubber additives derived from In the United States, the trademark "PlusRide" is used
scrap rubber and the chemical additive Verglimit to designate the Swedish formulation. PlusRide rubber
(Blackburn et al. 1978). is derived from granulating whole tires and tire buff-
Each year the United States disposes of about 200 ings, and contains chopped cords. The recommended
mi11ion passenger tires and 40 mi11ion truck tires. This specifications for PlusRide paving mixtures for different
represents about 2.1 mi11ion tons of scrap passenger tires levels of traffic are given in Table 1.
and roughly 1.9 mi11ion tonsofscraptruck tires (Taka11ou
et al. 1985). One method of disposing of this huge
amount of waste material is to recycle the ground rubber
tires into the asphalt pavement. Table 1. Recommended specifications for rubber-asphalt
In recent years, a growing number of state depart- (PlusRide) paving mixtures for different levels of traffic.
ments of transportation and municipalities have started
to use scrap rubber to modify asphaltic pavements. Two Plus Ride Plus Ride Plus Ride
Mix designation 9 12 16
different methods of incorporating scrap tire rubber into
paving mixes have been developed. The first type of Average daily traffic <2500 2500-10,000 >10,000
rubber modification uses finely ground rubber tire par- Thickness (in.) min. . 0.75 1.5 1.75
ticles that are mixed into the hot asphalt cement to create Aggregate % Passing
a rubberized asphalt binder. This binder is then added to Sieve sizes:
3/4 in. 100
a normal gradation of paving aggregate. This type of
5/8 in. 100
modification is called asphalt rubber concrete. l/2 in. 65-80
In the late 1960s, two Swedish companies, Skega AB 3/8 in. 100 60-80 50-60
and AB Vaegfoerbaettringar, developed a second type of 1/4 in. 60-80 30-42 30-42
rubber modification in a product named "Rubit." The no. 10 23-38 19-32 19-32
no.30 15-27 13-25 12-23
Swedish design incorporated 3 to 4% rubber (by weight
no.200 7-11 8-12 6-10
of a mixture) into an asphalt pavement surface mixture Preliminary mix design:
to increase skid resistance. The mixture provided a new Rubber, % of total mix
form of ice control as well as reduced pavement/tire by weight 3.0 3.0 3.0
interaction noise (Stuart and Mogawer 1988). The over- by volume (approx.) 6.7 6.7 6.7
Asphalt, % of total mix
an mix consists of blending the larger rubber particles,
by weight 7.5 7.5 7.5
[1/16 in. (0.16-cm) up to 3/8-in. (0.95-cm] into a gap- by volume (approx.) 20.2 20.2 20.2
graded aggregate mix, substituting the rubber particles Maximum voids(%) 2 3 4
 Many states and municipalities have placed, or are in RESEARCH PROCEDURE
the process of placing, test sections containing PlusRide.
At least 18 states across the nation, from Rhode Island to The initial investigation of rubber-modified asphalt
Alaska, have built test sections using PlusRide (Stuart concrete consisted of making laboratory samples of the
and Mogawer 1988, Takallou et al. 1985). PlusRide mix with 3% by weight of rubber particles
Because of the widespread interest in the PlusRide substituted for the respective gradation of stone aggre-
paving mixtures, we decided to further concentrate the gate. Additional samples were then made by increasing
investigation of passive pavement modifications using the rubber content to 6 and 12% by weight (see Table 2).
rubber aggregate. In particular, it was of interest to study The Marshall method of mix design was used to
the addition of higher concentrations of rubber to en- determine the optimum asphalt content according to
hance the ice disbanding characteristics of this type of ASTM D-1559. Five sets of samples were prepared for
pavement modification. each of the four mix designs using varying asphalt
Verglimit was developed in the 1970s by Chemische content percentages. Three specimens were made for
Fabrik Kalk in Cologne, West Germany, and tested in each percentage of asphalt content, for a total of 15
Europe as a means of improving ice control. Verglimit samples for each mix design. The typical asphalt content
consists of particles (0.1 to 5 mm) of calcium chloride range used for each mix design was 5, 5.5, 6, 6.5, and 7%
with a small amount of sodium hydroxide. This mixture of total aggregate weight. After testing, the data were
is coated with a water-resistant layer of either linseed oil plotted and the optimum asphalt content was determined
or polyvinyl acetate and is used as an integral [Link] the by the curves generated. If the optimum asphalt content
wearing course. The encapsulation keeps the material level could not be made from the first set of samples,
inactive until the particles break under the action of more samples were made using more or less asphalt
traffic. The additive then mixes with moisture in the air content until the optimum was determined. Guidelines
or on the pavement to form a salt solution on the and previous lab mix design experience were then used
pavement surface. The material has been used with in selecting the best mix from the materials considered.
mixed results in Europe, Japan, Canada, and in the The final mix design for each rubber content selected
United States. There have been unsubstantiated reports was a balance of optimum asphalt content, air voids,
that the skid resistance of the pavement with Verglimit stability, and flow.
may be low in dry weather and that the life of the
pavement may be reduced up to 50% (Stuart and Mogawer TEST MATERIALS
1988). Because of these drawbacks and the general
concern about chemical additives, it was decided that the The stone aggregate used in the study was a rounded
investigation of [Link] pavement modifications would Maryland gravel with the plus no. 4 sieve size put
focus on the use of rubber additives in asphaltic concrete. through a crusher. The sand was a natural sand. The
Consequently, no work was done with chemical addi- material was obtained from the Strategic Highway Re-
tives to pavement materials. search Program aggregate storage facility in Texas.

Table 2. Mix designs using PlusRide concept.

