The evaluation of epoxy asphalt and epoxy asphalt mixtures

The Evaluation of Epoxy Asphalt and Epoxy Asphalt Mixtures

Jack Youtcheff

TITLE

Nelson Gibson TITLE

Federal Highway Administration

McLean, Virginia

Aroon Shenoy TITLE

Ghazi Al-Khateeb

TITLE

Soil and Land Use Technology, Inc. McLean, Virginia

Acknowledgements

The authors wish to express their gratitude to Dr. Mihai Marasteanu (University of Minnesota), and Dr. William Buttlar (University of Illinois) for the valuable fracture data they provided, Bob Gaul of ChemCo Systems for providing the epoxy asphalt, and the laboratory support staff, Terry Arnold, Susan Needham, Scott Parobeck, and Frank Davis at the Turner-Fairbank Highway Research Centre.

Proceedings of the Canadian Technical Asphalt Association, Vol. 51, pp. 351-368 (2006)

352 EVALUATION OF EPOXY ASPHALT AND EPOXY ASPHALT MIXTURES

ABSTRACT The paper describes the results obtained in our laboratory as part of an international project under the Organization for Economic Co-operation and Development (OECD) Road Transport Research program. The main objective is to determine whether a surface pavement with radically extended and maintenance-free service life is economically viable on high trafficked highways, when costs of road user delays due to road maintenance and rehabilitation are considered. Toward this objective, the characteristics of an epoxy asphalt binder are compared with those of a number of polymer-modified binders using the SuperpaveTM binder protocols. In addition, asphalt mixes were prepared with each binder type utilizing similar aggregates and mix designs. Tests conducted with the Superpave Shear Tester (SST), Hamburg Wheel Tracking Device (WTD), Pine Wheel Tester, Thermal Stress Restrained Specimen Test (TSRST), and Simple Performance Test (SPT) indicated that the epoxy asphalt mix is a stiffer material and exhibits excellent resistance to rutting, moisture damage, and fatigue. The Superpave binder low temperature tests were considerably higher than those from the TSRST however, results from fracture testing with the Compact Tension Test and Semicircular Bending Test qualitatively suggest the epoxy asphalt to be more resistant to low temperature thermal cracking.

RÉSUMÉ L’exposé décrit les résultats obtenus dans nos laboratoires comme partie d’un projet international du programme de recherche en route et transport de l’Organisation de Coopération et de Développement Économique (OCDE). L’objectif principal est de déterminer si un revêtement de surface avec une vie de service radicalement prolongée et sans entretien est économiquement viable sur les routes à forte circulation quand on considère les coûts des retards des usagers causés par l’entretien et la réhabilitation de la route. Dans ce but, les caractéristiques d’un bitume époxy sont comparées au moyen des protocoles des bitumes SuperpaveTM avec celles d’un certain nombre de bitumes polymères. En outre, des enrobés bitumineux ont été préparés avec chacun des types de bitume en utilisant des granulats et des formulations semblables. Les essais exécutés avec le test de cisaillement Superpave (SST), le manège à roue Hamburg WTD), le test à la roue Pine, le test de contrainte thermique à retrait empêché (TSRST) et le test de performance simple (SPT) indiquent que l’enrobé au bitume époxy est un matériau plus rigide et montre une excellente résistance à l’orniérage, aux détériorations à l’humidité et à la fatigue. Les essais à basse température des bitumes Superpave étaient considérablement plus élevés que ceux du TSRST. Cependant, les résultats du test de rupture à l’essai compact de traction et à l’essai de flexion semi-circulaire semblent indiquer que le bitume époxy est plus résistant à la fissuration à basse température.

