Fatigue Behavior of Rubberized Asphalt Concrete Mixtures Containing Warm

Asphalt Additives

Feipeng Xiao1*, Ph.D., P.E., Wenbin Zhao2, and Serji N. Amirkhanian3, Ph.D.
1
Research Assistant Professor, Asphalt Rubber Technology Service, Department of Civil
Engineering, Clemson University, Clemson, SC 29634-0911 U.S.A., Tel: 001-864-6566799,
Fax: 001-864-6566186, E-mail: feipenx@clemson.edu
2
Graduate Research Assistant, Department of Civil Engineering, Clemson University, Clemson,
SC 29634-0911
3
Professor, Department of Civil Engineering, Clemson University, Clemson, SC 29634-0911

ABSTRACT:
The long-term performance of pavement is associated with various factors such as pavement
structure, materials, traffic loading, and environmental conditions. Improving the understanding
of the fatigue behavior of the specific rubberized warm mix asphalt (WMA) is helpful in
recycling the scrap tires and saving energy. This study explores the utilization of the
conventional fatigue analysis approach in investigating the fatigue life of rubberized asphalt
concrete mixtures containing the WMA additive. The fatigue beams were made with one rubber
type (-40 mesh ambient crumb rubber), two aggregate sources, two WMA additives
(Asphamin® and Sasobit®), and tested at 20ºC. A total of 8 mixtures were performed and 29
fatigue beams were tested in this study. The test results indicated that the addition of crumb
rubber and WMA additive not only reduced the mixing and compaction temperatures of
rubberized asphalt mixtures offset by crumb rubber but also effectively extended the long-term
performance of pavement when compared with conventional asphalt pavement. In addition, the
exponential function forms are efficient in achieving the correlations between the dissipated
energy and load cycle as well as mixture stiffness and load cycle.

Keywords: Rubberized asphalt concrete; Warm asphalt additive; Mixing and compaction
temperature; Stiffness; Dissipated energy; Fatigue Life.

*: Corresponding author
 Xiao et al. (2008)

INTRODUCTION

Fatigue cracking, called alligator cracking and associated with repetitive traffic loading, is

considered to be one of the most significant distress modes in flexible pavements. The fatigue

life of an asphalt pavement is directly related to various engineering properties of a typical hot

mix asphalt (HMA). The complicated microstructure of asphalt concrete is related to the

gradation of aggregate, the properties of aggregate-binder interface, the void size distribution,

and the interconnectivity of voids. As a result, the fatigue property of asphalt mixtures is very

complicated and sometimes difficult to predict (1-3).

Understanding the ability of an asphalt pavement to resist fractures from repeated loading

condition is essential for developing superior HMA pavement designs. Previous studies have

been conducted to understand the occurrence of fatigue and how to extend pavement life under

repetitive traffic loading (3-4). However, reaching a better understanding of fatigue behavior of

asphalt pavements continues to challenge researchers worldwide, particularly as newer materials

with more complex properties are being used in the field.

The recycling of scrap tires has been of interest to the domestic and international asphalt

industry for over 40 years. The utilization of crumb rubber modifier (CRM) in asphalt binders

has proven to be beneficial from many stand points. The use of CRM, expanded to HMA,

continues to evolve since the CRM binders enhance the performance of asphalt mixtures by

increasing the resistance of the pavements to permanent deformation and thermal and fatigue

cracking. Many researchers have found that utilizing crumb rubber in pavement construction is

both effective and economical (5-9).

Recently, the “warm mix asphalt” (WMA) is widely being used in the hot HMA industry as

a mean of reducing energy requirements and lowering emissions. WMA can significantly reduce

2
 Xiao et al. (2008)

the mixing and compacting temperatures of asphalt mixtures, by either lowering the viscosity of

asphalt binders, or causing foaming in the binders. Reduced mixing and paving temperatures

decreases the energy required to produce HMA, reduces emissions and odors from plants, and

makes for better working conditions at both the plant and the paving site (10-15).

However, the influence of crumb rubber and WMA additives mixed with virgin mixtures

together has not yet been identified clearly. The interaction of modified mixtures is not well

understood from the standpoint of binder properties and field performance. It has been shown

that the WMA additives reduce the mixing and compaction temperatures and achieve ideal

workability of HMA without significantly affecting the engineering properties of the mixtures

(10-13). While the addition of crumb rubber increases the demand of asphalt binder and

increases the mixing and compacting temperatures, it is helpful in resisting the high temperature

deformation and extending the long-term performance of HMA. Because of the complicated

relationships of these two materials in the modified mixtures, detailed information will be

beneficial to help obtain an optimum balance in the use of these materials. Very few fatigue

studies of modified asphalt mixtures, including crumb rubber and warm asphalt additives, have

been performed in recent years (16). However, the utilization of these materials will enable the

engineers to find an environmentally friendly method to deal with these materials, save money,

energy, and furthermore, protect the environment.