Oo/o Rubber 3% Rubber 6% Rubber 12% Rubber

6% A. C. 6.5% A. C. 7% A. C. 9.5 o/oA.C.
Stone Rubber Stone Rubber Stone Rubber Stone
Sieve %passing (g) % (g) % (g) % (g) % (g) % (g) % (g)

1/2 in. 100 0 100 0 100 0 100 0

3/8 in. 80 240 70 300 69 285 80 21.6 71 229.6
no.4 60 240 36 340 34 323 60 21.6 39 253:4
no. 10 40 240 10 25 110 3 28.5 26 76 40 21.6 28 87.1
no. 30 20 240 10 19 60 3 28.5 20 57 20 21.6 20 63.4
no.200 10 120 10 10 90 II 85 0 21.6 10 79.2
PAN 10 120 100 95 10 79.2

Composition (g)
Stone 1200 1000 921 792
Rubber 0 30 57 108
Asphalt 72 67.0 68.5 86

2
 The rubber aggregate was obtained from Baker Rub- The Marshall stability values were obtained for the
ber, Inc. It consisted of ground rubber produced from four different rubber-aggregate asphfllt mixes following
passenger and truck tires with a majority of the fabric ASTM D-1559. The average Marshall stability results
removed. The maximum fabric content by weight was are shown in Table 3. The optimum Marshall values for
0.5%. the four mix designs, consisting of 0, 3, 6, and 12%
The asphalt cement was an Oklahoma Crude AC20 rubber content, are presented in Table 4. The stability
from the SHRP storage facility supplied in 5-gallon values for the 3% rubber-aggregate asphalt in Tables 3
buckets and delivered cold. It is one of the SHRP and 4 are greater than the average value of 411 obtained
cataloged asphalts and we were required to use it in this from'field test section results.
project.

Table 4. Optimum Marshall values (lbf).

DEVELOPMENT OF
RUBBER-AGGREGATE Asphalt content Rubber content (o/o by weight)
ASPHALT CONCRETE MIX DESIGN (o/o by weight) 0 3 6 12
BASED ON THE "GAP GRADED"
PLUSRIDE CONCEPT 6.0 1888
6.5 617
The stone aggregate was sieved into the following 7.0 407
9.5 350
sizes: 3/8 in., no. 4, no. 10, no. 30, no. 200, Pan. The .31
8 in. and no. 4 aggregates were then washed to remove Air voids(%) 4.0 4.6 4.7 4.0
fines. The aggregate was weighed, combined in a metal
pan and placed in an oven set at 350°F (177°C). The
stone was placed in the oven in the afternoon for use the
next morning, and the asphalt was heated just prior to The resilient modulus was also determined for the
mixing in a seamless covered tin. When the asphalt four different rubber-modified asphalt mixes following
temperature reached 280°F (138°C), the rubber aggre- ASTM D-4123. This repeated-load indirect tension test
gate was combined with the stone aggregate. The stone method is conducted by applying compressive loads
and rubber mixture was then put in the oven for 10 with a prescribed sinusoidal waveform and can be used
minutes. The material was removed from the oven, to study effects of temperature, loading rate, and rest
placed on a balance, and tared. The asphalt cement was periods. Consequently, the values of resilient modulus
added to the mix to the desired percentage. The mixture can be used to evaluate the relative performance of
was mixed until uniform and then compacted following bituminous mixtures as well as to generate performance
the standard 50-blow Marshall procedure for a 4-in.- input for pavement design or pavement evaluation and
(1 0.2-cm)-diam. mold. The samples were allowed to analysis.
cool overnight and were then extracted from the molds The resilient modulus was computed using the fol-
for testing. lowing expression:

Table 3. Average Marshall stability results (lbf). P (v + 0.2734}

MR = -----'"----~
tD
Asphalt content Rubber content (o/o by weight)
(o/o by weight) 0 3 6 12 where p vertical load
v Poisson's ratio
4 890 t = specimen thickness
5 1596 380
530
D = horizontal deformation.
5.5
6 1888 540 398
6.5 617 212 Mean values of MR for the four mix designs are given
7 1478 583 407 in Table 5 as a function of temperature and pulse (load)
7.5 458 258 time. The mean resilient modulus is also plotted in
8 327
8.5 297
Figures 1 and 2 as a function of temperature. In Table 5,
9 298 a 0.05-sec load time simulates 65- to 80-krn/hr traffic
9.5 350 conditions, and a 0.10-sec load time simulates a 25- to
10 40-krnlhr traffic condition. A Poisson's ratio value of
10.5 277 0.35 was used in all the computations.

3
 Table 5. Mean resilient modulus.

MR (kg/cm 2 X 1oJ)
Rubber 40°F 77°F 100°F
content (4°C) (25°C) (38°C}
(o/oby Pulse time (sec.)
weight) 0.05 0.10 0.05 0.10 0.05 0.10

0 12.142 7.251 1.647 1.319 0.277 0.214

3 6.030 4.717 0.665 0.526 0.113 0.089
6 3.544 3.592 0.476 0.384 0.077 0.061
12 2.096 1.742 0.250 0.194 0.048 0.035

Table 5 and Figures 1 and 2 show that the mean more under a longer loading time. Similar results were
resilient modulus for the 3% rubber mix is roughly half reported by Takallou et al. (1985).
of that for the mix with no rubber (0% ). Likewise, the The creep modulus was also determined for the four
mean resilient modulus for the 12% rubber mix is roughly mix designs. The creep modulus Me is basically the same
half that for the mix with 6% rubber content. Basically, kind of measurement as MR. The term is used as a
this trend remains the same as the load time is increased convenience to indicate loading times that are long
from 0.05 to 0.10 sec. The resilient modulus of the 0% compared to those used in the resilient modulus tests. For
rubber asphalt concrete shows the largest decrease in the creep test, 1000-sec load times (approximately 16
strength between the two loading times, especially at min) are used.
40°F (4°C). Table 6 and Figures 3 and 4 show the creep test
The resilient modulus increased (the mix got stron-. results. Tests were not conducted at 40°F (4°C) as they
ger) with a decrease in temperature; also, as the load time were with the resilient modulus tests because the readout
increased, the resilient modulus decreased or yielded device was not sensitive enough.