YOUCHEFF, GIBSON, SHENOY & AL-KHATEEB 353

1.0 INTRODUCTION Roadwork associated with the maintenance and rehabilitation of pavements can cause considerable travel delays and congestion. Long life pavements would significantly reduce the need for such activities, as well as the associated inconveniences and user costs. Critical highways with large amounts of traffic are the top candidates for the kind of pavement materials that offer the highest performance and longest life. Such applications may justify the use of premium materials and their associated costs to offset any future user and maintenance and rehabilitation costs, and thus generate overall savings. Epoxy asphalt is such a premium material. Historically, the high cost associated with its placement has limited its use to orthotropic bridge decks. Balala [1] reported on the rational for selecting this binder for the San Mateo Bridge in San Francisco, while Gaul [2] reported on its continued good performance 30 years after placement. Through an international collective of researchers, the Organization for Economic Co-operation and Development (OECD) / European Conference of Ministers of Transport (ECMT) is designing and testing two candidate materials that are likely to meet the technological and economic requirements of long life wearing courses. The first phase of this study was a paper study to evaluate available high performance pavement materials and cost benefits associated with expected performance [3]. Two candidate materials were selected for further study; epoxy asphalt concrete for flexible pavement applications and ultra high performance hydraulic cement concrete for rigid pavement applications. The purpose of this paper is to present laboratory material characterization of epoxy asphalt binder and epoxy asphalt mixture in comparison to other conventional materials for qualitative assessment of expected field performance. A quantitative assessment of performance is planned to follow once the associated accelerated pavement testing is completed. In this study, the epoxy asphalt characterization was integrated into ongoing polymer modification research as part of Transportation Pooled Fund (TPF) study, TPF 5(019) “Full Scale Accelerated Performance Testing for Superpave and Structural Validation” [4, 5]. In this study, a single aggregate and volumetric mix design is used with different modified asphalts. Comparisons are made between these modified asphalt mixtures, as well as other mixtures available to collaborative researchers. Among the issues that ultimately need to be addressed is the performance relative to that of a standard premium mix design, as well as the application and extension of existing models to long life pavements (35-45 years). 2.0 PROJECT DESCRIPTION The approach taken by the OECD working team was to obtain a local source of epoxy asphalt and identify a reference mix; that is a mix that has an extensive history in testing or field evaluation. Mixes made with the epoxy asphalt were then compared with the performance of the reference mix. ChemCo Systems of California supplied the epoxy asphalt used in this study. The SuperpaveTM (Superpave) mixture control was an unmodified Performance Graded (PG) 70-22 asphalt binder with a diabase aggregate of 12.5 mm nominal maximum aggregate size gradation. The optimum asphalt binder content was 5.3 percent by total mass.


Preliminary testing was conducted on the binder properties to become familiar with working with a two-part binder and its curing characteristics. This was followed by a wide range of binder and mix tests that evaluated the performance of epoxy asphalt with regard to rutting, moisture damage, and thermal cracking. As crack propagation, fatigue, and thermal cracking were considered potential shortcomings, several approaches to evaluating fracture properties were investigated. 3.0 BINDER TESTING The Superpave conditioning schemes using the Rolling Thin Film Oven (RTFO) and Pressure Aging Vessel (PAV) are not appropriate for evaluating epoxy asphalt. The uncured epoxy asphalt is too fluid, while the cured binder is more akin to a flexible plastic. Nevertheless, the intermediate and high temperature characteristics of epoxy asphalt can be evaluated using a Dynamic Shear Rheometer (DSR) and the low temperature properties measured on the Bending Beam Rheometer (BBR). A TA Instruments DSR was used for generating dynamic data at different temperatures (16°C, 22°C, 28°C, 34°C, 58°C, 64°C, and 70°C) with a set of torsion bar fixtures. The samples for the test were prefabricated using the BBR mould, then cut into two, such that rectangular sample bars for the DSR torsion tests were obtained measuring 6.3 x 12 x 50 mm in dimension. The data were generated using frequency sweeps covering the range from 0.1 to 100 radians/s at very low levels of strain of about 10-3 to be within the linear viscoelastic range of response. The data at all temperatures and frequencies were consolidated to the reference temperature of 28°C by appropriate shifting using the time-temperature superposition principle.