The objective of this study was to gain an improved understanding in the long-term

performance characteristics (fatigue behavior) of the rubberized asphalt concrete mixtures

containing WMA additives through a series of experimental tests. Experiments were carried out

to evaluate rheological properties of the modified binder (unaged and aged binders) as well as

3
 Xiao et al. (2008)

the engineering properties of the mixture, such as the stiffness and fatigue life performed by

HMA flexural testing.

BACKGROUND

The fatigue characteristics of asphalt mixtures are usually expressed as relationships

between the initial stress or strain and the number of load repetitions to failure-determined by

using repeated flexure, direct tension, or diametral tests performed at several stress or strain

levels. The fatigue behavior of a specific mixture, characterized by the slope and relative level of

the stress or strain versus the number of load repetitions to failure, may be defined using the

following equation (17).

N f  a(1 /  0 ) b (1 / S 0 ) c or N f  a (1 /  0 ) b (1 / S 0 ) c (1)

Where N f = number of load application or crack initiation,  0 ,  0 = tensile strain and stress,

respectively, So = initial mix stiffness, and a, b, c = experimentally determined coefficients.

In recent years, several researchers have used the energy approach for predicting the fatigue

behavior of the asphalt mixtures. The dissipated energy per cycle, Wi, for a linearly viscoelastic

material is given by the following equation (18-22):

n n
W  Wi     i i sin( i ) (2)
i 1 i 1

Where, W = cumulative dissipated energy to failure, Wi = dissipated energy at load cycle i,

 i = stress amplitude at load cycle i,  i = strain amplitude at load cycle i, and  i = phase

shift between stress and strain at load cycle i.

4
 Xiao et al. (2008)

Research has shown that the dissipated energy approach makes it possible to predict the

fatigue behavior of mixtures in the laboratory over a wide range of conditions based on the

results of a few simple fatigue tests. Such a relationship can be characterized in the form of the

following equation (18-22):

W  A (N f )Z (3)

Where, Nf = fatigue life, W= cumulative dissipated energy to failure, and A, Z =

experimentally determined coefficients.

EXPERIMENTAL PROGRAM AND PROCEDURES

Materials

One virgin binder (PG 64-22) and one crumb rubber modified (CRM) binder (PG 64-22 +

10% -40 mesh rubber) were used in this study. The PG 64-22 binder was a mixture of several

sources that could not be identified by the supplier. One type of rubber, -40 mesh ambient rubber,

was used in this study. Previous research and field projects conducted in South Carolina

indicated that the -40 mesh ambient rubber is effective in improving the engineering properties

of rubberized mixtures.

Asphamin® and Sasobit® were used in this study as two WMA additives. Aspha-min® is

Sodium–Aluminum–Silicate which is hydro thermally crystallized as a very fine powder. It

contains approximately 21% crystalline water by weight. By adding it to an asphalt mix, the fine

water spray is created as all the crystalline water is released, which results in volume expansion

in the binder, therefore increasing the workability and compactability of the mix at lower

temperatures. Sasobit® is a long chain of aliphatic hydrocarbons obtained from coal gasification

5
 Xiao et al. (2008)

using the Fischer-Tropsch process. After crystallization, it forms a lattice structure in the binder

which is the basis of the structural stability of the binder containing Sasobit® (10-11).

Two aggregate sources (A and B) were used for preparing the samples (Table 1). Aggregate

A, a type of granite, is predominantly composed of quartz and potassium feldspar while

aggregate B (schist) is a metamorphic rock. Hydrated lime, used as an anti-strip additive, was

added at a rate of 1% by dry mass of aggregate. A total number of eight mixtures were evaluated

in this research. In this paper, the mixtures made from aggregates A and B without rubber and

WMA additive are referred to as ACO and BCO; and the mixtures with rubber but no WMA

additive are referred to as ARO and BRO. In addition, the mixtures with rubber and Asphamin®

are designated as ARA and BRA, and the mixtures with rubber and Sasobit® are labeled as ARS

and BRS, respectively.

Superpave mix design

The combined aggregate gradations for the 12.5 mm mixtures were selected in accordance

with the specification set by the South Carolina Department of Transportation (SCDOT). The

gradations for each aggregate source (A and B) are shown in Figure 1, which shows that the

design aggregate gradations for each aggregate source are the same when using different WMA

additives (Asphamin® and Sasobit® ) at the same percentages of rubber (0% or 10% rubber),

while the gradations are similar when comparing mixtures from both aggregate sources.

Superpave mix design defines that the laboratory mixing and compaction temperatures can

be determined by using a plot of viscosity versus temperature. While there are no previous

specifications available regarding the mixing and compaction temperatures for rubberized

mixture containing WMA additives, some researchers have developed guidelines for mixing and

6
 Xiao et al. (2008)

compaction temperatures when using either WMA or rubber (10-11, 23-24). The temperatures,

shown in Table 3, were determined in accordance with previous research projects (16, 24).

Though the mixing and compaction temperatures increase as the percentage of crumb rubber

increases, these can be reduced by adding either Asphamin® or Sasobit®.