;;- 12
E
~ C\1
Ol
:::.. E
(/) ~
:::J Ol 8
:; ~
"0 (/)
0 :::J
~ 8 :;
c "0
0
~ ~
·c;;
Q)
c
a: ~
c ·c;;
co
Q) a:
Q)
4
~ c
4 co
Q)
~

60 80 100 60 80 100
Temperature (°F) Temperature (°F)

Figure 1. Mean resilient modulus for a 0.05- Figure 2. Mean resilient modulus for a 0.1 0-second load
second load time offour mix designs based on the time offour mix designs based on the Plus Ride concept.
PlusRide concept.

4
 Table 6 shows that the mean creep modulus is re- The resilient modulus increased or the mix got stiffer
duced, for a given temperature, by adding rubber to the with a decrease in temperature .and, as the load time
mix. By adding just 3% rubber, the mean creep modulus increased, the resilient modulus decreased or the mix
or bearing capacity is reduced by a third from 0.05 to yielded more under longer loading times. Similar results
0.016 (kg/cm2x103) at 77°F (25°C). The presence of were reported by Takallou et al. (1985).
rubber in the mix at higher temperatures also reduces the The creep modulus was also determined for the four
bearing capacity, but it is affected less as shown by the mix designs. The creep modulus Me is basically tpe same
flatter slopes in Figure 4 of the rubber mixes vs. the 0% kind of measurement as MR. The term is used as a
mtx. convenience to indicate long loading times compared to
the loading times used in the resilient modulus tests. For
the creep test, 1000-sec load times (approximately 16
min) are used.
Table 6. Creep tests, 1000-second load time.
The total creep, however, as shown in Figure 3, is
Mean creep Mean creep higher for the rubber mixes, pointing out the benefits of
Load Rubber temperature Total creep modulus their performance at lower temperatures, that is, greater
(kg) % OF (em x Jo-4) (kglcm 2 x lrP) elasticity and better resistance to thermal cracking.
Simulated traffic (wheel passage) tests of ice grown
4.536 0 77 8.3 0.0508
100 20.7 0.0204 on the PlusRide rubber-aggregate asphalt samples did
not show significant ice disbanding but did suggest a
3 77 28.7 0.0157 way that the mix design could be altered to improve the
100 46.5 0.0097 ice disbanding performan·ce under traffic conditions.
We decided to increase the size of rubber aggregate to
6 77 40.5 0.0114
100 58.5 -0.0079. increase the potential for ice disbanding under wheel
loadings. Consequently, further testing of the PlusRide
12 77 58.8 0.0076 mix concept was halted in favor of testing larger rubber
t'oo 84.2 0.0053 aggregate mix designs.

80

C\1

E
~ 0.040
E 60 ~
~ (/)
0.. ::I
Q) :;
Q) "0
0
0 ~
cti 0..
0 40 Q)

t- ~
u
c: 0.020
ro
Q)
~

20

o~-------'-----'-----..L.-----'
o~--- _ __._____.....______.________.
60 80 100 60 80 100
Temperature (°F) Temperature (°F)

Figure 3. Total creep offour mix designs based on Figure 4. Mean creep modulus offour mix designs
the PlusRide concept. based on the PlusRide concept.

5
 REVISED CHUNK RUBBER ASPHALT results (Foster 1985), we decided to design for a mini-
CONCRETE (CRAC) MIX DESIGN mum of 3% air voids.
This finer mix may solve some of the surface aggre-
A conclusion of the first part of the study was that the gate loss experienced by the California Department of
-~maximum size of rubber aggregate should be increased Transportation (CALTRANS) on their Route 395 test
to promote more of an area-wide flexure of the ice/ sections south ofRavendale (VanKirk 1989, Doty 1988).
substrate interface under traffic loadings. A new rubber- The higher air voids may also solve bleeding problems
modified asphalt concrete mix design was developed, experienced by the Alaska Department of Transporta-
and various maximum sizes of rubber aggregate were tion.*
tested. Based upon laboratory results and simulated Table 8 presents the CRAC Marshall stability results
traffic testing, a new mix design was defined (see Table and shows that the 3, 6, and 12% rubber mixes are more
7). than twice as strong as the original mixes. Table 9 shows
Our Chunk Rubber Asphalt Concrete (CRAC) mix the final CRAC mix designs for 0, 3, 6, 12, 25, 57, and
design is denser than the original PlusRide mix designs 100% rubber content.
first tested; however, the rubber aggregate is larger. Table 10 shows that the control (0% rubber content)
Based upon current rutting problems and prior research CRAC mix design is denser than the corresponding
PlusRide mix.
The mean resilient modulus results shown in Table 11
and in Figures 5 and 6 show that the CRAC mixes have
higher resilient moduli at all temperatures and at both
Table 7. Chunk Rubber Asphalt Concrete
loading rates. This is due to the finer, denser mix. The
control mix design-no rubber.
stronger mix will support higher load~. better resist
Desired rutting, and provide a stronger matrix above which the
percentage rubber particles will project for better ice disbonding
Sieve passing Range performance. The results of the ice disbonding tests
under wheel passage conditions are discussed later in
l/2 100 100
3/8 85 80-90 this section.
no.4 60 55-65
no. 10 40 35-45
no.30 20 15-25 * Personal communication with D. C. Esch, Alaska Depart-
no.200 5 0-10 ment of Transportation, Fairbanks, .1990 and, J. L. Van Kirk,
Asphalt cement 6 5.5-6.5
R. Doty and R. Page, CALTRANS, Sacramento, California,
(%by wt)
1990.

Table 8. Average Marshall stability results (lbf).