1.00E+00

1.00E+01

1.00E+02

1.00E+03

1.00E+04

1.00E+05

1.00E+06

1.00E+07

1.00E+08

1.00E+09

1E-07 1E-05 0.001 0.1 10 1000 100000

Reduced Frequency (radians/s)

|G*|/

sin δ

, Pa

Epoxy AsphaltunagedPG 70-22 Controlunaged

Figure 1. Master Plots of the Superpave Complex Modulus ( |G*| ) divided by the Sine of the Phase Angle (δ) Parameter ( |G*|/sin δ ) vs. the Reduced Frequency Curves at a Reference Temperature of

28°C for Epoxy Asphalt and Performance Graded (PG) 70-22


The master plots compared in Figure 1 show that the values of the rheological parameters for the epoxy asphalt are in a very different range than what is normally obtained for conventional asphalts or other polymer-modified asphalts. In order to get some idea of the difference in the rheological behaviour of the epoxy asphalt, comparisons of the properties of the epoxy asphalt are done vis-à-vis a conventional unmodified high performance grade asphalt, namely a PG 70-22 control. It can be seen from the plots in Figure 1, that while the Superpave, Complex Modulus ( |G*| ) divided by the Sine of the Phase Angle (δ) parameter, ( |G*|/sin δ) values are close to each other and follow the same trend at high frequencies or equivalently low temperatures, the same is not the case at low frequencies or equivalently, high temperatures. The epoxy asphalt retains its high Superpave parameter |G*|/sin δ values even at high temperatures, indicating that epoxy asphalts would provide excellent rutting resistance. To assess the effect of Ultraviolet (UV) oxidation, cured beams were conditioned in a Q-Panel Xenon Test Chamber. Conditioning cycles entailed the application of UV light for 55 minutes followed by 5 minutes of misting with chilled water. The surface temperature during the “sunlit cycle” was 90°C. Specimens were conditioned for 24 hr, 1 week, 2 weeks, 4 weeks, 8 weeks, 12 weeks, and 16 weeks. Visual examination of the side exposed to UV showed evidence of hairline cracks and the surface gradually attained a dull weathered look following long exposure times. The cracks do not appear to penetrate very far; presumably the crack growth is arrested by the cross-linking nature of the epoxy system. The rheological results are shown in Figures 2 to 4, respectively for three parameters: the Complex Modulus ( |G*| ), the Phase Angle (δ), and the Superpave parameter |G*|/sin δ at various aging conditions.

1.00E+05

1.00E+06

1.00E+07

1.00E+08

1.00E+09

1E-07 1E-05 0.001 0.1 10 1000 100000


|G*|,

Pa

Epoxy Asphalt unagedEpoxy 24HrEpoxy 1 w eekEpoxy 2 w eeks Epoxy 4 w eeks Epoxy 8 w eeks Epoxy 12 w eeks Epoxy 16 w eeks

Figure 2. Master Plot of the Complex Modulus ( |G*| ) vs. the Reduced Frequency Curves at Reference Temperature of 28°C for Epoxy Asphalt Ultraviolet (UV) Aged for Different Intervals


10

100

0.1 1 10 100Reduced Frequency (radians/s)

δ, d

egre

esEpoxy Asphalt unaged Epoxy 24Hr

Epoxy 1 w eek Epoxy 2 w eeks

Epoxy 4 w eeks Epoxy 8 w eeks

Epoxy 12 w eeks Epoxy 16 w eeks

Figure 3. Master Plot of the Phase Angle (δ) vs. the Reduced Frequency Curves at a Reference Temperature of 28°C for Epoxy Asphalt Ultraviolet (UV) Aged for Different Intervals

1.00E+05

1.00E+06

1.00E+07

1.00E+08

1.00E+09

1E-07 1E-05 0.001 0.1 10 1000 100000


|G*|/

sin δ

, Pa

Epoxy Asphalt unagedEpoxy 24HrEpoxy 1 w eekEpoxy 2 w eeks Epoxy 4 w eeks Epoxy 8 w eeks Epoxy 12 w eeks Epoxy 16 w eeks

Figure 4. Master Plot of the Superpave Parameter ( |G*| ) divided by the Sine of the Phase Angle (δ) Parameter ( |G*|/sin δ ) vs. the Reduced Frequency Curves at Reference Temperature of 28°C for