Fatigue beam fabrication and test procedures

Fatigue beams were made in the laboratory and two-four beams of each mixture were tested

for this study (Figure 2). All tests were performed in a temperature-controlled chamber at 20 ±

0.5°C. In this study, a repeated sinusoidal loading at a frequency of 10 Hz was used; in addition,

the controlled strain mode was employed. The control and data acquisition software measured

the deflection of the beam specimen, computed the strain in the specimen and adjusted the load

applied by the loading device (AASHTO T321).

The test apparatus also recorded load cycles, applied load, and beam deflections. Failure is

assumed to occur when the stiffness reaches half of its initial value, which is determined from

the load at approximately 50 repetitions; the test is terminated automatically when this load has

diminished by 50 percent. The flexural stiffness and dissipated energy of fatigue beam are

determined as follows (AASHTO T321):

1. Flexural stiffness (Pa):

aP(3l 2  4a 2 )
S  /  (4)
4bh3

Where,  = tensile stress, in Pa;  = maximum tensile stain, in m/m; P = applied peak-

to-peak load, in Newton; a = space between inside clamps, in meters; b = average beam

width, in meters; h = average beam height, in meters; Δ= beam deflection at neutral axis,

in meters; and l = length of beam between outside clamps, in meters.

7
 Xiao et al. (2008)

2. Dissipated energy (J/m3) per cycle:

D   sin( 360 f ) (5)

Where, f = load frequency, in Hz; and  = time lag between Pmax and  max , in second

3. Cumulative dissipated energy (J/m3):

i n
W=  Di (6)
i 1

Where, Di = D for the ith load cycle

ANALYSIS OF TEST RESULTS

Statistical considerations

Results of the stiffness, cumulative dissipated energy, and fatigue life values were

statistically analyzed with 5% level of significance (0.05 probability of a Type I error) with

respect to the effects of aggregate sources and WMA additive types. For these comparisons, it

should be noted that all specimens were produced at optimum binder content.

Binder analysis

Table 3 shows that the viscosity of rubberized asphalt binder decreases while the high

temperature performance (G*/sinδ) of overall binders increase with the addition of WMA

additive. The unaged binder test result shows that the Asphamin® and Sasobit® can improve

the workability (viscosity) and rutting resistance (G*/sinδ) of mixtures. While the aged

rubberized binders show that the G*sinδ values decrease with the addition of rubber, these values

increase slightly as the WMA additives are added. It also can be seen that the stiffness values of

binders have similar trends with G*sinδ values due to the addition of these materials. Aged

8
 Xiao et al. (2008)

binder properties show that the WMA additives produce a slightly effect on the long-term

performance of asphalt binder.

Analysis of fatigue test results

Testing data were analyzed using Equations 4-6 presented earlier to compute the stress,

strain, stiffness, phase angle, the dissipated energy per cycle as the function of the number of

load cycles, and the cumulative dissipated energy to a given load cycle. In this study, fatigue life

was defined as the number of repeated cycles corresponding to a 50 percent reduction in initial

stiffness, which was measured at the 50th load cycle. Several fatigue beam specimens were

utilized to characterize the fatigue behavior of a mixture in order to avoid too much or too little

loss in stiffness. This procedure involved testing control specimens (ACO and BCO samples) at

a 500 micro strain level with the controlled strain mode of loading at a frequency of 10 Hz.

Table 4 presents a typical analyzed fatigue test results which were computed at various

cycles from the raw data. It can be seen that the stress value and dissipated energy per cycle

generally decrease as the number of cycle increases. That is, at the same strain level, the greater

stress is needed to reach the desired strain values at the beginning of fatigue test than at the end

of the test. At the same time, the dissipated energy per cycle during the first thousands of cycles

is remarkably greater than those during the final cycles (50% loss of initial stiffness). As

expected, the asphalt pavement in the field rapidly releases the potential energy within the first

several years, followed by the further reduction of pavement performance caused by micro-

cracking under repeated traffic loading conditions. Previous research also presents this similar

long-term performance process (3-4, 21, 22).

9
 Xiao et al. (2008)

The test results presented in Figure 3(a) show that the fatigue life of fatigue beams made

from aggregate A has a greater value than those made from aggregate B, though the aggregate B

has a lower LA abrasion loss and absorption values. Conversely, the standard deviations of the

fatigue test results for each mixture are large since the variability of fatigue life is generally

based upon the micro-structure of beams (e.g. the aggregate-binder interface, the void size

distribution, the interconnectivity of voids, distribution of aggregate particles, film thickness and

the aged status of binder). Through previous research, which also determined that large

variabilities exist in the fatigue test results, the authors found that increasing the number of the

repeated specimens reduced the variability (24). Moreover, in comparison with the control

fatigue beam (without rubber and WMA additive), the rubberized fatigue beam without WMA

additive or with Sasobit® additive has a slightly greater fatigue life while the rubberized fatigue

beam with Asphamin® additive has a slightly lower fatigue life, regardless of aggregate sources.