Asphalt
content Rubber content(% by weight)
(%by weight) 0 3 6 12

5.5 1600 (5.6)*

6.0 1755 (4.6) 950 (5.8)
6.5 2025 (3.0)t 1120 (4.9) 530 (6.3)
7.0 2270 (2.2) 1310(3.5)t 605 (5.9)
7.5 1885 (1.5) 1130 (3.0) 850 (4.1)t
8.0 1125 (0.6) 825 (3.3) 470 (5.3)
. 8.5 800 (2.7) 670 (4.2)
9.0 690 (3.l)t
9.5 . 650 (2.8)
0.0 640 (1.8)

* Air voids-percentage of total mix

tOptimum

6
 Table 9. Chunk Rubber Asphalt Concrete mix designs.

%Rubber 0 3 6 12 25* 57* 100*

Stone Rubber Stone Rubber Stone Rubber Stone Rubber Stone Rubber Stone Rubber Stone
passing retained passing retained passing retained passing retained passing retained passing retained passing
Sieve (%) (%) (g) (%) (%) (g) (%) (%) (g) . (o/o) (%) (g) (%) (%) (g) (%) (%) (g) (%)

112 in. 100: 100 100 100 100 100 100

3/8 in. 85 2 21.6 87 3 31.1 89 6 57.4 93 3 24.0 100 10 61.7 100 85 68.8 100

no.4 ) 60 10.8 65 .3 31.1 69 6 57.4 80 9 78.9 100 20 123.5 100 60 114.7 100

no. 10 40 40 40 40 13 102.9 73 20 123.5 100 40 91.7 100

no. 30 20 20 20 20 20 6.7 41.4 36 20 91.7 100

no.200 5 5 5' 5 5 9 5 68.8 0.1 0.59**

Pan 22.9 0

Asphalt cement 6.0 6.5 7.5 9.5 12 16 20

(%total weight
of aggregate)

Rubber(% total 0 3 6 12 25 57 100

weight of
aggregate)

Air voids 3.0 3.5 4.1 3.1 8.7t Did not measure Couldn't measure
(%total mix)

Note: All figures are for standard 4-in.-diam. Marshall pucks

* No Marshall mix design done for these three mixes
t Average of 3 samples
**No stone added
 Table 10. Control mix designs.

PlusRide CRREL Chunk Rubber Asphalt Concrete

mix 12 control %passing by weight
%Rubber 3 0 0 3 6 12
Sieve

5/8 in. 100 100 100 100 100 100

l/2 in. 95-100
3/8 in. 60-80 85-95 85 87 (2)* 89(3) 93(6)
l/4 in. 30-42 60-75 60 65 (I) 69(3) 80(6)
no. 10 19-32 38-50 40 40 40 40
no.30 13-25 19-27 20 20 20 20
no. 200 8-12 2-6 5 5 5 5
Asphalt
(%total mix 6.0 7.5 6.4 6.5 7.0 9.5
by weight)

( )*-% rubber is % total mix weight.

Table 11. Mean resilient modulus Chunk Rubber Asphalt Concrete (CRA C).

MR (kg!cm 2 x1rP)

40°F 7rF 100°F

(40C) (25°C) (38°C)
Rubber content Pulse time (seconds)
(o/o by weight) 0.05 0.10 0.05 0.10. 0.05 0.10

0 16.978 10.799 2.353 1.920 0.520 0.401

3 8.321 6.462 0.705 0.533 0.244 0.193.
6 4.997 3.698 0.591 0.465 0.198 0.147
12 2.934 2.352 0.300 0.228 0.094 0.069

C\1

E
Q
C> N
~
(/)
E
::l Q
::; C> 10
u ~
0 (/)
~ 10 ::l
c ::;
"0
~ 0
·u; ~
a:
(])
c
c ~
ClS
(])
·u;
(])
~ a: 5
c
ClS
(])
~

60 80 100 60 80 100
Temperature (°F) Temperature (°F)

Figure 5. Mean resilient modulus for a 0.05- .Figure 6. Mean resilient modulus for a 0.10-second
second load time of various CRA C mix designs. load time of various CRAC mix designs.

8
 ROLLING RESISTANCE ON
RUBBER-AGGREGATE
ASPHALT PAVEMENTS

A thorough review of the literature was undertaken Control

concerning the addition of rubber particles to asphalt
41° F
pavement mixes, with a particular emphasis on the
potential increase in rolling resistance on rubber aggre- 10 3 /"-------------------
.... PlusRide
gate surfaces.
Rubber-modified asphalt surfaces have been con-
structed for evaluation purposes in at least 18 states,
n
including Alaska, California, Kansas, Massachusetts, n
Minnesota, Missouri, Montana, Nevada, New Jersey, 77
New Mexico, New York, Oklahoma, Rhode Island,
South Dakota, Tennessee, Texas, Utah, and Washington
(Stuart and Mogawer 1988, France 1989, Doty 1988,
Esch 1982, Civil Engineering 1990, Dvorak 1990). An ,.... _____
extensive evaluation program of rubber-aggregate as-
---
. phalt surfaces with rubber contents from 1 to 3% rubber
by weight is currently being conducted by the California
Department of Transportation (CALTRANS). The sur-
,/
'*' ------
104

....

faces constructed by CALTRANS with a rubber content 10 1 ~~--~~--~~--~--~~--~~

0 20 40 60 80 100
of 3% were made with the PlusRide material (France Aging Time (days)
1989, Doty 1988).
No data were found in the literature that directly Figure 7. PlusRide: Resilient modulus vs aging time
evaluated the rolling resistance of a rubber-aggregate and test temperature (after Stewart and Mogawer
asphalt surface. Laboratory data on the properties of a 1988).
PlusRide mixture, including the modulus of resiliency,
were found in a recent evaluation by the Federal High-
way Administration (FHWA) (Stuart and Mogawer resistance of an asphalt pavement surface. While most
1988). Table 12 and Figure 7 show the resilient modulus U.S. evaluations have focused on pavements with rubber
ofa PlusRide pavement specimen (3% rubber content by contents of 3% or less, the Texas Department of State
weight), as a function oftemperature and sample age, in Highways and Public Transportation reports using as-
comparison to a conventional asphalt pavement speci- phalt containing up to 25% crumb rubber (Civil Engi-
men. However, no relationship between the modulus of neering 1990). The increased rubber content would be
resiliency and ·the rolling resistance of a pavement sur- expected to improve the deicing characteristics of the
face was found in the literature. pavement surface.
There does not appear to be any method short of full- CRREL has estimated that the rolling resistance of a
scale testing to reliably quantify the effect of rubber- rubber-aggregate asphalt pavement surface may be 2 to
aggregate asphalt on vehicle rolling resistance. The 3% higher than a conventional asphalt surface. This
literature is also devoid of evaluations on the effect of the estimate, while based on engineering judgment, seems
rubber content on the strength, serviceability, and rolling reasonable given the available data on rolling resistance.