Epoxy Asphalt UV Aged for Different Times


Over this aging interval, three recognizable changes took place with regard to the properties. The first change occurred after a week of aging, which is most likely the period when the epoxy has gone through the final stages of curing. Following two weeks of aging, there did not appear to be any difference from the one week of aging however, the next radical change occurred after four weeks of UV conditioning. The epoxy then showed virtually no change, even after 16 weeks of UV conditioning. 4.0 MIXTURE TESTING 4.1 High Temperature Permanent Deformation The Simple Performance Test (SPT) protocol for the Flow Number test was carried out at 64°C [6]. This test applies repeated haversine loads with rest periods. This test is intrinsically designed to characterize the viscoplastic properties of the mixture intended for a rutting performance evaluation. The results, accumulated permanent strain as a function of loading cycle, are shown in Figure 5. The ALF mixtures experience much higher permanent deformations than the epoxy asphalt mixtures. The figure also demonstrates the significant differences between the unmodified mixture (PG 70-22) and the modified mixtures (CR-TB and SBS LG), as well as between the modified mixtures and those with epoxy asphalt. The new ALF epoxy refers to the same gradation and volumetric conditions as the other mixtures. The old ALF epoxy refers to a more coarse graded mixture. For both epoxy asphalt mixes, very little permanent deformation is seen relative to the other ALF mixtures.

0

20000

40000

60000

80000

100000

0 2000 4000 6000 8000 10000 12000

Cycle Number

Perm

anen

t Stra

in

(mic

rost

rain

s)

PG 70-22SBS LGCR-TBOld EpoxyNew Epoxy

Figure 5. Accumulated Permanent Strains vs. Load Cycles for ALF and Epoxy Asphalt Mixtures from the Simple Performance Test Flow Number

A second type of SPT test protocol for Dynamic Modulus ( |E*| ) was completed on the ALF and epoxy asphalt mixtures with a slight deviation from the specified temperatures and frequencies. The Dynamic


Modulus ( |E*| ) test was conducted at a range of temperatures and frequencies to characterize the ALF mixtures, as well as the epoxy asphalt mixtures. As can be seen in the master curves in Figure 6, epoxy asphalt was significantly stiffer at all temperatures and frequencies. The implications of this again reflect that negligible rutting would be expected with this material. However, one must take caution when using these properties in any existing mechanistic empirical models such as those in the National Cooperative Highway Research Program (NCHRP) 1-37A methodology [7].

1

10

100

1,000

10,000

100,000

-10 -8 -6 -4 -2 0 2 4 6Reduced Frequency (Hz)

|E*|

(Mpa

)

Epoxy Asphalt

Control

Air blown

Terpolymer

SBS 64-40

CR-TB

SBS LG

Figure 6. Comparison of Dynamic Modulus ( |E*| ) for Epoxy Asphalt and ALF Mixtures

While the |E*| follows time temperature superposition concept, another fundamental viscoelastic property - the phase angle - does not, as shown in Figure 7. This indicates a degree of cross-linking in the polymer matrix of the epoxy. The implications of this have little ramifications toward secondary distress or rutting prediction, but any advanced viscoelastic material characterization and prediction across multiple temperatures will rely upon the inherent viscoelastic relaxation modulus. Further, the master curve shape indirectly indicates damage and fracture development can be much different and less susceptible than for convention unmodified or modified asphalt mixture when inspecting the maximum slope of the |E*| master curve (in log-log space). Kim et al. [8] studied, among others, a material property related to the viscoelastic crack speed and energy releaser rate. It was found to be strongly associated, a Coefficient of Determination (R2) of 0.82, with the slope of the viscoelastic modulus master curve through a large series of experiments on fatigue damage of mastics and assorted binders.


0

5

10

15

20

25

-8 -6 -4 -2 0 2 4

Reduced Frequency (Hz)

Phas

e A

ngle

(deg

)

4oC

19oC31oC

46oC

58oC

Figure 7. Dynamic Modulus ( |E*| ) Phase Angle of Epoxy Asphalt Mixture 4.2 Moisture Damage Two moisture damage torture tests were used to evaluate the susceptibility of the epoxy asphalt mix. Rutting induced by the Hamburg Wheel Tracking Device run at 64°C for 40,000 passes was less than 2 mm and these conditions were 6°C higher and 20,000 passes longer than typical tests [9, 10]. Rutting induced by the second method, the Pine Wheel Tester at 60°C for 40,000 passes, was less than 1 mm. Here too, the test conditions used were considerably more severe than standard [10]. Essentially negligible amounts of rutting and moisture damage were observed. It can be inferred that stripping and rutting distresses are of little concern with epoxy asphalt mixtures, although further confirmation is required with additional aggregate types and gradations. 4.3 Fracture