Figure 3(a) shows that the addition of crumb rubber and/or Sasobit® slightly benefits the long-

term performance of asphalt pavement while the Asphamin® results in a slight decrease of the

fatigue life, though these additives are critical in reducing the mixing and compaction

temperatures of mixture. In addition, the statistical analysis (t-statistics) in Table 5 indicates that,

with respect to the effect of aggregate source, there is a significant different fatigue life value

between any two aggregate sources regardless of mix types. As shown in Table 6, the influence

of a WMA additive on the fatigue life is generally not significant (p-value > 0.05) for overall

mixtures

The flexural stiffness of an asphalt pavement, associated with repetitive traffic loading

and pavement thickness, is related to the various aspects of HMA, such as rutting, resilient

modulus, and fatigue life. In this study, the fatigue beams were made with a height of

10
 Xiao et al. (2008)

approximate 50 mm and the values were competed from Equation (3), defined as the ratio of

tensile stress-to-tensile strain. The test results shown in Figure 3(b) show that the aggregate A

mixture has greater stiffness values since under the repeated loading the induced micro-strain of

the mixture from aggregate A is smaller. This greater stiffness may be the result of different

aggregate sources producing different interfaces among the binder, voids, and aggregate, thus,

affecting the corresponding fatigue behavior of the pavement. Previous research indicates that

while the initial stiffness of rubberized mixture is less than the conventional mixture (19), the

initial stiffness values of mixtures in this study showed no obvious trend when the additional

crumb rubber and WMA additive were blended together. Moreover, the statistical analysis in

Tables 5 and 6 indicates that aggregate source has a significant influence on the stiffness values

generally while the effects of rubber and WMA additive is not significant for all four types of

mixtures.

The dissipated energy, computed from Equations 5 and 6 was used as an indicator of

fatigue cracking in the asphalt layer (19-22). As shown in Figure 3(c), the cumulative dissipated

energy of mixture made from aggregate B is slightly higher than that of mixture from aggregate

A. However, the statistical results in Table 5 indicate that, except for the mixture with Sasobit

additive, other mixtures from two aggregate sources have no significant different cumulative

dissipated energy values. With respect to the effect of rubber and WMA additive, Table 6 shows

that there is a significant different value between control and rubberized mixture in general, but

the influence of WMA additive on cumulative dissipated energy is not significant for all

rubberized mixture.

11
 Xiao et al. (2008)

Correlation analysis of fatigue test factors

G*sinδ, strongly associated with fatigue life of the mixture, has become a basic parameter

used to describe fatigue characteristics of asphalt binder. Thus, the study of G*sinδ is beneficial

for researchers and engineers to analyze fatigue behavior of asphalt pavements. Table 3 indicates

the effect of rubber and WMA additive on G*sinδ of binder. In Figure 4(a), it can been seen that

the fatigue life decreases remarkably with a corresponding increase of G*sinδ regardless of

aggregate source. Figure 4 indicates that the binder aging process, for the materials used in this

research, does shorten the fatigue life of asphalt pavement. However, the addition of crumb

rubber enhances the long-term performance of asphalt pavement.

The correlations between the stiffness of the beam and binder are shown in Figure 4(b).

Similar to Figure 4(a), it can be seen that the stiffness values of the beam do not have a large

alteration with an increase in binder stiffness. Table 3 shows that the rubberized binders have

lower stiffness values than original asphalt binder. However, the stiffness values of the

rubberized mixture do not exhibit a similar trend. In addition, as shown in Figure 4(b), two

aggregate sources also show the different effects on mixture stiffness in terms of binder stiffness

in this study.

AASHTO T321 assumes that the fatigue life depends on the accumulation of dissipated

energy from each load cycle. Thus, the dissipated energy may be plotted against load cycles for

the particular load cycles where the data was collected. As shown in Figure 5, the correlations

between the dissipated energy per cycles with load cycles indicate that the dissipated energy

increases at a negative exponential growth as the number of load cycles increase, in other words,

the dissipated energy decreases insignificantly initially and then it reduces rapid prior to reaching

the 50% stiffness. For example, as shown in Figure 5(a), the dissipated energy of three fatigue

12
 Xiao et al. (2008)

beams exhibits a slightly decrease before the number of the repeated loads is less than 10,000

cycles, after that, the dissipated energy decreases quickly until the final load cycle accomplishes.

Figure 5 indicates that the individual fatigue beam from each mixture has different

dissipated energy values per load cycle, regardless of the mixture types (i.e. ACO, ARO, etc) and

these values are greater when using aggregate B. The results in Figure 5 also show that the

crumb rubber and WMA additive do not affect the dissipated energy per cycle.

The initial stiffness of fatigue beam, determined by the initial tensile stress and strain, can

be plotted using stiffness (S) against load cycles (n) and best fitting the data to exponential

function of the form shown below:

S  A ebn (7)

Where, e = natural logarithm to the base e, and A, b = experimentally determined

coefficients.

As shown in Figure 6, it can be noted that, in most cases, the stiffness values of various fatigue

beams from same mixture present similar results, and the mixture from aggregate B has a greater

stiffness value. However, the addition of crumb rubber and WMA additive do not exhibit a

significant effect on the stiffness values regardless of aggregate sources.