Table 12. Effect of aging: resilient modulus test results (Stuart and Mogawer 1988).

Control Plus Ride

Temperature Days Days
(oF) 2 7 14 28 90 2 7 14 28 90
Resilient modulus (ksi)
41 2.140 2.100 1.970 2.110 2.130 0.904 0.909 [Link] 1.160 1,030
77 0.248 0.283 0.239 0.255 0.301 0.222 0.165 0.212- 0.177 0.125
104 0.040 0.044 0.043 0.058 0.054 0.041 0.036 0.030 0.046 0.018

9
Table 13. Rolling resistances of passenger cars (St. John bond. The investigation was divided into four areas: 1)
and Kobert 1978, Institute of Transportation Engineers development of testing equipment, 2) test specimen
1982).
preparations, 3) initial static loading tests, and 4) wheel
Rolling resistance passage tests. Each of these areas will be discussed in the .
(lb_flton (~{vehicle weight) following sections.
Vehicle Badly broken Dry,
speed Smooth and patched well-packed Loose Development of testing equipment
(mph) pavement asphalt gravel sand A brief search of the literature was made to determine
2o 25 29 31 35 if any test apparatus had been previously developed to
30 27 34 35 40 test the effects of wheel passages on ice-pavement
40 29 40 50 57 dis bonding in the laboratory. The literature search did
50 31 51 62 76 not identify any test machine that met the requirements
60 34 of the proposed experiments. Therefore, a machine was
designed and built to simulate wheel passages on pave-
ment surfaces. This machine was a modification of the
Table 13 compares the rolling resistance of passenger standard circular track polishing machine (ASTM E660)
cars on smooth pavements and various surfaces of lower used in pavement surface wear research.
quality. These values are based on the work of Claffey, The wheel passage machine was designed with four
who measured the increased fuel consumption for ve- rubber-tired wheels following a circular track around a
hicle operation on lower quality surfaces (St. John and · central axis with a radius of 1.5 ft (45.7 em). The surface
Kobett 1978, Institute ofTransportation Engineers 1982, of the track could accommodate up to 12 cylindrical
Claffey 1971). The table shows that, at a vehicle speed pavement specimens, whose wearing surface was flush
of 20 mph (32 km/hr), rolling resistance is increased by with the surface of the track. The design of the machine
16% on a badly broken and patched asphalt pavement, in was such that each test surface could be up to 6 in. in
comparison to a smooth asphalt pavement. This increase diameter, with a maximum thickness of 2 in. When the
in rolling resistance is 38% for a vehicle speed of 40 mph machine was rotating at 31.8 rpm around its central axis,
(64 km/hr), and 65% for a vehicle speed of 50 mph (80 the wheels were traveling over the test pavement sur-
km/hr ). However, the increased rolling resistance caused faces at an equivalent speed of3.4 mph (5.5 kmlhr). The
by small increases in deformation for a rubber-aggregate wheels were loaded via calibrated compression springs
asphalt pavement would be expected to be much smaller that could be adjusted to apply a specified normal load.
than the increase in rolling resistance for a broken and The experiments were conducted using non-treaded
patched pavement. tires to eliminate ambiguous effects that the presence of
There does not appear to be any method short of full- tire treads could impose. Therefore, smooth pneumatic
scale testing to reliably quantify the effect of rubber- cart tires (4.10/3.50-5) with a width of 3.5 in. were
aggregate asphalt on vehicle rolling resistance. The initially used on the apparatus. Typical loadings for non-
literature is also devoid of evaluations on the effect of the treaded tires were found in the literature.
rubber content on the strength, serviceability, and rolling Preliminary runs indicated, however, that the cart
resistance of an asphalt pavement surface. While most tires were not behaving in the same manner as conven-
U.S. evaluations have focused on pavements with rubber tional automobile tires. The cart tires were carrying the
contents of 3% or less, the Texas Department of State majority of the normal load in the center of each tire's
Highways and Public Transportation reports using liq- footprint rather than at the sidewalls. Further, heat buildup
uid asphalt containing up to 25% crumb rubber by at the tire/ice interface from the rubbing action of the
weight of asphalt cement (Civil Engineering 1990). The untreaded tire turning in a tight radius caused ablation of
increased rubber content would be expected to improve the ice. The ablation of the ice and the heat buildup were
the deicing characteristics of the pavement surface. substantially reduced by modifying the wheel assem-
blies to allow more independent movement.
After several additional test runs, however, the wide
LABORATORY TESTING OF pneumatic tires were replaced by narrow hard rubber
PAVEMENT MODIFICATIONS wheels with a diameter of 9.875 in. (25.1 em) and width of
1.625 in. (4.13 em). The forces imposed on the ice under
Laboratory experiments were conducted to investi- the narrow tires were equivalent to the forces imposed
gate the effectiveness of varibus passive pavement modi- on the ice under the sidewall of a conventionally loaded
fications that might be used in combination with pas- automobile tire. The narrow tires also eliminated, to a
sages of wheeled vehicles to break the ice-pavement great extent, the ice ablation due to tire rubbing.