4.3.1 Fatigue to Fracture Initiation Direct axial, reversed tension-compression, cyclic loading can be used to evaluate an asphalt mixture’s susceptibly to fatigue damage. These tests were performed at 19°C on 150 mm tall by 71 mm diameter cylindrical specimens with specially designed grips and deformation characterization very similar in spirit to that advanced by other researchers [11]. As part of initial stages of fatigue characterization under the United States Federal Highway Administration (FHWA) Accelerated Loading Facility (ALF) research program, these tests were preformed on the epoxy asphalt and also the Terminally Blended Crumb Rubber


mixture (CR-TB) from the ALF for comparison, as well. This particular test applies a prescribed strain at the platens, which are controlled by the actuator and resulting strains are measured from three Linear Variable Displacement Transducers (LVDTs) mounted over the centre of the specimen. This test is neither strain nor stress controlled when observing the strain over the centre of the specimen so as to be away from any edge effects at the adhesive bond between the asphalt mixture and the loading platens. Both the CR-TB and epoxy asphalt mixture were controlled at a platen-to-platen strain level of about 1,300 microstrains. The target was to obtain 1,000 microstrains, the mean transverse and longitudinal tensile strain level observed at the bottom of the Asphalt Cement (AC) layer under the ALF. The CR-TB mixture achieved an on-specimen strain level of about 1,000 microstrains and fracture was induced in the specimen at about 6,200 loading cycles where the test was stopped. The epoxy asphalt specimens achieve approximately 250 microstrains at the centre on-specimen LVDT location from 1,300 microstrains at the platens. The fatigue hysteresis loops of the stress-strain for both the CR-TB and the epoxy asphalt mixtures are shown in Figure 8. The CR-TB experiences some fatigue deterioration at 6,200 loading cycles as noted by the changes in the hysteresis loop as the test progresses. On the other hand, the hysteresis loop of the epoxy mixture at 100,000 loading cycles was identical to that at the beginning of the fatigue test, which indicates that the epoxy mixture did not show any fatigue deterioration. Negligible reduction in stress was observed at 100,000 loading cycles. This indicates a significant strain gradient is tolerated by the epoxy asphalt and can contribute to the relatively large stiffness and non-linear material response. It is difficult to make estimations of fatigue properties of the epoxy asphalt mixture when comparing it to the laboratory result of the CR-TB or its fatigue cracking performance under the ALF.

-1,500

-1,000

-500

0

500

1,000

1,500

-1,500 -1,000 -500 0 500 1,000Strain (µε )

Stre

ss (k

Pa)

Terminal Blend Crumb Rubber, Cycles 1-3Terminal Blend Crumb Rubber, Cycles 157-160Terminal Blend Crumb Rubber, Cycles 6,197- 6,200Epoxy Asphalt, Cycles 1-3Epoxy Asphalt, Cycles 999,997- 100,000

Figure 8. Fatigue Hysteresis Loops of Epoxy Asphalt and Terminally Blended Crumb Rubber

Mixtures (Same Aggregate and Volumetric Properties)

Figure 9 shows a plot of the cumulative crack length versus the number of ALF wheel passes [12]. Qualitatively, it can be said that the epoxy asphalt mixture should have better fatigue performance than the polymer-modified mixtures given that no damage could be ascertained in the mixture under 100,000 cycles of 250 microstrain loading; a condition non-epoxy asphalt mixtures should sustain damage in the


form of a reduction in modulus. Much more intensive fatigue characterization continues under laboratory experiments.