The correlations between the repetition number of fatigue beam and cumulative

dissipated energy are shown in Figure 7(a). Although these two linear models can be used to

determine the predicted values, it was hard to obtain accurate results due to the limited test

specimens and variability of materials. Similarly, as shown in Figure 7(b), the loss of stiffness

under repeated loading, as expected, is also related to the cumulative dissipated energy though

they are not highly correlated with each other.

13
 Xiao et al. (2008)

CONCLUSIONS

The following conclusions were determined based upon the limited experimental data

presented regarding the fatigue life of the modified binder and mixtures for the materials tested

for this research project:

 The combination of the crumb rubber and WMA additive in asphalt binder is beneficial

for improving the rheological properties of both the unaged and aged binders (e.g.

increase G*sin δ and reduce G*/sin δ values), The increase in the mixing and

compaction temperatures due to the addition of crumb rubber can be offset by adding the

warm asphalt additives, which lowers the mixing and compaction temperatures of

rubberized mixtures comparable to conventional HMA.

 The experimental results indicated that fatigue life and stiffness of the rubberized WMA

mixture from aggregate A is greater than aggregate B while the cumulative dissipated

energy of mixtures made from aggregate A is slightly lower. Moreover, the fatigue life

of the mixtures made with crumb rubber and WMA additive is greater than the control

mixtures (no rubber and WMA additive), except the mixtures containing Asphamin®

additive.

 Statistical analysis results illustrated that there are no significant differences in the

stiffness and cumulative dissipated energy values for overall mixtures (control,

rubberized, or WMA mixtures) while fatigue life values from control mixtures are

significantly different with other rubberized mixtures. In addition, statistical results

presented the aggregate sources play a key role in determining fatigue life, stiffness and

cumulative dissipated energy values of mixtures.

14
 Xiao et al. (2008)

 There are good correlations between the fatigue life and G*sin δ as well as mixture and

binder stiffness values. The exponential function forms are efficient in achieving the

correlations between the dissipated energy and load cycle as well as mixture stiffness

and load cycles.

ACKNOWLEDGMENTS

The financial support of South Carolina Department of Health and Environmental

Control (SC DHEC) is greatly appreciated. However, the results and opinions presented in this

paper do not necessarily reflect the view and policy of the SC DHEC.

15
 Xiao et al. (2008)

REFERENCES
1. Wang, L.B., Wang, X., Mohammad L., and Wang, Y.P. (2004) “Application of Mixture
Theory in the Evaluation of Mechanical Properties of Asphalt Concrete.” Journal of
Materials in Civil Engineering, Vol. 16, 167-174.

2. You, A., and Buttlar, W.G. (2004) “Discrete Element Modeling to Predict the Modulus of
Asphalt Concrete Mixtures.” Journal of Material in Civil Engineering, Vol. 16, 140-146.

3. Xiao F., Amirkhanian S.N. and Juang H.C. (2009) “Prediction of Fatigue Life of
Rubberized Asphalt Concrete Mixtures Containing Reclaimed Asphalt Pavement Using
Artificial Neural Networks” Journal of Materials in Civil Engineering, Vol. 21, No. 6, pp.
253-261, 2009

4. Daniel, J. S., and Kim, R. Y. (2001) “Laboratory Evaluation of Fatigue Damage and
Healing of Asphalt Mixtures.” Journal of Materials in Civil Engineering, Vol. 13, 434-
440.

5. Huang, B., Li, G., Pang, S.S., and Eggers, J.H., (2004) “Investigation into waste tire
rubber-filled concrete,” Journal of Materials in Civil Engineering, Vol. 16, pp. 187-194

6. Xiao, F., Amirkhanian, S.N., and Juang, H.C. (2007) “Rutting Resistance of the Mixture
Containing Rubberized Concrete and Reclaimed Asphalt Pavement” Journal of Materials
in Civil Engineering, Vol. 19, 475-483

7. Xiao F. and Amirkhanian S.N. (2009 “Asphalt Binder Rheology Sensitivity Investigation
on Resilient Modulus of Rubberized Mixtures Using Artificial Neural Network
Approach” Journal of Testing and Evaluation (ASTM), Vol. 37, No.2, pp. 129-138, 2009

8. Xiao F., Amirkhanian S.N., Shen J., and Putman B.J. (2009) “Influences of Crumb
Rubber Size and Type on Reclaimed Asphalt Pavement (RAP) Mixtures” Construction
and Building Materials, Vol. 23(2), pp.1028-1034

9. Xiao F., Amirkhanian S.N., and Shen J.N. (2009) “Effects of Long Term Aging on
Laboratory Prepared Rubberized Asphalt Binders” Journal of Testing and Evaluation
(ASTM), DOI: 10.1520/JTE101706.