10
Preparation of coating from the surface of the aggregate and rubber
pavement surface specimens particles located at the specimen surface. Hand sanding·
The pavement surface specimens used for testing with coarse sandpaper, sandblasting, wire brushing by
were core samples taken from various parts of the hand and machine, and cutting the sample with a dia-
country and laboratory-produced test specimens devel- mond saw to expose a new surface all produced unsatis-
oped by CRREL using the Marshall method. The pave- factory results. The initial attempts at sandblasting the
ment surface specimens were provided by the highway surface of the pavement specimen to remove the asphalt
or transportation departments of California, Connecti- coating failed because the sandblasting left particles of
cut, and New York and .included several types of port- sand embedded in the soft asphalt. However, these initial
land cement concrete, conventional asphalt and rubber- attempts were performed at room temperature. We found
modified asphalt surfaces. The portland cement concrete that when the sample was frozen, the sand particles did not
specimens included both conventional smooth~ finished become embedded in the asphalt. Therefore, this method
surfaces and surfaces with 114-in. grooves. The conven- of freezing the specimens and then sandblasting them
tional asphalt specimens included both dense-graded was used to remove the asphalt coating from the surface.
and open-graded mixes. The laboratory-produced speci- The surface of each specimen was cleaned using a
mens were rubber-modified asphalt samples in the form procedure developed under SHRP contract H-203. This
of 4- and 6-in. circular pavement specimens with vary- cleaning procedure involved first rinsing each specimen
ing percentages of rubber added. with ethyl alcohol, scrubbing it with a stiff brush, rinsing
Several preliminary 4-in.- (10.2-cm-) diam. rubber- it three more times with ethyl alcohol, drying it in a
modified asphalt specimens were produced in the labo- vacuum chamber for 1 hr, rinsing it again three times
ratory by CRREL. The asphalt content of these speci- with deionized water, and letting it air dry. Ice was then
mens ranged from 6 to 9.5% by weight (see Table 4) and grown on the cleaned surfaces_ in a bottom-up mode,
the maximum rubber particle size was 1/2 in. (1.3 em). again using the procedures developed under SHRP Con-
CRREL found that increased asphalt percentages were tract H-203.
required as the amount of fine rubber particles used in the
mix was increased. Based on the evaluation of the Preliminary loading experiments
samples, it was hypothesized that the fine rubber par- Initial deflection and single event load tests were
ticles contribute only to the elasticity of the pavement. A performed on saw cut rubber-modified CRAC asphalt
concern was expressed that a pavement surface con- surfaces with a thin ice layer [1/16 in. (0.16 em) thick-
structed of a material with the necessary elasticity to ness] to gain an understanding of the force required to
cause destruction of the ice-pavement bond might in- cause ice fracture on the specimens. An Instron tensile/
crease tire-pavement rolling resistance and reduce pave- compression machine was used in these tests. The Instron
ment durability. It was further hypothesized that destruc- machine was equipped with an environmental chamber
tion of the ice-pavement bond could be achieved in a less that encompassed the loading base and tup. The loading
elastic pavement if localized ice deflections at sites of tup, which induced the forces onto the specimens, was
rubber particles could induce crack propagation. There- constructed of a 2-in.- (5.1-cm-) diameter rubber stop-
fore, it was decided that CRREL should develop addi- per with approximately the same durometer reading
tional mix designs for rubber-modified asphalt mixes by as that of a conventional tire. This tup was attached with
increasing the percentage of large rubber particles and epoxy to a steel ram of the same diameter. The tests
reducing the percentage of fine rubber particles to as were conducted at a temperature of approximately
small a level as possible. The maximum rubber particle 15°F (-9°C).
size selected was 3/8 in. (0.95 em) rather than the 1/2-in. Pressure loadings ranging up to 150 psi (1.03 x 106
(1.3 em) size used in the earlier specimens. Pa) were imposed with the rubber tup on ice-covered
The revised CRAC mix design was used to produce rubber-modified asphalt surfaces containing 12 and 25%
thenextsetofrubber-modifiedasphaltspecimens. These rubber. Cracks in the ice were not detected on any of
specimens were 4 in. (10.2 em) and 6 in. (15.2 em) in these specimens. The thickness of the ice layer was then
diameter and 2.5 in. thick. Pavement specimens were reduced from l/16 in. thickness (0.16 em) to a very thin
made with the following rubber contents: 0, 3, 6, 12, 25, layer. Again, the 150-psi ( 1.03,.MPa) pressure loadings
57, and 100 percent (by weight). were applied, but only one small crack developed (on the
During preparation, the coarse aggregate and rubber specimen with the 25% rubber content). This crack
particles at the surfaces of the specimens became cov- seemed to heal itself in a short amount of time. From
ered with a thin coating of asphalt, a condition not typical these facts, it was evident that conventional ice fractur-
of roadway surfaces' that are open to traffic. Various ing on rubbet-modified asphalt surfaces would have to
techniques were used in an attempt to remove the asphalt be generated through repeated loadings and fatigue, not