0

20

40

60

80

100

120

0 50000 100000 150000 200000 250000 300000 350000

Number of ALF Passes

Cum

ulat

ive

Cra

ck L

engt

h (m

) L3S3 (Air Blown)L2S3 (Control)L5S3 (CR-TB)L6S3 (Terpolymer)L4S3 (SBS LG)L7S3 (Fibers)L1S2 (CR-AZ)

Figure 9. Cumulative Crack Length vs. Number of Accelerated Loading Facility (ALF) Passes

Indirect Tensile (IDT) strength and resilient modulus were completed on the ALF mixtures (cores directly taken from the ALF lanes) and the epoxy asphalt mixture at 19°C to examine their fracture energy ratios [13]. The fracture energy is calculated from the stress-strain curves that are corrected up to the point of fracture, where the resilient modulus curve is used to remove any post-fracture effects. While strength values in the range of 550 to 1,350 kPa were achieved with the ALF polymer modified mixtures, the epoxy asphalt mixture reached the capacity of the test machine (load of 22,240 Newtons) reaching a stress of 1,885 kPa but did not reach the tensile strength of the epoxy asphalt mixture. The maximum tensile strain in the epoxy asphalt mixture before reaching the load capacity was on the order of 865 microstrains, whereas the ALF mixtures reached fracture between 1,250 and 2,000 microstrains. This is a reflection of the reversed tension-compression axial fatigue tests indicating that the epoxy asphalt concrete can sustain larger stresses and strains before fracture appears in the mixture. This lends itself to, qualitatively, significantly larger fatigue life and fracture properties than conventional modified and unmodified asphalt mixtures. 4.3.2 Direct Fracture and Low Temperature Cracking Tests Three different laboratory characterization tests were performed on the epoxy asphalt mixture. • Thermal Stress Restrained Specimen Test (TSRST). • Compact Tension Tests. • Semicircular Bending Tests.


The TSRST was performed on the epoxy asphalt mixture following AASHTO TP10-93 [14] as a means to understand the material’s performance under extreme cooling events where low temperature thermal cracking becomes an issue. A cooling rate of 15°C/hr was used. The TSRST finding for epoxy asphalt mixtures shows a lower fracture temperature of -26°C. This is considerably lower than the Superpave binder criteria from the slope of the creep stiffness curve in the BBR (m) and the Flexural Creep Stiffness (S )values, which grade the un-aged cured epoxy asphalt a PG X-10. The disc-shaped Compact Tension Test takes cues from an established metal fracture test and was pioneered in its application to asphalt concrete by the work of Wagoner, Buttlar, and Paulino [15] and Wagoner et al. [16]. An overview of the test specimen geometry is shown in Figure 10. A pre-cut notch is placed in the specimen and fractures as the specimen is pulled apart. Load and crack mouth opening displacement are measured and the fracture energy is calculated as the area under the load-crack mouth opening displacement curve divided by the area of the fracture face (Figure 11). The tests were performed over a range of temperatures; the test results are shown in Table 1.

Figure 10. Disc Shaped Compact Tension Sample and Test Configuration (Courtesy Bill Buttlar, University of Illinois)

As can be seen, the fracture energy decreases with decreasing temperature. The magnitudes of the epoxy asphalt fracture energy at -18°C are compared to those from a field section studied by Wagoner et al. [16]. Mixtures with PG 64-22, PG 58-28, and PG 58-34 binder used in the different sections indicated fracture energies at -20°C of about 220 Joules/square metre (J/m2) to 190 J/m2 (PG 64-22 mixtures), about 305 J/m2 (PG 58-22) and about 305 J/m2 to 350 J/m2 (PG 58-34). The epoxy asphalt mixture at -18°C exhibited fracture energies on the order of 610 J/m2, which is significantly larger than conventional mixtures indicating better resistance to low temperature thermal cracking. An engineered interlayer in the study exhibited fracture energy of 1,400 J/m2 at -20°C, but at -10°C the mixture did not truly fracture as a single crack did not propagate but smaller distributed cracking or blunting occurred ahead of the notch. Similar behaviour was observed for the epoxy asphalt at warmer temperatures. When the epoxy asphalt truly fractured, the bond between the aggregate and the epoxy asphalt binder remained strong as visually observed at the fracture face where little epoxy asphalt binder de-bonding from the bare face of stones was observed. The straight fracture of the epoxy asphalt indicated only “Mode I” or opening type of fracture occurred, which is important for performance prediction models.