10. Hurley, G., and Prowell, B., (2005a) “Evaluation of Aspha-Min® for use in Warm Mix
Asphalt,” NCAT Report 05-04, Auburn

11. Hurley, G., and Prowell, B., (2005b) “Evaluation of Sasobit® ® for use in Warm Mix
Asphalt,” NCAT Report 05-06, Auburn

12. Gandhi, T. and Amirkhanian, S., (2007) “Laboratory Investigation of Warm Asphalt
Binder Properties – A Preliminary Investigation,” MAIREPAV5 Proceedings, Vol. 5, pp.
475-480, Park City, Utah

16
 Xiao et al. (2008)

13. Kristjansdottir, O., Muench, S, Michael, L, and Burke, G., (2007) “Assessing Potential
for Warm-Mix Asphalt Technology Adoption,” Transportation Research Record: Journal
of the Transportation Research Board, No. 2040, pp. 91-99

14. Wasiuddin, N., Selvamohan, S., Zaman, M., and Guegan, M., (2007) “Comparative
Laboratory Study of Sasobit® and Aspha-min Additives in Warm-Mix Asphalt,”
Transportation Research Record: Journal of the Transportation Research Board, No. 1998,
pp. 82-88

15. Prowell, B., Hurley, G., and Crews, E., (2007) “Field Performance of Warm-Mix Asphalt
at National Center for Asphalt Center for Asphalt Technology Test Track,”
Transportation Research Record: Journal of the Transportation Research Board, No. 1998,
pp. 96-102

16. Akisetty C. K., (2008) “Evaluation of Warm Asphalt Additives on Performance

Properties of CRM Binders and Mixtures,” Ph.D. dissertation, Clemson University,
Clemson, SC

17. Monismith, C.L., Epps, J.A., and Finn, F.N. (1985) “Improved Asphalt Mix Design.”
Proceedings of Association of Asphalt Paving Technologists, Vol. 54. 347-406

18. Tayebali, A.A. (1992) “Re-calibration of Surrogate Fatigue Models Using all Applicable
A-003A Fatigue data.” Technical memorandum prepared for SHRP Project A-003A.
Institute of Transportation Studies, University of California, Berkeley.

19. Rowe, G.M. (1993) “Performance of Asphalt Mixtures in the Trapezoidal Fatigue Test.”
Proceedings of Association of Asphalt Paving Technologists, Vol. 62, 344–384.

20. Hossain, M., Swartz, S., and Hoque, E. (1999) “Fracture and Tensile Characteristics of
Asphalt Rubber Concrete.” Journal of Materials in Civil Engineering, Vol. 11, 287-294.

21. Birgisson, B., Soranakom, C., Napier, J. A. L., and Roque, R. (2004) “Microstructure and
Facture in Asphalt Mixtures Using a Boundary Element Approach.” Journal of Materials
in Civil Engineering, Vol. 16, 116-121.

22. Shen S. and Carpenter S., (2005) “Application of the dissipated energy concept in fatigue
endurance limit testing,” Transportation Research Record: Journal of the Transportation
Research Board, No. 21929, pp. 165-173

23. Xiao F.P., Amirkhanian S.N., Shen J.N., and Putman B.J. (2008) “Influences of Crumb
Rubber Size and Type on Reclaimed Asphalt Pavement (RAP) Mixtures” Construction
and Building Materials, (In press)

24. Xiao, F.P. (2006) “Development of Fatigue Predictive Models of Rubberized Asphalt
Concrete (RAC) Containing Reclaimed Asphalt Pavement (RAP) Mixtures.” Ph. D
dissertation, Clemson University.

17
 Xiao et al. (2008)

Table 1 Aggregate property of mixtures

Aggregate LA Absorption Sand
Abrasion Specific Gravity Soundness % Loss at 5 Cycles Hardness
Source Loss (%) (%) Equivalent
Dry SSD 11/2
Apparent 3/4 to3/8 3/8 to #4
(BLK) (BLK) to3/4
A 51 0.80 2.740 2.770 2.800 0.2 0.1 0.1 - 5
B 34 0.60 2.780 2.800 2.830 0.4 0.6 0.9 35 5

Table 2 Mixing and compaction temperatures of mixtures

Mixing temperature Compaction temperature
(°C) (°C)
ACO/BCO 152-158 132-138
ARO/BRO 170-176 152-158
ARA/BRA 155-165 155-165
ARS/BRS 155-165 155-165
Note: ACO/BCO; ARO/BRO; ARA/BRA; and ARS/BRS-control; rubberized; rubberized
Asphamin® ; and rubberized Sasobit® from aggregates A and B, respectively

Table 3 Binder properties

Unaged binder Aged binder (RTFO+PAV)
Viscosity Std. G*/sinδ Std. G*sinδ Std. Stiffness Std.
Pa.s (135°C) kPa (64°C) kPa (25°C) Mpa (-12°C)
PG 64-22 0.41 0.0 1.2 212.1 2970.0 572.8 221.0 20.5
PG 64-22R 1.60 0.0 3.7 60.8 1705.6 66.1 128.5 2.5
PG 64-22RA 1.48 0.1 4.7 631.9 2042.1 3.9 148.0 1.2
RS
PG 64-22 1.44 0.1 5.2 469.4 2160.3 170.2 150.5 0.6
Note: R-rubberized; A-Asphamin® ; S-Sasobit®

18
 Xiao et al. (2008)

Table 4 Typical analyzed fatigue test results (BCO-B)