11
through the application of single loads. Consequently, it particles on the surface of the specimen, but with no
was decided to postpone further Instron testing. regularity or consistency. One specimen developed a
crack near its center, while another specimen developed
Wheel passage tests a crack near its edge. Both of these cracks were in the
Initial runs with the wheel passage device were made wheel path and both were at the site of a rubber particle
with the various portland cement concrete, conventional in the pavement surface. Neither temperature nor ice
asphalt specimens, and a rubber-modified (PlusRide) thickness appeared to make a significant difference in
asphalt specimen acquired from California. The port- results.
land cement concrete specimens included smooth and For the CRAC specimens with 6% rubber content, ice
grooved surfaces. The asphalt specimens included both cracking in the wheel path was observed to be more
dense-graded and open-graded asphalt types. The rubber consistent and extensive than for the 3% specimens. The
content in the California specimens was unknown, but portion of the ice surfaces at the edges of the specimens,
from observation of the specimens it was estimated to be where the wheel passages began and ended, showed
less than 3%. consistent cracking where the rubber particles were
The initial tests were conducted at a temperature of present. The cracking was observed at both temperatures
25°F (--4 °C) and with an ice thickness of about 114 in. and both ice thicknesses. The degree of cracking in the
(0.64 em). Each specimen was subjected to 400 wheel wheel path increased with increasing wheel passes and
passages. No significant cracks were detected in the ice increased much more markedly at 25°F (4°C) than it did
layers, and the ice-pavement bond remained undam- for the same number of wheel passes at 15°F (-9°C).
aged. The tests were terminated after the tires wore the For the CRAC specimens with 12% rubber content, a
ice down to the substrates with no visible cracking much greater occurrence of cracks was observed at both
having taken place on most of the specimens. A few very temperatures and ice thicknesses. Cracking developed
small cracks were thought to occur around some of the around the rubber particles at the edges of the specimens
exposed rubber particles of the PlusRide asphalt speci- in the wheel path after only a very few wheel passages.
mens. This ended the testing with the conventional As the number of wheel passages increased, further
portland cement concrete and asphalt specimens. Atten- cracking occurred at the sites of the rubber particles
tion was then directed toward testing rubber-modified throughout the wheel path. During these tests, it ap-
asphalt surfaces. peared that the ice became fatigued and the disbanding
Wheel passage tests of rubber-modified asphalt sur- at the ice/specimen interface began to take place. Deteri-
faces were designed to investigate four factors that may oration of the ice-pavement bond continued to the point
contribute to crack initiation and propagation at the ice/ that the ice became fully disbanded from the pavement
pavement interface. These factors were the percentage surface, leaving the specimen surface exposed.
of rubber content (Blackburn et al. 1978), the ambient Ice grown on the CRAC surfaces with 25 and 57%
temperature (Stuart and Mogawer 1988), the ice thick- rubber content experienced cracki~g with as few as 10
ness (Takallou et al. 1985) and the number of wheel wheel passes. After 6000 passes on the surface with 25%
passages over the specimen surface (Foster 1985). These rubber content, as much as 50% of the wheel path area
additional wheel passage tests were conducted with ice was cleared of all ice, while the remaining ice in the
layers grown on 6-in.- ( 15 .2-cm-) diam. specimens made wheel path showed signs of severe deterioration of the
according to the Chunk Rubber Asphalt Concrete mix ice-pavement bond. All CRAC specimens with rubber
designs given in Tables 2~9. The rubber content of these contents higher than 25% showed signs of severe crack-
CRAC test specimens varied from zero to 100%. Tests ing and ice-pavement bond deterioration after only few
were performed at two temperatures [ 15°F (-9°C) and wheel passes. The specimen with 100% rubber content
25°F (--4°C)] and two ice thicknesses [1/16 in. (0.16 em) experienced 50% ice removal in the wheel path after
and 118 in. (0.32 em)]. The number of wheel passes only 400 wheel passages and total ice removal after only
ranged from 1 to over 6000. Multiple tests of selected 1000 wheel passages.
combinations of percentage rubber content, tempera- We concluded from the wheel passage tests that the
ture, and ice thickness were run to verify ice cracking and occurrence and frequency of cracking are directly re-
disbanding characteristics. lated to the surface condition of the specimen. The extent
Specimens with 0% rubber did not develop cracks in of the ice cracking on the specimens varied with rubber
the ice layer or undergo any obvious deterioration of the content and number of wheel passages from no cracking
ice-pavement bond. These results were the same for to total disbandment. The results clearly indicate that
each temperature-ice thickness combination. increased rubber content (i.e., an increased presence of
CRAC specimens with 3% rubber content developed larger rubber particles on the pavement surface) in-
occasional cracks in the ice at the locations of rubber creased the incidence of cracking. Surface characteris-

12
tics other than the presence of rubber particles did not A new material, Chunk Rubber Asphalt Concrete
appear to affect cracking. (CRAC) was developed, using rubber particles of3/8 in.
The size and origin of the cracks for CRAC surfaces and larger than the no. 4 sieve, vs the no. 10 and no. 30
with lower rubber contents (<12%) indicate that these sieves for the PlusRide. CRAC would fit under the new
surfaces rely on localized deflection around the rubber industry description for Rubber Modified Asphalt Con-
particles to induce cracking, while asphalt surfaces with crete or RUMAC.
rubber content over 12% experienced disbandment The Marshall Stability values for CRAC doubled in
through area deflection. strength for the 3, 6, and 12% rubber contents. Mean
The location of the rubber particles in the asphalt resilient modulus values were also greater for the CRAC
surfaces with lower rubber contents (12% or less) was mixes, indicating a stronger, more dense mix to support
found to make a substantial difference in crack propaga- the rubber particles protruding above the pavement
tion. Testing indicated that cracks tended to develop at surface.
particles of rubber located on the pavement surface. There does not appear to be any method to reliably
Further, it was found that increasing the rubber content quantify the effect of rubber aggregate pavement on
within the range of rubber content below 12% does not vehicle rolling resistance short of constructing full-scale
necessarily ensure a proportionate increase in crack test sections.
propagation. For example, one surf~ce with a 6% rubber The literature is devoid of evaluations on the effect of
content actually had more exposed rubber particles than rubber content on the strength, serviceability, and rolling
a similar surface with 12% rubber content. Under the resistance of an asphalt pavement surface. Increased
same loading, the 6% specimen developed more cracks rubber content would be expected to improve the deicing
in fewer wheel passes than the 12% specimen. Further- characteristics of the pavement surface; however, strength
more, it should be understood that the cracks which and serviceability must also be evaluated to determine
developed in the wheel path at the edges of the speci- optimum rubber contents for various levels and types of
mens cannot be attributed exclusively to edge effects. traffic.
Even at these locations, cracks developed only at loca- Although field verification was not possible within
tions where rubber particles were present. the time constraints of the SHRP program, laboratory
The area deflection experienced on CRAC surfaces wheel testing results clearly indicate that increased rub-
with higher rubber contents results from the increasing ber content (i.e., an increased presence of larger rubber
elasticity of the surface and decreased ability of ice particles on the pavement surface) increased the inci-
formed on that surface to support the wheel loads. As the dence of ice cracking. CRAC surfaces with lower rubber.
amount of support provided to the ice decreases, due to contents indicate that ice cracking relies on localized
the increased presence of elastic rubber particles, the deflection around the rubber particles. Surfaces with
ability to support the loading diminishes and the wheel over I 2% rubber experienced disbandment through area
load must increasingly be supported by the ice layer. deflection.
Thicker ice layers have much greater strength and bridg-
ing ability, but the ice layers of 1116- and 118-in. (0.16-
and 0.32-cm) thickness cannot support the loading im- RECOMMENDATIONS
posed by a typical automobile tire on CRAC surfaces
with high rubber content. Consequently, these thin ice Full-scale field test sections are required to determine
layers begin to crack and fatigue'after repeated loading the optimum rubber content for the new Chunk Rubber
and are eventually disbanded and separated from the Asphalt Concrete (CRAC) materials. Strength, service-
asphalt surface. ability, and rolling resistance need to'be measured under
real-world conditions with vehicles trafficking the test
sections.
SUMMARY Funding is being sought to construct field test sec-
tions at CRREL for evaluation with its test vehicle and
Rubber aggregate was selected as the best method for conduct needed laboratory and field tests to define
producing an asphalt concrete wearing surface from CRAC mixes for different levels and types of traffic.
which ice would dis bond under traffic, and the PlusRide
concept using 3, 6, and 12% rubber by weight was
evaluated. Based upon laboratory wheel loading tests LITERATURE CITED
conducted at the Midwest Research Institute, it was
decided to use larger pieces of rubber aggregate to Blackburn, R.R., A.D. St. John and P.J. Heenan
increase ice breakup under traffic. ( 1978) Physical alternatives to chemicals for highway