CMOD is Crack Mouth Opening Displacement

Figure 11. Typical Disc Shaped Compact Tension Test Data

(Courtesy Bill Buttlar, University of Illinois)

Table 1. Test Results from Compact Tension Test (Courtesy Bill Buttlar, University of Illinois)

Test Temperature

(°C)

Fracture Energy (J/m2)

Peak Load (kN) Notes

+19 Did not fracture, no data collected

+10 4354.3 5.08 Crack mouth opening displacement gage reached maximum value stopped test early, one test

-6 931.4 7.24 One sample hit load limit, had to restart test, did not include in this summary

-18 610.9 7.74 Two good tests Semicircular bending tests on asphalt concrete have recently been studied [17] to evaluate the resistance asphalt concrete to fracture at temperatures below -20°C. A schematic of the test is shown in Figure 12, where the specimen can be created from Superpave gyratory compactor specimens or field cores. Much like the disc shaped compact tension specimen, the semicircular bending specimen is fabricated with a pre-cut notch. The specimen is loaded vertically in a direction parallel to the crack propagation. This is in contrast to the disc shaped compact tension test where the specimen is loaded perpendicular to the crack. Based on the specimen size and geometry and the peak load, the critical stress intensity factor (fracture toughness) can be calculated.


CMOD is Crack Mouth Opening Displacement

Figure 12. Schematic of Semicircular Bending Test

The semicircular bending test was conducted on the epoxy asphalt and a traditional mixture with PG 64-22 asphalt cement at -18 and -30°C. The results are summarized in Table 2, while the load versus load line displacement chart for the -30°C testing is shown in Figure 13.

Table 2. Summary of Semi-Circular Bending Test Data (Courtesy Mihai Marasteanu, University of Minnesota)

Temperature (°C) Mixture Fracture Energy (N/m) [COV %]

Fracture Toughness, KIC (Mpa m ) [COV %]

-18 Epoxy Asphalt 690 [19%] 2.3 [10%] -18 PG 64-22 506 [22%] 0.92 [7%] -30 Epoxy Asphalt 447 [14%] 2.4 [4%] -30 PG 64-22 274 [6 %] 1.0 [3%]

Note: PG is Performance Graded, COV is Coefficient of Variance


0123456789

-0.05 0 0.05 0.1 0.15 0.2 0.25 0.3LLD (mm)

Load

(kN

)Epoxy1-2-1-30Epoxy2-2-1-30Epoxy3-20-1-3064-2864-2864-28

Figure 13. Semicircular Bending Test Results at -30°C (Courtesy Mihai Marasteanu, University of Minnesota)

For the two mixtures compared, the fracture toughness tends to offer a more discriminative material property and less variation than the fracture energy, which is also seen graphically amongst the three replicates. These tests qualitatively support the findings of the other tests that epoxy asphalt mixture is much more resistant to low temperature thermal cracking than conventional asphalt concrete mixtures. 5.0 CONCLUSIONS It has been observed that properties of the epoxy asphalt binder cannot be completely characterized by the Superpave binder specification due to the significantly higher level of polymer cross-linking. It is apparent from the characterization tests completed thus far that epoxy asphalt concrete exhibits: • Very little rutting potential as evidenced by comparison with polymer modified dense graded Hot

Mix Asphalt (HMA). • Negligible stripping / moisture damage as per the Hamburg Wheel Tracking Device and Pine Rut

Tester. • Significantly larger resistance to crack initiation and fracture propagation when compared to

conventional asphalt concrete mixtures. Stresses that remain to be evaluated are delamination when epoxy asphalt is applied as an asphalt concrete overlay on top of distressed flexible pavement and/or rigid pavements.


REFERENCES 1. Balala B. “Studies Leading to Choice of Epoxy Asphalt for Pavement on Steel Orthotropic Bridge

Deck of San Mateo-Hayward Bridge”, Highway Research Record 287, Highway Research Board, National Academy of Science, Washington, D.C., 12-18 (1969).