Period Number Stress Strain Dynamic Stiffness Phase Angle Dissipated Energy Cumulative Engergy
3 3
Cycles Pa m/m Pa Degree J/m J/m
50 1860.89 1.42E-04 1.31E+07 72 0.21 63.70
100 2110.50 1.53E-04 1.37E+07 72 0.26 130.03
250 2050.47 1.73E-04 1.19E+07 72 0.28 207.84
500 1706.07 1.25E-04 1.36E+07 72 0.17 308.91
1000 1864.05 1.72E-04 1.08E+07 72 0.26 361.95
1600 2074.17 1.74E-04 1.19E+07 72 0.29 382.90
2500 1875.11 1.65E-04 1.14E+07 72 0.25 445.51
5000 1750.30 1.66E-04 1.06E+07 72 0.23 484.11
10000 1728.18 1.79E-04 9.68E+06 72 0.25 537.07
15850 1497.53 1.91E-04 7.84E+06 72 0.23 567.47
19954 1289.00 1.66E-04 7.78E+06 72 0.17 581.37
25120 980.96 1.56E-04 6.29E+06 72 0.12 594.70

Table 5 Statistical analysis of mechanical properties in terms of aggregate sources

Test properties (Agg. A & B)
P-value
Cumulative energy Stiffness Fatigue life
Control 0.431 0.059 0.048
Rubberized 0.297 0.011 0.030
Rubberized+Asphmin 0.293 0.005 0.002
Rubberized+Sasobit 0.036 0.205 0.033
Note: P-value < α = 0.05 (significant difference); P-value > α = 0.05 (No significant difference)

Table 6 Statistical analysis of mechanical properties in terms of mixture types

Mixture type (0-control, 1-rubberized, 2-rubberized+Asphmin, 3-rubberized+Sasobit)
P-value
0~1 0~2 0~3 1~2 1~3 2~3
Cumulative energy 0.017 0.065 0.036 0.149 0.346 0.141
Stiffness 0.372 0.073 0.236 0.129 0.394 0.163
Fatigue life 0.194 0.194 0.226 0.125 0.145 0.308
Note: P-value < α = 0.05 (significant difference); P-value > α = 0.05 (No significant difference)

19
 Xiao et al. (2008)

100

80

Percent passing (%)

60

40
Agg. A
20 Agg. B
Low Range
Up Range
0
0.075 0.15 0.60 2.36 4.75 9.5 12.5 19.0
Sieve size (mm)

Figure 1 Gradations of aggregate A and B

20
 Xiao et al. (2008)

Figure 2 Fatigue beam and testing

21
 Xiao et al. (2008)

800000
8x10 5
Aggregate A Aggregate B

600000

Fatigue life (cycle)

6x10 5

400000
4x10 5

2x10 5
200000

00
CO RO RA RS
Mixture type
(a)
2.0x10 7
20000000
Aggregate A Aggregate B

1.5x10 7
15000000
Stiffness (kPa)

1.0x10 7
10000000

5000000
0.5x10 7

00
CO RO RA RS
Mixture type
(b)

1,000
Cumulative dissipated energy (J/m 3 )

Aggregate A Aggregate B
800

600

400

200

0
CO RO RA RS
Mixture type
(c)

Figure 3 Mechanical properties (a) Fatigue life; (b) Stiffness; (c) cumulative energy

22
 Xiao et al. (2008)

1000000
10 6
100000
10 5

Fa tigue life (cycle)

10000
10 4

10 3 y A = -141.59x + 562241
1000
y B = -19.604x + 96617
100
10 2
10
10
Aggregate A Aggregate B
10
1000 2000 3000 4000
G*sinδ (kPa)
(a)
100000000
10 8
Stiffness (kPa)

10000000
10 7

y A = 4095.5x + 8E+06
10 6
1000000 y B = -15759x + 2E+07

Aggregate A Aggregate B
100000
10 5
100 150 200 250
Binder stiffness (MPa)
(b)
Figure 4 Correlations, (a) fatigue life and G*sinδ; (b) mixture stiffnes and binder stiffness

23
 Xiao et al. (2008)

0.3 0.3 0.3 0.3

y A = 0.1508e-7E-06x y A = 0.1108e-6E-06x ARO-A y A = 0.1454e-4E-06x y A = 0.1663e-3E-06x
y B = 0.1468e-4E-06x y B = 0.117e-5E-06x ARO-B y B = 0.1647e -8E-06x y B = 0.1776e-8E-06x

Dissipated energy (J/m 3 )

Dissipated energy (J/m 3 )

Dissipated energy (J/m 3 )
Dissipated energy (J/m 3 ) ARO-C
y C = 0.1249e-2E-06x
0.2 0.2 ARO-D 0.2 0.2

0.1 0.1 0.1 0.1

ACO-A
ACO-B y C = 0.1372e-2E-06x
ACO-C y D = 0.1499e -2E-06x ARA-A ARA-B ARS-A ARS-B
0.0 0.0 0.0 0.0
2 10 10 2 10 4 100000 10 6
10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 10
10 100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10
10 10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 100 10 3 10000
1000 10 5 1000000
Cycles Cycles Cycles Cycles