13
deicing. Final Report ofMRI to FHWA, U.S. Depart- Technologists Symposium on Asphalt Mix Design, Feb-
ment of Transportation. ruary.
Civil Engineering (I 990) Old tires bounce back. News France, A. ( 1989) Asphalt rubber tested in Caltrans 1-80
Briefs, June. Project. Roads and Bridges, March.
Claffey, P. (1971) Running costs of motor vehicles as Institute of Transportation Engineers ( 1982) Trans-
affected by road design and traffic. NCHRP Report Ill, portation and Traffic Engineering Handbook. Second
Transportation Research Board. Edition, Englewood Cliffs, N.J.: Prentice-Hall.
Doty, R.N. (1988) Flexible pavement rehabilitation St. John, A.D. and D.R. Kobett ( 1978) Grade effects on
using asphalt-rubber combinations-A progress report. traffic flow stability and capacity. NCHRP Report 185,
California Department of Transportation, Presented at Transportation Research Board.
the 67th Annual Meeting ofthe Transportation Research Stuart, K.D., and W.S. Mogawer (1988) Laboratory
Board, January. evaluation ofVerglimit and PlusRide. Draft Version of
Doty, R.N. (1988) Flexible pavement rehabilitation, Report No. FHWA-RD-88-173, Federal Highway Ad-
using asphalt-rubber combinations-A progress report. ministration.
Transportation Research Record 1196. Takallou, H.B., J. McQuillen, Jr., and R.G. Hicks
Dvorak, J.A. (1990) Old tires cover the road. Kansas (1985) Effect of mix ingredients on performance of
City Star, August 18. rubber modified asphalt mixtures. FHWA-AK-RD-86-
Esch, D.C. ( 1982) Construction and benefits of rubber- 05, Federal Highway Administration.
modified asphalt pavements. Transportation Research VanKirk, J.L. (1989) CALTRANS experience with
Record 860. asphalt-rubber concrete-An overview and future di-
Foster, C.R.( 1985) The strength of asp~alt pavements. rection. Paper presented at the National Seminar on
Paper submitted to Association of Asphalt Pavement Asphalt-Rubber, Kansas City, Missouri, October 30-31.

14
 Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
1-'uouc reporting ouroen lor tn1s couect1on Ollnlormat1on IS est1mated to average 1 hour per response. mcluding the t1me tor rev1ewmg 1nstruct1ons, searching ex1s11ng aata sources, gathenng and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestion for reducing this burden, to Washington Headquarters Services. Directorate for Information Operations and Reports. 1215 Jefferson Davis Highway, Suite 1204, Arlington,
VA 22202·4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188). Washington, DC 20503.

1. AGENCY USE ONLY (Leave blank) ,2. REPORT DATE ,3. REPORT TYPE AND DATES COVERED
· December 1991
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Use of Scrap Rub~e~ inAsphalt Pavement Surfaces

SHRP Contract H-204
6. AUTHORS

Robert A. Eaton, Richard J. Roberts and Robert R. Blackburn

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

REPORT NUMBER
U.S. Army Cold Regions [Link] Midwest Research Institute
and Engineering Laboratory Kansas City, Missouri
72 Lyme Road Special Report 91-27
Hanover, New Hampshire 03755-1290
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING
AGENCY REPORT NUMBER

Michigan Technological University Strategic Highway Research Program

Houghton, Michigan Washington, D.C.

11. SUPPLEMENTARY NOTES

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release; distribution is unlimited.

Available from NTIS, Springfield, Virginia 22161

13. ABSTRACT (Maximum 200words)

Scrap tire rubber was mixed into an asphalt concrete wearing course to study the effect of ice dis bonding from the pavement
surface under traffic. Rubber contents ofO, 3, 6, and 12% by weight were studied. Initial laboratory ice disbanding test results
led to the development of a new paving material, Chunk Rubber Asphalt Concrete (CRAC), that uses larger pieces of rubber
in a much denser asphalt concrete mix. Strength values doubled and ice disbanding performance was enhanced.

14. SUBJECT TERMS 15. NUMBER OF PAGES

21
Asphalt concrete Pavements Tires 16. PRICE CODE
Ice-pavement bonding Scrap rubber
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF REPORT OF THIS PAGE OF ABSTRACT
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED UL
NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. Z39-18
298-102