2. Gaul RW. “Epoxy Asphalt Concrete–A Polymer Concrete with 25 Years Experience”, American

Society for Testing and Materials, Special Publication 166-13, 233-251 (1993). 3. Organization for Economic Co-operation and Development (OECD). “Economic Evaluation of

Long-Life Pavements: Phase 1”, 2005(3), Paris, France, March (2005). 4. Qi X, Al-Khateeb G, Mitchell T, Stuart K. “Construction of Accelerated Loading Facility (ALF)

Pooled Fund Study Pavements”, TPF 5(019), Final Report, Turner-Fairbank Highway Research Centre (TFHRC), November (2003).

5. Qi X, Shenoy A, Al-Khateeb G, Arnold T, Gibson N, Youtcheff J, Harman T. “Laboratory

Characterization and Full-Scale Accelerated Performance Testing Crumb Rubber Asphalts and Other Modified Asphalt Systems”, Journal of Road Materials and Pavements Design (accepted for publication 2006).

6. Bonaquist RF. “Simple Performance Tester for Superpave Mix Design”, Quarterly Progress Report

(Appendix B), National Cooperative Highway Research Program Project 9-29, Transportation Research Board, National Research Council, National Academies, Washington, D.C. (2003).

7. Andrei D, Witczak MW, Mirza MW. “Development of a Revised Predictive Model for the

Dynamic (Complex) Modulus of Asphalt Mixtures”, National Cooperative Highway Research Program 1-37A Inter Team Report, University of Maryland, March (1999).

8. Kim Y, Lee HJ, Little DN, Kim YR. “A Simple Testing Method to Evaluate Fatigue Fracture and

Damage Performance of Asphalt Mixtures”, Journal, Association of Asphalt Paving Technologists, 75 (2006).

9. Stuart KD, Youtcheff JS, Mogawer WS. “Understanding the Performance of Modified Asphalt

Binders in Mixtures: Evaluation of Moisture Sensitivity”, Report FHWA-RD-02-029, Federal Highway Administration (FHWA), McLean, Virginia (2002).

10. Pine Instrument Company. “Rotary Asphalt Wheel Tester Operation Manual-LMAFW1A (Rev

002)”, Grove City, Pennsylvania (2003). 11. Chehab GR, Kim YR, Schapery RA, Witczak MW, Bonaquist R. “Characterization of Asphalt

Concrete in Uniaxial Tension using a Viscoelastoplastic Continuum Damage Model”, Journal, Association of Asphalt Paving Technologists 72, 315-355 (2003).

12. Al-Khateeb G, Shenoy A. “A Distinctive Fatigue Failure Criterion”, Journal, Association of

Asphalt Paving Technologists, 73, 585-622 (2004).


13. Roque R, Bjorn B, Christos D. “Development and Field Evaluation of Energy-Based Criteria for Top-Down Cracking Performance of Hot-Mix Asphalt”, Journal, Association of Asphalt Paving Technologists, 73, 229-260 (2004).

14. American Association of State Highway and Transportation Officials (AASHTO) TP10-93,

“Standard Test Method for Thermal Stress Restrained Specimen Tensile Strength”, AASHTO Provisional Standards, Interim Edition, Washington, D.C., April (2001).

15. Wagoner MP, Buttlar WG, Paulino GH. “Disk-Shaped Compact Tension Test for Asphalt Concrete

Fracture”, Journal, Society for Experimental Mechanics, 45(3), 270-277, June (2005). 16. Wagoner MP, Buttlar WG, Paulino GH, Blankenship P. “Laboratory Testing Suite for

Characterization of Asphalt Concrete Mixtures Obtained from Field Cores”, Journal, Association of Asphalt Paving Technologists, 75 (2006).

17. Li X, Marasteanu MO. “Cohesive Modeling of Fracture in Asphalt Mixtures at Low

Temperatures”, International Journal of Fracture (submitted 2004). 18. Li X, Marasteanu MO, Iverson N, Lubuz JF. “Observation of Crack Propagation in Asphalt

Mixtures Using Acoustic Emission”, Pre-Print CD-ROM, 85th Annual Meeting of the Transportation Research Board, National Research Council, National Academies, Washington, D.C. (2006).

The evaluation of epoxy asphalt and epoxy asphalt mixtures

Documents