(a) (b) (c) (d)

0.3 0.4 0.4 0.4
y A = 0.1726e -1E-05x y A = 0.2175e-4E-05x y A = 0.2516e-2E-05x
y B= 0.1912e-2E-05x y B= 0.2378e-1E-05x y B = 0.2455e-6E-06x

Dissipated energy (J/m 3 )

Dissipated energy (J/m 3 )
Dissipated energy (J/m 3 )

Dissipated energy (J/m 3 )

0.3 y C = 0.2157e-1E-05x 0.3 0.3

0.2 y D = 0.2639e-4E-05x

0.2 0.2 0.2

0.1 y A = 0.1566e-1E-05x BRO-A

y B = 0.2253e-1E-05x 0.1 BRO-B 0.1 0.1
BRO-C
BCO-A BCO-B BRO-D BRS-A BRS-B BRA-A BRA-B
0.0 0.0 0.0 0.0
2
10 10100 10 4 10000 10 6
10 10 31000 10 5 100000 10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 2
10 10100 10 4 10000
10 31000 10 6
10 5 100000 10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000
Cycles Cycles Cycles Cycles
(e) (f) (g) (h)

Figure 5 Dissipated energy versus load cycles (repeatition) (a),(e) control beam (aggregate A
and B); (b),(f) rubberized beam (aggregate A and B); (c),(g) rubberized Asphamin® beam
(aggregate A and B); (d),(h) rubberized Sasobit® beam (aggregate A and B)

24
 Xiao et al. (2008)

10 8
1.0E+08 10 8
1.0E+08 10 8
1.0E+08 10 8
1.0E+08
y A = 9E+06e-6E-06x y A = 9E+06e-6E-06x y A = 9E+06e-5E-06x y A = 1E+07e-3E-06x
y B = 8E+06e -5E-06x
y B = 8E+06e-4E-06x y B = 8E+06e-7E-06x y B = 9E+06e-5E-06x
y C = 8E+06e-1E-06x

Stiffness (kPa)
Stiffness (kPa)

Stiffness (kPa)
Stiffness (kPa)
y C = 8E+06e-2E-06x
y D = 9E+06e-6E-06x

10 7
1.0E+07 10 7
1.0E+07 10 7
1.0E+07 10 7
1.0E+07

ARO-A
ACO-A ARO-B
ACO-B ARO-C
ARO-D ARA-A ARA-B ARS-A ARS-B
ACO-C
10 6
1.0E+06 10 6
1.0E+06 10 6
1.0E+06 10 6
1.0E+06
10 10 2 10 4 100000 10 6
10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10
10 10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 10 3 10000
100 1000 10 5 1000000
Cycles Cycles Cycles Cycles

(a) (b) (c) (d)

10 8
1.0E+08 10 8
1.0E+08
-7E-06x
10 8
1.0E+08 10 8
1.0E+08
y A = 1E+07e-1E-05x y A = 8E+06e y A = 1E+07e-6E-05x y A = 1E+07e-2E-05x
y B = 1E+07e-3E-05x y B = 1E+07e -3E-05x y B = 2E+07e-2E-05x y B = 1E+07e-7E-06x
y C = 1E+07e-9E-06x

Stiffness (kPa)

Stiffness (kPa)
Stiffness (kPa)
Stiffness (kPa)

y D = 2E+07e-3E-05x

10 7
1.0E+07 10 7
1.0E+07 10 7
1.0E+07 10 7
1.0E+07

BRO-A
BRO-B
BRO-C
BCO-A BCO-B BRO-D BRA-A BRA-B BRS-A BRS-B
10 6
1.0E+06 10 6
1.0E+06 10 6
1.0E+06 10 6
1.0E+06
2
10 10100 10 4 10000 10 6
10 10 31000 10 5 100000 10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000 10 2
10 10100 10 4 10000
10 31000 10 6
10 5 100000 10 10
10 2
100 1000 10 4 100000
10 3 10000 10 6
10 5 1000000
Cycles Cycles Cycles Cycles

(e) (f) (g) (h)

Figure 6 Stiffness versus load cycles (repeatition) (a),(e) control beam (aggregate A and B);
(b),(f) rubberized beam (aggregate A and B); (c),(g) rubberized Asphamin® beam (aggregate A
and B); (d),(h) rubberized Sasobit® beam (aggregate A and B)

25
 Xiao et al. (2008)

1000000
10 6

Repetition (cycle)
100000
10 5

10000
10 4 y A = 63.372x + 243145
y B = 163.28x - 40713
Aggregate A Aggregate B
10 3
1000
100 600 1100
Cumulative disspated energy (J/m 3 )
(a)
100000007
10
Loss of stiffness (kPa)

1000000
10 6
y A = 11.92x + 7E+06
y B = 2121.7x + 3E+06

Aggregate A Aggregate B
100000
10 5
100 600 1100
Cumulative dissipated energy (J/m 3 )
(b)

Figure 7 Cumulative dissipated energy versus (a) repetition and (b) stiffness

26